Removed now obsolete CVS substitutions.

nigel.stanger / ouexam

Browse code Removed now obsolete CVS substitutions. master OUEXAM_2_4_2 OUEXAM_2_4_1 OUEXAM_2_4
1 parent eeb1007 commit 4b60553ca9b31f68ed4694ce120ec725475e1b07 Nigel Stanger authored on 13 Jun 2013

Patch

Showing 8 changed files

Ignore Space Show notes View MANIFEST
CreateCluster.pdf - images for the example document CreateIndex.pdf CreateObjectType.pdf CreateTable1.pdf CreateTable2.pdf FAQ INSTALL MANIFEST Makefile README TODO example.tex - example document source files example1.tex example2.tex lstlang0.sty - listings language definitions for the example ouexam.dtx - ouexam class and documentation ouexam.ins - ouexam installation driver ****************************************************************************** File: $Id$ ******************************************************************************** CreateCluster.pdf - images for the example document CreateIndex.pdf CreateObjectType.pdf CreateTable1.pdf CreateTable2.pdf FAQ INSTALL MANIFEST Makefile README TODO example.tex - example document source files example1.tex example2.tex lstlang0.sty - listings language definitions for the example ouexam.dtx - ouexam class and documentation ouexam.ins - ouexam installation driver

Ignore Space Show notes View

MANIFEST

Ignore Space Show notes View Makefile
###################################################################### # # Makefile for ouexam document class. # ###################################################################### SHELL=/bin/sh .PHONY: class doc example install tidy clean # files EXAMPLE1_IMAGES = eg1-1.pdf eg1-2.pdf eg1-3.pdf EXAMPLE2_IMAGES = eg2-2.pdf eg2-3.pdf eg2-4.pdf eg2-5.pdf # Build everything. all: class doc example # Build the class file only. class: ouexam.cls ouexam.cls: ouexam.dtx latex ouexam.ins # Build the documentation only. doc: ouexam.pdf ouexam.pdf: ouexam.dtx $(EXAMPLE1_IMAGES) $(EXAMPLE2_IMAGES) pdflatex $< pdflatex $< pdflatex $< $(EXAMPLE1_IMAGES): example1.pdf pdfjam --outfile $@ $< `echo $@ \| cut -c 5` $(EXAMPLE2_IMAGES): example2.pdf pdfjam --outfile $@ $< `echo $@ \| cut -c 5` # Build the example only. example: example1.pdf example2.pdf example%.pdf: example%.tex example.tex ouexam.cls pdflatex $< pdflatex $< # Install files in the appropriate locations in the texmf tree rooted at # $(TEXMFHOME). Either define this as a shell variable or pass it as # command line argument, e.g., make install TEXMFHOME=/usr/local/texmf. install: class doc example mkdir -p $(TEXMFHOME)/tex/latex/ouexam cp ouexam.cls $(TEXMFHOME)/tex/latex/ouexam mkdir -p $(TEXMFHOME)/doc/latex/ouexam/example cp eg.pdf ouexam.pdf ouexam.dtx ouexam.ins \ FAQ INSTALL MANIFEST README TODO \ $(TEXMFHOME)/doc/latex/ouexam cp Create.pdf example.tex example.pdf lstlang0.sty \ $(TEXMFHOME)/doc/latex/ouexam/example texhash # Clean up: remove temporary files. tidy: rm -f .tmp .aux .out .log .glo .toc .ps \ $(EXAMPLE1_IMAGES) $(EXAMPLE2_IMAGES) # Clean up: remove everything except the original source. clean: tidy rm -f ouexam.pdf eg.pdf example.pdf .cls ###################################################################### # # File: $Id$ # # Makefile for ouexam document class. # ###################################################################### SHELL=/bin/sh .PHONY: class doc example install tidy clean # files EXAMPLE1_IMAGES = eg1-1.pdf eg1-2.pdf eg1-3.pdf EXAMPLE2_IMAGES = eg2-2.pdf eg2-3.pdf eg2-4.pdf eg2-5.pdf # Build everything. all: class doc example # Build the class file only. class: ouexam.cls ouexam.cls: ouexam.dtx latex ouexam.ins # Build the documentation only. doc: ouexam.pdf ouexam.pdf: ouexam.dtx $(EXAMPLE1_IMAGES) $(EXAMPLE2_IMAGES) pdflatex $< pdflatex $< pdflatex $< $(EXAMPLE1_IMAGES): example1.pdf pdfjam --outfile $@ $< `echo $@ \| cut -c 5` $(EXAMPLE2_IMAGES): example2.pdf pdfjam --outfile $@ $< `echo $@ \| cut -c 5` # Build the example only. example: example1.pdf example2.pdf example%.pdf: example%.tex example.tex ouexam.cls pdflatex $< pdflatex $< # Install files in the appropriate locations in the texmf tree rooted at # $(TEXMFHOME). Either define this as a shell variable or pass it as # command line argument, e.g., make install TEXMFHOME=/usr/local/texmf. install: class doc example mkdir -p $(TEXMFHOME)/tex/latex/ouexam cp ouexam.cls $(TEXMFHOME)/tex/latex/ouexam mkdir -p $(TEXMFHOME)/doc/latex/ouexam/example cp eg.pdf ouexam.pdf ouexam.dtx ouexam.ins \ FAQ INSTALL MANIFEST README TODO \ $(TEXMFHOME)/doc/latex/ouexam cp Create.pdf example.tex example.pdf lstlang0.sty \ $(TEXMFHOME)/doc/latex/ouexam/example texhash # Clean up: remove temporary files. tidy: rm -f .tmp .aux .out .log .glo .toc .ps \ $(EXAMPLE1_IMAGES) $(EXAMPLE2_IMAGES) # Clean up: remove everything except the original source. clean: tidy rm -f ouexam.pdf eg.pdf example.pdf .cls

Ignore Space Show notes View

Makefile

######################################################################
#
# Makefile for ouexam document class.
#
######################################################################

SHELL=/bin/sh

.PHONY: class doc example install tidy clean

# files
EXAMPLE1_IMAGES = eg1-1.pdf eg1-2.pdf eg1-3.pdf
EXAMPLE2_IMAGES = eg2-2.pdf eg2-3.pdf eg2-4.pdf eg2-5.pdf

# Build everything.
all: class doc example

# Build the class file only.
class: ouexam.cls
ouexam.cls: ouexam.dtx
	latex ouexam.ins

# Build the documentation only.
doc: ouexam.pdf

ouexam.pdf:	ouexam.dtx $(EXAMPLE1_IMAGES) $(EXAMPLE2_IMAGES)
	pdflatex $<
	pdflatex $<
	pdflatex $<

$(EXAMPLE1_IMAGES): example1.pdf
	pdfjam --outfile $@ $< `echo $@ | cut -c 5`

$(EXAMPLE2_IMAGES): example2.pdf
	pdfjam --outfile $@ $< `echo $@ | cut -c 5`

# Build the example only.
example: example1.pdf example2.pdf

example%.pdf: example%.tex example.tex ouexam.cls
	pdflatex $<
	pdflatex $<

# Install files in the appropriate locations in the texmf tree rooted at
# $(TEXMFHOME). Either define this as a shell variable or pass it as
# command line argument, e.g., make install TEXMFHOME=/usr/local/texmf.
install: class doc example
	mkdir -p $(TEXMFHOME)/tex/latex/ouexam
	cp ouexam.cls $(TEXMFHOME)/tex/latex/ouexam
	mkdir -p $(TEXMFHOME)/doc/latex/ouexam/example
	cp eg*.pdf ouexam.pdf ouexam.dtx ouexam.ins \
		FAQ INSTALL MANIFEST README TODO \
		$(TEXMFHOME)/doc/latex/ouexam
	cp Create*.pdf example*.tex example*.pdf lstlang0.sty \
		$(TEXMFHOME)/doc/latex/ouexam/example
	texhash

# Clean up: remove temporary files.
tidy:
	rm -f *.tmp *.aux *.out *.log *.glo *.toc *.ps \
		$(EXAMPLE1_IMAGES) $(EXAMPLE2_IMAGES)

# Clean up: remove everything except the original source.
clean: tidy
	rm -f ouexam.pdf eg*.pdf example*.pdf *.cls

######################################################################
#
# File: $Id$
#
# Makefile for ouexam document class.
#
######################################################################

SHELL=/bin/sh

.PHONY: class doc example install tidy clean

# files
EXAMPLE1_IMAGES = eg1-1.pdf eg1-2.pdf eg1-3.pdf
EXAMPLE2_IMAGES = eg2-2.pdf eg2-3.pdf eg2-4.pdf eg2-5.pdf

# Build everything.
all: class doc example

# Build the class file only.
class: ouexam.cls
ouexam.cls: ouexam.dtx
	latex ouexam.ins

# Build the documentation only.
doc: ouexam.pdf

ouexam.pdf:	ouexam.dtx $(EXAMPLE1_IMAGES) $(EXAMPLE2_IMAGES)
	pdflatex $<
	pdflatex $<
	pdflatex $<

$(EXAMPLE1_IMAGES): example1.pdf
	pdfjam --outfile $@ $< `echo $@ | cut -c 5`

$(EXAMPLE2_IMAGES): example2.pdf
	pdfjam --outfile $@ $< `echo $@ | cut -c 5`

# Build the example only.
example: example1.pdf example2.pdf

example%.pdf: example%.tex example.tex ouexam.cls
	pdflatex $<
	pdflatex $<

# Clean up: remove temporary files.
tidy:
	rm -f *.tmp *.aux *.out *.log *.glo *.toc *.ps \
		$(EXAMPLE1_IMAGES) $(EXAMPLE2_IMAGES)

# Clean up: remove everything except the original source.
clean: tidy
	rm -f ouexam.pdf eg*.pdf example*.pdf *.cls

Ignore Space Show notes View example.tex
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % % Note: No document class --- this is handled by the appropriate % "driver" file, which makes it easier to switch class options % such as markingschedule. Doing it this way means that you don't % wipe out the original exam paper when you want to generate a % marking schedule. % %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% \usepackage{graphicx} \usepackage{listings} \examyear{2001} \department{Information Science} \papernumber{INFO 321} \papertitle{Database Systems} \instructions{Answer \underline{ALL} questions.} \allowcalculators[approved] \begin{document} \lstloadlanguages{ODL,OQL,[Oracle8]SQL,[Oracle8]PLSQL} \lstset% { basicstyle=\sffamily, showstringspaces=false, tabsize=2, xleftmargin=1cm, belowskip=0pt, commentstyle=\itshape, numbers=left, numberstyle=\scriptsize, numbersep=5pt, columns=fullflexible, } \maketitlepage %====================================================================== % PHYSICAL TUNING SECTION %====================================================================== \begin{examsection}{25}{}{Questions in this section (total 25 marks) relate to physical database design and tuning.} %====================================================================== \begin{question}{10} A major U.S. mail order firm is having performance problems with queries on the \textsf{Customer} table in its Oracle8 order-processing database. Consider the following information about the \textsf{Customer} table, then answer the questions below: \\ \textsf{\textbf{Customer}(\underline{customer\_no}, name, address, city, state, country, phone, email, status)} \begin{itemize} \item \textsf{customer\_no} is the primary key and because this is an Oracle8 database, a unique B-tree index has been automatically created on this column. \item \textsf{status} is a single character column that holds one of the values `I' (inactive), `B' (bronze), `S' (silver) or `G' (gold). The value of this column is determined by a customer's order history. \item About 10\% of customers are inactive, about 50\% have bronze status, about 35\% have silver status and about 5\% have gold status. \item The average row size is about 200 bytes and there are approximately 5,000,000 rows. \item The \textsf{Customer} table is heavily queried (averaging thirty new queries per minute) by four different groups of employees. Group~1 queries only inactive customers, group~2 queries only bronze status customers and groups~3 and~4 both query only silver and gold status customers. Members of any group may query the table at any time and queries often overlap. \item UPDATE, INSERT and DELETE operations on the table are uncommon. \item Very few queries are for individual customers; rather, most queries are based on various combinations of values from \textsf{status}, \textsf{city}, \textsf{state} and \textsf{country}. \end{itemize} \medskip %---------------------------------------------------------------------- \begin{subquestion}{3} Describe and justify an \emph{index-based} physical tuning solution that will improve the performance of queries on the \textsf{Customer} table. (No code is required.) \begin{marking} Since queries tend to be on combinations of several low-cardinality columns [1] and the table is very large [1], the most effective solution would be to place bitmap indexes on status, city, state and country [1]. B-tree indexes on any of the individual columns would not be ideal because of their low cardinality. A possible alternative is a composite B-tree index on all four columns [1 for suggesting this if appropriate], but this is not ideal and would be much larger than the four bitmap indexes [1 for this explanation if appropriate]. Hashing is definitely \emph{not} an option, because very few queries are exact match. \end{marking} \end{subquestion} %---------------------------------------------------------------------- \begin{subquestion}{7} \begin{subsubquestion}{5} \label{nonindex} Assuming that there are two identical fast disks available, describe and justify two alternative \emph{non index-based} physical tuning solutions that will improve the performance of queries on the \textsf{Customer} table. For each solution include details of how the two disks will be used. (No code is required.) \begin{marking} Alternative 1: partition the table across the two disks based on status (four partitions, one for each status value) [1]. This will allow simultaneous parallel access to different parts of the table by the different groups [1]. Since two groups both access silver and gold customers, it makes sense to put the silver customers on disk one and the gold customers on disk two. Bronze customers account for 50\% of the table, so it makes sense to also put them on disk two (as the gold customers are the smallest group), and put the inactive customers on disk one with the silver customers. [1 for any sensible partitioning scheme] Alternative 2: replicate the table across both disks (one replica on each disk) [1]. This will also provide simultaneous parallel access to the data (hard to tell whether this would be better or worse than the partitioning above) [1]. Note that any solution must be capable of being implemented in Oracle8. \end{marking} \end{subsubquestion} \begin{subsubquestion}{2} Briefly discuss the relative advantages and/or disadvantages of the two alternative solutions you described in part~\ref{nonindex} above. \begin{marking} The partitioning solution given above doesn't really help with the bronze customer rows---we would need random horizontal partitioning to really help with this, which Oracle8 doesn't support. More generally, Oracle8 does not migrate rows into the correct partition if a customer's status changes, so the data will need to be reloaded occasionally [1]. For replication, we have the usual problem of keeping the replicas synchronised if they are both read/write. This is somewhat mitigated by the fact that there are only two copies [1]. \end{marking} \end{subsubquestion} \end{subquestion} %---------------------------------------------------------------------- \end{question} \newpage %====================================================================== \begin{question}{10} There are two generic parameters that affect the structure of a B-tree index. %---------------------------------------------------------------------- \begin{subquestion}{6} Identify these two parameters and describe what each parameter specifies. Use diagrams to illustrate your answer. \begin{marking} \emph{Node size} [1] specifies the maximum number of pointers in each index node [1]. Suitable diagram [1]. \emph{Percentage fill} [1] specifies the number of pointers in each index node that are allocated when a node is created [1]. In effect, it determines the amount of free space left in each node. Suitable diagram [1]. \end{marking} \end{subquestion} %---------------------------------------------------------------------- \begin{subquestion}{4} For each of the two parameters, discuss the practical effect(s) of: \begin{subsubquestion}{2} high values of the parameter; and \begin{marking} Large node size reduces the height of the B-tree, resulting in potentially fewer I/Os to access a leaf node [$\frac{1}{2}$]. It can however increase the number of rows locked in a single leaf node [$\frac{1}{2}$]. High percent fill means that most pointers in a node will be allocated when the node is created [$\frac{1}{2}$]. This leaves little free space in the B-tree for new key values or key values that change [$\frac{1}{2}$]. \end{marking} \end{subsubquestion} \begin{subsubquestion}{2} low values of the parameter. \begin{marking} Small node size increases the height of the B-tree, resulting in potentially more I/Os to access a leaf node [$\frac{1}{2}$]. The number of rows locked in a single leaf node will be reduced [$\frac{1}{2}$]. Low percent fill means that most pointers in a node will not be allocated when the node is created [$\frac{1}{2}$]. This leaves little free space in the B-tree for new key values or key values that change [$\frac{1}{2}$]. \end{marking} \end{subsubquestion} \end{subquestion} %---------------------------------------------------------------------- \end{question} %====================================================================== \begin{question}{5}\label{clusterq} %---------------------------------------------------------------------- \begin{subquestion}{2} Define what is meant by the term ``clustering'' in the context of physical database design and tuning, and explain why it is beneficial. \begin{marking} Clustering is the grouping of related rows physically near to each other on disk [1] so that they may be retrieved in as few I/Os as possible (ideally one)---beneficial because of the slowness of disk I/O compared to memory I/O [1]. \end{marking} \end{subquestion} %---------------------------------------------------------------------- \begin{subquestion}{3}\label{clusterqcode} Consider the following Oracle8 SQL table definition: \begin{lstlisting}[language={[Oracle8]SQL}]{} create table enrolment ( student_id char(7), paper_code char(7), enrol_year number(4), primary key (student_id, paper_code, enrol_year) ); \end{lstlisting} The \textsf{Enrolment} table is often queried for all rows relating to a particular student in a particular year (for example, student 1234567 in 1998, student 9876543 in 2000, etc.). Write appropriate Oracle8 SQL code that will cluster the rows of the table in such a way as to support these types of query (you do not need to calculate cluster sizes). Relevant Oracle8 syntax diagrams are given on the next page. \begin{marking} \begin{lstlisting}[language={[Oracle8]SQL}]{} create cluster enrolment_cluster (student_id char(7), year number(4)); create index enrol_cluster_index on cluster enrolment_cluster; create table enrolment ( student_id char(7), paper_code char(7), enrol_year number(4), primary key (student_id, paper_code, enrol_year) ) cluster enrolment_cluster(student_id, enrol_year); \end{lstlisting} \end{marking} [1] each for correct cluster, cluster index and table clustering. \end{subquestion} \end{question} \end{examsection} \newpage % CREATE CLUSTER (first part), p 4-207 \begin{center} \textbf{Oracle8 syntax diagrams for use in question~\ref{clusterq}(\ref{clusterqcode})} \bigskip \includegraphics{CreateCluster} \vfill % CREATE INDEX (parts 1 and 2), p. 4-238 \includegraphics{CreateIndex} % CREATE TABLE (parts 1 and 2), p. 4-308 \vfill \includegraphics{CreateTable1} \bigskip \includegraphics{CreateTable2} \end{center} %====================================================================== % OBJECT DBMS SECTION %====================================================================== \begin{examsection}{25}{}{Questions in this section (total 25 marks) relate to object database management systems and the Object Data Management Group's object model (ODMG 3.0).} %====================================================================== \begin{question}{10}\label{odlq} Examine the following Object Definition Language (ODL) schema, then answer the questions below. \begin{lstlisting}[language=ODL]{} module ExamODL { interface Person { attribute string name; attribute date birth_date; attribute struct Address { string street; string city; string country; } home_address; }; class Employee : Person (extent Staff key emp_ID) { attribute string emp_ID; attribute string position; attribute string department; attribute unsigned long salary; }; }; \end{lstlisting} %---------------------------------------------------------------------- \begin{subquestion}{2} How many attributes will instances of the \textsf{Employee} class have? Explain your answer. \begin{marking} This depends on your interpretation of what constitutes an attribute. I prefer seven [1]. It has four of its own and three inherited from the \textsf{Person} class (\textsf{home\_address} being counted as a single embedded literal attribute) [1]. Nine is a possible alternative answer [1], because you could count \textsf{home\_address} as three attributes, giving four local and five inherited attributes [1]. No marks for any other number \emph{unless} it has a sensible explanation, in which case award marks as appropriate. \end{marking} \end{subquestion} %---------------------------------------------------------------------- \begin{subquestion}{3} Examine the clause ``\lstinline[language={[Oracle8]SQL}]+(extent Staff key emp_ID)+'' on line 14 of the ODL listing above. \begin{subsubquestion}{2} Explain the effect(s) of this clause. \begin{marking} \lstinline[language={[Oracle8]SQL}]+extent Staff+ defines the name of the \emph{extent} (collection of all instances) of the \textsf{Employee} class [1]. \lstinline[language={[Oracle8]SQL}]+key emp_ID+ defines \textsf{emp\_ID} to be an \emph{object key} for the \textsf{Employee} class [1]. \end{marking} \end{subsubquestion} \begin{subsubquestion}{1} Identify the relational equivalents of the items specified in this clause. \begin{marking} Extent = relation [$\frac{1}{2}$] and object key = primary key [$\frac{1}{2}$]. \end{marking} \end{subsubquestion} \end{subquestion} %---------------------------------------------------------------------- \begin{subquestion}{5}\label{oracleodl} Write \emph{all} the code required to implement an Oracle8 object type that is as similar as possible to the \textsf{Employee} class. The syntax diagram for the \textsf{\textbf{create type}} statement is given on the next page. \begin{marking} \begin{lstlisting}[language={[Oracle8]SQL}]{} create type address_type as object ( street varchar2(50), -- the data types aren't that important city varchar2(30), -- as long as they roughly match country varchar2(30) ); create type employee_type as object ( emp_id char(7), name varchar2(50), birth_date date, home_address address_type, position char(10), department char(20), salary number ); \end{lstlisting} [--1] for major errors, or for not including \textsf{address\_type} as a separate object type. [--$\frac{1}{2}$] for each attribute missing or in the wrong place. \end{marking} \end{subquestion} \end{question} \newpage \begin{center} \textbf{Oracle8 syntax diagram for use in question~\ref{odlq}(\ref{oracleodl})} \bigskip \includegraphics{CreateObjectType} \bigskip\bigskip \end{center} %====================================================================== \begin{question}{15} %---------------------------------------------------------------------- \begin{subquestion}{4} Compare in detail the ODMG concept of an \emph{object name} with the relational concept of a \emph{primary key}. \begin{marking} An object name is a meaningful user-generated identifier [$\frac{1}{2}$] for a single, specific object instance [$\frac{1}{2}$]. It is not part of the object's state [$\frac{1}{2}$] and a particular instance can have multiple object names associated with it [$\frac{1}{2}$]. In comparison, a primary key is also user-defined (but may not always be meaningful) [$\frac{1}{2}$]. While a primary key uniquely identifies individual rows (which are the equivalents of object instances), it is associated with \emph{all} rows in a table, not just a single row [$\frac{1}{2}$]. A primary key comprises attributes of the relation, so it is part of the relation's ``state'' [$\frac{1}{2}$]. An individual row has only one primary key associated with it [$\frac{1}{2}$]. \end{marking} \end{subquestion} %---------------------------------------------------------------------- \begin{subquestion}{4} Compare in detail the ODMG concept of a \emph{relationship} with the relational concept of a \emph{foreign key}, with particular emphasis on any differences between the two concepts. Use an appropriate diagram or diagrams to illustrate your answer. \begin{marking} An ODMG relationship is a binary association between two object type, comprising a pair of \emph{traversal paths}, which are effectively pointers linking the two related instances [1]. ODMG relationships may be one-to-one, one-to-many or many-to-many, but we cannot specify participation [1]. A relational foreign key is also a binary association between two relations, comprising an attribute stored in one of the relations that refers to the primary key of the other relation; in this respect, it is really equivalent to a traversal path rather than a full ODMG relationship [1]. Relationships defined by foreign keys cannot be many-to-many, but we can partially specify participation [1]. Useful diagram(s). \end{marking} \end{subquestion} %---------------------------------------------------------------------- \begin{subquestion}{6} Define in detail the ODMG concepts of \emph{interface}, \emph{class} and \emph{literal}. \begin{marking} An interface is an ODMG object type that defines behaviour only [1]. You cannot create instances of interfaces (cf. abstract classes) [1]. A class is an ODMG object type that can define both state and behaviour [1]. You can create instances of classes [1]. A literal is an ODMG type (non-object) that defines state only [1]. They have no identity and are effectively just structured data types [1]. \end{marking} \end{subquestion} %---------------------------------------------------------------------- \begin{subquestion}{1} Date and Darwen claim in their \emph{Third Manifesto} that the relational model can support objects if it is implemented correctly. Briefly explain how they claim this can be achieved. \begin{marking} If we implement domains correctly/fully then they are effectively the equivalent of an object class [1]. \end{marking} \end{subquestion} \end{question} \end{examsection} %====================================================================== % DISTRIBUTED DBMS SECTION %====================================================================== \begin{examsection}{25}{}{Questions in this section (total 25 marks) relate to distributed database systems.} %====================================================================== \begin{question}{4} Describe what is meant by the terms ``distributed data'' and ``distributed database'' and explain the key difference(s) between the two. Use an example to illustrate the difference(s). \begin{marking} Distributed data are spread across multiple nodes on a network but are neither centrally managed nor highly integrated [1]. A distributed database combines distributed data with distributed database processing [1] (including management, access and transaction processing)---this is the key difference [1]. Reasonable example [1]. \end{marking} \end{question} %====================================================================== \begin{question}{10} Date's ``objective zero'' for distributed database management systems states ``a distributed database looks and behaves like a non-distributed database to any user''. Describe two \emph{different} ways in which this objective may be achieved, and for each discuss in detail how well it is implemented in Oracle8. \begin{marking} Basically a detailed description of any two of Date's objectives. Example: Location transparency is the ability to query data without needing to know on which network node the data are physically located. It is implemented reasonably well in Oracle8 by way of synonyms and views, after some initial setup work. [2 marks for description, 3 marks for implementation] \end{marking} \end{question} %====================================================================== \begin{question}{5} A company currently has a distributed database spread across two sites. It is necessary to move some data from one site to the other. %---------------------------------------------------------------------- \begin{subquestion}{3} Describe the general process that allows us to complete distributed transactions such as that described above with reasonable safety. \begin{marking} This is a distributed transaction, so the process involved is two-phase commit. In the first (\emph{prepare}) phase, the transaction coordinator tells all sites involved in the transaction to prepare to commit (by flushing buffers, locking data, etc.). Each site then responds ``OK'' or ``abort''. In the second (\emph{commit}) phase, if all sites respond ``OK'' the coordinator tells all sites to commit, otherwise it tells all sites to roll back. [1] for identifying 2PC, [1] each for both phases. \end{marking} \end{subquestion} %---------------------------------------------------------------------- \begin{subquestion}{2} Assuming an Oracle8 distributed database, give an example of a situation in which it may not be possible to automatically recover from a failed distributed transaction, and explain how such a situation might be dealt with. \begin{marking} If a site fails for a short period, in-doubt transactions will automatically complete when the site comes back up. If however the outage is for a significant period of time, then the DBMS cannot automatically commit in-doubt transactions as the data context may have changed considerably in the interim [1]. The DBA can commit or roll back in-doubt transactions manually, but this is risky [1]. \end{marking} \end{subquestion} \end{question} %====================================================================== \begin{question}{6} %---------------------------------------------------------------------- \begin{subquestion}{2} Consider the following Oracle8 SQL statement: \begin{lstlisting}[language={[Oracle8]SQL},numbers=none]{} create synonym exam_synonym for sales.customer@db_chch.world; \end{lstlisting} Explain the purpose of this statement. \begin{marking} It creates (in the local schema) an ``alias'' to the \textsf{customer} table in the \textsf{sales} schema of the \textsf{db\_chch.world} database [1]. This synonym can be queried as if it were a local copy of the table (i.e., it provides location transparency) [1]. \end{marking} \end{subquestion} %---------------------------------------------------------------------- \begin{subquestion}{4} Consider the following Oracle8 SQL statement: \begin{lstlisting}[language={[Oracle8]SQL}]{} create snapshot exam_snapshot refresh fast start with sysdate next sysdate + 1/12 as select * from sales.customer@db_chch.world; \end{lstlisting} Explain in detail what this statement does. \begin{marking} It creates a read only replica [1] of the \textsf{customer} table from the \textsf{sales} schema in the \textsf{db\_chch.world} database [1]. Changes to the original table only [1] are copied across every two hours starting now [1]. \end{marking} \end{subquestion} \end{question} %====================================================================== \end{examsection} %====================================================================== % DECISION SUPPORT SYSTEMS SECTION %====================================================================== \begin{examsection}{25}{}{Questions in this section (total 25 marks) relate to data warehouses and online analytical processing (OLAP) tools.} %====================================================================== \begin{question}{8} Compare in detail the characteristics of operational data and decision support data with respect to the following four areas: \begin{itemize} \item the types of decisions they are used to support; \item their timespan; \item their granularity; and \item their dimensionality. \end{itemize} \begin{marking} (Note: ``in detail'' implies more than just a table of brief points---only award half marks for such answers.) \begin{itemize} \item \textbf{Decision type:} operational data support short term day-to-day decision making [1] whereas decision support data support tactical and strategic decision making (longer term) [1]. \item \textbf{Timespan:} operational data tend to focus on \emph{now} (current transactions over maybe a few days) [1] whereas DS data range over much larger time scales (months, years, decades), both backwards and forwards [1]. \item \textbf{Granularity:} operational data are generally very fine-grained (individual transactions) [1], whereas DS data are aggregated at varying levels of summarisation [1]. \item \textbf{Dimensionality:} operational data tend to be ``flat'' (only one variable of interest associated with a given data value) [1], whereas DS data can have many dimensions (variables of interest) associated with a data value [1]. \end{itemize} \end{marking} \end{question} %====================================================================== \begin{question}{6} Discuss in detail the key differences between databases designed for online transaction processing and databases designed for data warehousing, in terms of both: \begin{subquestion}{2} the pattern of data use; and \begin{marking} OLTP databases are typically not queried heavily and the queries tend to be relatively simple [$\frac{1}{2}$]; they also have a high rate of online concurrent updates [$\frac{1}{2}$]. Conversely, data warehouses typically have very high query rates and the queries are very complex [$\frac{1}{2}$]; there are usually few updates (often read-only) and those updates that do occur are often carried out as batch operations [$\frac{1}{2}$]. ([$\frac{1}{2}$] for saying that OLTP typically have many concurrent users, or that data warehouses typically have few users.) \end{marking} \end{subquestion} \begin{subquestion}{4} physical design and tuning strategy. \begin{marking} OLTP databases are queried and updated by many users simultaneously, so we need to trade off between optimising for concurrency, updates or query performance [1]. The usual tuning techniques may be applied, but we should be particularly careful with the use of indexes (typically B-tree) because of the relatively high update rates [1]. ([$\frac{1}{2}$] for saying that the database is often relatively small.) In a data warehouse, updates are few but queries are many and complex, so we need to optimise aggessively for query performance [1]. The usual tuning techniques may be applied, but in particular we can heavily index the data (bitmap indexes and index-organised tables really come into their own here) [1].([$\frac{1}{2}$] for saying that the database is typically very large.) \end{marking} \end{subquestion} \end{question} %====================================================================== \begin{question}{4} Describe in detail the steps involved in the process of importing data into a data warehouse from external sources. \begin{marking} \begin{itemize} \item Extraction: obtaining data from various sources (electronically or otherwise) [1]. The data are often loaded into a special \emph{staging database} for further processing [1]. \item Transformation: data are cleaned up and manipulated into standard formats, so that data may be queried consistently regardless of source [1]. \item Loading: the transformed data are loaded into the data warehouse by some form of bulk loading tool [1]. \end{itemize} \end{marking} \end{question} %====================================================================== \begin{question}{1} Explain what is meant by the term ``data mart''. \begin{marking} A data mart is a small \emph{specialised} data warehouse, often a subset of a larger enterprise-wide data warehouse [1]. \end{marking} \end{question} %====================================================================== \begin{question}{6} Describe in detail two ways in which an OLAP tool can be used to change the way in which data are presented. Use diagrams and/or examples to illustrate your answer. \begin{marking} Any two of: \begin{itemize} \item Drill down/roll up allows us to move from more summarised data to more specific data and vice versa. \item ``Slice and dice'' allows us to reduce our view of a multidimensional data set to specific values of one or more dimensions. \item Pivoting allows us to change the layout of the data being output to give a different perspective on the same data set, e.g., viewing sales of products by region vs. sales in a region by product. \end{itemize} [1] each for technique, [1] each for description, [1] each for useful diagrams or examples. Any of the following are worth half marks: \begin{itemize} \item graphs/visualisation tools \item highlighting exceptions \end{itemize} \end{marking} \end{question} \end{examsection} \end{document} %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % % File: $Id$ % % Note: No document class --- this is handled by the appropriate % "driver" file, which makes it easier to switch class options % such as markingschedule. Doing it this way means that you don't % wipe out the original exam paper when you want to generate a % marking schedule. % %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% \usepackage{graphicx} \usepackage{listings} \examyear{2001} \department{Information Science} \papernumber{INFO 321} \papertitle{Database Systems} \instructions{Answer \underline{ALL} questions.} \allowcalculators[approved] \begin{document} \lstloadlanguages{ODL,OQL,[Oracle8]SQL,[Oracle8]PLSQL} \lstset% { basicstyle=\sffamily, showstringspaces=false, tabsize=2, xleftmargin=1cm, belowskip=0pt, commentstyle=\itshape, numbers=left, numberstyle=\scriptsize, numbersep=5pt, columns=fullflexible, } \maketitlepage %====================================================================== % PHYSICAL TUNING SECTION %====================================================================== \begin{examsection}{25}{}{Questions in this section (total 25 marks) relate to physical database design and tuning.} %====================================================================== \begin{question}{10} A major U.S. mail order firm is having performance problems with queries on the \textsf{Customer} table in its Oracle8 order-processing database. Consider the following information about the \textsf{Customer} table, then answer the questions below: \\ \textsf{\textbf{Customer}(\underline{customer\_no}, name, address, city, state, country, phone, email, status)} \begin{itemize} \item \textsf{customer\_no} is the primary key and because this is an Oracle8 database, a unique B-tree index has been automatically created on this column. \item \textsf{status} is a single character column that holds one of the values `I' (inactive), `B' (bronze), `S' (silver) or `G' (gold). The value of this column is determined by a customer's order history. \item About 10\% of customers are inactive, about 50\% have bronze status, about 35\% have silver status and about 5\% have gold status. \item The average row size is about 200 bytes and there are approximately 5,000,000 rows. \item The \textsf{Customer} table is heavily queried (averaging thirty new queries per minute) by four different groups of employees. Group~1 queries only inactive customers, group~2 queries only bronze status customers and groups~3 and~4 both query only silver and gold status customers. Members of any group may query the table at any time and queries often overlap. \item UPDATE, INSERT and DELETE operations on the table are uncommon. \item Very few queries are for individual customers; rather, most queries are based on various combinations of values from \textsf{status}, \textsf{city}, \textsf{state} and \textsf{country}. \end{itemize} \medskip %---------------------------------------------------------------------- \begin{subquestion}{3} Describe and justify an \emph{index-based} physical tuning solution that will improve the performance of queries on the \textsf{Customer} table. (No code is required.) \begin{marking} Since queries tend to be on combinations of several low-cardinality columns [1] and the table is very large [1], the most effective solution would be to place bitmap indexes on status, city, state and country [1]. B-tree indexes on any of the individual columns would not be ideal because of their low cardinality. A possible alternative is a composite B-tree index on all four columns [1 for suggesting this if appropriate], but this is not ideal and would be much larger than the four bitmap indexes [1 for this explanation if appropriate]. Hashing is definitely \emph{not} an option, because very few queries are exact match. \end{marking} \end{subquestion} %---------------------------------------------------------------------- \begin{subquestion}{7} \begin{subsubquestion}{5} \label{nonindex} Assuming that there are two identical fast disks available, describe and justify two alternative \emph{non index-based} physical tuning solutions that will improve the performance of queries on the \textsf{Customer} table. For each solution include details of how the two disks will be used. (No code is required.) \begin{marking} Alternative 1: partition the table across the two disks based on status (four partitions, one for each status value) [1]. This will allow simultaneous parallel access to different parts of the table by the different groups [1]. Since two groups both access silver and gold customers, it makes sense to put the silver customers on disk one and the gold customers on disk two. Bronze customers account for 50\% of the table, so it makes sense to also put them on disk two (as the gold customers are the smallest group), and put the inactive customers on disk one with the silver customers. [1 for any sensible partitioning scheme] Alternative 2: replicate the table across both disks (one replica on each disk) [1]. This will also provide simultaneous parallel access to the data (hard to tell whether this would be better or worse than the partitioning above) [1]. Note that any solution must be capable of being implemented in Oracle8. \end{marking} \end{subsubquestion} \begin{subsubquestion}{2} Briefly discuss the relative advantages and/or disadvantages of the two alternative solutions you described in part~\ref{nonindex} above. \begin{marking} The partitioning solution given above doesn't really help with the bronze customer rows---we would need random horizontal partitioning to really help with this, which Oracle8 doesn't support. More generally, Oracle8 does not migrate rows into the correct partition if a customer's status changes, so the data will need to be reloaded occasionally [1]. For replication, we have the usual problem of keeping the replicas synchronised if they are both read/write. This is somewhat mitigated by the fact that there are only two copies [1]. \end{marking} \end{subsubquestion} \end{subquestion} %---------------------------------------------------------------------- \end{question} \newpage %====================================================================== \begin{question}{10} There are two generic parameters that affect the structure of a B-tree index. %---------------------------------------------------------------------- \begin{subquestion}{6} Identify these two parameters and describe what each parameter specifies. Use diagrams to illustrate your answer. \begin{marking} \emph{Node size} [1] specifies the maximum number of pointers in each index node [1]. Suitable diagram [1]. \emph{Percentage fill} [1] specifies the number of pointers in each index node that are allocated when a node is created [1]. In effect, it determines the amount of free space left in each node. Suitable diagram [1]. \end{marking} \end{subquestion} %---------------------------------------------------------------------- \begin{subquestion}{4} For each of the two parameters, discuss the practical effect(s) of: \begin{subsubquestion}{2} high values of the parameter; and \begin{marking} Large node size reduces the height of the B-tree, resulting in potentially fewer I/Os to access a leaf node [$\frac{1}{2}$]. It can however increase the number of rows locked in a single leaf node [$\frac{1}{2}$]. High percent fill means that most pointers in a node will be allocated when the node is created [$\frac{1}{2}$]. This leaves little free space in the B-tree for new key values or key values that change [$\frac{1}{2}$]. \end{marking} \end{subsubquestion} \begin{subsubquestion}{2} low values of the parameter. \begin{marking} Small node size increases the height of the B-tree, resulting in potentially more I/Os to access a leaf node [$\frac{1}{2}$]. The number of rows locked in a single leaf node will be reduced [$\frac{1}{2}$]. Low percent fill means that most pointers in a node will not be allocated when the node is created [$\frac{1}{2}$]. This leaves little free space in the B-tree for new key values or key values that change [$\frac{1}{2}$]. \end{marking} \end{subsubquestion} \end{subquestion} %---------------------------------------------------------------------- \end{question} %====================================================================== \begin{question}{5}\label{clusterq} %---------------------------------------------------------------------- \begin{subquestion}{2} Define what is meant by the term ``clustering'' in the context of physical database design and tuning, and explain why it is beneficial. \begin{marking} Clustering is the grouping of related rows physically near to each other on disk [1] so that they may be retrieved in as few I/Os as possible (ideally one)---beneficial because of the slowness of disk I/O compared to memory I/O [1]. \end{marking} \end{subquestion} %---------------------------------------------------------------------- \begin{subquestion}{3}\label{clusterqcode} Consider the following Oracle8 SQL table definition: \begin{lstlisting}[language={[Oracle8]SQL}]{} create table enrolment ( student_id char(7), paper_code char(7), enrol_year number(4), primary key (student_id, paper_code, enrol_year) ); \end{lstlisting} The \textsf{Enrolment} table is often queried for all rows relating to a particular student in a particular year (for example, student 1234567 in 1998, student 9876543 in 2000, etc.). Write appropriate Oracle8 SQL code that will cluster the rows of the table in such a way as to support these types of query (you do not need to calculate cluster sizes). Relevant Oracle8 syntax diagrams are given on the next page. \begin{marking} \begin{lstlisting}[language={[Oracle8]SQL}]{} create cluster enrolment_cluster (student_id char(7), year number(4)); create index enrol_cluster_index on cluster enrolment_cluster; create table enrolment ( student_id char(7), paper_code char(7), enrol_year number(4), primary key (student_id, paper_code, enrol_year) ) cluster enrolment_cluster(student_id, enrol_year); \end{lstlisting} \end{marking} [1] each for correct cluster, cluster index and table clustering. \end{subquestion} \end{question} \end{examsection} \newpage % CREATE CLUSTER (first part), p 4-207 \begin{center} \textbf{Oracle8 syntax diagrams for use in question~\ref{clusterq}(\ref{clusterqcode})} \bigskip \includegraphics{CreateCluster} \vfill % CREATE INDEX (parts 1 and 2), p. 4-238 \includegraphics{CreateIndex} % CREATE TABLE (parts 1 and 2), p. 4-308 \vfill \includegraphics{CreateTable1} \bigskip \includegraphics{CreateTable2} \end{center} %====================================================================== % OBJECT DBMS SECTION %====================================================================== \begin{examsection}{25}{}{Questions in this section (total 25 marks) relate to object database management systems and the Object Data Management Group's object model (ODMG 3.0).} %====================================================================== \begin{question}{10}\label{odlq} Examine the following Object Definition Language (ODL) schema, then answer the questions below. \begin{lstlisting}[language=ODL]{} module ExamODL { interface Person { attribute string name; attribute date birth_date; attribute struct Address { string street; string city; string country; } home_address; }; class Employee : Person (extent Staff key emp_ID) { attribute string emp_ID; attribute string position; attribute string department; attribute unsigned long salary; }; }; \end{lstlisting} %---------------------------------------------------------------------- \begin{subquestion}{2} How many attributes will instances of the \textsf{Employee} class have? Explain your answer. \begin{marking} This depends on your interpretation of what constitutes an attribute. I prefer seven [1]. It has four of its own and three inherited from the \textsf{Person} class (\textsf{home\_address} being counted as a single embedded literal attribute) [1]. Nine is a possible alternative answer [1], because you could count \textsf{home\_address} as three attributes, giving four local and five inherited attributes [1]. No marks for any other number \emph{unless} it has a sensible explanation, in which case award marks as appropriate. \end{marking} \end{subquestion} %---------------------------------------------------------------------- \begin{subquestion}{3} Examine the clause ``\lstinline[language={[Oracle8]SQL}]+(extent Staff key emp_ID)+'' on line 14 of the ODL listing above. \begin{subsubquestion}{2} Explain the effect(s) of this clause. \begin{marking} \lstinline[language={[Oracle8]SQL}]+extent Staff+ defines the name of the \emph{extent} (collection of all instances) of the \textsf{Employee} class [1]. \lstinline[language={[Oracle8]SQL}]+key emp_ID+ defines \textsf{emp\_ID} to be an \emph{object key} for the \textsf{Employee} class [1]. \end{marking} \end{subsubquestion} \begin{subsubquestion}{1} Identify the relational equivalents of the items specified in this clause. \begin{marking} Extent = relation [$\frac{1}{2}$] and object key = primary key [$\frac{1}{2}$]. \end{marking} \end{subsubquestion} \end{subquestion} %---------------------------------------------------------------------- \begin{subquestion}{5}\label{oracleodl} Write \emph{all} the code required to implement an Oracle8 object type that is as similar as possible to the \textsf{Employee} class. The syntax diagram for the \textsf{\textbf{create type}} statement is given on the next page. \begin{marking} \begin{lstlisting}[language={[Oracle8]SQL}]{} create type address_type as object ( street varchar2(50), -- the data types aren't that important city varchar2(30), -- as long as they roughly match country varchar2(30) ); create type employee_type as object ( emp_id char(7), name varchar2(50), birth_date date, home_address address_type, position char(10), department char(20), salary number ); \end{lstlisting} [--1] for major errors, or for not including \textsf{address\_type} as a separate object type. [--$\frac{1}{2}$] for each attribute missing or in the wrong place. \end{marking} \end{subquestion} \end{question} \newpage \begin{center} \textbf{Oracle8 syntax diagram for use in question~\ref{odlq}(\ref{oracleodl})} \bigskip \includegraphics{CreateObjectType} \bigskip\bigskip \end{center} %====================================================================== \begin{question}{15} %---------------------------------------------------------------------- \begin{subquestion}{4} Compare in detail the ODMG concept of an \emph{object name} with the relational concept of a \emph{primary key}. \begin{marking} An object name is a meaningful user-generated identifier [$\frac{1}{2}$] for a single, specific object instance [$\frac{1}{2}$]. It is not part of the object's state [$\frac{1}{2}$] and a particular instance can have multiple object names associated with it [$\frac{1}{2}$]. In comparison, a primary key is also user-defined (but may not always be meaningful) [$\frac{1}{2}$]. While a primary key uniquely identifies individual rows (which are the equivalents of object instances), it is associated with \emph{all} rows in a table, not just a single row [$\frac{1}{2}$]. A primary key comprises attributes of the relation, so it is part of the relation's ``state'' [$\frac{1}{2}$]. An individual row has only one primary key associated with it [$\frac{1}{2}$]. \end{marking} \end{subquestion} %---------------------------------------------------------------------- \begin{subquestion}{4} Compare in detail the ODMG concept of a \emph{relationship} with the relational concept of a \emph{foreign key}, with particular emphasis on any differences between the two concepts. Use an appropriate diagram or diagrams to illustrate your answer. \begin{marking} An ODMG relationship is a binary association between two object type, comprising a pair of \emph{traversal paths}, which are effectively pointers linking the two related instances [1]. ODMG relationships may be one-to-one, one-to-many or many-to-many, but we cannot specify participation [1]. A relational foreign key is also a binary association between two relations, comprising an attribute stored in one of the relations that refers to the primary key of the other relation; in this respect, it is really equivalent to a traversal path rather than a full ODMG relationship [1]. Relationships defined by foreign keys cannot be many-to-many, but we can partially specify participation [1]. Useful diagram(s). \end{marking} \end{subquestion} %---------------------------------------------------------------------- \begin{subquestion}{6} Define in detail the ODMG concepts of \emph{interface}, \emph{class} and \emph{literal}. \begin{marking} An interface is an ODMG object type that defines behaviour only [1]. You cannot create instances of interfaces (cf. abstract classes) [1]. A class is an ODMG object type that can define both state and behaviour [1]. You can create instances of classes [1]. A literal is an ODMG type (non-object) that defines state only [1]. They have no identity and are effectively just structured data types [1]. \end{marking} \end{subquestion} %---------------------------------------------------------------------- \begin{subquestion}{1} Date and Darwen claim in their \emph{Third Manifesto} that the relational model can support objects if it is implemented correctly. Briefly explain how they claim this can be achieved. \begin{marking} If we implement domains correctly/fully then they are effectively the equivalent of an object class [1]. \end{marking} \end{subquestion} \end{question} \end{examsection} %====================================================================== % DISTRIBUTED DBMS SECTION %====================================================================== \begin{examsection}{25}{}{Questions in this section (total 25 marks) relate to distributed database systems.} %====================================================================== \begin{question}{4} Describe what is meant by the terms ``distributed data'' and ``distributed database'' and explain the key difference(s) between the two. Use an example to illustrate the difference(s). \begin{marking} Distributed data are spread across multiple nodes on a network but are neither centrally managed nor highly integrated [1]. A distributed database combines distributed data with distributed database processing [1] (including management, access and transaction processing)---this is the key difference [1]. Reasonable example [1]. \end{marking} \end{question} %====================================================================== \begin{question}{10} Date's ``objective zero'' for distributed database management systems states ``a distributed database looks and behaves like a non-distributed database to any user''. Describe two \emph{different} ways in which this objective may be achieved, and for each discuss in detail how well it is implemented in Oracle8. \begin{marking} Basically a detailed description of any two of Date's objectives. Example: Location transparency is the ability to query data without needing to know on which network node the data are physically located. It is implemented reasonably well in Oracle8 by way of synonyms and views, after some initial setup work. [2 marks for description, 3 marks for implementation] \end{marking} \end{question} %====================================================================== \begin{question}{5} A company currently has a distributed database spread across two sites. It is necessary to move some data from one site to the other. %---------------------------------------------------------------------- \begin{subquestion}{3} Describe the general process that allows us to complete distributed transactions such as that described above with reasonable safety. \begin{marking} This is a distributed transaction, so the process involved is two-phase commit. In the first (\emph{prepare}) phase, the transaction coordinator tells all sites involved in the transaction to prepare to commit (by flushing buffers, locking data, etc.). Each site then responds ``OK'' or ``abort''. In the second (\emph{commit}) phase, if all sites respond ``OK'' the coordinator tells all sites to commit, otherwise it tells all sites to roll back. [1] for identifying 2PC, [1] each for both phases. \end{marking} \end{subquestion} %---------------------------------------------------------------------- \begin{subquestion}{2} Assuming an Oracle8 distributed database, give an example of a situation in which it may not be possible to automatically recover from a failed distributed transaction, and explain how such a situation might be dealt with. \begin{marking} If a site fails for a short period, in-doubt transactions will automatically complete when the site comes back up. If however the outage is for a significant period of time, then the DBMS cannot automatically commit in-doubt transactions as the data context may have changed considerably in the interim [1]. The DBA can commit or roll back in-doubt transactions manually, but this is risky [1]. \end{marking} \end{subquestion} \end{question} %====================================================================== \begin{question}{6} %---------------------------------------------------------------------- \begin{subquestion}{2} Consider the following Oracle8 SQL statement: \begin{lstlisting}[language={[Oracle8]SQL},numbers=none]{} create synonym exam_synonym for sales.customer@db_chch.world; \end{lstlisting} Explain the purpose of this statement. \begin{marking} It creates (in the local schema) an ``alias'' to the \textsf{customer} table in the \textsf{sales} schema of the \textsf{db\_chch.world} database [1]. This synonym can be queried as if it were a local copy of the table (i.e., it provides location transparency) [1]. \end{marking} \end{subquestion} %---------------------------------------------------------------------- \begin{subquestion}{4} Consider the following Oracle8 SQL statement: \begin{lstlisting}[language={[Oracle8]SQL}]{} create snapshot exam_snapshot refresh fast start with sysdate next sysdate + 1/12 as select * from sales.customer@db_chch.world; \end{lstlisting} Explain in detail what this statement does. \begin{marking} It creates a read only replica [1] of the \textsf{customer} table from the \textsf{sales} schema in the \textsf{db\_chch.world} database [1]. Changes to the original table only [1] are copied across every two hours starting now [1]. \end{marking} \end{subquestion} \end{question} %====================================================================== \end{examsection} %====================================================================== % DECISION SUPPORT SYSTEMS SECTION %====================================================================== \begin{examsection}{25}{}{Questions in this section (total 25 marks) relate to data warehouses and online analytical processing (OLAP) tools.} %====================================================================== \begin{question}{8} Compare in detail the characteristics of operational data and decision support data with respect to the following four areas: \begin{itemize} \item the types of decisions they are used to support; \item their timespan; \item their granularity; and \item their dimensionality. \end{itemize} \begin{marking} (Note: ``in detail'' implies more than just a table of brief points---only award half marks for such answers.) \begin{itemize} \item \textbf{Decision type:} operational data support short term day-to-day decision making [1] whereas decision support data support tactical and strategic decision making (longer term) [1]. \item \textbf{Timespan:} operational data tend to focus on \emph{now} (current transactions over maybe a few days) [1] whereas DS data range over much larger time scales (months, years, decades), both backwards and forwards [1]. \item \textbf{Granularity:} operational data are generally very fine-grained (individual transactions) [1], whereas DS data are aggregated at varying levels of summarisation [1]. \item \textbf{Dimensionality:} operational data tend to be ``flat'' (only one variable of interest associated with a given data value) [1], whereas DS data can have many dimensions (variables of interest) associated with a data value [1]. \end{itemize} \end{marking} \end{question} %====================================================================== \begin{question}{6} Discuss in detail the key differences between databases designed for online transaction processing and databases designed for data warehousing, in terms of both: \begin{subquestion}{2} the pattern of data use; and \begin{marking} OLTP databases are typically not queried heavily and the queries tend to be relatively simple [$\frac{1}{2}$]; they also have a high rate of online concurrent updates [$\frac{1}{2}$]. Conversely, data warehouses typically have very high query rates and the queries are very complex [$\frac{1}{2}$]; there are usually few updates (often read-only) and those updates that do occur are often carried out as batch operations [$\frac{1}{2}$]. ([$\frac{1}{2}$] for saying that OLTP typically have many concurrent users, or that data warehouses typically have few users.) \end{marking} \end{subquestion} \begin{subquestion}{4} physical design and tuning strategy. \begin{marking} OLTP databases are queried and updated by many users simultaneously, so we need to trade off between optimising for concurrency, updates or query performance [1]. The usual tuning techniques may be applied, but we should be particularly careful with the use of indexes (typically B-tree) because of the relatively high update rates [1]. ([$\frac{1}{2}$] for saying that the database is often relatively small.) In a data warehouse, updates are few but queries are many and complex, so we need to optimise aggessively for query performance [1]. The usual tuning techniques may be applied, but in particular we can heavily index the data (bitmap indexes and index-organised tables really come into their own here) [1].([$\frac{1}{2}$] for saying that the database is typically very large.) \end{marking} \end{subquestion} \end{question} %====================================================================== \begin{question}{4} Describe in detail the steps involved in the process of importing data into a data warehouse from external sources. \begin{marking} \begin{itemize} \item Extraction: obtaining data from various sources (electronically or otherwise) [1]. The data are often loaded into a special \emph{staging database} for further processing [1]. \item Transformation: data are cleaned up and manipulated into standard formats, so that data may be queried consistently regardless of source [1]. \item Loading: the transformed data are loaded into the data warehouse by some form of bulk loading tool [1]. \end{itemize} \end{marking} \end{question} %====================================================================== \begin{question}{1} Explain what is meant by the term ``data mart''. \begin{marking} A data mart is a small \emph{specialised} data warehouse, often a subset of a larger enterprise-wide data warehouse [1]. \end{marking} \end{question} %====================================================================== \begin{question}{6} Describe in detail two ways in which an OLAP tool can be used to change the way in which data are presented. Use diagrams and/or examples to illustrate your answer. \begin{marking} Any two of: \begin{itemize} \item Drill down/roll up allows us to move from more summarised data to more specific data and vice versa. \item ``Slice and dice'' allows us to reduce our view of a multidimensional data set to specific values of one or more dimensions. \item Pivoting allows us to change the layout of the data being output to give a different perspective on the same data set, e.g., viewing sales of products by region vs. sales in a region by product. \end{itemize} [1] each for technique, [1] each for description, [1] each for useful diagrams or examples. Any of the following are worth half marks: \begin{itemize} \item graphs/visualisation tools \item highlighting exceptions \end{itemize} \end{marking} \end{question} \end{examsection} \end{document}

Ignore Space Show notes View

example.tex

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%
% Note: No document class --- this is handled by the appropriate
% "driver" file, which makes it easier to switch class options
% such as markingschedule. Doing it this way means that you don't
% wipe out the original exam paper when you want to generate a
% marking schedule.
%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

\usepackage{graphicx}
\usepackage{listings}

\examyear{2001}
\department{Information Science}
\papernumber{INFO 321}
\papertitle{Database Systems}
\instructions{Answer \underline{ALL} questions.}
\allowcalculators[approved]

\begin{document}

\lstloadlanguages{ODL,OQL,[Oracle8]SQL,[Oracle8]PLSQL}
\lstset%
{	basicstyle=\sffamily,
	showstringspaces=false,
	tabsize=2,
	xleftmargin=1cm,
	belowskip=0pt,
	commentstyle=\itshape,
	numbers=left,
	numberstyle=\scriptsize,
	numbersep=5pt,
	columns=fullflexible,
}

\maketitlepage

%======================================================================
% PHYSICAL TUNING SECTION
%======================================================================
\begin{examsection}{25}{}{Questions in this section (total 25 marks) relate to
physical database design and tuning.}

%======================================================================

\begin{question}{10}

A major U.S. mail order firm is having performance problems with queries on
	the \textsf{Customer} table in its Oracle8 order-processing database.
	Consider the following information about the \textsf{Customer} table, then
	answer the questions below:	\\

\textsf{\textbf{Customer}(\underline{customer\_no}, name, address, city,
	state, country, phone, email, status)}

\begin{itemize}

\item \textsf{customer\_no} is the primary key and because this is an
		Oracle8 database, a unique B-tree index has been automatically created
		on this column.

\item \textsf{status} is a single character column that holds one of
		the values `I' (inactive), `B' (bronze), `S' (silver) or `G' (gold).
		The value of this column is determined by a customer's order history.

\item About 10\% of customers are inactive, about 50\% have bronze
		status, about 35\% have silver status and about 5\% have gold status.

\item The average row size is about 200 bytes and there are
		approximately 5,000,000 rows.

\item The \textsf{Customer} table is heavily queried (averaging thirty
		new queries per minute) by four different groups of employees. Group~1
		queries only inactive customers, group~2 queries only bronze status
		customers and groups~3 and~4 both query only silver and gold status
		customers. Members of any group may query the table at any time and
		queries often overlap.

\item UPDATE, INSERT and DELETE operations on the table are uncommon.

\item Very few queries are for individual customers; rather, most
		queries are based on various combinations of values from
		\textsf{status}, \textsf{city}, \textsf{state} and \textsf{country}.

\end{itemize}
	\medskip

%----------------------------------------------------------------------

\begin{subquestion}{3}
		Describe and justify an \emph{index-based} physical tuning solution
		that will improve the performance of queries on the \textsf{Customer}
		table. (No code is required.)
		\begin{marking}
			Since queries tend to be on combinations of several low-cardinality
			columns [1] and the table is very large [1], the most effective
			solution would be to place bitmap indexes on status, city, state
			and country [1].

B-tree indexes on any of the individual columns would not be ideal
			because of their low cardinality. A possible alternative is a
			composite B-tree index on all four columns [1 for suggesting this
			if appropriate], but this is not ideal and would be much larger
			than the four bitmap indexes [1 for this explanation if
			appropriate].

Hashing is definitely \emph{not} an option, because very few
			queries are exact match.
		\end{marking}
	\end{subquestion}
		
%----------------------------------------------------------------------

\begin{subquestion}{7}
		\begin{subsubquestion}{5}
			\label{nonindex}
			Assuming that there are two identical fast disks available,
			describe and justify two alternative \emph{non index-based}
			physical tuning solutions that will improve the performance of
			queries on the \textsf{Customer} table. For each solution include
			details of how the two disks will be used. (No code is required.)
			\begin{marking}
				Alternative 1: partition the table across the two disks based
				on status (four partitions, one for each status value) [1].
				This will allow simultaneous parallel access to different parts
				of the table by the different groups [1]. Since two groups both
				access silver and gold customers, it makes sense to put the
				silver customers on disk one and the gold customers on disk
				two. Bronze customers account for 50\% of the table, so it
				makes sense to also put them on disk two (as the gold customers
				are the smallest group), and put the inactive customers on disk
				one with the silver customers. [1 for any sensible partitioning
				scheme]
	
				Alternative 2: replicate the table across both disks (one
				replica on each disk) [1]. This will also provide simultaneous
				parallel access to the data (hard to tell whether this would be
				better or worse than the partitioning above) [1].
	
				Note that any solution must be capable of being implemented in
				Oracle8.
			\end{marking}
		\end{subsubquestion}

\begin{subsubquestion}{2}
			Briefly discuss the relative advantages and/or disadvantages of the
			two alternative solutions you described in part~\ref{nonindex}
			above.
			\begin{marking}
				The partitioning solution given above doesn't really help with
				the bronze customer rows---we would need random horizontal
				partitioning to really help with this, which Oracle8 doesn't
				support. More generally, Oracle8 does not migrate rows into the
				correct partition if a customer's status changes, so the data
				will need to be reloaded occasionally [1].

For replication, we have the usual problem of keeping the
				replicas synchronised if they are both read/write. This is
				somewhat mitigated by the fact that there are only two copies
				[1].
			\end{marking}
		\end{subsubquestion}
	\end{subquestion}
	
%----------------------------------------------------------------------

\end{question}

\newpage

%======================================================================

\begin{question}{10}
	There are two generic parameters that affect the structure of a B-tree
	index.

%----------------------------------------------------------------------

\begin{subquestion}{6}
		Identify these two parameters and describe what each parameter
		specifies. Use diagrams to illustrate your answer.
		\begin{marking}
			\emph{Node size} [1] specifies the maximum number of pointers in
			each index node [1]. Suitable diagram [1].

\emph{Percentage fill} [1] specifies the number of pointers in each
			index node that are allocated when a node is created [1]. In
			effect, it determines the amount of free space left in each node.
			Suitable diagram [1].
		\end{marking}
	\end{subquestion}
	
%----------------------------------------------------------------------

\begin{subquestion}{4}
		For each of the two parameters, discuss the practical effect(s) of:
		\begin{subsubquestion}{2}
			high values of the parameter; and
			\begin{marking}
				Large node size reduces the height of the B-tree, resulting in
				potentially fewer I/Os to access a leaf node [$\frac{1}{2}$].
				It can however increase the number of rows locked in a single
				leaf node [$\frac{1}{2}$].
	
				High percent fill means that most pointers in a node will be
				allocated when the node is created [$\frac{1}{2}$]. This leaves
				little free space in the B-tree for new key values or key
				values that change [$\frac{1}{2}$].
			\end{marking}
		\end{subsubquestion}
		\begin{subsubquestion}{2}
			low values of the parameter.
			\begin{marking}
				Small node size increases the height of the B-tree, resulting
				in potentially more I/Os to access a leaf node [$\frac{1}{2}$].
				The number of rows locked in a single leaf node will be reduced
				[$\frac{1}{2}$].
	
				Low percent fill means that most pointers in a node will not be
				allocated when the node is created [$\frac{1}{2}$]. This leaves
				little free space in the B-tree for new key values or key
				values that change [$\frac{1}{2}$].
			\end{marking}
		\end{subsubquestion}
	\end{subquestion}
	
%----------------------------------------------------------------------

\end{question}
	
%======================================================================

\begin{question}{5}\label{clusterq}

%----------------------------------------------------------------------

\begin{subquestion}{2}
		Define what is meant by the term ``clustering'' in the context of
		physical database design and tuning, and explain why it is beneficial.
		\begin{marking}
			Clustering is the grouping of related rows physically near to each
			other on disk [1] so that they may be retrieved in as few I/Os as
			possible (ideally one)---beneficial because of the slowness of disk
			I/O compared to memory I/O [1].
		\end{marking}
	\end{subquestion}
	
%----------------------------------------------------------------------

\begin{subquestion}{3}\label{clusterqcode}
		Consider the following Oracle8 SQL table definition:
		
		\begin{lstlisting}[language={[Oracle8]SQL}]{}
create table enrolment
(	student_id char(7),
	paper_code char(7),
	enrol_year number(4),
  
	primary key (student_id, paper_code, enrol_year)
);
		\end{lstlisting}

The \textsf{Enrolment} table is often queried for all rows relating to
		a particular student in a particular year (for example, student 1234567
		in 1998, student 9876543 in 2000, etc.). Write appropriate Oracle8 SQL
		code that will cluster the rows of the table in such a way as to
		support these types of query (you do not need to calculate cluster
		sizes). Relevant Oracle8 syntax diagrams are given on the next page.
		\begin{marking}
			\begin{lstlisting}[language={[Oracle8]SQL}]{}
create cluster enrolment_cluster (student_id char(7), year number(4));

create index enrol_cluster_index on cluster enrolment_cluster;

create table enrolment
(	student_id char(7),
	paper_code char(7),
	enrol_year number(4),
  
	primary key (student_id, paper_code, enrol_year)
) cluster enrolment_cluster(student_id, enrol_year);
			\end{lstlisting}
		\end{marking}
		[1] each for correct cluster, cluster index and table clustering.
	\end{subquestion}

\end{question}

\end{examsection}

\newpage

% CREATE CLUSTER (first part), p 4-207

\begin{center}
	\textbf{Oracle8 syntax diagrams for use in
	question~\ref{clusterq}(\ref{clusterqcode})}
	
	\bigskip

\includegraphics{CreateCluster}

\vfill

% CREATE INDEX (parts 1 and 2), p. 4-238

\includegraphics{CreateIndex}

% CREATE TABLE (parts 1 and 2), p. 4-308

\vfill

\includegraphics{CreateTable1}
	
	\bigskip	
	
	\includegraphics{CreateTable2}
\end{center}

%======================================================================
% OBJECT DBMS SECTION
%======================================================================
\begin{examsection}{25}{}{Questions in this section (total 25 marks) relate to object
database management systems and the Object Data Management Group's object model
(ODMG 3.0).}

%======================================================================

\begin{question}{10}\label{odlq}
	Examine the following Object Definition Language (ODL) schema, then answer
	the questions below.

\begin{lstlisting}[language=ODL]{}
module ExamODL
{
	interface Person
	{
		attribute string name;
		attribute date birth_date;
		attribute struct Address
		{	string street;
			string city;
			string country;
		} home_address;
	};

class Employee : Person (extent Staff key emp_ID)
	{
		attribute string emp_ID;
		attribute string position;
		attribute string department;
		attribute unsigned long salary;
	};
};
	\end{lstlisting}
	
%----------------------------------------------------------------------

\begin{subquestion}{2}
		How many attributes will instances of the \textsf{Employee} class have?
		Explain your answer.
		\begin{marking}
			This depends on your interpretation of what constitutes an
			attribute. I prefer seven [1]. It has four of its own and three
			inherited from the \textsf{Person} class (\textsf{home\_address}
			being counted as a single embedded literal attribute) [1].

Nine is a possible alternative answer [1], because you could count
			\textsf{home\_address} as three attributes, giving four local and
			five inherited attributes [1].

No marks for any other number \emph{unless} it has a sensible
			explanation, in which case award marks as appropriate.
		\end{marking}
	\end{subquestion}
	
%----------------------------------------------------------------------

\begin{subquestion}{3}
		Examine the clause
		``\lstinline[language={[Oracle8]SQL}]+(extent Staff key emp_ID)+''
		on line 14 of the ODL listing above.

\begin{subsubquestion}{2}
			Explain the effect(s) of this clause.
			\begin{marking}
				\lstinline[language={[Oracle8]SQL}]+extent Staff+ defines the
				name of the \emph{extent} (collection of all instances) of the
				\textsf{Employee} class [1].

\lstinline[language={[Oracle8]SQL}]+key emp_ID+ defines
				\textsf{emp\_ID} to be an \emph{object key} for the
				\textsf{Employee} class [1].
			\end{marking}
		\end{subsubquestion}
		
		\begin{subsubquestion}{1}
			Identify the relational equivalents of the items specified in this
			clause.
			\begin{marking}
				Extent = relation [$\frac{1}{2}$] and object key = primary key
				[$\frac{1}{2}$].
			\end{marking}
		\end{subsubquestion}
	\end{subquestion}

%----------------------------------------------------------------------

\begin{subquestion}{5}\label{oracleodl}
		Write \emph{all} the code required to implement an Oracle8 object type
		that is as similar as possible to the \textsf{Employee} class. The
		syntax diagram for the \textsf{\textbf{create type}} statement is given
		on the next page.
		\begin{marking}
			\begin{lstlisting}[language={[Oracle8]SQL}]{}
create type address_type as object
(	street varchar2(50),	-- the data types aren't that important
	city varchar2(30),		-- as long as they roughly match
	country varchar2(30)
);

create type employee_type as object
(	emp_id char(7),
	name varchar2(50),
	birth_date date,
	home_address address_type,
	position char(10),
	department char(20),
	salary number
);
			\end{lstlisting}
			[--1] for major errors, or for not including \textsf{address\_type}
			as a separate object type. [--$\frac{1}{2}$] for each attribute
			missing or in the wrong place.
		\end{marking}
	\end{subquestion}
	
\end{question}

\newpage

\begin{center}
	\textbf{Oracle8 syntax diagram for use in
	question~\ref{odlq}(\ref{oracleodl})}
	
	\bigskip
	\includegraphics{CreateObjectType}
	\bigskip\bigskip
\end{center}

%======================================================================

\begin{question}{15}

%----------------------------------------------------------------------

\begin{subquestion}{4}
		Compare in detail the ODMG concept of an \emph{object name} with the
		relational concept of a \emph{primary key}.
		\begin{marking}
			An object name is a meaningful user-generated identifier
			[$\frac{1}{2}$] for a single, specific object instance
			[$\frac{1}{2}$]. It is not part of the object's state
			[$\frac{1}{2}$] and a particular instance can have multiple object
			names associated with it [$\frac{1}{2}$].

In comparison, a primary key is also user-defined (but may not
			always be meaningful) [$\frac{1}{2}$]. While a primary key uniquely
			identifies individual rows (which are the equivalents of object
			instances), it is associated with \emph{all} rows in a table, not
			just a single row [$\frac{1}{2}$]. A primary key comprises
			attributes of the relation, so it is part of the relation's
			``state'' [$\frac{1}{2}$]. An individual row has only one primary
			key associated with it [$\frac{1}{2}$].
		\end{marking}
	\end{subquestion}
	
%----------------------------------------------------------------------

\begin{subquestion}{4}
		Compare in detail the ODMG concept of a \emph{relationship} with the
		relational concept of a \emph{foreign key}, with particular emphasis on
		any differences between the two concepts. Use an appropriate diagram or
		diagrams to illustrate your answer.
		\begin{marking}
			An ODMG relationship is a binary association between two object
			type, comprising a pair of \emph{traversal paths}, which are
			effectively pointers linking the two related instances [1]. ODMG
			relationships may be one-to-one, one-to-many or many-to-many, but
			we cannot specify participation [1].

A relational foreign key is also a binary association between two
			relations, comprising an attribute stored in one of the relations
			that refers to the primary key of the other relation; in this
			respect, it is really equivalent to a traversal path rather than a
			full ODMG relationship [1]. Relationships defined by foreign keys
			cannot be many-to-many, but we can partially specify participation
			[1].

Useful diagram(s).
		\end{marking}
	\end{subquestion}
	
%----------------------------------------------------------------------

\begin{subquestion}{6}
		Define in detail the ODMG concepts of \emph{interface}, \emph{class}
		and \emph{literal}.
		\begin{marking}
			An interface is an ODMG object type that defines behaviour only
			[1]. You cannot create instances of interfaces (cf. abstract
			classes) [1].

A class is an ODMG object type that can define both state and
			behaviour [1]. You can create instances of classes [1].

A literal is an ODMG type (non-object) that defines state only [1].
			They have no identity and are effectively just structured data
			types [1].
		\end{marking}
	\end{subquestion}
	
%----------------------------------------------------------------------

\begin{subquestion}{1}
		Date and Darwen claim in their \emph{Third Manifesto} that the
		relational model can support objects if it is implemented correctly.
		Briefly explain how they claim this can be achieved.
		\begin{marking}
			If we implement domains correctly/fully then they are effectively
			the equivalent of an object class [1].
		\end{marking}
	\end{subquestion}

\end{question}

\end{examsection}

%======================================================================
% DISTRIBUTED DBMS SECTION
%======================================================================
\begin{examsection}{25}{}{Questions in this section (total 25 marks) relate to
distributed database systems.}

%======================================================================

\begin{question}{4}
	Describe what is meant by the terms ``distributed data'' and ``distributed
	database'' and explain the key difference(s) between the two. Use an
	example to illustrate the difference(s).
	\begin{marking}
		Distributed data are spread across multiple nodes on a network but are
		neither centrally managed nor highly integrated [1]. A distributed
		database combines distributed data with distributed database processing
		[1] (including management, access and transaction processing)---this is
		the key difference [1]. Reasonable example [1].
	\end{marking}
\end{question}

%======================================================================

\begin{question}{10}
	Date's ``objective zero'' for distributed database management systems
	states ``a distributed database looks and behaves like a non-distributed
	database to any user''. Describe two \emph{different} ways in which this
	objective may be achieved, and for each discuss in detail how well it is
	implemented in Oracle8.
	\begin{marking}
		Basically a detailed description of any two of Date's objectives.
		Example: Location transparency is the ability to query data without
		needing to know on which network node the data are physically located.
		It is implemented reasonably well in Oracle8 by way of synonyms and
		views, after some initial setup work. [2 marks for description, 3 marks
		for implementation]
	\end{marking}
\end{question}

%======================================================================

\begin{question}{5}
	A company currently has a distributed database spread across two sites. It
	is necessary to move some data from one site to the other.

%----------------------------------------------------------------------

\begin{subquestion}{3}
		Describe the general process that allows us to complete distributed
		transactions such as that described above with reasonable safety.
		\begin{marking}
			This is a distributed transaction, so the process involved is
			two-phase commit. In the first (\emph{prepare}) phase, the
			transaction coordinator tells all sites involved in the transaction
			to prepare to commit (by flushing buffers, locking data, etc.).
			Each site then responds ``OK'' or ``abort''. In the second
			(\emph{commit}) phase, if all sites respond ``OK'' the coordinator
			tells all sites to commit, otherwise it tells all sites to roll
			back.

[1] for identifying 2PC, [1] each for both phases.
		\end{marking}
	\end{subquestion}
	
%----------------------------------------------------------------------

\begin{subquestion}{2}
		Assuming an Oracle8 distributed database, give an example of a
		situation in which it may not be possible to automatically recover from
		a failed distributed transaction, and explain how such a situation
		might be dealt with.
		\begin{marking}
			If a site fails for a short period, in-doubt transactions will
			automatically complete when the site comes back up. If however the
			outage is for a significant period of time, then the DBMS cannot
			automatically commit in-doubt transactions as the data context may
			have changed considerably in the interim [1]. The DBA can commit or
			roll back in-doubt transactions manually, but this is risky [1].
		\end{marking} 
	\end{subquestion}

\end{question}

%======================================================================

\begin{question}{6}

%----------------------------------------------------------------------

\begin{subquestion}{2}
		Consider the following Oracle8 SQL statement:
		\begin{lstlisting}[language={[Oracle8]SQL},numbers=none]{}
create synonym exam_synonym for sales.customer@db_chch.world;
		\end{lstlisting}
		Explain the purpose of this statement.
		\begin{marking}
			It creates (in the local schema) an ``alias'' to the
			\textsf{customer} table in the \textsf{sales} schema of the
			\textsf{db\_chch.world} database [1]. This synonym can be queried
			as if it were a local copy of the table (i.e., it provides location
			transparency) [1].
		\end{marking}
	\end{subquestion}

%----------------------------------------------------------------------

\begin{subquestion}{4}
		Consider the following Oracle8 SQL statement:
		\begin{lstlisting}[language={[Oracle8]SQL}]{}
create snapshot exam_snapshot
	refresh fast
		start with sysdate
		next sysdate + 1/12
	as select * from sales.customer@db_chch.world;
		\end{lstlisting}
		Explain in detail what this statement does.
		\begin{marking}
			It creates a read only replica [1] of the \textsf{customer} table
			from the \textsf{sales} schema in the \textsf{db\_chch.world}
			database [1]. Changes to the original table only [1] are copied
			across every two hours starting now [1].
		\end{marking}
	\end{subquestion}

\end{question}
	
%======================================================================

\end{examsection}

%======================================================================
% DECISION SUPPORT SYSTEMS SECTION
%======================================================================
\begin{examsection}{25}{}{Questions in this section (total 25 marks) relate to data
warehouses and online analytical processing (OLAP) tools.}

%======================================================================

\begin{question}{8}
	Compare in detail the characteristics of operational data and decision
	support data with respect to the following four areas:
	\begin{itemize}
		\item the types of decisions they are used to support;
		\item their timespan;
		\item their granularity; and
		\item their dimensionality.
	\end{itemize}
	\begin{marking}
		(Note: ``in detail'' implies more than just a table of brief
		points---only award half marks for such answers.)
		\begin{itemize}

\item \textbf{Decision type:} operational data support short term
			day-to-day decision making [1] whereas decision support data
			support tactical and strategic decision making (longer term) [1].

\item \textbf{Timespan:} operational data tend to focus on
			\emph{now} (current transactions over maybe a few days) [1] whereas
			DS data range over much larger time scales (months, years,
			decades), both backwards and forwards [1].

\item \textbf{Granularity:} operational data are generally very
			fine-grained (individual transactions) [1], whereas DS data are
			aggregated at varying levels of summarisation [1].

\item \textbf{Dimensionality:} operational data tend to be ``flat''
			(only one variable of interest associated with a given data value)
			[1], whereas DS data can have many dimensions (variables of
			interest) associated with a data value [1].

\end{itemize}
	\end{marking}
\end{question}

%======================================================================

\begin{question}{6}
	Discuss in detail the key differences between databases designed for online
	transaction processing and databases designed for data warehousing, in
	terms of both:

\begin{subquestion}{2}
		the pattern of data use; and
		\begin{marking}
			OLTP databases are typically not queried heavily and the queries
			tend to be relatively simple [$\frac{1}{2}$]; they also have a high
			rate of online concurrent updates [$\frac{1}{2}$]. Conversely, data
			warehouses typically have very high query rates and the queries are
			very complex [$\frac{1}{2}$]; there are usually few updates (often
			read-only) and those updates that do occur are often carried out as
			batch operations [$\frac{1}{2}$]. ([$\frac{1}{2}$] for saying that
			OLTP typically have many concurrent users, or that data warehouses
			typically have few users.)
		\end{marking}
	\end{subquestion}

\begin{subquestion}{4}
		physical design and tuning strategy.
		\begin{marking}
			OLTP databases are queried and updated by many users
			simultaneously, so we need to trade off between optimising for
			concurrency, updates or query performance [1]. The usual tuning
			techniques may be applied, but we should be particularly careful
			with the use of indexes (typically B-tree) because of the
			relatively high update rates [1]. ([$\frac{1}{2}$] for saying that
			the database is often relatively small.)

In a data warehouse, updates are few but queries are many and
			complex, so we need to optimise aggessively for query performance
			[1]. The usual tuning techniques may be applied, but in particular
			we can heavily index the data (bitmap indexes and index-organised
			tables really come into their own here) [1].([$\frac{1}{2}$] for
			saying that the database is typically very large.)
		\end{marking}
	\end{subquestion}
\end{question}

%======================================================================

\begin{question}{4}
	Describe in detail the steps involved in the process of importing data into
	a data warehouse from external sources.
	\begin{marking}
		\begin{itemize}

\item Extraction: obtaining data from various sources
			(electronically or otherwise) [1]. The data are often loaded into a
			special \emph{staging database} for further processing [1].

\item Transformation: data are cleaned up and manipulated into
			standard formats, so that data may be queried consistently
			regardless of source [1].

\item Loading: the transformed data are loaded into the data
			warehouse by some form of bulk loading tool [1].

\end{itemize}
	\end{marking}
\end{question}

%======================================================================

\begin{question}{1}
	Explain what is meant by the term ``data mart''.
	\begin{marking}
		A data mart is a small \emph{specialised} data warehouse, often a
		subset of a larger enterprise-wide data warehouse [1].
	\end{marking}
\end{question}

%======================================================================

\begin{question}{6}
	Describe in detail two ways in which an OLAP tool can be used to change the
	way in which data are presented. Use diagrams and/or examples to illustrate
	your answer.
	\begin{marking}
		Any two of:
		\begin{itemize}

\item Drill down/roll up allows us to move from more summarised
			data to more specific data and vice versa.

\item ``Slice and dice'' allows us to reduce our view of a
			multidimensional data set to specific values of one or more
			dimensions.

\item Pivoting allows us to change the layout of the data being
			output to give a different perspective on the same data set, e.g.,
			viewing sales of products by region vs. sales in a region by
			product.

\end{itemize}
		[1] each for technique, [1] each for description, [1] each for useful
		diagrams or examples. Any of the following are worth half marks:
		\begin{itemize}
			\item graphs/visualisation tools
			\item highlighting exceptions
		\end{itemize}
	\end{marking}
\end{question}

\end{examsection}

\end{document}

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%
% File: $Id$
%
% Note: No document class --- this is handled by the appropriate
% "driver" file, which makes it easier to switch class options
% such as markingschedule. Doing it this way means that you don't
% wipe out the original exam paper when you want to generate a
% marking schedule.
%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

\usepackage{graphicx}
\usepackage{listings}

\examyear{2001}
\department{Information Science}
\papernumber{INFO 321}
\papertitle{Database Systems}
\instructions{Answer \underline{ALL} questions.}
\allowcalculators[approved]

\begin{document}

\maketitlepage

%======================================================================

\begin{question}{10}

\textsf{\textbf{Customer}(\underline{customer\_no}, name, address, city,
	state, country, phone, email, status)}

\begin{itemize}

\item \textsf{customer\_no} is the primary key and because this is an
		Oracle8 database, a unique B-tree index has been automatically created
		on this column.

\item About 10\% of customers are inactive, about 50\% have bronze
		status, about 35\% have silver status and about 5\% have gold status.

\item The average row size is about 200 bytes and there are
		approximately 5,000,000 rows.

\item UPDATE, INSERT and DELETE operations on the table are uncommon.

\item Very few queries are for individual customers; rather, most
		queries are based on various combinations of values from
		\textsf{status}, \textsf{city}, \textsf{state} and \textsf{country}.

\end{itemize}
	\medskip

%----------------------------------------------------------------------

Hashing is definitely \emph{not} an option, because very few
			queries are exact match.
		\end{marking}
	\end{subquestion}
		
%----------------------------------------------------------------------

\end{question}

\newpage

%======================================================================

\begin{question}{10}
	There are two generic parameters that affect the structure of a B-tree
	index.

%----------------------------------------------------------------------

\end{question}
	
%======================================================================

\begin{question}{5}\label{clusterq}

%----------------------------------------------------------------------

create index enrol_cluster_index on cluster enrolment_cluster;

\end{question}

\end{examsection}

\newpage

% CREATE CLUSTER (first part), p 4-207

\begin{center}
	\textbf{Oracle8 syntax diagrams for use in
	question~\ref{clusterq}(\ref{clusterqcode})}
	
	\bigskip

\includegraphics{CreateCluster}

\vfill

% CREATE INDEX (parts 1 and 2), p. 4-238

\includegraphics{CreateIndex}

% CREATE TABLE (parts 1 and 2), p. 4-308

\vfill

\includegraphics{CreateTable1}
	
	\bigskip	
	
	\includegraphics{CreateTable2}
\end{center}

%======================================================================

\begin{question}{10}\label{odlq}
	Examine the following Object Definition Language (ODL) schema, then answer
	the questions below.

Nine is a possible alternative answer [1], because you could count
			\textsf{home\_address} as three attributes, giving four local and
			five inherited attributes [1].

\begin{subquestion}{3}
		Examine the clause
		``\lstinline[language={[Oracle8]SQL}]+(extent Staff key emp_ID)+''
		on line 14 of the ODL listing above.

%----------------------------------------------------------------------

\newpage

\begin{center}
	\textbf{Oracle8 syntax diagram for use in
	question~\ref{odlq}(\ref{oracleodl})}
	
	\bigskip
	\includegraphics{CreateObjectType}
	\bigskip\bigskip
\end{center}

%======================================================================

\begin{question}{15}

%----------------------------------------------------------------------

Useful diagram(s).
		\end{marking}
	\end{subquestion}
	
%----------------------------------------------------------------------

A class is an ODMG object type that can define both state and
			behaviour [1]. You can create instances of classes [1].

\end{question}

\end{examsection}

%======================================================================

\begin{question}{5}
	A company currently has a distributed database spread across two sites. It
	is necessary to move some data from one site to the other.

%----------------------------------------------------------------------

[1] for identifying 2PC, [1] each for both phases.
		\end{marking}
	\end{subquestion}
	
%----------------------------------------------------------------------

\end{question}

%======================================================================

\begin{question}{6}

%----------------------------------------------------------------------

\end{question}
	
%======================================================================

\end{examsection}

%======================================================================

\item \textbf{Decision type:} operational data support short term
			day-to-day decision making [1] whereas decision support data
			support tactical and strategic decision making (longer term) [1].

\item \textbf{Granularity:} operational data are generally very
			fine-grained (individual transactions) [1], whereas DS data are
			aggregated at varying levels of summarisation [1].

\end{itemize}
	\end{marking}
\end{question}

%======================================================================

\begin{question}{6}
	Discuss in detail the key differences between databases designed for online
	transaction processing and databases designed for data warehousing, in
	terms of both:

%======================================================================

\begin{question}{4}
	Describe in detail the steps involved in the process of importing data into
	a data warehouse from external sources.
	\begin{marking}
		\begin{itemize}

\item Extraction: obtaining data from various sources
			(electronically or otherwise) [1]. The data are often loaded into a
			special \emph{staging database} for further processing [1].

\item Transformation: data are cleaned up and manipulated into
			standard formats, so that data may be queried consistently
			regardless of source [1].

\item Loading: the transformed data are loaded into the data
			warehouse by some form of bulk loading tool [1].

\end{itemize}
	\end{marking}
\end{question}

%======================================================================

\item Drill down/roll up allows us to move from more summarised
			data to more specific data and vice versa.

\item ``Slice and dice'' allows us to reduce our view of a
			multidimensional data set to specific values of one or more
			dimensions.

\end{examsection}

\end{document}

Ignore Space Show notes View example1.tex
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % % Generates a plain examination paper. % %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% \documentclass{ouexam} \input{example} %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % % File: $Id$ % % Generates a plain examination paper. % %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% \documentclass{ouexam} \input{example}

Ignore Space Show notes View

example1.tex

Ignore Space Show notes View example2.tex
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % % Generates an examination paper with marking schedule information. % %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% \documentclass[markingschedule]{ouexam} \input{example} %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % % File: $Id$ % % Generates an examination paper with marking schedule information. % %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% \documentclass[markingschedule]{ouexam} \input{example}

Ignore Space Show notes View

example2.tex

Ignore Space Show notes View lstlang0.sty
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % % Listings language definitions used in the ouexam examples. % %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% \lst@definelanguage{ODL}% { morekeywords={class,extent,key,keys,module,interface,const,true,false,% typedef,date,time,interval,timestamp,dictionary,set,list,bag,float,% double,long,short,unsigned,char,boolean,octet,any,struct,union,switch,% case,default,enum,array,sequence,string,readonly,attribute,relationship,% inverse,exception,oneway,void,in,out,inout,raises,context,extends},% sensitive=true,% morecomment=[l]//,% morecomment=[s]{/}{/},% morestring=[b]",% morestring=[b]'% }[keywords,comments,strings] \lst@definelanguage{OQL}% { morekeywords={abs,all,and,andthen,any,array,as,asc,avg,bag,boolean,by,% char,count,date,define,desc,dictionary,distinct,double,element,enum,% except,exists,false,first,flatten,float,for,from,group,having,import,% in,intersect,interval,is_defined,is_undefined,last,like,list,listtoset,% long,max,min,mod,nil,not,octet,or,order,orelse,query,select,set,short,% some,string,struct,sum,time,timestamp,true,undefine,undefined,union,% unique,unsigned,where},% sensitive=true,% morecomment=[l]//,% morecomment=[s]{/}{/},% morestring=[b]",% morestring=[b]'% }[keywords,comments,strings] \lst@definelanguage[Oracle8]{SQL}% { morekeywords={access,account,activate,add,admin,after,all,allocate,% all_rows,alter,analyze,and,any,archive,archivelog,array,as,asc,% at,audit,authenticated,authorization,autoextend,automatic,avg,backup,% become,before,begin,between,bfile,bitmap,blob,block,body,by,cache,% cache_instances,cancel,cascade,cast,cfile,chained,change,char,character,% char_cs,check,checkpoint,choose,chunk,clear,clob,clone,close,% close_open_cached_cursors,cluster,coalesce,column,columns,comment,% commit,committed,compatibility,compile,complete,composite_limit,% compress,compute,connect,connect_time,constraint,constraints,contents,% continue,controlfile,convert,cost,count,cpu_per_call,cpu_per_session,% create,current,current_schema,current_user,cursor,cycle,dangling,% database,datafile,datafiles,dataobjno,date,dba,deallocate,debug,dec,% decimal,declare,default,deferrable,deferred,degree,delete,deref,desc,% directory,disable,disconnect,dismount,distinct,distributed,dml,double,% drop,dump,each,else,enable,end,enforce,entry,escape,estimate,events,% except,exceptions,exchange,excluding,exclusive,execute,exists,expire,% explain,extent,extents,externally,failed_login_attempts,false,fast,% file,first_rows,flagger,float,flush,for,force,foreign,freelist,% freelists,from,full,function,global,globally,global_name,grant,% group,groups,hash,hashkeys,having,header,heap,identified,idle_time,% if,immediate,in,including,increment,index,indexed,indexes,indicator,% ind_partition,initial,initially,initrans,insert,instance,instances,% instead,int,integer,intermediate,intersect,into,is,isolation,% isolation_level,keep,key,kill,layer,less,level,library,like,limit,% link,list,lob,local,lock,log,logfile,logging,logical_reads_per_call,% logical_reads_per_session,long,manage,master,max,maxarchlogs,% maxdatafiles,maxextents,maxinstances,maxlogfiles,maxloghistory,% maxlogmembers,maxsize,maxtrans,maxvalue,member,min,minextents,% minimum,minus,minvalue,mode,modify,mount,move,mts_dispatchers,% multiset,national,nchar,nchar_cs,nclob,needed,nested,network,new,% next,nls_calendar,nls_characterset,nls_iso_currency,nls_language,% nls_numeric_characters,nls_sort,nls_special_chars,nls_territory,% noarchivelog,noaudit,nocache,nocompress,nocycle,noforce,nologging,% nomaxvalue,nominvalue,none,noorder,nooveride,noparallel,noresetlogs,% noreverse,normal,nosort,not,nothing,nowait,null,number,numeric,% nvarchar2,object,objno,objno_reuse,of,off,offline,oid,oidindex,% old,on,online,only,opcode,open,optimal,optimizer_goal,option,or,% order,organization,overflow,own,package,parallel,partition,% password,password_grace_time,password_life_time,password_lock_time,% password_reuse_max,password_reuse_time,password_verify_function,% pctfree,pctincrease,pctthreshold,pctused,pctversion,percent,% permanent,plan,plsql_debug,post_transaction,precision,preserve,% primary,prior,private,private_sga,privilege,privileges,procedure,% profile,public,purge,queue,quota,range,raw,rba,read,real,rebuild,% recover,recoverable,recovery,ref,references,referencing,refresh,% rename,replace,reset,resetlogs,resize,resource,restricted,return,% returning,reuse,reverse,revoke,role,roles,rollback,row,rowid,rowlabel,% rownum,rows,rule,sample,savepoint,scan_instances,schema,scn,scope,% sd_all,sd_inhibit,sd_show,segment,seg_block,seg_file,select,sequence,% serializable,session,sessions_per_user,session_cached_cursors,set,% share,shared,shared_pool,shrink,size,skim_unusable_indexes,smallint,% snapshot,some,sort,specification,split,sqlcode,sqlerror,sql_trace,% standby,start,statement_id,statistics,stop,storage,store,structure,% successful,sum,switch,synonym,sysdate,sysdba,sysoper,system,table,% tables,tablespace,tablespaces,tablespace_no,tabno,temporary,than,% the,then,thread,time,timestamp,to,toplevel,trace,tracing,transaction,% transitional,trigger,triggers,true,truncate,tx,type,uba,uid,unarchived,% under,undo,union,unique,unlimited,unlock,unrecoverable,until,unusable,% unused,updatable,update,usage,use,user,using,validate,validation,value,% values,varchar,varchar2,varray,varying,view,when,whenever,where,with,% without,work,write,xid},% sensitive=false,% morecomment=[l]--,% morecomment=[s]{/}{/},% morestring=[b]",% morestring=[b]'% }[keywords,comments,strings] \lst@definelanguage[Oracle8]{PLSQL}[Oracle8]{SQL} { morekeywords={ abort,accept,arraylen,assert,assign,base_table,binary_integer,% boolean,case,char_base,clusters,colauth,constant,crash,currval,% data_base,debugoff,debugon,definition,delay,digits,dispose,do,% elsif,exception,exit,fetch,form,generic,goto,interface,limited,% loop,mlslabel,mod,natural,naturaln,nextval,number_base,others,out,% pls_integer,positive,positiven,pragma,raise,record,release,remr,% rowtype,run,separate,signtype,space,sql,sqlerrm,statement,stddev,% subtype,tabauth,task,terminate,variance,views,while,xor},% sensitive=false% }[keywords] \endinput %% %% End of file `lstlang0.sty'. %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % % File: $Id$ % % Listings language definitions used in the ouexam examples. % %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% \lst@definelanguage{ODL}% { morekeywords={class,extent,key,keys,module,interface,const,true,false,% typedef,date,time,interval,timestamp,dictionary,set,list,bag,float,% double,long,short,unsigned,char,boolean,octet,any,struct,union,switch,% case,default,enum,array,sequence,string,readonly,attribute,relationship,% inverse,exception,oneway,void,in,out,inout,raises,context,extends},% sensitive=true,% morecomment=[l]//,% morecomment=[s]{/}{/},% morestring=[b]",% morestring=[b]'% }[keywords,comments,strings] \lst@definelanguage{OQL}% { morekeywords={abs,all,and,andthen,any,array,as,asc,avg,bag,boolean,by,% char,count,date,define,desc,dictionary,distinct,double,element,enum,% except,exists,false,first,flatten,float,for,from,group,having,import,% in,intersect,interval,is_defined,is_undefined,last,like,list,listtoset,% long,max,min,mod,nil,not,octet,or,order,orelse,query,select,set,short,% some,string,struct,sum,time,timestamp,true,undefine,undefined,union,% unique,unsigned,where},% sensitive=true,% morecomment=[l]//,% morecomment=[s]{/}{/},% morestring=[b]",% morestring=[b]'% }[keywords,comments,strings] \lst@definelanguage[Oracle8]{SQL}% { morekeywords={access,account,activate,add,admin,after,all,allocate,% all_rows,alter,analyze,and,any,archive,archivelog,array,as,asc,% at,audit,authenticated,authorization,autoextend,automatic,avg,backup,% become,before,begin,between,bfile,bitmap,blob,block,body,by,cache,% cache_instances,cancel,cascade,cast,cfile,chained,change,char,character,% char_cs,check,checkpoint,choose,chunk,clear,clob,clone,close,% close_open_cached_cursors,cluster,coalesce,column,columns,comment,% commit,committed,compatibility,compile,complete,composite_limit,% compress,compute,connect,connect_time,constraint,constraints,contents,% continue,controlfile,convert,cost,count,cpu_per_call,cpu_per_session,% create,current,current_schema,current_user,cursor,cycle,dangling,% database,datafile,datafiles,dataobjno,date,dba,deallocate,debug,dec,% decimal,declare,default,deferrable,deferred,degree,delete,deref,desc,% directory,disable,disconnect,dismount,distinct,distributed,dml,double,% drop,dump,each,else,enable,end,enforce,entry,escape,estimate,events,% except,exceptions,exchange,excluding,exclusive,execute,exists,expire,% explain,extent,extents,externally,failed_login_attempts,false,fast,% file,first_rows,flagger,float,flush,for,force,foreign,freelist,% freelists,from,full,function,global,globally,global_name,grant,% group,groups,hash,hashkeys,having,header,heap,identified,idle_time,% if,immediate,in,including,increment,index,indexed,indexes,indicator,% ind_partition,initial,initially,initrans,insert,instance,instances,% instead,int,integer,intermediate,intersect,into,is,isolation,% isolation_level,keep,key,kill,layer,less,level,library,like,limit,% link,list,lob,local,lock,log,logfile,logging,logical_reads_per_call,% logical_reads_per_session,long,manage,master,max,maxarchlogs,% maxdatafiles,maxextents,maxinstances,maxlogfiles,maxloghistory,% maxlogmembers,maxsize,maxtrans,maxvalue,member,min,minextents,% minimum,minus,minvalue,mode,modify,mount,move,mts_dispatchers,% multiset,national,nchar,nchar_cs,nclob,needed,nested,network,new,% next,nls_calendar,nls_characterset,nls_iso_currency,nls_language,% nls_numeric_characters,nls_sort,nls_special_chars,nls_territory,% noarchivelog,noaudit,nocache,nocompress,nocycle,noforce,nologging,% nomaxvalue,nominvalue,none,noorder,nooveride,noparallel,noresetlogs,% noreverse,normal,nosort,not,nothing,nowait,null,number,numeric,% nvarchar2,object,objno,objno_reuse,of,off,offline,oid,oidindex,% old,on,online,only,opcode,open,optimal,optimizer_goal,option,or,% order,organization,overflow,own,package,parallel,partition,% password,password_grace_time,password_life_time,password_lock_time,% password_reuse_max,password_reuse_time,password_verify_function,% pctfree,pctincrease,pctthreshold,pctused,pctversion,percent,% permanent,plan,plsql_debug,post_transaction,precision,preserve,% primary,prior,private,private_sga,privilege,privileges,procedure,% profile,public,purge,queue,quota,range,raw,rba,read,real,rebuild,% recover,recoverable,recovery,ref,references,referencing,refresh,% rename,replace,reset,resetlogs,resize,resource,restricted,return,% returning,reuse,reverse,revoke,role,roles,rollback,row,rowid,rowlabel,% rownum,rows,rule,sample,savepoint,scan_instances,schema,scn,scope,% sd_all,sd_inhibit,sd_show,segment,seg_block,seg_file,select,sequence,% serializable,session,sessions_per_user,session_cached_cursors,set,% share,shared,shared_pool,shrink,size,skim_unusable_indexes,smallint,% snapshot,some,sort,specification,split,sqlcode,sqlerror,sql_trace,% standby,start,statement_id,statistics,stop,storage,store,structure,% successful,sum,switch,synonym,sysdate,sysdba,sysoper,system,table,% tables,tablespace,tablespaces,tablespace_no,tabno,temporary,than,% the,then,thread,time,timestamp,to,toplevel,trace,tracing,transaction,% transitional,trigger,triggers,true,truncate,tx,type,uba,uid,unarchived,% under,undo,union,unique,unlimited,unlock,unrecoverable,until,unusable,% unused,updatable,update,usage,use,user,using,validate,validation,value,% values,varchar,varchar2,varray,varying,view,when,whenever,where,with,% without,work,write,xid},% sensitive=false,% morecomment=[l]--,% morecomment=[s]{/}{/},% morestring=[b]",% morestring=[b]'% }[keywords,comments,strings] \lst@definelanguage[Oracle8]{PLSQL}[Oracle8]{SQL} { morekeywords={ abort,accept,arraylen,assert,assign,base_table,binary_integer,% boolean,case,char_base,clusters,colauth,constant,crash,currval,% data_base,debugoff,debugon,definition,delay,digits,dispose,do,% elsif,exception,exit,fetch,form,generic,goto,interface,limited,% loop,mlslabel,mod,natural,naturaln,nextval,number_base,others,out,% pls_integer,positive,positiven,pragma,raise,record,release,remr,% rowtype,run,separate,signtype,space,sql,sqlerrm,statement,stddev,% subtype,tabauth,task,terminate,variance,views,while,xor},% sensitive=false% }[keywords] \endinput %% %% End of file `lstlang0.sty'.

Ignore Space Show notes View

lstlang0.sty

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%
% Listings language definitions used in the ouexam examples.
%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

\lst@definelanguage{ODL}%
{	morekeywords={class,extent,key,keys,module,interface,const,true,false,%
		typedef,date,time,interval,timestamp,dictionary,set,list,bag,float,%
		double,long,short,unsigned,char,boolean,octet,any,struct,union,switch,%
		case,default,enum,array,sequence,string,readonly,attribute,relationship,%
		inverse,exception,oneway,void,in,out,inout,raises,context,extends},%
	sensitive=true,%
	morecomment=[l]//,%
	morecomment=[s]{/*}{*/},%
	morestring=[b]",%
	morestring=[b]'%
}[keywords,comments,strings]

\lst@definelanguage{OQL}%
{	morekeywords={abs,all,and,andthen,any,array,as,asc,avg,bag,boolean,by,%
		char,count,date,define,desc,dictionary,distinct,double,element,enum,%
		except,exists,false,first,flatten,float,for,from,group,having,import,%
		in,intersect,interval,is_defined,is_undefined,last,like,list,listtoset,%
		long,max,min,mod,nil,not,octet,or,order,orelse,query,select,set,short,%
		some,string,struct,sum,time,timestamp,true,undefine,undefined,union,%
		unique,unsigned,where},%
	sensitive=true,%
	morecomment=[l]//,%
	morecomment=[s]{/*}{*/},%
	morestring=[b]",%
	morestring=[b]'%
}[keywords,comments,strings]

\lst@definelanguage[Oracle8]{SQL}%
{	morekeywords={access,account,activate,add,admin,after,all,allocate,%
		all_rows,alter,analyze,and,any,archive,archivelog,array,as,asc,%
		at,audit,authenticated,authorization,autoextend,automatic,avg,backup,%
		become,before,begin,between,bfile,bitmap,blob,block,body,by,cache,%
		cache_instances,cancel,cascade,cast,cfile,chained,change,char,character,%
		char_cs,check,checkpoint,choose,chunk,clear,clob,clone,close,%
		close_open_cached_cursors,cluster,coalesce,column,columns,comment,%
		commit,committed,compatibility,compile,complete,composite_limit,%
		compress,compute,connect,connect_time,constraint,constraints,contents,%
		continue,controlfile,convert,cost,count,cpu_per_call,cpu_per_session,%
		create,current,current_schema,current_user,cursor,cycle,dangling,%
		database,datafile,datafiles,dataobjno,date,dba,deallocate,debug,dec,%
		decimal,declare,default,deferrable,deferred,degree,delete,deref,desc,%
		directory,disable,disconnect,dismount,distinct,distributed,dml,double,%
		drop,dump,each,else,enable,end,enforce,entry,escape,estimate,events,%
		except,exceptions,exchange,excluding,exclusive,execute,exists,expire,%
		explain,extent,extents,externally,failed_login_attempts,false,fast,%
		file,first_rows,flagger,float,flush,for,force,foreign,freelist,%
		freelists,from,full,function,global,globally,global_name,grant,%
		group,groups,hash,hashkeys,having,header,heap,identified,idle_time,%
		if,immediate,in,including,increment,index,indexed,indexes,indicator,%
		ind_partition,initial,initially,initrans,insert,instance,instances,%
		instead,int,integer,intermediate,intersect,into,is,isolation,%
		isolation_level,keep,key,kill,layer,less,level,library,like,limit,%
		link,list,lob,local,lock,log,logfile,logging,logical_reads_per_call,%
		logical_reads_per_session,long,manage,master,max,maxarchlogs,%
		maxdatafiles,maxextents,maxinstances,maxlogfiles,maxloghistory,%
		maxlogmembers,maxsize,maxtrans,maxvalue,member,min,minextents,%
		minimum,minus,minvalue,mode,modify,mount,move,mts_dispatchers,%
		multiset,national,nchar,nchar_cs,nclob,needed,nested,network,new,%
		next,nls_calendar,nls_characterset,nls_iso_currency,nls_language,%
		nls_numeric_characters,nls_sort,nls_special_chars,nls_territory,%
		noarchivelog,noaudit,nocache,nocompress,nocycle,noforce,nologging,%
		nomaxvalue,nominvalue,none,noorder,nooveride,noparallel,noresetlogs,%
		noreverse,normal,nosort,not,nothing,nowait,null,number,numeric,%
		nvarchar2,object,objno,objno_reuse,of,off,offline,oid,oidindex,%
		old,on,online,only,opcode,open,optimal,optimizer_goal,option,or,%
		order,organization,overflow,own,package,parallel,partition,%
		password,password_grace_time,password_life_time,password_lock_time,%
		password_reuse_max,password_reuse_time,password_verify_function,%
		pctfree,pctincrease,pctthreshold,pctused,pctversion,percent,%
		permanent,plan,plsql_debug,post_transaction,precision,preserve,%
		primary,prior,private,private_sga,privilege,privileges,procedure,%
		profile,public,purge,queue,quota,range,raw,rba,read,real,rebuild,%
		recover,recoverable,recovery,ref,references,referencing,refresh,%
		rename,replace,reset,resetlogs,resize,resource,restricted,return,%
		returning,reuse,reverse,revoke,role,roles,rollback,row,rowid,rowlabel,%
		rownum,rows,rule,sample,savepoint,scan_instances,schema,scn,scope,%
		sd_all,sd_inhibit,sd_show,segment,seg_block,seg_file,select,sequence,%
		serializable,session,sessions_per_user,session_cached_cursors,set,%
		share,shared,shared_pool,shrink,size,skim_unusable_indexes,smallint,%
		snapshot,some,sort,specification,split,sqlcode,sqlerror,sql_trace,%
		standby,start,statement_id,statistics,stop,storage,store,structure,%
		successful,sum,switch,synonym,sysdate,sysdba,sysoper,system,table,%
		tables,tablespace,tablespaces,tablespace_no,tabno,temporary,than,%
		the,then,thread,time,timestamp,to,toplevel,trace,tracing,transaction,%
		transitional,trigger,triggers,true,truncate,tx,type,uba,uid,unarchived,%
		under,undo,union,unique,unlimited,unlock,unrecoverable,until,unusable,%
		unused,updatable,update,usage,use,user,using,validate,validation,value,%
		values,varchar,varchar2,varray,varying,view,when,whenever,where,with,%
		without,work,write,xid},%
	sensitive=false,%
	morecomment=[l]--,%
	morecomment=[s]{/*}{*/},%
	morestring=[b]",%
	morestring=[b]'%
}[keywords,comments,strings]

\lst@definelanguage[Oracle8]{PLSQL}[Oracle8]{SQL}
{	morekeywords={ abort,accept,arraylen,assert,assign,base_table,binary_integer,%
		boolean,case,char_base,clusters,colauth,constant,crash,currval,%
		data_base,debugoff,debugon,definition,delay,digits,dispose,do,%
		elsif,exception,exit,fetch,form,generic,goto,interface,limited,%
		loop,mlslabel,mod,natural,naturaln,nextval,number_base,others,out,%
		pls_integer,positive,positiven,pragma,raise,record,release,remr,%
		rowtype,run,separate,signtype,space,sql,sqlerrm,statement,stddev,%
		subtype,tabauth,task,terminate,variance,views,while,xor},%
	sensitive=false%
}[keywords]
\endinput
%%
%% End of file `lstlang0.sty'.

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%
% File: $Id$
%
% Listings language definitions used in the ouexam examples.
%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

Ignore Space Show notes View ouexam.dtx
\def\ouexamdate{13 June 2013} \def\ouexamversion{2.4} \def\ouexamshortdate{2013/06/13} % \iffalse meta-comment %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% %% %% Copyright 1999--2013 Nigel Stanger and University of Otago %% %% You may use this package freely, and also distribute it %% provided that you don't change it, make any money off %% it or pretend that you wrote it. %% %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% %<driver> \documentclass[a4paper]{ltxdoc} \usepackage{doc} \DisableCrossrefs \CodelineNumbered \RecordChanges \usepackage{graphicx} % change hyperref options as required \usepackage[colorlinks,pdfpagemode=None,urlcolor=blue, linkcolor=red,pdfauthor={Nigel Stanger}]{hyperref} \title{The \textsf{ouexam} document class, v\ouexamversion} \author{Nigel Stanger\\nstanger@infoscience.otago.ac.nz} \date{\ouexamdate} \begin{document} \maketitle \DocInput{ouexam.dtx} \end{document} %</driver> % % \fi % %% \CheckSum{896} %% %% \CharacterTable %% {Upper-case \A\B\C\D\E\F\G\H\I\J\K\L\M\N\O\P\Q\R\S\T\U\V\W\X\Y\Z %% Lower-case \a\b\c\d\e\f\g\h\i\j\k\l\m\n\o\p\q\r\s\t\u\v\w\x\y\z %% Digits \0\1\2\3\4\5\6\7\8\9 %% Exclamation \! Double quote \" Hash (number) \# %% Dollar \$ Percent \% Ampersand \& %% Acute accent \' Left paren \( Right paren \) %% Asterisk \ Plus \+ Comma \, %% Minus \- Point \. Solidus \/ %% Colon \: Semicolon \; Less than \< %% Equals \= Greater than \> Question mark \? %% Commercial at \@ Left bracket \[ Backslash \\ %% Right bracket \] Circumflex \^ Underscore \_ %% Grave accent \` Left brace \{ Vertical bar \\| %% Right brace \} Tilde \~} %% % % \MakeShortVerb{\\|} % % \changes{1.0}{1999/04/15}{Initial version.} % % \begin{abstract} % This document class allows you to create Otago University final % examination papers using \LaTeXe\@. It implements the major formatting % requirements specified by the University, provides useful % macros to ease the process of building the title page and automatically % deals with fiddly little issues such as question and section numbering, % the number of pages in the paper, printing ``\textbf{TURN OVER}'' in the % bottom right corner of every page except the last one and ensuring that the % total number of marks for a question adds up to the expected number. % \end{abstract} % % % \tableofcontents % % % \section{Overview} % % \changes{2.1}{2004/04/10}{NJS Added requirements and installation instructions.} % \changes{2.3}{2012/09/04}{NJS Dropped vanilla \LaTeX\ for building docs.} % \subsection{Requirements and installation} % % The \textsf{ouexam} package requires the \textsf{verbatim}, % \textsf{fontenc}, \textsf{lmodern} and \textsf{textcomp} packages, all % of which should come standard with most \TeX\ installations. To build % the documentation and example files, you will need at least version 1.1 % of the \textsf{listings} package, and the \textsf{pdfjam} tool. % % Installation is relatively simple: % \begin{enumerate} % % \item Unpack the distribution archive and \texttt{cd} to the % distribution directory. % % \item \texttt{make}. % % \item \texttt{make install TEXMF\_INSTALL=/path/to/texmf}. % \texttt{/path/to/texmf} should be the root of your preferred % \texttt{texmf} tree (e.g., \texttt{/usr/share/texmf}). You may need % to do this as root depending on which \texttt{texmf} tree you are % installing into. You can also define \texttt{TEXMF\_INSTALL} as an % environment variable then simply type \texttt{make install}. % % \end{enumerate} % You can, of course, always install manually if you wish. % % % \changes{2.0}{2002/01/11}{NJS Added note on backwards compatibility.} % \subsection{Important note on backwards compatibility} % % Version 2.0 of \textsf{ouexam} introduced several major changes from earlier % versions. Consequently, documents created using an earlier version of % \textsf{ouexam} are \emph{incompatible} with version 2.0 or later. Versions 2.0 % and greater of \textsf{ouexam} will detect attempts to process a version 1.x % document and display an error message. % % The older version (1.2) is still available and can be downloaded from the same % location as the current version. To process older version documents, just copy % the version 1.2 \|ouexam.cls\| file into the same directory as the document file. % \TeX\ searches the current directory first, so this copy will take precedence % over any other installed version of \textsf{ouexam}. % % % \subsection{Class options} % % This document class is based on the \textsf{article} class and accepts % most of the options accepted by \textsf{article} (the \textsf{twoside}, % \textsf{10pt} and \textsf{11pt} options are ignored). The standard % options are \textsf{onecolumn}, \textsf{oneside}, \textsf{a4paper} and % \textsf{12pt}. In addition, \textsf{ouexam} accepts the following % ``native'' class options: % \begin{description} % \item[\textsf{draft}] \DescribeMacro{draft} This option has the usual % effects that you would expect for \textsf{draft} mode in the % \textsf{article} class, plus \fbox{DRAFT} is printed in % the footer of every page. Note that if you use the \textsf{graphicx} % package and turn on the \textsf{draft} option in \textsf{ouexam}, all % included graphics will be drawn in draft mode unless you specify the % \|draft=false\| option to \cs{includegraphics}. % % \item[\textsf{markingschedule}] \DescribeMacro{markingschedule} % Rather than write a separate marking schedule for an examination paper, you % can use the \textsf{marking} environment to embed marking information % within questions (see % \hyperref[Sec:Questions:Marking]{section~\ref{Sec:Questions:Marking}}). % By default this information is not printed (for obvious reasons!), but % when it comes time to print a marking schedule for the examination, % including the \textsf{markingschedule} class option in the % \cs{usepackage} command will cause \textsf{ouexam} to print the hidden % content. It will also print \fbox{MARKING SCHEDULE} in the header of % every page. % % \item[\textsf{times}] \DescribeMacro{times} % The Examinations Office prefers all examination papers to be set in 12pt % Palatino. However, if this is not available, they will accept 13pt Times % instead. The \textsf{times} class option enables the latter, but should % only be used if Palatino is not available. % \end{description} % % % \subsection{Required packages} % % This class requires the \textsf{verbatim} package in order to implement % the marking schedule functionality. It also expects modern \LaTeX\ font % handling capabilities, and therefore requires the \textsf{fontenc}, % \textsf{textcomp} and \textsf{lmodern} classes. % % % \subsection{Page margins} % % This class assumes A4 paper. You will probably get weird results if you try % to do anything different. The margins are set up as follows: top and bottom % 2cm (headers and footers are inside this margin), left and right 2.54cm (1in). % % % \changes{2.1.1}{2006/08/21}{NJS Updated \textsf{lastpage} style to display % ``\textbf{END}''.} % \subsection{Page styles} % % There are three page styles defined in this class: % \begin{description} % \item[\textsf{plain}] \DescribeMacro{plain} This is a slight modification % of the \textsf{plain} page style from \textsf{article}. It produces pages % that have the page number centered at the bottom, the paper number in the % top right corner and the text ``\textbf{TURN OVER}'' in the bottom right. % This is the default page style. % % \item[\textsf{lastpage}] \DescribeMacro{lastpage} This is similar to % \textsf{plain} but with ``\textbf{END}'' instead of ``\textbf{TURN % OVER}''. It is used for the last page of the examination. You % normally will not have to use this yourself---the class should take % care of it automatically. The class does however occasionally seem % to get confused, so there may times when you have to set the page % style of the last page manually. It will be fairly obvious when you % need to do this---the usual effect is that the last page % has ``\textbf{TURN OVER}'' printed on it when it should not. % % \item[\textsf{titlepage}] \DescribeMacro{titlepage} This is similar to % \textsf{plain} but without the page header, and is used for the title page % of the paper. As with \textsf{lastpage}, you will normally not use this % yourself, because the \cs{maketitlepage} macro handles this automatically % (see \hyperref[Sec:TitlePage]{section~\ref{Sec:TitlePage}}). % \end{description} % % % \section{Writing an examination paper} % % % \changes{2.0.2}{2002/05/11}{NJS Moved the title page subsection to the start % of this section.} % \subsection{The title page} % \label{Sec:TitlePage} % % This class defines a collection of macros that let you fill in the various % parts of the examination title page. This is analogous to the process you use % to generate the title of a document in, for example, the \textsf{article} class % (using \cs{title}, \cs{author}, \cs{date} and \cs{maketitle}). That is, you % issue the macros described below in the document preamble, then issue a % \cs{maketitlepage} in the document body. % % The \DescribeMacro{\examyear} \cs{examyear} macro lets you specify the year % in which the examination is being held, for example, \|\examyear{1999}\|. This % macro is mandatory. % % The \DescribeMacro{\department} \cs{department} macro lets you specify the % name of the department that produced the examination paper, for example, % \|\department{Information\| \|Science}\|. This macro is mandatory. % % The \DescribeMacro{\papernumber} \cs{papernumber} macro lets you specify % the paper number that the examination is for, for example, % \|\papernumber{COMP 101}\|. This macro is mandatory. Note that when you % specify semester information using \cs{semester}, this is appended to % the paper number in the page header. For example, if you specify % \|\papernumber{COMP 101}\| and \|\semester{2}\|, the header will contain % ``COMP 101 (Semester Two)'' rather than just ``COMP 101''. % % The \DescribeMacro{\papertitle} \cs{papertitle} macro lets you specify the % title of the paper, for example, % \|\papertitle{Systems Analysis and Design Methods}\|. This macro is mandatory. % % \changes{2.0.2}{2002/08/22}{NJS Added `FY' to the \cs{semester} macro % description.} % Some papers are offered in more than one semester. The % \DescribeMacro{\semester} \cs{semester} macro lets you specify which % semester the examination is for. The argument can be either ``1'' or ``2'' % for semesters 1 and 2, ``SS'' for summer school, ``FY'' for a full-year % paper or ``SP'' for a special examination, for example, \|\semester{2}\|. % Invalid values for the argument cause a warning to be raised. This macro is % optional---if you omit it, no semester information is generated. % % The \DescribeMacro{\timeallowed} \cs{timeallowed} macro lets you specify the % length of the examination in hours, for example, \|\timeallowed{2}\|. This % macro is optional---if you omit it, it defaults to three (3) hours. % % \changes{2.1}{2004/04/05}{NJS Modified \cs{allowcalculators} to conform % to the new University calculator regulations.} % \changes{2.1}{2004/04/05}{NJS Added \cs{permitcalculators} as a synonym % for \cs{allowcalculators}.} % If calculators are permitted in the examination, use the % \DescribeMacro{\allowcalculators} % \DescribeMacro{\permitcalculators} % \cs{allowcalculators} macro (\cs{permitcalculators} will also work). % This macro is optional---if you omit it, a sentence is inserted saying % that calculators are \emph{not} permitted. \cs{allowcalculators} has a % single optional argument that specifies the kind of calculators % permitted. If omitted, it defaults to ``any'' (see below). Otherwise, % this argument must be one of the following values: % \begin{description} % % \item[\textsf{none}] A sentence is inserted noting that no calculators are % permitted. % % \item[\textsf{approved}] A sentence is inserted noting that only approved % calculators are permitted. % % \item[\textsf{any}] (default) A sentence is inserted noting that any % calculator that does not have a communication capability is % permitted. % % \end{description} % % The \DescribeMacro{\instructions} \cs{instructions} macro lets you specify % instructions on how candidates should complete the % examination, for example, \|\instructions{Answer ALL\| \|questions.}\|. This macro % is optional. Note that the content of the instructions can be just about % anything. It is up to you to control formatting, such as how you want lines % broken, etc. This also true of the \cs{material}, \cs{copiesof} and % \cs{otherinstructions} macros described below. % % The \DescribeMacro{\material} \cs{material} macro lets you specify any % additional material that candidates are provided in addition to the % examination paper itself, for example, \|\material{SQL schema definition}\|. % This macro is optional. % % The \DescribeMacro{\copiesof} \cs{copiesof} macro lets you specify any % material that candidates are allowed to bring into the examination, for % example, \|\copiesof{McFadden &\| \|Hoffer, 5th edition}\|. This macro is % optional. % % The \DescribeMacro{\otherinstructions} \cs{otherinstructions} macro lets you % specify any other instructions not covered by any of the above, for example, % \|\otherinstructions{No smoking\| \|allowed.}\|. This macro is optional. % % Once you have specified the content of the title page using the above % macros, you simply issue a \DescribeMacro{\maketitlepage} \cs{maketitlepage} % to generate the front page of the examination (cf. \cs{maketitle}). The title % page generated meets University formatting requirements. % Note that the number of pages in the examination paper is generated % and inserted into the title page automatically---you do not need to specify % it manually. % % % \subsection{Sections} % \label{Sec:Sections} % % Examination papers may optionally have multiple sections, ``numbered'' A, B, \ldots. Rather than redefine the existing section macros, this class defines an \DescribeEnv{examsection} \textsf{examsection} environment that generates a new section. The \textsf{examsection} environment has three mandatory arguments: % \begin{enumerate} % \item The expected number of marks for the question. If you leave this % argument empty it defaults to zero. \textsf{ouexam} keeps a running total % of the \emph{actual} number of marks encountered within a section, which is % compared with the value of this argument when the environment is closed. % An error is raised if the values are not equal. The running total is also % typeset right-justified at the end of the section in the form % ``\textbf{[SECTION A TOTAL 20 MARKS]}'' as an additional verification. % % \item Instructions for answering questions in this section (such as % ``Answer any TWO questions.''). If you leave this argument empty it % defaults to ``ANSWER ALL QUESTIONS.''. % % \item A description of the topic of the section. This can be left % empty. % \end{enumerate} % Every section begins on a new page, and the section title is formatted as % ``\textbf{\underline{Section A}}'', in \cs{large} size. For example: \\ % % \noindent\|\begin{examsection}{25}{}{These questions are remarkably boring.}\| \\ % \hspace{1cm}\vdots \\ % \|\end{examsection}\| \\ % % \noindent will produce the following (assuming that the actual number of % marks in the question is correct): % % \begin{center} % \fbox{% % \begin{minipage}{0.9\columnwidth} % {\large\noindent\textbf{\underline{Section~A}}} \\[0.5\baselineskip] % ANSWER ALL QUESTIONS. \\ % These questions are remarkably boring. \\ % \hspace{2cm}\vdots \\ % \mbox{}\hfill\textbf{[SECTION A TOTAL 25 MARKS]} % \end{minipage} % } % \end{center} % % % \subsection{Questions} % \label{Sec:Questions} % % \textsf{ouexam} provides three environments for building examination % questions: \DescribeEnv{question} \textsf{question}, % \DescribeEnv{subquestion} \textsf{subquestion} and \textsf{subsubquestion}, % which correspond to top-level questions, parts of questions, and % \DescribeEnv{subsubquestion} sub-parts of questions respectively. They % produce questions that are numbered according to University examination % formatting requirements. Note that questions are normally numbered % sequentially throughout the entire paper regardless of any section % boundaries. % % All three environments have a single mandatory argument which is the % expected number of marks for the question, part or sub-part. If this % argument is left empty it will default to zero. This argument works in much % the same way as the first argument to the \textsf{examsection} environment % (see \hyperref[Sec:Sections]{section~\ref{Sec:Sections}}): \textsf{ouexam} % keeps a running total of the number of marks encountered within each % question part and sub-part, and compares this total with the expected value % when the environment closes. Where appropriate, the running total is % typeset right-justified in the form ``(5 marks)''. For example: % \begin{verbatim} % \begin{question}{5} % \begin{subquestion}{2} % Why is the sky blue? % \end{subquestion} % \begin{subquestion}{3} % Explain in detail how the sky can be made pink. % \end{subquestion} % \end{question} % \begin{question}{1} % Define the term ``floccinaucinihilipilification''. % \end{question} % \end{verbatim} % % \noindent will produce the following: % \begin{center} % \fbox{% % \begin{minipage}{0.9\columnwidth} % \begin{enumerate} % \item % \begin{enumerate} % \item Why is the sky blue? \hfill (2 marks) % % \item Explain in detail how the sky can be made pink. \hfill (3 % marks) % \end{enumerate} % % \item Define the term ``floccinaucinihilipilification''. \hfill (1 % mark) % \end{enumerate} % \end{minipage} % } % \end{center} % % The \textsf{question} environments handle formatting of marks totals % automatically, including considerations such as: % \begin{itemize} % \item whether or not to print the number of marks for a question at all % (they should not be printed if the question has sub-parts); % % \item whether to print ``mark'' or ``marks'' (e.g., question 2 above); and % % \item where to position the number of marks relative to the question, % depending on how full the last line of the question is. % \end{itemize} % % \changes{2.3.1}{2013/05/31}{NJS Enforced proper question environment nesting.} % The various \textsf{question} environments must be correctly nested in order for the features described above to work correctly, i.e., \textsf{question} at the top level, followed by \textsf{subquestion} and \textsf{subsubquestion}. Any variation to this nesting order will cause an error. % % % \changes{2.2}{2010/04/26}{NJS Added time allocation macros.} % \subsection{Time allocations} % \label{Sec-Questions-Time} % % If you wish to suggest a time allocation for completing a question, use the \DescribeMacro{\timeallocation} \cs{timeallocation} macro. The only argument is the number of minutes to be allocated for this question. The macro will output the time allocation in the form: \textbf{[The suggested time allocation for this question is \emph{t} minutes.]} % % If you wish to allow a certain amount of time for candidates to read some text, use the \DescribeMacro{\readingtime} \cs{readingtime} macro. The only argument is the number of minutes to be allocated for reading time. The macro will output the reading time in the form: \textbf{[Allow \emph{t} minutes reading time.]} % % Both macros assume the base unit to be minutes; if you wish to change this you will need to redefine the macros. Otherwise, the argument can be almost anything: text, numbers, etc. % % \textsf{ouexam} currently does not check that time allocations sum correctly. This may be implemented in a future version. % % % \subsection{Marking schedule information} % \label{Sec:Questions:Marking} % % Rather than writing a separate marking schedule for an examination paper, % you can use the \DescribeEnv{marking} \textsf{marking} environment to embed % marking information within questions. By default this information is not % printed (for obvious reasons!), but when it comes time to print a marking % schedule for the examination, use the \textsf{markingschedule} class option % to print the hidden content. The marking information will be set in % \emph{italics} to differentiate it from the main body of the examination, % and \fbox{MARKING SCHEDULE} will also be printed in the header of every % page. % % \changes{2.1}{2004/04/10}{NJS Added driver file tip.} % A useful tip for writing examination papers is to create two separate % ``driver'' documents that are set up the document class for the % ``plain'' and ``marking schedule'' versions of the examination, % respectively. Each should then include a separate document that contains % the actual content of the examination paper, for example: % % \begin{verbatim} % \documentclass[markingschedule]{ouexam} % % \input{exampaper} % contains the actual examination content % \end{verbatim} % % Doing this enables you to build the ``plain'' and ``marking schedule'' % versions independently, without overwriting each other or having to % change the original source file. Simply run \LaTeX\ on the appropriate % driver file to produce the version that you want. See the example files % that came with the \textsf{ouexam} distribution for an example of this % approach in action. % % % \subsection{Miscellaneous} % % Most examinationss are marked out of 100, and \textsf{ouexam} defaults to % this. However, if you want an examination that is marked out of some other % number, for example, 90 marks, you can specify this using the % \DescribeMacro{\examoutof} \cs{examoutof} macro. Thus, \|\examoutof{90}\| % will set the expected number of marks for the examination to 90. This value % is used by \textsf{ouexam} to verify that the marks for all the questions % add up to the expected number of marks for the examination. % % % \section{Errors and warnings} % % In this section are described the error messages and warning produced by % \textsf{ouexam}, and the reasons why they occur. % % % \subsection{Error messages} % % \noindent\texttt{INCOMPATIBLE (\ldots): this document was written for an % earlier} \\ % \texttt{version of ouexam \ldots} \\ % You are trying to use v2.0 or later of \textsf{ouexam} with a document that was written for \textsf{ouexam} v1.2 or earlier. Versions 2.0 and later of \textsf{ouexam} are \emph{fundamentally incompatible} with earlier versions. The only solutions here are either to revert to \textsf{ouexam} v1.2 or earlier (you can put a copy of the class file in the same directory as the document), or rewrite the document to conform to the current version of \textsf{ouexam}. You should be able to download a copy of v1.2 from the same location that you found the current version. \\ % % \noindent\texttt{no \cs{examyear} was specified} \\ % \noindent\texttt{no \cs{department} was specified} \\ % \noindent\texttt{no \cs{papernumber} was specified} \\ % \noindent\texttt{no \cs{papertitle} was specified} \\ % You have not specified one of these macros. All four of these macros are mandatory and must be included in the preamble of any \textsf{ouexam} document. \\ % % % \noindent\texttt{question environment nested inside another question, subquestion or \\ % \hspace{1em}subsubquestion environment} \\ % \noindent\texttt{subquestion environment either outside a question environment, or \\ % \hspace{1em}nested inside another subquestion or subsubquestion environment} \\ % \noindent\texttt{subsubquestion environment either outside a subquestion environment, \\ % \hspace{1em}or nested inside another subsubquestion environment} \\ % These errors all mean that you have incorrectly nested your \textsf{question} environments in some way. \textsf{question}s must always appear at the outermost level, \textsf{subquestion}s must always appear inside \textsf{question}s, and \textsf{subsubquestion}s must always appear inside \textsf{subquestion}s. Any other kind of nesting is an error. \\ % % % \subsection{Warnings} % % \noindent\texttt{actual number of marks for exam (\ldots) does not match expected \\ % \hspace{1em}number of marks (\ldots)} \\ % The actual number of marks supplied for the whole examination (calculated by summing the marks for all the questions) is not the same as the expected total for the examination. Adjust the marks for the questions until the totals match. \\ % % \noindent\texttt{actual mark (\ldots) for section \ldots\ does not match expected \\ % \hspace{1em}mark (\ldots)} \\ % The actual number of marks supplied for a particular section (calculated by summing the marks for all the questions in that section) is not the same as the expected total for that section. Adjust the marks for the questions in the section concerned until the totals match. The section number is given in the warning. \\ % % \noindent\texttt{actual mark (\ldots) for question \ldots\ does not match expected \\ % \hspace{1em}mark (\ldots)} \\ % The actual number of marks supplied for a particular question (calculated by summing the marks of all its sub-parts) is not the same as the expected total for that question. Adjust the marks for the sub-parts of the question until the totals match. The question number is given in the warning. \\ % % \changes{2.0.2}{2002/08/22}{NJS Added `FY' to \cs{semester} error description.} % \noindent\texttt{invalid value `\ldots' for \cs{semester}; valid values are `1', `2', \\ % \hspace{1em}`SS', `FY' and `SP'. No semester information will be printed} \\ % You have provided an invalid value as the argument to the \cs{semester} macro. The allowed values are ``1'' (semester one), ``2'' (semester two), ``SS'' (summer school), ``FY'' (full-year) and ``SP'' (special examination). Any other values will be ignored. Note that these are case-sensitive, for example, `sp' is invalid. \\ % % \changes{2.1}{2004/04/05}{NJS Added \cs{allowcalculators} error description.} % \noindent\texttt{invalid value `\ldots' for \cs{allowcalculators}; valid values are \\ % \hspace{1em}`none', `any' and `approved'. No calculators will be permitted} \\ % You have provided an invalid value as the argument to the \cs{allowcalculators} macro. The allowed values are ``none'' (no calculators permitted), ``any'' (any calculator permitted) and ``approved'' (only approved calculators permitted). Any other values will be ignored, and \textsf{ouexam} will assume that no calculators are permitted. Note that these are case-sensitive, for example, `NONE' is invalid. \\ % % % \section{Example} % % \changes{2.1}{2004/04/10}{NJS Updated example to conform to 2.1 changes.} % \changes{2.0}{2002/01/25}{NJS Updated example to conform to 2.0 changes.} % \changes{1.1}{1999/04/20}{NJS Updated example to conform to 1.1 changes.} % The code given below is for the first three pages of the INFO~321 2001 % final examination. The output produced by this source is shown in % \hyperref[Fig.Example]{figure~\ref{Fig.Example} on % page~\pageref{Fig.Example}}. Particular points to note in this example % are: % \begin{itemize} % % \item The number of hours is not specified, so it defaults to three. % % \item This examination allows only approved calculators. % % \item The \textsf{listings} package (v1.1 or later) is used to % typeset code listings. % % \item The argument of the \cs{instructions} macro can be just about % anything, as can the arguments of the \cs{material}, \cs{copiesof} and % \cs{otherinstructions} macros. In this example, the \cs{instructions} % macro has been omitted, causing \textsf{ouexam} to output the default % instructions. % % \item Marking schedule information has been embedded within the % questions. \hyperref[Fig.Example2]{Figure~\ref{Fig.Example2} on % page~\pageref{Fig.Example2}} shows the output produced by this source % when the \textsf{markingschedule} class option is used. % % \end{itemize} % The full source for this example may be found in the % \texttt{example.tex} files that came with this distribution. % % \small % \begin{verbatim} % \documentclass{ouexam} % % \usepackage{graphicx} % \usepackage{listings} % % \examyear{2001} % \department{Information Science} % \papernumber{INFO 321} % \papertitle{Database Systems} % \allowcalculators[approved] % % \begin{document} % % \lstloadlanguages{ODL,OQL,[Oracle8]SQL,[Oracle8]PLSQL} % \lstset{basicstyle=\sffamily, showstringspaces=false, tabsize=2, % xleftmargin=1cm, belowskip=0pt, commentstyle=\itshape, % numbers=left, numberstyle=\scriptsize, numbersep=5pt, % columns=fullflexible,} % % \maketitlepage % % \begin{examsection}{25}{}{Questions in this section (total 25 marks) % relate to physical database design and tuning.} % % % % \begin{question}{10} % % A major U.S. mail order firm is having performance problems with queries % on the \textsf{Customer} table in its Oracle8 order-processing database. % Consider the following information about the \textsf{Customer} table, % then answer the questions below: \\ % % \textsf{\textbf{Customer}(\underline{customer\_no}, name, address, city, % state, country, phone, email, status)} % % \begin{itemize} % \item \textsf{customer\_no} is the primary key and because this is an % Oracle8 database, a unique B-tree index has been automatically created % on this column. % % \item \textsf{status} is a single character column that holds one of % the values `I' (inactive), `B' (bronze), `S' (silver) or `G' (gold). % The value of this column is determined by a customer's order history. % % \item About 10\% of customers are inactive, about 50\% have bronze % status, about 35\% have silver status and about 5\% have gold status. % % \item The average row size is about 200 bytes and there are % approximately 5,000,000 rows. % % \item The \textsf{Customer} table is heavily queried (averaging thirty % new queries per minute) by four different groups of employees. Group~1 % queries only inactive customers, group~2 queries only bronze status % customers and groups~3 and~4 both query only silver and gold status % customers. Members of any group may query the table at any time and % queries often overlap. % % \item UPDATE, INSERT and DELETE operations on the table are uncommon. % % \item Very few queries are for individual customers; rather, most % queries are based on various combinations of values from % \textsf{status}, \textsf{city}, \textsf{state} and \textsf{country}. % \end{itemize}\medskip % % \begin{subquestion}{3} % Describe and justify an \emph{index-based} physical tuning solution % that will improve the performance of queries on the \textsf{Customer} % table. (No code is required.) % \begin{marking} % Since queries tend to be on combinations of several low-cardinality % columns [1] and the table is very large [1], the most effective % solution would be to place bitmap indexes on status, city, state and % country [1]. % % B-tree indexes on any of the individual columns would not be ideal % because of their low cardinality. A possible alternative is a % composite B-tree index on all four columns [1 for suggesting this if % appropriate], but this is not ideal and would be much larger than % the four bitmap indexes [1 for this explanation if appropriate]. % Hashing is definitely \emph{not} an option, because very few queries % are exact match. % \end{marking} % \end{subquestion} % % \begin{subquestion}{7} % \begin{subsubquestion}{5}\label{nonindex} % Assuming that there are two identical fast disks available, describe % and justify two alternative \emph{non index-based} physical tuning % solutions that will improve the performance of queries on the % \textsf{Customer} table. For each solution include details of how % the two disks will be used. (No code is required.) % \begin{marking} % Alternative 1: partition the table across the two disks based on % status (four partitions, one for each status value) [1]. This will % allow simultaneous parallel access to different parts of the table % by the different groups [1]. Since two groups both access silver % and gold customers, it makes sense to put the silver customers on % disk one and the gold customers on disk two. Bronze customers % account for 50\% of the table, so it makes sense to also put them % on disk two (as the gold customers are the smallest group), and % put the inactive customers on disk one with the silver customers. % [1 for any sensible partitioning scheme] % % Alternative 2: replicate the table across both disks (one replica % on each disk) [1]. This will also provide simultaneous parallel % access to the data (hard to tell whether this would be better or % worse than the partitioning above) [1]. % % Note that any solution must be capable of being implemented in % Oracle8. % \end{marking} % \end{subsubquestion} % % \begin{subsubquestion}{2} % Briefly discuss the relative advantages and/or disadvantages of the % two alternative solutions you described in part~(\ref{nonindex}) % above. % \begin{marking} % The partitioning solution given above doesn't really help with the % bronze customer rows---we would need random horizontal % partitioning to really help with this, which Oracle8 doesn't % support. More generally, Oracle8 does not migrate rows into the % correct partition if a customer's status changes, so the data will % need to be reloaded occasionally [1]. % % For replication, we have the usual problem of keeping the replicas % synchronised if they are both read/write. This is somewhat % mitigated by the fact that there are only two copies [1]. % \end{marking} % \end{subsubquestion} % \end{subquestion} % \end{question} % % \newpage % \begin{question}{10} % There are two generic parameters that affect the structure of a B-tree % index. % % \begin{subquestion}{6} % Identify these two parameters and describe what each parameter % specifies. Use diagrams to illustrate your answer. % \begin{marking} % \emph{Node size} [1] specifies the maximum number of pointers in % each index node [1]. Suitable diagram [1]. % % \emph{Percentage fill} [1] specifies the number of pointers in each % index node that are allocated when a node is created [1]. In effect, % it determines the amount of free space left in each node. Suitable % diagram [1]. % \end{marking} % \end{subquestion} % % \begin{subquestion}{4} % For each of the two parameters, discuss the practical effect(s) of: % \begin{subsubquestion}{2} % high values of the parameter; and % \begin{marking} % Large node size reduces the height of the B-tree, resulting in % potentially fewer I/Os to access a leaf node [$\frac{1}{2}$]. It % can however increase the number of rows locked in a single leaf % node [$\frac{1}{2}$]. % % High percent fill means that most pointers in a node will be % allocated when the node is created [$\frac{1}{2}$]. This leaves % little free space in the B-tree for new key values or key values % that change [$\frac{1}{2}$]. % \end{marking} % \end{subsubquestion} % \begin{subsubquestion}{2} % low values of the parameter. % \begin{marking} % Small node size increases the height of the B-tree, resulting in % potentially more I/Os to access a leaf node [$\frac{1}{2}$]. The % number of rows locked in a single leaf node will be reduced % [$\frac{1}{2}$]. % % Low percent fill means that most pointers in a node will not be % allocated when the node is created [$\frac{1}{2}$]. This leaves % little free space in the B-tree for new key values or key values % that change [$\frac{1}{2}$]. % \end{marking} % \end{subsubquestion} % \end{subquestion} % \end{question} % % % % \begin{question}{5}\label{clusterq} % % \begin{subquestion}{2} % Define what is meant by the term ``clustering'' in the context of % physical database design and tuning, and explain why it is beneficial. % % \begin{marking} % Clustering is the grouping of related rows physically near to each % other on disk [1] so that they may be retrieved in as few I/Os as % possible (ideally one)---beneficial because of the slowness of disk % I/O compared to memory I/O [1]. % \end{marking} % \end{subquestion} % % \begin{subquestion}{3}\label{clusterqcode} % Consider the following Oracle8 SQL table definition: % % \begin{lstlisting}[language={[Oracle8]SQL}]{} % create table enrolment % ( student_id char(7), % paper_code char(7), % enrol_year number(4), % % primary key (student_id, paper_code, enrol_year) % ); % \end{lstlisting} % % The \textsf{Enrolment} table is often queried for all rows relating to % a particular student in a particular year (for example, student % 1234567 in 1998, student 9876543 in 2000, etc.). Write appropriate % Oracle8 SQL code that will cluster the rows of the table in such a way % as to support these types of query (you do not need to calculate % cluster sizes). Relevant Oracle8 syntax diagrams are given on the next % page. % \begin{marking} % \begin{lstlisting}[language={[Oracle8]SQL}]{} % create cluster enrolment_cluster (student_id char(7), year number(4)); % % create index enrol_cluster_index on cluster enrolment_cluster; % % create table enrolment % ( student_id char(7), % paper_code char(7), % enrol_year number(4), % % primary key (student_id, paper_code, enrol_year) % ) cluster enrolment_cluster(student_id, enrol_year); % \end{lstlisting} % \end{marking} % [1] each for correct cluster, cluster index and table clustering. % \end{subquestion} % % \end{question} % % \end{examsection} % . % . % . % \end{document} % \end{verbatim} % \normalsize % % \begin{figure} % \fbox{\includegraphics[scale=0.28]{eg1-1}} % \hfill % \fbox{\includegraphics[scale=0.28]{eg1-2}} % % \begin{center} % \fbox{\includegraphics[scale=0.28]{eg1-3}} % \end{center} % % \caption{Output produced by \textsf{ouexam}.} % \label{Fig.Example} % \end{figure} % % \begin{figure} % \fbox{\includegraphics[scale=0.28]{eg2-2}} % \hfill % \fbox{\includegraphics[scale=0.28]{eg2-3}} % % \bigskip % % \fbox{\includegraphics[scale=0.28]{eg2-4}} % \hfill % \fbox{\includegraphics[scale=0.28]{eg2-5}} % % \caption{Output produced by \textsf{ouexam} when % \textsf{markingschedule} is used.} % \label{Fig.Example2} % \end{figure} % % % \StopEventually{} % % \changes{2.0}{2002/01/10}{NJS Revamped and tided up documentation and code.} % % \section{The code} % % \subsection{Preamble} % % \begin{macrocode} \NeedsTeXFormat{LaTeX2e}[1998/06/01] \ProvidesClass{ouexam}[\ouexamshortdate\space % v\ouexamversion\space Otago % University examination paper] % \end{macrocode} % % \begin{macro}{\@unsupported} % \changes{2.0}{2002/01/10}{NJS New \cs{@unsupported} macro.} % This macro is used to handle documents written for older versions of % \textsf{ouexam}. All obsolete macros map to this macro, which prints out an % appropriate error message. It needs to be defined early because it is used % to trap use of the obsolete \textsf{multichoice} class option. % % \begin{macrocode} \def\@unsupported#1#2{\ClassError{ouexam}{% ^^JINCOMPATIBLE (#1)^^J% This document was written for an earlier version of ouexam and is^^J% not compatible with ouexam v\ouexamversion. Please use ouexam v#2 or earlier^^J% to process this document. Version #2 can be downloaded from the^^J% same location as version\space\ouexamversion}} % \end{macrocode} % \end{macro} % % \begin{macro}{multichoice} % \changes{1.1}{1999/04/20}{NJS New \textsf{multichoice} class option.} % \changes{2.0}{2000/09/04}{NJS Removed support for \textsf{multichoice} option.} % The \textsf{multichoice} class option is no longer supported, but rather than % just breaking older documents, we can at least try to present a reasonable % error message. % % \begin{macrocode} \DeclareOption{multichoice}{% \OptionNotUsed\@unsupported{multichoice class option}{1.2}% } % \end{macrocode} % \end{macro} % % \begin{macro}{twoside} % \changes{1.1}{1999/04/20}{NJS Turned off \textsf{twoside} option.} % This document class is based on the \textsf{article} class and accepts any of % the options accepted by \textsf{article}. The \textsf{twoside} option does % not really make sense, however---all Otago examination papers are printed % single-sided anyway, and the format is such that two-sided printing would % look no different. The \textsf{twoside} option is therefore not used in this % class. % % \begin{macrocode} \DeclareOption{twoside}{\OptionNotUsed} % \end{macrocode} % \end{macro} % % \begin{macro}{10pt} % \begin{macro}{11pt} % \changes{2.3}{1999/04/20}{NJS Turned off \textsf{10pt} and \textsf{11pt} % options.} % The Examinations Office specifies a minimum of 12pt for the font size, so the % \textsf{10pt} and \textsf{11pt} options are not used in this class. % % \begin{macrocode} \DeclareOption{10pt}{\OptionNotUsed} \DeclareOption{11pt}{\OptionNotUsed} % \end{macrocode} % \end{macro} % \end{macro} % % \begin{macro}{draft} % \changes{2.0}{2001/05/15}{NJS New \textsf{draft} class option.} % \begin{macro}{\if@draft} % \changes{2.0}{2001/05/15}{NJS New switch for \textsf{draft}.} % The \textsf{draft} option determines whether the examination is to be printed % in draft mode (default off). The \cs{if@draft} switch is used to determine % whether this is in effect (default false). Draft mode has the usual effects % that you would expect for draft mode in the \textsf{article} class, plus % \fbox{DRAFT} is printed in the footer of every page. % % \begin{macrocode} \newif\if@draft \@draftfalse \DeclareOption{draft}{\@drafttrue\PassOptionsToClass{\CurrentOption}{article}} % \end{macrocode} % % Note that if you use the \textsf{graphicx} package and turn on the % \textsf{draft} option in \textsf{ouexam}, all included graphics will % be drawn in draft mode unless you specify the \|draft=false\| option to the % \cs{includegraphics} macro. The amount of effort required to fix what is % a relatively small issue isn't really worth it, so this will not change. % \end{macro} % \end{macro} % % \begin{macro}{markingschedule} % \changes{2.0}{2000/09/05}{NJS New \textsf{markingschedule} class option.} % \begin{macro}{\if@markingschedule} % \changes{2.0}{2000/09/05}{NJS New switch for \textsf{markingschedule}.} % The \textsf{markingschedule} option determines whether marking schedule % information is printed in addition to the questions (default off). The % \cs{if@markingschedule} switch is used to determine whether this is in effect % (default false). % % \begin{macrocode} \newif\if@markingschedule \@markingschedulefalse \DeclareOption{markingschedule}{\@markingscheduletrue} % \end{macrocode} % \end{macro} % \end{macro} % % \begin{macro}{times} % \changes{2.3}{2012/09/03}{NJS New \textsf{times} class option.} % \begin{macro}{\if@times} % \changes{2.3}{2012/09/03}{NJS New switch for \textsf{times}.} % The \textsf{times} option formats the exam paper using 13pt Times. If this % option is not specified, then the exam paper is formatted using 12pt Palatino. % % \begin{macrocode} \newif\if@times \@timesfalse \DeclareOption{times}{\@timestrue} % \end{macrocode} % \end{macro} % \end{macro} % % All other options are passed directly to the \textsf{article} class: % % \begin{macrocode} \DeclareOption{\PassOptionsToClass{\CurrentOption}{article}} \ProcessOptions % \end{macrocode} % % \changes{1.1}{1999/04/20}{NJS Changed default point size to 12pt.} % The defaults for this class are \textsf{onecolumn}, \textsf{oneside}, % \textsf{a4paper} and \textsf{12pt}. The font size is the only one that might % need to change. % % \begin{macrocode} \LoadClass[onecolumn,oneside,a4paper,12pt]{article} % \end{macrocode} % % % \subsection{Required packages} % % \changes{2.0}{2001/05/08}{NJS Removed requirement for \textsf{calc} and % \textsf{ifthen} packages.} % The \textsf{verbatim} package, is used to implement the \textsf{marking} % environment: % % \begin{macrocode} \RequirePackage{verbatim} % \end{macrocode} % % \changes{2.3}{2012/09/04}{NJS Added requirement for \textsf{fontenc}, % \textsf{textcomp} and \textsf{lmodern} packages.} % % Running under vanilla OT1 encoding often causes all sorts of font substitution % warnings, so the \textsf{fontenc} (with \textsf{T1} option), \textsf{textcomp} % and \textsf{lmodern} packages are used to bring \textsf{ouexam}'s font % handling up to more modern standards. % % \begin{macrocode} \RequirePackage[T1]{fontenc} \RequirePackage{textcomp} \RequirePackage{lmodern} % \end{macrocode} % % % \changes{2.3}{2012/09/04}{NJS Added font setup for \textsf{times} class % option.} % \subsection{Font setup} % % The Examinations Office allows for two standard font styles: 12pt Palatino % (preferred) or 13pt Times New Roman. The former is the default; the latter is % only used if the \textsf{times} class option is used, and we need to % explicitly define all the standard sizes. The following is extrapolated from % the definitions in the \textsf{article} class and seems about right: % % \begin{macrocode} \if@times% \usepackage{mathptmx}% \renewcommand{\normalsize}{\fontsize{13}{15.7}\selectfont}% \renewcommand{\tiny}{\fontsize{\@viipt}{\@viiipt}\selectfont}% \renewcommand{\scriptsize}{\fontsize{\@ixpt}{11}\selectfont}% \renewcommand{\footnotesize}{\fontsize{\@xpt}{\@xiipt}\selectfont}% \renewcommand{\small}{\fontsize{\@xipt}{13.6}\selectfont}% \renewcommand{\large}{\fontsize{\@xivpt}{18}\selectfont}% \renewcommand{\Large}{\fontsize{\@xviipt}{22}\selectfont}% \renewcommand{\LARGE}{\fontsize{\@xxpt}{25}\selectfont}% \renewcommand{\huge}{\fontsize{\@xxvpt}{30}\selectfont}% \let\Huge=\huge% \else% \usepackage{mathpazo}% \fi% % \end{macrocode} % % % \subsection{Page setup} % % \begin{macro}{\oddsidemargin} % \begin{macro}{\topmargin} % \begin{macro}{\textwidth} % \begin{macro}{\textheight} % A4 paper is assumed (since this class is intended only for Otago examination % papers, this seems a reasonable assumption). The margins are top and bottom % 2cm, left and right 2.54cm (1in). % % \begin{macrocode} \setlength{\oddsidemargin}{0cm} \setlength{\topmargin}{-0.54cm} \setlength{\textwidth}{15.92cm} \setlength{\textheight}{25.7cm} \advance\textheight by-\headheight \advance\textheight by-\headsep \advance\textheight by-\footskip % \end{macrocode} % \end{macro} % \end{macro} % \end{macro} % \end{macro} % % % \subsection{Page styles}\label{pagestyles} % % \changes{2.0}{2001/05/17}{NJS Added support to page styles for \textsf{draft} and \textsf{markingschedule} class options.} % \changes{2.0}{2002/01/15}{NJS Fixed headers so that the page number is always centered.} % \changes{2.1.1}{2006/08/21}{NJS Updated \textsf{lastpage} style to display ``\textbf{END}''.} % \changes{2.3}{2012/09/03}{NJS Rearranged headers and footers to conform to the latest formatting requirements. Swapped \fbox{MARKING SCHEDULE} and \fbox{DRAFT} between header and footer so that the page number doesn't get overprinted.} % \changes{2.4}{2013/06/06}{NJS Replaced the \texttt{lastpage} page style with direct manipulation of the footer contents, as \cs{pagestyle} and \cs{thispagestyle} don't get applied to automatically generated float pages on the last page of the document.} % %\begin{macro}{\if@lastpagetext} %\begin{macro}{\if@lastpagefloat} % These two switches are used in some slightly convoluted logic to ensure that ``\textbf{END}'' is printed on the last page of the document (see below for details). % \begin{macrocode} \newif\if@lastpagetext \@lastpagetextfalse \newif\if@lastpagefloat \@lastpagefloatfalse % \end{macrocode} % \end{macro} % \end{macro} % % \begin{macro}{\ps@plain} % The \textsf{plain} page style is based on that in the \textsf{article} class, and produces pages with the page number centered at the bottom, the paper number in the top right corner and the text ``\textbf{TURN OVER}'' in the bottom right. If semester information is specified, this is included in parentheses after the paper code, for example, ``COMP 101 (Semester Two)''. \fbox{MARKING SCHEDULE} is printed at top left if the \textsf{markingschedule} class option is used. % \begin{macrocode} \def\ps@plain{% \def\@oddhead{\@markingschedule\hfill\@pnumber% \ifx\@semester\@empty\else\ (\@semester)\fi}% \let\@evenhead\@oddhead% \def\@oddfoot{% % \end{macrocode} % % Note that we have to get a bit clever to ensure that the page number is properly centered in the page footer. The naive solution is just to use \cs{hfill}s, but this will cause the page number to shift around depending on the length of \cs{@pnumber}. The \textsf{fancyhdr} package will do this for us, but it seems a bit of overkill to require this package merely so that we can ensure that the page number remains centered. Instead, we've borrowed the relevant technique from that class, which draws the left, centre and right bits of the footer as three overlapping boxes. You just have to be careful that the value of \cs{@pnumber} isn't so long that it overwrites the page number. % \begin{macrocode} \rlap{\parbox[b]{\columnwidth}{\raggedright\@draft\strut}}\hfill% \parbox[b]{\columnwidth}{\centering\textrm{\thepage}\strut}\hfill% \llap{\parbox[b]{\columnwidth}{\raggedleft\textbf{% % \end{macrocode} % % An additional complication is that only the very last page of the document should have ``\textbf{END}'' displayed in the footer instead of ``\textbf{TURN OVER}''. This turns out to be a lot harder than you might think! Simply applying a page style \cs{AtEndDocument} works fine, \emph{except} when you have unprocessed floats at the end of the document. These are placed on an automatically-generated float page at the end, which \cs{pagestyle} and \cs{thispagestyle} don't affect. That is, a \textsf{lastpage} page style simply won't work in this situation. % % The solution is therefore to not use a page style, but rather check in the \textsf{plain} page style whether or not this is the last page. Again, this isn't as simple as it sounds! The logic goes like this: % % \begin{itemize} % % \item If there are no unprocessed floats, meaning that the last page of the document will be a text page (indicated by \cs{@lastpagetexttrue}), then output ``\textbf{END}''. % % \item If there are unprocessed floats, meaning that the last page of the document will be a float page (indicated by \cs{@lastpagefloattrue}), then: % % \begin{itemize} % % \item If the float page is currently being generated (indicated by \cs{@fcolmadetrue} from \texttt{latex.ltx}), output ``\textbf{END}''. % % \item Otherwise, output ``\textbf{TURN OVER}''. % % \end{itemize} % % \item Otherwise, output ``\textbf{TURN OVER}''. % % \end{itemize} % % Note that this currently fails if more than one float page needs to be generated at the end of the document. \emph{All} such pages will have ``\textbf{END}'' printed at the bottom. % \begin{macrocode} \if@lastpagetext END% \else\if@lastpagefloat\if@fcolmade END% \else TURN OVER\fi% \else TURN OVER\fi% \fi}}}% }% \let\@evenfoot\@oddfoot% } % \end{macrocode} % \end{macro} % % \textsf{plain} is the default page style: % \begin{macrocode} \pagestyle{plain} % \end{macrocode} % % \cs{AtEndDocument}, we need to set the \texttt{lastpage} switches accordingly. The \cs{@deferlist} macro from \texttt{latex.ltx} holds a list of any unprocessed floats, so if this is empty, we know that there's nothing more to do and can just set \cs{@lastpagetexttrue}. If the list still contains unprocessed floats, set \cs{@lastpagefloattrue} then \cs{clearpage} to force the floats out. We also need to decrement the page counter by one, otherwise it ends up being one too many, presumably as a result of the \cs{clearpage}. % % I suspect that this will fail if the document uses \textsf{twocolumn} and has double floats that span both columns, because there is a separate \cs{@dbldeferlist}. Also, as noted above, this currently doesn't account for multiple float pages at the end of the document. % \begin{macrocode} \AtEndDocument{% \ifx\@deferlist\@empty% \@lastpagetexttrue% \else% \@lastpagefloattrue% \clearpage% \addtocounter{page}{-1}% \fi% } % \end{macrocode} % % \begin{macro}{\ps@titlepage} % The \textsf{titlepage} page style is similar to \textsf{plain}, but without the header. % \begin{macrocode} \def\ps@titlepage{% \def\@oddhead{\@markingschedule}% \let\@evenhead\@oddhead% \def\@oddfoot{\@draft\hfill\textbf{TURN OVER}}% \let\@evenfoot\@oddfoot% } % \end{macrocode} % \end{macro} % % % \subsection{Various counters} % % Earlier versions of \textsf{ouexam} used the \textsf{enumerate} % environment to build questions, which meant that special counters were not % required. This version of the class, however, uses a different mechanism for % building questions, so we need to define custom counters for several things. % % \begin{macro}{question} % \begin{macro}{subquestion} % \changes{2.0}{2000/09/11}{NJS New \texttt{subquestion} counter.} % \changes{2.0.1}{2000/04/30}{NJS Removed prefixed question number from % \cs{labelsubquestion} macro so that references now print as just ``(a)'' % instead of ``1(a)''.} % \begin{macro}{subsubquestion} % \changes{2.0}{2000/09/11}{NJS New \texttt{subsubquestion} counter.} % \changes{2.0.1}{2000/04/30}{NJS The \texttt{subsubquestion} was incorrectly % printing as ``(a)'' rather than ``(i)''. Fixed.} % \changes{2.0.1}{2000/04/30}{NJS Removed prefixed question numbers from % \cs{labelsubsubquestion} macro so that references now print as just ``(i)'' % instead of ``1(a)(i)''.} % First, we define three counters for the three possible levels of question: % top-level question, second-level sub-question or third-level sub-sub-question: % \begin{macrocode} \newcounter{question} \newcounter{subquestion}[question] \newcounter{subsubquestion}[subquestion] % \end{macrocode} % The three question levels are numbered as 1., (a) and (i) respectively, so we % also need to redefine the associated \cs{the} and \cs{label} macros % appropriately: % \begin{macrocode} \renewcommand{\thequestion}{\arabic{question}} \newcommand{\labelquestion}{\hfil\arabic{question}.} \renewcommand{\thesubquestion}{(\alph{subquestion})} \newcommand{\labelsubquestion}{\hfil(\alph{subquestion})} \renewcommand{\thesubsubquestion}{(\roman{subsubquestion})} \newcommand{\labelsubsubquestion}{\hfil(\roman{subsubquestion})} % \end{macrocode} % \end{macro} % \end{macro} % \end{macro} % % \begin{macro}{examexpected} % \changes{2.0}{2000/09/11}{NJS New \texttt{examexpected} counter.} % The \|examexpected\| counter stores the total expected number of marks for the % examination, and defaults to 100: % \begin{macrocode} \newcounter{examexpected} \setcounter{examexpected}{100} % \end{macrocode} % \end{macro} % \begin{macro}{examoutof} % \changes{2.0}{2000/09/11}{NJS New \texttt{examoutof} macro.} % To specify that an examination is out of some number of marks other than % 100, use the \cs{examoutof} macro to set the value: % \begin{macrocode} \newcommand{\examoutof}[1]{\setcounter{examexpected}{#1}} % \end{macrocode} % \end{macro} % % \begin{macro}{sectexpected} % \changes{2.0}{2002/01/09}{NJS New \cs{sectexpected} counter.} % \begin{macro}{qexpected} % \changes{2.0}{2000/09/11}{NJS New \cs{qexpected} counter.} % \begin{macro}{subqexpected} % \changes{2.0}{2000/09/11}{NJS New \texttt{subqexpected} counter.} % \begin{macro}{subsubqexpected} % \changes{2.0}{2000/09/11}{NJS New \texttt{subsubqexpected} counter.} % The next four counters perform the same function as the \|examexpected\| counter % for sections, questions, sub-questions and sub-sub-questions respectively. No % macros are required to set these values as they are set automatically by the % question-building environments. % % \begin{macrocode} \newcounter{sectexpected} \newcounter{qexpected} \newcounter{subqexpected} \newcounter{subsubqexpected} % \end{macrocode} % \end{macro} % \end{macro} % \end{macro} % \end{macro} % % \begin{macro}{lastexpected} % \changes{2.0}{2000/09/11}{NJS New \texttt{lastexpected} counter.} % The \|lastexpected\| counter records the expected marks total for the last % question. This will eventually be used to verify that the marks total for a % marking schedule equals the marks total for the corresponding question, but % this hasn't been implemented yet. % % \begin{macrocode} \newcounter{lastexpected} % \end{macrocode} % \end{macro} % % \begin{macro}{examrunning} % \changes{2.0}{2000/09/11}{NJS New \texttt{examrunning} counter.} % \begin{macro}{sectrunning} % \changes{2.0}{2002/01/09}{NJS New \texttt{sectrunning} counter.} % \begin{macro}{qrunning} % \changes{2.0}{2000/09/11}{NJS New \texttt{qrunning} counter.} % \begin{macro}{subqrunning} % \changes{2.0}{2000/09/11}{NJS New \texttt{subqrunning} counter.} % These four counters keep track of the running total of marks for the % current examination, section, question and subquestion respectively. This % is later compared against the corresponding expected total. A running total % is not needed for sub-sub-questions because they do not have sub-parts. The % \|qrunning\| and \|subqrunning\| counters are reset when the associated % question counters are incremented. % % \begin{macrocode} \newcounter{examrunning} \setcounter{examrunning}{0} \newcounter{sectrunning} \newcounter{qrunning}[question] \newcounter{subqrunning}[subquestion] % \end{macrocode} % \end{macro} % \end{macro} % \end{macro} % \end{macro} % % \begin{macro}{hassubs} % \changes{2.0}{2000/09/11}{NJS New \texttt{hassubs} counter.} % \begin{macro}{hassubsubs} % \changes{2.0}{2000/09/11}{NJS New \texttt{hassubsubs} counter.} % These two counters are used to track whether questions and % sub-questions have sub-parts. Counters are used rather than booleans because % counters are set globally and booleans are not (or at least do not appear to % be). Both counters reset when the associated question counters are % incremented. % % \begin{macrocode} \newcounter{hassubs}[question] \newcounter{hassubsubs}[subquestion] % \end{macrocode} % \end{macro} % \end{macro} % % \begin{macro}{qdepth} % \changes{2.3.1}{2013/05/31}{NJS New \texttt{qdepth} counter.} % This counter is incremented every time a \textsf{question}, \textsf{subquestion} or \textsf{subsubquestion} environment is opened, and decremented every time these environments close. It measures the nesting depth of questions, and is used to check that we do not have things like a \texttt{subquestion} not enclosed by a \texttt{question}, or a \texttt{question} nested inside another \texttt{question}. Note that this counter does \emph{not} reset inside any other environment. % % \begin{macrocode} \newcounter{qdepth} \setcounter{qdepth}{0} % \end{macrocode} % \end{macro} % % At the end of the document, check that the running total of marks for the % examination matches what we were expecting: % % \begin{macrocode} \AtEndDocument{% \ifnum\theexamexpected=\theexamrunning% \else\ClassWarning{ouexam}{% actual number of marks for exam (\theexamrunning) does not^^J% match expected number of marks (\theexamexpected)}% \fi% } % \end{macrocode} % % \begin{macro}{xsection} % \changes{2.0}{2002/01/29}{NJS New \texttt{xsection} counter.} % Finally, we set up a counter for the section ``number'' and define it so % that it prints out as an upper case letter rather than a number. We could % just redefine the \|section\| counter, but that interferes with the section % numbering in the documentation, and we can't have that, can we? \|:)\| % % \begin{macrocode} \newcounter{xsection} \setcounter{xsection}{0} \renewcommand{\thexsection}{\Alph{xsection}} % \end{macrocode} % \end{macro} % % % \subsection{Question-building environments and associated items} % % \changes{2.0}{2000/09/11}{NJS \textsf{question} environment no longer a % redefinition of \textsf{enumerate}.} % \changes{2.0}{2001/05/15}{NJS Added penalty handling to ensure correct % placement of marks in output.} % \changes{2.0}{2001/05/15}{NJS Total marks for a question are now printed only % if the question has no sub-questions.} % \begin{environment}{question} % \changes{2.3.1}{2013/05/31}{NJS Check question depth when opening the % environment.} % The \textsf{question} environment specifies a top-level question, and produces % questions numbered in the form ``1.'', ``2.'', etc. It has one mandatory % argument, the number of marks allocated to the question, which is used to % initialise the \|qexpected\| counter\footnote{If it were not for the fact that % you can only refer to environment arguments in the environment preamble, the % \texttt{qexpected} counter would be unnecessary.}. If the argument is left % empty, default to zero for the number of marks: % % \begin{macrocode} \newenvironment{question}[1]{% \def\@nummarks{#1}% \ifx\@nummarks\@empty\setcounter{qexpected}{0}% \else\setcounter{qexpected}{#1}\fi% % \end{macrocode} % % Check that the environment is being opened at the correct depth. The \textsf{question} environment should normally only be opened at depth 0: % \begin{macrocode} \ifnum\theqdepth=0% \else\ClassError{ouexam}{% question environment nested inside another question,^^J% subquestion or subsubquestion environment}% \fi% \addtocounter{qdepth}{1}% % \end{macrocode} % % For each top-level question, we increment the \|question\| counter, which also % resets the \|hassubs\| and \|qrunning\| counters to zero: % \begin{macrocode} \refstepcounter{question}% % \end{macrocode} % The question itself is built using a single-item \textsf{list} environment. % \begin{macrocode} \begin{list}{\labelquestion}{\settowidth{\labelwidth}{88.}}% \item \ignorespaces% % \end{macrocode} % % When the environment closes, the following things happen: % \begin{enumerate} % \item If the question has no sub-questions, the value of the \|qexpected\| % counter is added to \|examrunning\| and \|sectrunning\|, and both \|qrunning\| % and \|lastexpected\| are set to the value of \|qexpected\|. % % \begin{macrocode} }{% \ifnum\thehassubs=0% \addtocounter{examrunning}{\value{qexpected}}% \addtocounter{sectrunning}{\value{qexpected}}% \setcounter{qrunning}{\value{qexpected}}% % \end{macrocode} % % The total number of marks for the question is then printed right-justified % on the line as ``(\emph{m} marks)'' where \emph{m} is the value of the % \|qrunning\| counter. The environment determines whether to print ``mark'' or % ``marks'' automatically, and figures out whether the number of marks will fit % on the last line of the question or needs to be placed on the next line. % The code for handling the line breaking is derived from an example on page % 106 of \emph{The \TeX{}book}: % % \begin{macrocode} \unskip\nobreak\hfil\penalty50\hskip2em\hbox{}\nobreak% \hfil(\theqrunning~\ifnum\theqrunning=1 mark\else marks\fi)% \parfillskip=0pt \finalhyphendemerits=0 \par% % \end{macrocode} % % \item If the question \emph{does} have sub-questions, then \|qrunning\|, % \|examrunning\| and \|sectrunning\| have already been set by the various % sub-environments. The value of \|qrunning\| is then compared with % \|qexpected\|, and a warning is raised if they do not match. The total number % of marks for the question is \emph{not} printed in this case. % % \begin{macrocode} \else% \ifnum\theqrunning=\theqexpected% \else\ClassWarning{ouexam}{% actual mark (\theqrunning) for question % \thequestion\space doesn't match^^J% expected mark (\theqexpected)}% \fi% \fi% % \end{macrocode} % % \item \|lastexpected\| is set to the value of \|qexpected\| so it can be used % in any subsequent \textsf{marking} environment. % % \begin{macrocode} \setcounter{lastexpected}{\value{qexpected}}% \end{list}% % \end{macrocode} % % \item \|qdepth\| is decremented. % % \begin{macrocode} \addtocounter{qdepth}{-1}% } % \end{macrocode} % % \end{enumerate} % \end{environment} % % \begin{environment}{subquestion} % \changes{2.0}{2000/09/11}{NJS \textsf{subquestion} environment no longer a % redefinition of \textsf{enumerate}.} % \changes{2.0}{2001/05/13}{NJS Added penalty handling to ensure correct placement % of marks in output.} % \changes{2.3.1}{2013/05/31}{NJS Check question depth when opening the % environment.} % The \textsf{subquestion} environment is the analogue of \textsf{question} for % sub-questions. On initialisation, it sets \|subqexpected\| to the value of the % mandatory argument (defaulting to zero if the argument is empty), increments % the \|hassubs\| counter so that the enclosing \textsf{question} environment can % react appropriately, increments \|subquestion\| (thus resetting \|hassubsubs\| and % \|subqrunning\| to zero) and opens a single-item list with numbering of the form % ``(a)'', ``(b)'', etc. % % \begin{macrocode} \newenvironment{subquestion}[1]{% \def\@nummarks{#1}% \ifx\@nummarks\@empty\setcounter{subqexpected}{0}% \else\setcounter{subqexpected}{#1}\fi% \setcounter{subqexpected}{#1}% \addtocounter{hassubs}{1}% \refstepcounter{subquestion}% % \end{macrocode} % The \textsf{subquestion} environment should normally only be opened at depth 1: % \begin{macrocode} \ifnum\theqdepth=1% \else\ClassError{ouexam}{% subquestion environment either outside a question^^J% environment, or nested inside another subquestion or subsubquestion^^J% environment}% \fi% \addtocounter{qdepth}{1}% \begin{list}{\labelsubquestion}{\settowidth{\labelwidth}{(m)}}% \item \ignorespaces% }{% % \end{macrocode} % % When the environment closes, if performs similar checks to the % \textsf{question} environment. If the sub-question has no sub-sub-questions, % the value of \|subqexpected\| is added to \|qrunning\| and \|sectrunning\|: % % \begin{macrocode} \ifnum\thehassubsubs=0% \addtocounter{examrunning}{\value{subqexpected}}% \addtocounter{sectrunning}{\value{subqexpected}}% \addtocounter{qrunning}{\value{subqexpected}}% \setcounter{subqrunning}{\value{subqexpected}}% % \end{macrocode} % Then the number of marks for the sub-question are typeset in a similar manner % to the \textsf{question} environment: % \begin{macrocode} \unskip\nobreak\hfil\penalty50\hskip2em\hbox{}\nobreak% \hfil(\thesubqrunning~\ifnum\thesubqrunning=1 mark\else marks\fi)% \parfillskip=0pt \finalhyphendemerits=0 \par% % \end{macrocode} % % If the sub-question \emph{does} have sub-sub-questions, check the running total % against the expected number and raise an error if they don't match. % \begin{macrocode} \else% \ifnum\thesubqrunning=\thesubqexpected% \else\ClassWarning{ouexam}{% actual mark (\thesubqrunning) for question % \thequestion\thesubquestion\space doesn't match^^J% expected mark (\thesubqexpected)}% \fi% \fi% % \end{macrocode} % Finally, set \|lastexpected\| to the value of \|qexpected\| so it can be used % in any subsequent \textsf{marking} environment. % \begin{macrocode} \setcounter{lastexpected}{\value{subqexpected}}% \end{list}% \addtocounter{qdepth}{-1}% } % \end{macrocode} % \end{environment} % % \begin{environment}{subsubquestion} % \changes{2.0}{2000/09/11}{NJS \textsf{subsubquestion} environment no longer a % redefinition of \textsf{enumerate}.} % \changes{2.0}{2001/05/13}{NJS Added penalty handling to ensure correct placement % of marks in output.} % \changes{2.3.1}{2013/05/31}{NJS Check question depth when opening the % environment.} % This is similar to both \textsf{question} and \textsf{subquestion}, and % generates an appropriately numbered sub-sub-question (``(i)'', ``(ii)'', etc.). % \begin{macrocode} \newenvironment{subsubquestion}[1]{% \def\@nummarks{#1}% \ifx\@nummarks\@empty\setcounter{subsubqexpected}{0}% \else\setcounter{subsubqexpected}{#1}\fi% % \end{macrocode} % Increment \|hassubsubs\| so that the enclosing \textsf{subquestion} environment % can react appropriately: % \begin{macrocode} \addtocounter{hassubsubs}{1}% \refstepcounter{subsubquestion}% % \end{macrocode} % The \textsf{subsubquestion} environment should normally only be opened at depth 2: % \begin{macrocode} \ifnum\theqdepth=2% \else\ClassError{ouexam}{% subsubquestion environment either outside a subquestion^^J% environment, or nested inside another subsubquestion environment}% \fi% \addtocounter{qdepth}{1}% \begin{list}{\labelsubsubquestion}{\settowidth{\labelwidth}{(viii)}}% \item \ignorespaces% }{% % \end{macrocode} % % Closing this environment is a bit simpler than the previous two because we % don't need to check whether the sub-sub-question has any sub-components (this % never happens). All we have to do is set the appropriate counters and typeset % the number of marks. % % \begin{macrocode} \addtocounter{examrunning}{\value{subsubqexpected}}% \addtocounter{sectrunning}{\value{subsubqexpected}}% \addtocounter{qrunning}{\value{subsubqexpected}}% \addtocounter{subqrunning}{\value{subsubqexpected}}% \setcounter{lastexpected}{\value{subsubqexpected}}% \unskip\nobreak\hfil\penalty50\hskip2em\hbox{}\nobreak% \hfil(\thesubsubqexpected~\ifnum\thesubsubqexpected=1 mark% \else marks\fi)% \parfillskip=0pt \finalhyphendemerits=0 \par% \end{list}% \addtocounter{qdepth}{-1}% } % \end{macrocode} % \end{environment} % % % \subsection{Draft examination printing} % % \begin{macro}{\@marking} % \changes{2.0}{2001/05/17}{NJS New \cs{@draft} macro.} % Specifying the \textsf{draft} option causes \textsf{ouexam} to print % \fbox{DRAFT} in large letters in the footer of every page. This is drawn by % the \cs{@draft} macro, which is included in the page footer definition of % all the page styles for this class (see % \hyperref[pagestyles]{section~\ref{pagestyles}}). \textbf{Note:} Because % the \cs{@draft} macro is included in the page footer, it will usually % overflow the page boundaries, causing ``\|Overfull \vbox\|'' warnings. Note % the extra set of braces to limit the scope of the \cs{Huge}. % % \begin{macrocode} \if@draft\def\@draft{{\Huge\fbox{DRAFT}}} \else\let\@draft\@empty \fi % \end{macrocode} % \end{macro} % % % \changes{2.2}{2010/04/26}{NJS Added time allocation macros.} % \subsection{Time allocations} % % \begin{macro}{\timeallocation} % The \cs{timeallocation} macro takes a single argument representing the number of minutes suggested time allocation for a question, and outputs it in the form: \textbf{[The suggested time allocation for this question is \emph{t} minutes.]} The argument is not checked and can be anything (numbers, text). The macro will need to be redefined if either units other than minutes or different phrasing are required. % \begin{macrocode} \newcommand{\timeallocation}[1]{% \textbf{[The suggested time allocation for answering this question % is #1 minutes.]}% } % \end{macrocode} % \end{macro} % % \begin{macro}{\readingtime} % The \cs{readingtime} macro takes a single argument representing the number of minutes reading time allocated to some part of the examination, and outputs it in the form: \textbf{[Allow \emph{t} minutes reading time.]} The argument is not checked and can be anything (numbers, text). The macro will need to be redefined if either units other than minutes or different phrasing are required. % \begin{macrocode} \newcommand{\readingtime}[1]{\textbf{[Allow #1 minutes reading time.]}} % \end{macrocode} % \end{macro} % % \textsf{ouexam} currently does not check that any of these time allocations sum to the correct total. If implemented, this would work in a similar way to how marks totals are checked, and would be keyed in some way off \cs{@hours}. % % % \subsection{Marking schedule information} % % \begin{environment}{marking} % \begin{macro}{\@markingschedule} % \changes{1.2}{1999/10/26}{NJS New \textsf{marking} environment for including % answers and/or marking schedule information.} % \changes{2.0}{2000/09/11}{NJS Now prints ``MARKING SCHEDULE'' at the bottom % of every page if the \textsf{markingschedule} class option is specified.} % The \textsf{marking} environment allows the specification of marking % schedule information in the same location as the questions. If the % \textsf{markingschedule} class option is specified, marking schedule % information is printed in italics. In addition, \fbox{MARKING SCHEDULE} is % printed in large letters in the header of every page. This is drawn by the % \cs{@markingschedule} macro, which is included in the page header % definition of all the page styles for this class (see % \hyperref[pagestyles]{section~\ref{pagestyles}}). Note the extra set of % braces to limit the scope of the \cs{Huge}. % % \begin{macrocode} \if@markingschedule \newenvironment{marking}{\itshape}{\normalfont} \def\@markingschedule{{\Huge\fbox{MARKING SCHEDULE}}} \else % \end{macrocode} % % Normally \textsf{marking} just maps to the \textsf{comment} environment from % Rainer Sch\"{o}pf's \textsf{verbatim} package, i.e., marking information is not % printed: % % \begin{macrocode} \let\marking\comment \let\endmarking\endcomment \let\@markingschedule\@empty \fi % \end{macrocode} % \end{macro} % \end{environment} % % % \subsection{Section handling} % % \changes{2.0}{2002/01/09}{NJS new \textsf{examsection} environment.} % \changes{2.0}{2002/01/09}{NJS New \cs{@defsecinst} macro.} % \begin{environment}{examsection} % \begin{macro}{\@defsecinst} % The \textsf{examsection} environment specifies a major section of an % examination paper. Sections are sequentially numbered ``A'', ``B'', etc. The % environment has three arguments: % % \begin{enumerate} % \item The expected number of marks for the section. % % \item The instructions for this section. This defaults to ``ANSWER ALL % QUESTIONS.'' if left blank. To change the default text, redefine the % \cs{@defsecinst} macro. % \begin{macrocode} \def\@defsecinst{ANSWER ALL QUESTIONS.} % \end{macrocode} % % \item A description of the contents of the section which may be left blank. % \end{enumerate} % % \begin{macrocode} \newenvironment{examsection}[3]{% % \end{macrocode} % Every section begins on a new page. % \begin{macrocode} \newpage% % \end{macrocode} % If the argument for the number of marks is left empty, default to zero. % \begin{macrocode} \def\@nummarks{#1}% \ifx\@nummarks\@empty\setcounter{sectexpected}{0}% \else\setcounter{sectexpected}{#1}\fi% \refstepcounter{xsection}% \setcounter{sectrunning}{0}% % \end{macrocode} % The section title is formatted as ``\textbf{\underline{Section A}}'', in % \cs{large} size. % \begin{macrocode} {\large\noindent\textbf{\underline{Section~\thexsection}}}% \\[0.5\baselineskip]% \def\@usersecinst{#2}% \ifx\@usersecinst\@empty\@defsecinst\else\@usersecinst\fi\par% \def\@usersectopic{#3}% \ifx\@usersectopic\@empty\else\par\noindent\@usersectopic\fi \\% }{% % \end{macrocode} % % When the environment closes, the value of \|sectrunning\| is compared against % \|sectexpected\|, and a warning is raised if they do not match. The total number % marks for the section is printed in the form ``\textbf{[SECTION A TOTAL % \emph{m} MARKS]}'' where \emph{m} is the value of \|sectrunning\|. % % \begin{macrocode} \ifnum\thesectrunning=\thesectexpected% \else\ClassWarning{ouexam}{% actual mark (\thesectrunning) for section % \thexsection\space doesn't match^^J% expected mark (\thesectexpected)}% \fi% \bigskip\hfill\textbf{[SECTION \thexsection\ TOTAL % \thesectrunning\ MARKS]}% } % \end{macrocode} % \end{macro} % \end{environment} % % % \subsection{Title page generation} % % \begin{macro}{\examyear} % \begin{macro}{\@year} % \changes{2.0.2}{2002/08/22}{NJS Put back \@eha at end of the \ClassError. Don't know what this does, but the macro crashes without it there. It seems to be required if there's a \cs{protect}.} % The \cs{examyear} macro specifies the year in which the examination is being held. It redefines the \cs{@year} macro which is used in \cs{@maketitlepage}: % \begin{macrocode} \newcommand{\examyear}[1]{\def\@year{#1}} % \end{macrocode} % This macro is required: % \begin{macrocode} \def\@year{% \ClassError{ouexam}{no \protect\examyear\space was specified}\@eha% } % \end{macrocode} % \end{macro} % \end{macro} % % \begin{macro}{\department} % \begin{macro}{\@dept} % \changes{2.0.2}{2002/08/22}{NJS Put back \@eha at end of the \ClassError. Don't know what this does, but the macro crashes without it there. It seems to be required if there's a \cs{protect}.} % The \cs{department} macro specifies the name of the department, and is required. It redefines the \cs{@dept} macro which is used in \cs{@maketitlepage}: % \begin{macrocode} \newcommand{\department}[1]{\def\@dept{#1}} \def\@dept{% \ClassError{ouexam}{no \protect\department\space was specified}\@eha% } % \end{macrocode} % \end{macro} % \end{macro} % % \begin{macro}{\papernumber} % \begin{macro}{\@pnumber} % \changes{2.0.2}{2002/08/22}{NJS Put back \@eha at end of the \ClassError. Don't know what this does, but the macro crashes without it there. It seems to be required if there's a \cs{protect}.} % The \cs{papernumber} macro specifies the paper number, and is required. It redefines the \cs{@pnumber} macro which is used in \cs{@maketitlepage}: % \begin{macrocode} \newcommand{\papernumber}[1]{\def\@pnumber{#1}} \def\@pnumber{% \ClassError{ouexam}{no \protect\papernumber\space was specified}\@eha% } % \end{macrocode} % \end{macro} % \end{macro} % % \begin{macro}{\papertitle} % \begin{macro}{\@ptitle} % \changes{2.0.2}{2002/08/22}{NJS Put back \@eha at end of the \ClassError. Don't know what this does, but the macro crashes without it there. It seems to be required if there's a \cs{protect}.} % The \cs{papertitle} macro specifies the title of the paper, and is required. It redefines the \cs{@ptitle} macro which is used in \cs{@maketitlepage}: % \begin{macrocode} \newcommand{\papertitle}[1]{\def\@ptitle{#1}} \def\@ptitle{% \ClassError{ouexam}{no \protect\papertitle\space was specified}\@eha% } % \end{macrocode} % \end{macro} % \end{macro} % % \begin{macro}{\semester} % \begin{macro}{\@semester} % \changes{2.0.2}{2002/08/22}{NJS Added full-year support to \cs{semester}.} % \changes{2.0}{2002/01/14}{NJS Rewrote \cs{semester} to support summer school % and special exams.} % The \cs{semester} macro specifies which semester the examination is for. % Legal values for the argument are ``1'', ``2'', ``SS'' for summer school, % ``FY'' for full-year or ``SP'' for a special examination---anything else is % ignored. % \begin{macrocode} \def\@ONE{1} \def\@TWO{2} \def\@SS{SS} \def\@FY{FY} \def\@SP{SP} % \end{macrocode} % This macro redefines the \cs{@semester} macro which is used in % \cs{@maketitlepage}: % \begin{macrocode} \newcommand{\semester}[1]{% \def\@sem{#1}% \ifx\@sem\@ONE\def\@semester{Semester One}% \else\ifx\@sem\@TWO\def\@semester{Semester Two}% \else\ifx\@sem\@SS\def\@semester{Summer School}% \else\ifx\@sem\@FY\def\@semester{Full Year}% \else\ifx\@sem\@SP\def\@semester{Special Examination}% \else\ClassWarning{ouexam}{% invalid value `#1' for \protect\semester;^^J% valid values are `1', `2', `SS', `FY' and `SP'. No semester^^J% information will be printed}% \fi\fi\fi\fi\fi% } % \end{macrocode} % \cs{semester} is optional---if you omit it or give it an invalid argument, % \cs{@semester} remains empty. % \begin{macrocode} \let\@semester\@empty % \end{macrocode} % \end{macro} % \end{macro} % % \begin{macro}{\timeallowed} % \begin{macro}{\@hours} % The \cs{timeallowed} macro specifies the length of the examination in hours. % It redefines the \cs{@hours} macro which is used in \cs{@maketitlepage}: % \begin{macrocode} \newcommand{\timeallowed}[1]{\def\@hours{#1}} % \end{macrocode} % \cs{timeallowed} is optional---if you omit it, \cs{@hours} defaults to ``3'': % \begin{macrocode} \def\@hours{3} % \end{macrocode} % \end{macro} % \end{macro} % % \begin{macro}{num.pages} % \changes{1.1}{1999/04/20}{NJS Renamed \texttt{page.last} label to \texttt{num.pages}.} % \changes{2.4}{2013/06/05}{NJS Added check for \textsf{hyperref} being loaded.} % There is no macro to specify the number of pages in the examination because it is done automatically by inserting a \cs{newlabel} that refers to the last page directly into the \|.aux\| file. The \textsf{hyperref} package redefines the \cs{newlabel} macro (in particular it increases the number of arguments), so we need to test whether \textsf{hyperref} is loaded and write the appropriate string to the \|.aux\| file. We borrow a technique from the \textsf{pageslts} package of checking for the existence of \cs{Hy@Warning}. (Ideally we would use \cs{if@packageloaded}, but this can't be used after \cs{AtBeginDocument}, and can be foiled by a package loading \textsf{hyperref} \cs{AtBeginDocument}.) % \begin{macrocode} \AtEndDocument{% \@ifundefined{Hy@Warning}{% hyperref not loaded \immediate\write\@auxout{\string\newlabel{num.pages}{{}{\thepage}}}% }{% hyperref loaded \immediate\write\@auxout{\string\newlabel{num.pages}{{}{\thepage}{}{}{}}}% }% } % \end{macrocode} % This label is then referenced in \cs{@maketitlepage}. % \end{macro} % % \begin{macro}{num.questions} % \changes{1.1}{1999/04/20}{NJS Added \cs{num.questions} label representing the number of questions.} % \changes{2.4}{2013/06/05}{NJS Added check for \textsf{hyperref} being loaded.} % Similarly, there is no macro to specify the number of questions in the examination. This is calculated using the same method as for the number of pages. % \begin{macrocode} \AtEndDocument{% \@ifundefined{Hy@Warning}{% \immediate\write\@auxout{\string\newlabel{num.questions}{{}{\thequestion}}}% }{% \immediate\write\@auxout{\string\newlabel{num.questions}{{}{\thequestion}{}{}{}}}% }% } % \end{macrocode} % \end{macro} % % \begin{macro}{\instructions} % \begin{macro}{\@instructions} % The \cs{instructions} macro specifies instructions on how candidates should % answer questions. It redefines the \cs{@instructions} macro that is used in % \cs{@maketitlepage}: % \begin{macrocode} \def\@instructions{Answer \underline{ALL} questions.} % \end{macrocode} % \cs{instructions} is optional. If you omit it, \cs{@instructions} defaults % to ``Answer \underline{ALL} questions.'' % \begin{macrocode} \newcommand{\instructions}[1]{\def\@instructions{#1}} % \end{macrocode} % \end{macro} % \end{macro} % % \begin{macro}{\@NA} % \changes{2.1.3}{2009/04/28}{NJS Added \cs{@NA} macro.} % The \cs{@NA} macro enables us to easily assign and test for the default value % (``N/A'') for some of the title page macros. (There is certain text that % should not be output when particular macros are allowed to default.) % \begin{macrocode} \def\@NA{N/A} % \end{macrocode} % \end{macro} % % \begin{macro}{\material} % \begin{macro}{\@material} % \changes{2.1.3}{2009/04/28}{NJS Changed default to ``N/A'' for \cs{material}.} % The \cs{material} macro lets specifies any additional material that candidates % are provided in addition to the examination paper itself, and is optional. % It redefines the \cs{@material} macro that is used in \cs{@maketitlepage}: % \begin{macrocode} \let\@material\@NA \newcommand{\material}[1]{\def\@material{#1}} % \end{macrocode} % If you omit it, it defaults to ``N/A''. % \end{macro} % \end{macro} % % \begin{macro}{\allowcalculators} % \begin{macro}{\@calculators} % \changes{2.1}{2004/04/05}{NJS Modified calculator macros to conform to % new University calculator regulations.} % The \cs{allowcalculators} macro specifies whether or not calculators are % allowed in the examination (\cs{permitcalculators} is a synonym). If % \cs{allowcalculators} is omitted, \cs{@calculators} defaults to ``No % calculators are permitted.'': % \begin{macrocode} \def\@calculators{No calculators are permitted.} % \end{macrocode} % Issuing an \cs{allowcalculators[none]} does nothing, as the default is % to disallow calculators anyway. Issuing an \cs{allowcalculators} or % \cs{allowcalculators[any]} redefines \cs{@calculators} to ``No % restriction on the model of calculator to be used, but no device with % communication capability shall be accepted as a calculator.'' Issuing an % \cs{allowcalculators[approved]} redefines \cs{@calculators} to ``Only % calculators on the University of Otago list of approved calculators are % permitted.'': % \begin{macrocode} \def\@CalcNONE{none} \def\@CalcANY{any} \def\@CalcAPPROVED{approved} % \end{macrocode} % \changes{2.1.2}{2008/08/14}{NJS Ensured that the correct default is set % regardless of whether the calculators macro is called.} % \begin{macrocode} \let\@calc\@CalcNONE \newcommand{\allowcalculators}[1][any]{% \def\@calc{#1}% \ifx\@calc\@CalcNONE% \else\ifx\@calc\@CalcANY\def\@calculators{No restriction on the % model of calculator to be used, but no device with communication % capability shall be accepted as a calculator.}% \else\ifx\@calc\@CalcAPPROVED\def\@calculators{Only calculators on % the University of Otago list of approved calculators are permitted.}% \else\ClassWarning{ouexam}{% invalid argument `#1' for^^J% \protect\allowcalculators; valid values are `none', `any' and `approved'.^^J% No calculators will be permitted}% \fi\fi\fi% } % \end{macrocode} % \cs{@calculators} is used in \cs{@maketitlepage}. % \end{macro} % \end{macro} % % \begin{macro}{\permitcalculators} % \changes{2.1}{2004/04/05}{NJS Added \cs{permitcalculators} as a synonym % for \cs{allowcalculators}.} % \cs{permitcalculators} is a synonym for \cs{allowcalculators}. % \begin{macrocode} \let\permitcalculators\allowcalculators % \end{macrocode} % \end{macro} % % \begin{macro}{\copiesof} % \begin{macro}{\@copiesof} % \changes{2.1.3}{2009/04/28}{NJS Changed default to ``N/A'' for \cs{copiesof}.} % The \cs{copiesof} macro specifies any material that candidates are allowed to % bring into the examination, and is optional. It redefines the \cs{@copiesof} % macro that is used in \cs{@maketitlepage}: % \begin{macrocode} \let\@copiesof\@NA \newcommand{\copiesof}[1]{\def\@copiesof{#1}} % \end{macrocode} % \end{macro} % \end{macro} % % \begin{macro}{\otherinstructions} % \begin{macro}{\@otherinst} % \changes{2.1.3}{2009/04/28}{NJS Changed default to ``N/A'' for % \cs{otherinstructions}.} % The \cs{otherinstructions} macro specifies any other instructions not covered % by any of the above, and is optional. It redefines the \cs{@otherinst} % macro that is used in \cs{@maketitlepage}: % \begin{macrocode} \let\@otherinst\@NA \newcommand{\otherinstructions}[1]{\def\@otherinst{#1}} % \end{macrocode} % \end{macro} % \end{macro} % % \begin{macro}{\maketitlepage} % The \cs{maketitlepage} macro is the \textsf{ouexam} analogue of \cs{maketitle}. % It takes all of the information provided by the above macros and generates a % properly formatted examination title page. If you want to design your own title % page styles, redefine \cs{@maketitlepage} (this is not normally recommended % however). % \begin{macrocode} \newcommand{\maketitlepage}{\@maketitlepage} % \end{macrocode} % \end{macro} % % \begin{macro}{\@maketitlepage} % \changes{1.1}{1999/04/20}{NJS Renamed macro from \cs{maketitlepage}.} % \changes{2.4}{2013/06/05}{NJS Added check for \textsf{hyperref} being loaded.} % The \cs{@maketitlepage} macro generates an examination title page that meets the Otago University requirements for final examination papers. You can redefine this if you want to change the format, but normally you would not do this. % \begin{macrocode} \def\@maketitlepage{% \thispagestyle{titlepage}% % \end{macrocode} % The header information for the examination is printed centered at the top % of the page. The department, paper number and paper title (and the semester % information if required) are placed inside a double box. % \begin{macrocode} \begin{center}% {\Large \textbf{UNIVERSITY OF OTAGO EXAMINATIONS \@year}}% \\[\baselineskip]% \fbox{\framebox[\linewidth]{% \begin{tabular}{c}% \\% {\large \@dept} \\% \\% {\large Paper \@pnumber} \\% \\% {\large \@ptitle} \\% \ifx\@semester\@empty\else\@semester \\ \fi% \\% \end{tabular}% }}% \mbox{}\\[\baselineskip]% % \end{macrocode} % Next, the time allowed for completing the examination: % \begin{macrocode} \textbf{(TIME ALLOWED: \@hours\ HOURS)}% \\[\baselineskip]% \end{center}% % \end{macrocode} % The number of pages is automatically calculated, as described earlier. All we need to do is \cs{pageref} the \|num.pages\| label that we wrote into the \|.aux\| file, or \cs{pageref} if \textsf{hyperref} is loaded to inhibit hyperlink generation. % \begin{macrocode} \underline{This examination paper comprises % \@ifundefined{Hy@Warning}{\pageref{num.pages}}{\pageref{num.pages}} pages.% }% \\[\baselineskip]% % \end{macrocode} % Print out the instructions for answering questions: % \begin{macrocode} \underline{Candidates should answer questions as follows:}% \\[\baselineskip]% \hspace{1cm}\begin{minipage}{13.5cm}\@instructions\end{minipage}% \\[\baselineskip]% % \end{macrocode} % Print out any provided material: % \begin{macrocode} \underline{The following material is provided:}% \\[\baselineskip]% \hspace{1cm}\begin{minipage}{13.5cm}\@material\end{minipage}% \\[\baselineskip]% % \end{macrocode} % Print out use of calculators section: % \begin{macrocode} \underline{Use of calculators:} \\[\baselineskip]% \hspace{1cm}\begin{minipage}{13.5cm}\@calculators\end{minipage}% \\[\baselineskip]% % \end{macrocode} % \changes{2.1.3}{2009/04/28}{NJS Fixed vertical spacing.} % \changes{2.1.2}{2008/08/14}{NJS Ensure that the ``subject to inspection'' % line for calculators is only printed when calculators actually are allowed.} % Only print out the ``subject to inspection'' line if calculators actually % are allowed (irrelevant otherwise): % \begin{macrocode} \ifx\@calc\@CalcNONE% \else\mbox{}\hfill(Subject to inspection by the examiners.)% \\[\baselineskip]% \fi% % \end{macrocode} % Print out anything that candidates are allowed to bring: % \begin{macrocode} \underline{Candidates are permitted copies of:}% \\[\baselineskip]% \hspace{1cm}\begin{minipage}{13.5cm}\@copiesof\end{minipage}% \\[\baselineskip]% % \end{macrocode} % \changes{2.1.3}{2009/04/28}{NJS Fixed vertical spacing.} % \changes{2.1.2}{2008/08/14}{NJS Ensure that the ``subject to inspection'' % line for other material is only printed when there actually \emph{is} % other material.} % Only print out the ``subject to inspection'' line if there actually % \emph{is} other material (irrelevant otherwise): % \begin{macrocode} \ifx\@copiesof\@NA% \else\mbox{}\hfill(Subject to inspection by the examiners.)% \\[\baselineskip]% \fi% % \end{macrocode} % \changes{2.1.3}{2009/04/28}{NJS Removed superfluous blank line.} % Print out other instructions, and finish. % \begin{macrocode} \underline{Other Instructions:}% \\[\baselineskip]% \hspace{1cm}\begin{minipage}{13.5cm}\@otherinst\end{minipage}% \newpage% } % \end{macrocode} % \end{macro} % % % \subsection{Obsolete macros and environments} % % \changes{2.0}{2002/01/09}{NJS All \textsf{multichoice} macros and environments % made obsolete.} % \changes{2.0}{2002/01/09}{NJS \cs{markingschedule}, \cs{marks}, \cs{totalmarks} % macros made obsolete.} % \changes{2.0}{2002/01/09}{NJS \textsf{questions}, \textsf{subquestions}, % \textsf{subsubquestions} environments made obsolete.} % The following macros and environments are no longer supported as of version % 2.0. Documents that use these macros can be processed using ouexam v1.2 or % earlier. The obsolete multiple-choice examination support is dealt with % during option processing, so we can ignore the obsolete multiple-choice % macros. % % \begin{environment}{questions} % \begin{environment}{subquestions} % \changes{1.1}{1999/04/20}{NJS Now disabled when \textsf{multichoice} is % active.} % \begin{environment}{subsubquestions} % \changes{1.1}{1999/04/20}{NJS Now disabled when \textsf{multichoice} is % active.} % \begin{macro}{markingschedule} % \changes{1.2}{1999/10/26}{NJS New macro to turn marking information on and % off.} % The \textsf{questions}, \textsf{subquestions} and \textsf{subsubquestions} % environments have been replaced by the \textsf{question}, % \textsf{subquestion} and \textsf{subsubquestion} environments respectively. % \begin{macrocode} \newenvironment{questions}{% \def\item{\@unsupported{questions environment}{1.2}}}{} \newenvironment{subquestions}{% \def\item{\@unsupported{subquestions environment}{1.2}}}{} \newenvironment{subsubquestions}{% \def\item{\@unsupported{subsubquestions environment}{1.2}}}{} % \end{macrocode} % The \cs{markingschedule} macro has been replaced by the % \textsf{markingschedule} class option. We can't map this to the % \cs{@unsupported} macro because \cs{markingschedule} occurs in the document % preamble. Mapping it to \cs{@unsupported} causes \TeX\ to blow up for some % reason. It's safe enough to ignore it. % \begin{macrocode} \let\markingschedule\@empty % \end{macrocode} % \end{macro} % \end{environment} % \end{environment} % \end{environment} % % \changes{2.1}{2004/04/16}{NJS Removed the \texttt{marks} macro entirely % because it conflicts with a macro in e-TeX.} % \changes{1.1}{1999/04/20}{NJS This macro now increments the \texttt{marks} % counter.} % \begin{macro}{\totalmarks} % \changes{1.1}{1999/04/20}{NJS Removed the parameter for the number of marks; % Both these macros have been replaced by automatic calculations within the % question-building environments.} % The \cs{marks} macro has been replaced by the argument to the various % question-building environments. The \cs{totalmarks} macro has been replaced % by automatic calculations within these environments. % \begin{macrocode} \newcommand{\totalmarks}{\@unsupported{totalmarks macro}{1.2}} % \end{macrocode} % \end{macro} % % % % \Finale % % \PrintChanges \def\ouexamdate{4 September 2012} \def\ouexamversion{2.3} \def\ouexamshortdate{2012/09/04} % \iffalse meta-comment %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% %% %% File: $Id$ %% Copyright 1999--2006 Nigel Stanger and University of Otago %% %% You may use this package freely, and also distribute it %% provided that you don't change it, make any money off %% it or pretend that you wrote it. %% %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% %<driver> \documentclass[a4paper]{ltxdoc} \usepackage{doc} \DisableCrossrefs \CodelineNumbered \RecordChanges \usepackage{graphicx} % change hyperref options as required \usepackage[colorlinks,pdfpagemode=None,urlcolor=blue, linkcolor=red,pdfauthor={Nigel Stanger}]{hyperref} \title{The \textsf{ouexam} document class, v\ouexamversion} \author{Nigel Stanger\\nstanger@infoscience.otago.ac.nz} \date{\ouexamdate} \begin{document} \maketitle \DocInput{ouexam.dtx} \end{document} %</driver> % % \fi % %% \CheckSum{896} %% %% \CharacterTable %% {Upper-case \A\B\C\D\E\F\G\H\I\J\K\L\M\N\O\P\Q\R\S\T\U\V\W\X\Y\Z %% Lower-case \a\b\c\d\e\f\g\h\i\j\k\l\m\n\o\p\q\r\s\t\u\v\w\x\y\z %% Digits \0\1\2\3\4\5\6\7\8\9 %% Exclamation \! Double quote \" Hash (number) \# %% Dollar \$ Percent \% Ampersand \& %% Acute accent \' Left paren \( Right paren \) %% Asterisk \ Plus \+ Comma \, %% Minus \- Point \. Solidus \/ %% Colon \: Semicolon \; Less than \< %% Equals \= Greater than \> Question mark \? %% Commercial at \@ Left bracket \[ Backslash \\ %% Right bracket \] Circumflex \^ Underscore \_ %% Grave accent \` Left brace \{ Vertical bar \\| %% Right brace \} Tilde \~} %% % % \MakeShortVerb{\\|} % % \changes{1.0}{1999/04/15}{Initial version.} % % \begin{abstract} % This document class allows you to create Otago University final % examination papers using \LaTeXe\@. It implements the major formatting % requirements specified by the University, provides useful % macros to ease the process of building the title page and automatically % deals with fiddly little issues such as question and section numbering, % the number of pages in the paper, printing ``\textbf{TURN OVER}'' in the % bottom right corner of every page except the last one and ensuring that the % total number of marks for a question adds up to the expected number. % \end{abstract} % % % \tableofcontents % % % \section{Overview} % % \changes{2.1}{2004/04/10}{NJS Added requirements and installation instructions.} % \changes{2.3}{2012/09/04}{NJS Dropped vanilla \LaTeX\ for building docs.} % \subsection{Requirements and installation} % % The \textsf{ouexam} package requires the \textsf{verbatim}, % \textsf{fontenc}, \textsf{lmodern} and \textsf{textcomp} packages, all % of which should come standard with most \TeX\ installations. To build % the documentation and example files, you will need at least version 1.1 % of the \textsf{listings} package, and the \textsf{pdfjam} tool. % % Installation is relatively simple: % \begin{enumerate} % % \item Unpack the distribution archive and \texttt{cd} to the % distribution directory. % % \item \texttt{make}. % % \item \texttt{make install TEXMF\_INSTALL=/path/to/texmf}. % \texttt{/path/to/texmf} should be the root of your preferred % \texttt{texmf} tree (e.g., \texttt{/usr/share/texmf}). You may need % to do this as root depending on which \texttt{texmf} tree you are % installing into. You can also define \texttt{TEXMF\_INSTALL} as an % environment variable then simply type \texttt{make install}. % % \end{enumerate} % You can, of course, always install manually if you wish. % % % \changes{2.0}{2002/01/11}{NJS Added note on backwards compatibility.} % \subsection{Important note on backwards compatibility} % % Version 2.0 of \textsf{ouexam} introduced several major changes from earlier % versions. Consequently, documents created using an earlier version of % \textsf{ouexam} are \emph{incompatible} with version 2.0 or later. Versions 2.0 % and greater of \textsf{ouexam} will detect attempts to process a version 1.x % document and display an error message. % % The older version (1.2) is still available and can be downloaded from the same % location as the current version. To process older version documents, just copy % the version 1.2 \|ouexam.cls\| file into the same directory as the document file. % \TeX\ searches the current directory first, so this copy will take precedence % over any other installed version of \textsf{ouexam}. % % % \subsection{Class options} % % This document class is based on the \textsf{article} class and accepts % most of the options accepted by \textsf{article} (the \textsf{twoside}, % \textsf{10pt} and \textsf{11pt} options are ignored). The standard % options are \textsf{onecolumn}, \textsf{oneside}, \textsf{a4paper} and % \textsf{12pt}. In addition, \textsf{ouexam} accepts the following % ``native'' class options: % \begin{description} % \item[\textsf{draft}] \DescribeMacro{draft} This option has the usual % effects that you would expect for \textsf{draft} mode in the % \textsf{article} class, plus \fbox{DRAFT} is printed in % the footer of every page. Note that if you use the \textsf{graphicx} % package and turn on the \textsf{draft} option in \textsf{ouexam}, all % included graphics will be drawn in draft mode unless you specify the % \|draft=false\| option to \cs{includegraphics}. % % \item[\textsf{markingschedule}] \DescribeMacro{markingschedule} % Rather than write a separate marking schedule for an examination paper, you % can use the \textsf{marking} environment to embed marking information % within questions (see % \hyperref[Sec:Questions:Marking]{section~\ref{Sec:Questions:Marking}}). % By default this information is not printed (for obvious reasons!), but % when it comes time to print a marking schedule for the examination, % including the \textsf{markingschedule} class option in the % \cs{usepackage} command will cause \textsf{ouexam} to print the hidden % content. It will also print \fbox{MARKING SCHEDULE} in the header of % every page. % % \item[\textsf{times}] \DescribeMacro{times} % The Examinations Office prefers all examination papers to be set in 12pt % Palatino. However, if this is not available, they will accept 13pt Times % instead. The \textsf{times} class option enables the latter, but should % only be used if Palatino is not available. % \end{description} % % % \subsection{Required packages} % % This class requires the \textsf{verbatim} package in order to implement % the marking schedule functionality. It also expects modern \LaTeX\ font % handling capabilities, and therefore requires the \textsf{fontenc}, % \textsf{textcomp} and \textsf{lmodern} classes. % % % \subsection{Page margins} % % This class assumes A4 paper. You will probably get weird results if you try % to do anything different. The margins are set up as follows: top and bottom % 2cm (headers and footers are inside this margin), left and right 2.54cm (1in). % % % \changes{2.1.1}{2006/08/21}{NJS Updated \textsf{lastpage} style to display % ``\textbf{END}''.} % \subsection{Page styles} % % There are three page styles defined in this class: % \begin{description} % \item[\textsf{plain}] \DescribeMacro{plain} This is a slight modification % of the \textsf{plain} page style from \textsf{article}. It produces pages % that have the page number centered at the bottom, the paper number in the % top right corner and the text ``\textbf{TURN OVER}'' in the bottom right. % This is the default page style. % % \item[\textsf{lastpage}] \DescribeMacro{lastpage} This is similar to % \textsf{plain} but with ``\textbf{END}'' instead of ``\textbf{TURN % OVER}''. It is used for the last page of the examination. You % normally will not have to use this yourself---the class should take % care of it automatically. The class does however occasionally seem % to get confused, so there may times when you have to set the page % style of the last page manually. It will be fairly obvious when you % need to do this---the usual effect is that the last page % has ``\textbf{TURN OVER}'' printed on it when it should not. % % \item[\textsf{titlepage}] \DescribeMacro{titlepage} This is similar to % \textsf{plain} but without the page header, and is used for the title page % of the paper. As with \textsf{lastpage}, you will normally not use this % yourself, because the \cs{maketitlepage} macro handles this automatically % (see \hyperref[Sec:TitlePage]{section~\ref{Sec:TitlePage}}). % \end{description} % % % \section{Writing an examination paper} % % % \changes{2.0.2}{2002/05/11}{NJS Moved the title page subsection to the start % of this section.} % \subsection{The title page} % \label{Sec:TitlePage} % % This class defines a collection of macros that let you fill in the various % parts of the examination title page. This is analogous to the process you use % to generate the title of a document in, for example, the \textsf{article} class % (using \cs{title}, \cs{author}, \cs{date} and \cs{maketitle}). That is, you % issue the macros described below in the document preamble, then issue a % \cs{maketitlepage} in the document body. % % The \DescribeMacro{\examyear} \cs{examyear} macro lets you specify the year % in which the examination is being held, for example, \|\examyear{1999}\|. This % macro is mandatory. % % The \DescribeMacro{\department} \cs{department} macro lets you specify the % name of the department that produced the examination paper, for example, % \|\department{Information\| \|Science}\|. This macro is mandatory. % % The \DescribeMacro{\papernumber} \cs{papernumber} macro lets you specify % the paper number that the examination is for, for example, % \|\papernumber{COMP 101}\|. This macro is mandatory. Note that when you % specify semester information using \cs{semester}, this is appended to % the paper number in the page header. For example, if you specify % \|\papernumber{COMP 101}\| and \|\semester{2}\|, the header will contain % ``COMP 101 (Semester Two)'' rather than just ``COMP 101''. % % The \DescribeMacro{\papertitle} \cs{papertitle} macro lets you specify the % title of the paper, for example, % \|\papertitle{Systems Analysis and Design Methods}\|. This macro is mandatory. % % \changes{2.0.2}{2002/08/22}{NJS Added `FY' to the \cs{semester} macro % description.} % Some papers are offered in more than one semester. The % \DescribeMacro{\semester} \cs{semester} macro lets you specify which % semester the examination is for. The argument can be either ``1'' or ``2'' % for semesters 1 and 2, ``SS'' for summer school, ``FY'' for a full-year % paper or ``SP'' for a special examination, for example, \|\semester{2}\|. % Invalid values for the argument cause a warning to be raised. This macro is % optional---if you omit it, no semester information is generated. % % The \DescribeMacro{\timeallowed} \cs{timeallowed} macro lets you specify the % length of the examination in hours, for example, \|\timeallowed{2}\|. This % macro is optional---if you omit it, it defaults to three (3) hours. % % \changes{2.1}{2004/04/05}{NJS Modified \cs{allowcalculators} to conform % to the new University calculator regulations.} % \changes{2.1}{2004/04/05}{NJS Added \cs{permitcalculators} as a synonym % for \cs{allowcalculators}.} % If calculators are permitted in the examination, use the % \DescribeMacro{\allowcalculators} % \DescribeMacro{\permitcalculators} % \cs{allowcalculators} macro (\cs{permitcalculators} will also work). % This macro is optional---if you omit it, a sentence is inserted saying % that calculators are \emph{not} permitted. \cs{allowcalculators} has a % single optional argument that specifies the kind of calculators % permitted. If omitted, it defaults to ``any'' (see below). Otherwise, % this argument must be one of the following values: % \begin{description} % % \item[\textsf{none}] A sentence is inserted noting that no calculators are % permitted. % % \item[\textsf{approved}] A sentence is inserted noting that only approved % calculators are permitted. % % \item[\textsf{any}] (default) A sentence is inserted noting that any % calculator that does not have a communication capability is % permitted. % % \end{description} % % The \DescribeMacro{\instructions} \cs{instructions} macro lets you specify % instructions on how candidates should complete the % examination, for example, \|\instructions{Answer ALL\| \|questions.}\|. This macro % is optional. Note that the content of the instructions can be just about % anything. It is up to you to control formatting, such as how you want lines % broken, etc. This also true of the \cs{material}, \cs{copiesof} and % \cs{otherinstructions} macros described below. % % The \DescribeMacro{\material} \cs{material} macro lets you specify any % additional material that candidates are provided in addition to the % examination paper itself, for example, \|\material{SQL schema definition}\|. % This macro is optional. % % The \DescribeMacro{\copiesof} \cs{copiesof} macro lets you specify any % material that candidates are allowed to bring into the examination, for % example, \|\copiesof{McFadden &\| \|Hoffer, 5th edition}\|. This macro is % optional. % % The \DescribeMacro{\otherinstructions} \cs{otherinstructions} macro lets you % specify any other instructions not covered by any of the above, for example, % \|\otherinstructions{No smoking\| \|allowed.}\|. This macro is optional. % % Once you have specified the content of the title page using the above % macros, you simply issue a \DescribeMacro{\maketitlepage} \cs{maketitlepage} % to generate the front page of the examination (cf. \cs{maketitle}). The title % page generated meets University formatting requirements. % Note that the number of pages in the examination paper is generated % and inserted into the title page automatically---you do not need to specify % it manually. % % % \subsection{Sections} % \label{Sec:Sections} % % Examination papers may optionally have multiple sections, ``numbered'' A, B, \ldots. Rather than redefine the existing section macros, this class defines an \DescribeEnv{examsection} \textsf{examsection} environment that generates a new section. The \textsf{examsection} environment has three mandatory arguments: % \begin{enumerate} % \item The expected number of marks for the question. If you leave this % argument empty it defaults to zero. \textsf{ouexam} keeps a running total % of the \emph{actual} number of marks encountered within a section, which is % compared with the value of this argument when the environment is closed. % An error is raised if the values are not equal. The running total is also % typeset right-justified at the end of the section in the form % ``\textbf{[SECTION A TOTAL 20 MARKS]}'' as an additional verification. % % \item Instructions for answering questions in this section (such as % ``Answer any TWO questions.''). If you leave this argument empty it % defaults to ``ANSWER ALL QUESTIONS.''. % % \item A description of the topic of the section. This can be left % empty. % \end{enumerate} % Every section begins on a new page, and the section title is formatted as % ``\textbf{\underline{Section A}}'', in \cs{large} size. For example: \\ % % \noindent\|\begin{examsection}{25}{}{These questions are remarkably boring.}\| \\ % \hspace{1cm}\vdots \\ % \|\end{examsection}\| \\ % % \noindent will produce the following (assuming that the actual number of % marks in the question is correct): % % \begin{center} % \fbox{% % \begin{minipage}{0.9\columnwidth} % {\large\noindent\textbf{\underline{Section~A}}} \\[0.5\baselineskip] % ANSWER ALL QUESTIONS. \\ % These questions are remarkably boring. \\ % \hspace{2cm}\vdots \\ % \mbox{}\hfill\textbf{[SECTION A TOTAL 25 MARKS]} % \end{minipage} % } % \end{center} % % % \subsection{Questions} % \label{Sec:Questions} % % \textsf{ouexam} provides three environments for building examination % questions: \DescribeEnv{question} \textsf{question}, % \DescribeEnv{subquestion} \textsf{subquestion} and \textsf{subsubquestion}, % which correspond to top-level questions, parts of questions, and % \DescribeEnv{subsubquestion} sub-parts of questions respectively. They % produce questions that are numbered according to University examination % formatting requirements. Note that questions are normally numbered % sequentially throughout the entire paper regardless of any section % boundaries. % % All three environments have a single mandatory argument which is the % expected number of marks for the question, part or sub-part. If this % argument is left empty it will default to zero. This argument works in much % the same way as the first argument to the \textsf{examsection} environment % (see \hyperref[Sec:Sections]{section~\ref{Sec:Sections}}): \textsf{ouexam} % keeps a running total of the number of marks encountered within each % question part and sub-part, and compares this total with the expected value % when the environment closes. Where appropriate, the running total is % typeset right-justified in the form ``(5 marks)''. For example: % \begin{verbatim} % \begin{question}{5} % \begin{subquestion}{2} % Why is the sky blue? % \end{subquestion} % \begin{subquestion}{3} % Explain in detail how the sky can be made pink. % \end{subquestion} % \end{question} % \begin{question}{1} % Define the term ``floccinaucinihilipilification''. % \end{question} % \end{verbatim} % % \noindent will produce the following: % \begin{center} % \fbox{% % \begin{minipage}{0.9\columnwidth} % \begin{enumerate} % \item % \begin{enumerate} % \item Why is the sky blue? \hfill (2 marks) % % \item Explain in detail how the sky can be made pink. \hfill (3 % marks) % \end{enumerate} % % \item Define the term ``floccinaucinihilipilification''. \hfill (1 % mark) % \end{enumerate} % \end{minipage} % } % \end{center} % % The \textsf{question} environments handle formatting of marks totals % automatically, including considerations such as: % \begin{itemize} % \item whether or not to print the number of marks for a question at all % (they should not be printed if the question has sub-parts); % % \item whether to print ``mark'' or ``marks'' (e.g., question 2 above); and % % \item where to position the number of marks relative to the question, % depending on how full the last line of the question is. % \end{itemize} % % \changes{2.3.1}{2013/05/31}{NJS Enforced proper question environment nesting.} % The various \textsf{question} environments must be correctly nested in order for the features described above to work correctly, i.e., \textsf{question} at the top level, followed by \textsf{subquestion} and \textsf{subsubquestion}. Any variation to this nesting order will cause an error. % % % \changes{2.2}{2010/04/26}{NJS Added time allocation macros.} % \subsection{Time allocations} % \label{Sec-Questions-Time} % % If you wish to suggest a time allocation for completing a question, use the \DescribeMacro{\timeallocation} \cs{timeallocation} macro. The only argument is the number of minutes to be allocated for this question. The macro will output the time allocation in the form: \textbf{[The suggested time allocation for this question is \emph{t} minutes.]} % % If you wish to allow a certain amount of time for candidates to read some text, use the \DescribeMacro{\readingtime} \cs{readingtime} macro. The only argument is the number of minutes to be allocated for reading time. The macro will output the reading time in the form: \textbf{[Allow \emph{t} minutes reading time.]} % % Both macros assume the base unit to be minutes; if you wish to change this you will need to redefine the macros. Otherwise, the argument can be almost anything: text, numbers, etc. % % \textsf{ouexam} currently does not check that time allocations sum correctly. This may be implemented in a future version. % % % \subsection{Marking schedule information} % \label{Sec:Questions:Marking} % % Rather than writing a separate marking schedule for an examination paper, % you can use the \DescribeEnv{marking} \textsf{marking} environment to embed % marking information within questions. By default this information is not % printed (for obvious reasons!), but when it comes time to print a marking % schedule for the examination, use the \textsf{markingschedule} class option % to print the hidden content. The marking information will be set in % \emph{italics} to differentiate it from the main body of the examination, % and \fbox{MARKING SCHEDULE} will also be printed in the header of every % page. % % \changes{2.1}{2004/04/10}{NJS Added driver file tip.} % A useful tip for writing examination papers is to create two separate % ``driver'' documents that are set up the document class for the % ``plain'' and ``marking schedule'' versions of the examination, % respectively. Each should then include a separate document that contains % the actual content of the examination paper, for example: % % \begin{verbatim} % \documentclass[markingschedule]{ouexam} % % \input{exampaper} % contains the actual examination content % \end{verbatim} % % Doing this enables you to build the ``plain'' and ``marking schedule'' % versions independently, without overwriting each other or having to % change the original source file. Simply run \LaTeX\ on the appropriate % driver file to produce the version that you want. See the example files % that came with the \textsf{ouexam} distribution for an example of this % approach in action. % % % \subsection{Miscellaneous} % % Most examinationss are marked out of 100, and \textsf{ouexam} defaults to % this. However, if you want an examination that is marked out of some other % number, for example, 90 marks, you can specify this using the % \DescribeMacro{\examoutof} \cs{examoutof} macro. Thus, \|\examoutof{90}\| % will set the expected number of marks for the examination to 90. This value % is used by \textsf{ouexam} to verify that the marks for all the questions % add up to the expected number of marks for the examination. % % % \section{Errors and warnings} % % In this section are described the error messages and warning produced by % \textsf{ouexam}, and the reasons why they occur. % % % \subsection{Error messages} % % \noindent\texttt{INCOMPATIBLE (\ldots): this document was written for an % earlier} \\ % \texttt{version of ouexam \ldots} \\ % You are trying to use v2.0 or later of \textsf{ouexam} with a document that was written for \textsf{ouexam} v1.2 or earlier. Versions 2.0 and later of \textsf{ouexam} are \emph{fundamentally incompatible} with earlier versions. The only solutions here are either to revert to \textsf{ouexam} v1.2 or earlier (you can put a copy of the class file in the same directory as the document), or rewrite the document to conform to the current version of \textsf{ouexam}. You should be able to download a copy of v1.2 from the same location that you found the current version. \\ % % \noindent\texttt{no \cs{examyear} was specified} \\ % \noindent\texttt{no \cs{department} was specified} \\ % \noindent\texttt{no \cs{papernumber} was specified} \\ % \noindent\texttt{no \cs{papertitle} was specified} \\ % You have not specified one of these macros. All four of these macros are mandatory and must be included in the preamble of any \textsf{ouexam} document. \\ % % % \noindent\texttt{question environment nested inside another question, subquestion or \\ % \hspace{1em}subsubquestion environment} \\ % \noindent\texttt{subquestion environment either outside a question environment, or \\ % \hspace{1em}nested inside another subquestion or subsubquestion environment} \\ % \noindent\texttt{subsubquestion environment either outside a subquestion environment, \\ % \hspace{1em}or nested inside another subsubquestion environment} \\ % These errors all mean that you have incorrectly nested your \textsf{question} environments in some way. \textsf{question}s must always appear at the outermost level, \textsf{subquestion}s must always appear inside \textsf{question}s, and \textsf{subsubquestion}s must always appear inside \textsf{subquestion}s. Any other kind of nesting is an error. \\ % % % \subsection{Warnings} % % \noindent\texttt{actual number of marks for exam (\ldots) does not match expected \\ % \hspace{1em}number of marks (\ldots)} \\ % The actual number of marks supplied for the whole examination (calculated by summing the marks for all the questions) is not the same as the expected total for the examination. Adjust the marks for the questions until the totals match. \\ % % \noindent\texttt{actual mark (\ldots) for section \ldots\ does not match expected \\ % \hspace{1em}mark (\ldots)} \\ % The actual number of marks supplied for a particular section (calculated by summing the marks for all the questions in that section) is not the same as the expected total for that section. Adjust the marks for the questions in the section concerned until the totals match. The section number is given in the warning. \\ % % \noindent\texttt{actual mark (\ldots) for question \ldots\ does not match expected \\ % \hspace{1em}mark (\ldots)} \\ % The actual number of marks supplied for a particular question (calculated by summing the marks of all its sub-parts) is not the same as the expected total for that question. Adjust the marks for the sub-parts of the question until the totals match. The question number is given in the warning. \\ % % \changes{2.0.2}{2002/08/22}{NJS Added `FY' to \cs{semester} error description.} % \noindent\texttt{invalid value `\ldots' for \cs{semester}; valid values are `1', `2', \\ % \hspace{1em}`SS', `FY' and `SP'. No semester information will be printed} \\ % You have provided an invalid value as the argument to the \cs{semester} macro. The allowed values are ``1'' (semester one), ``2'' (semester two), ``SS'' (summer school), ``FY'' (full-year) and ``SP'' (special examination). Any other values will be ignored. Note that these are case-sensitive, for example, `sp' is invalid. \\ % % \changes{2.1}{2004/04/05}{NJS Added \cs{allowcalculators} error description.} % \noindent\texttt{invalid value `\ldots' for \cs{allowcalculators}; valid values are \\ % \hspace{1em}`none', `any' and `approved'. No calculators will be permitted} \\ % You have provided an invalid value as the argument to the \cs{allowcalculators} macro. The allowed values are ``none'' (no calculators permitted), ``any'' (any calculator permitted) and ``approved'' (only approved calculators permitted). Any other values will be ignored, and \textsf{ouexam} will assume that no calculators are permitted. Note that these are case-sensitive, for example, `NONE' is invalid. \\ % % % \section{Example} % % \changes{2.1}{2004/04/10}{NJS Updated example to conform to 2.1 changes.} % \changes{2.0}{2002/01/25}{NJS Updated example to conform to 2.0 changes.} % \changes{1.1}{1999/04/20}{NJS Updated example to conform to 1.1 changes.} % The code given below is for the first three pages of the INFO~321 2001 % final examination. The output produced by this source is shown in % \hyperref[Fig.Example]{figure~\ref{Fig.Example} on % page~\pageref{Fig.Example}}. Particular points to note in this example % are: % \begin{itemize} % % \item The number of hours is not specified, so it defaults to three. % % \item This examination allows only approved calculators. % % \item The \textsf{listings} package (v1.1 or later) is used to % typeset code listings. % % \item The argument of the \cs{instructions} macro can be just about % anything, as can the arguments of the \cs{material}, \cs{copiesof} and % \cs{otherinstructions} macros. In this example, the \cs{instructions} % macro has been omitted, causing \textsf{ouexam} to output the default % instructions. % % \item Marking schedule information has been embedded within the % questions. \hyperref[Fig.Example2]{Figure~\ref{Fig.Example2} on % page~\pageref{Fig.Example2}} shows the output produced by this source % when the \textsf{markingschedule} class option is used. % % \end{itemize} % The full source for this example may be found in the % \texttt{example.tex} files that came with this distribution. % % \small % \begin{verbatim} % \documentclass{ouexam} % % \usepackage{graphicx} % \usepackage{listings} % % \examyear{2001} % \department{Information Science} % \papernumber{INFO 321} % \papertitle{Database Systems} % \allowcalculators[approved] % % \begin{document} % % \lstloadlanguages{ODL,OQL,[Oracle8]SQL,[Oracle8]PLSQL} % \lstset{basicstyle=\sffamily, showstringspaces=false, tabsize=2, % xleftmargin=1cm, belowskip=0pt, commentstyle=\itshape, % numbers=left, numberstyle=\scriptsize, numbersep=5pt, % columns=fullflexible,} % % \maketitlepage % % \begin{examsection}{25}{}{Questions in this section (total 25 marks) % relate to physical database design and tuning.} % % % % \begin{question}{10} % % A major U.S. mail order firm is having performance problems with queries % on the \textsf{Customer} table in its Oracle8 order-processing database. % Consider the following information about the \textsf{Customer} table, % then answer the questions below: \\ % % \textsf{\textbf{Customer}(\underline{customer\_no}, name, address, city, % state, country, phone, email, status)} % % \begin{itemize} % \item \textsf{customer\_no} is the primary key and because this is an % Oracle8 database, a unique B-tree index has been automatically created % on this column. % % \item \textsf{status} is a single character column that holds one of % the values `I' (inactive), `B' (bronze), `S' (silver) or `G' (gold). % The value of this column is determined by a customer's order history. % % \item About 10\% of customers are inactive, about 50\% have bronze % status, about 35\% have silver status and about 5\% have gold status. % % \item The average row size is about 200 bytes and there are % approximately 5,000,000 rows. % % \item The \textsf{Customer} table is heavily queried (averaging thirty % new queries per minute) by four different groups of employees. Group~1 % queries only inactive customers, group~2 queries only bronze status % customers and groups~3 and~4 both query only silver and gold status % customers. Members of any group may query the table at any time and % queries often overlap. % % \item UPDATE, INSERT and DELETE operations on the table are uncommon. % % \item Very few queries are for individual customers; rather, most % queries are based on various combinations of values from % \textsf{status}, \textsf{city}, \textsf{state} and \textsf{country}. % \end{itemize}\medskip % % \begin{subquestion}{3} % Describe and justify an \emph{index-based} physical tuning solution % that will improve the performance of queries on the \textsf{Customer} % table. (No code is required.) % \begin{marking} % Since queries tend to be on combinations of several low-cardinality % columns [1] and the table is very large [1], the most effective % solution would be to place bitmap indexes on status, city, state and % country [1]. % % B-tree indexes on any of the individual columns would not be ideal % because of their low cardinality. A possible alternative is a % composite B-tree index on all four columns [1 for suggesting this if % appropriate], but this is not ideal and would be much larger than % the four bitmap indexes [1 for this explanation if appropriate]. % Hashing is definitely \emph{not} an option, because very few queries % are exact match. % \end{marking} % \end{subquestion} % % \begin{subquestion}{7} % \begin{subsubquestion}{5}\label{nonindex} % Assuming that there are two identical fast disks available, describe % and justify two alternative \emph{non index-based} physical tuning % solutions that will improve the performance of queries on the % \textsf{Customer} table. For each solution include details of how % the two disks will be used. (No code is required.) % \begin{marking} % Alternative 1: partition the table across the two disks based on % status (four partitions, one for each status value) [1]. This will % allow simultaneous parallel access to different parts of the table % by the different groups [1]. Since two groups both access silver % and gold customers, it makes sense to put the silver customers on % disk one and the gold customers on disk two. Bronze customers % account for 50\% of the table, so it makes sense to also put them % on disk two (as the gold customers are the smallest group), and % put the inactive customers on disk one with the silver customers. % [1 for any sensible partitioning scheme] % % Alternative 2: replicate the table across both disks (one replica % on each disk) [1]. This will also provide simultaneous parallel % access to the data (hard to tell whether this would be better or % worse than the partitioning above) [1]. % % Note that any solution must be capable of being implemented in % Oracle8. % \end{marking} % \end{subsubquestion} % % \begin{subsubquestion}{2} % Briefly discuss the relative advantages and/or disadvantages of the % two alternative solutions you described in part~(\ref{nonindex}) % above. % \begin{marking} % The partitioning solution given above doesn't really help with the % bronze customer rows---we would need random horizontal % partitioning to really help with this, which Oracle8 doesn't % support. More generally, Oracle8 does not migrate rows into the % correct partition if a customer's status changes, so the data will % need to be reloaded occasionally [1]. % % For replication, we have the usual problem of keeping the replicas % synchronised if they are both read/write. This is somewhat % mitigated by the fact that there are only two copies [1]. % \end{marking} % \end{subsubquestion} % \end{subquestion} % \end{question} % % \newpage % \begin{question}{10} % There are two generic parameters that affect the structure of a B-tree % index. % % \begin{subquestion}{6} % Identify these two parameters and describe what each parameter % specifies. Use diagrams to illustrate your answer. % \begin{marking} % \emph{Node size} [1] specifies the maximum number of pointers in % each index node [1]. Suitable diagram [1]. % % \emph{Percentage fill} [1] specifies the number of pointers in each % index node that are allocated when a node is created [1]. In effect, % it determines the amount of free space left in each node. Suitable % diagram [1]. % \end{marking} % \end{subquestion} % % \begin{subquestion}{4} % For each of the two parameters, discuss the practical effect(s) of: % \begin{subsubquestion}{2} % high values of the parameter; and % \begin{marking} % Large node size reduces the height of the B-tree, resulting in % potentially fewer I/Os to access a leaf node [$\frac{1}{2}$]. It % can however increase the number of rows locked in a single leaf % node [$\frac{1}{2}$]. % % High percent fill means that most pointers in a node will be % allocated when the node is created [$\frac{1}{2}$]. This leaves % little free space in the B-tree for new key values or key values % that change [$\frac{1}{2}$]. % \end{marking} % \end{subsubquestion} % \begin{subsubquestion}{2} % low values of the parameter. % \begin{marking} % Small node size increases the height of the B-tree, resulting in % potentially more I/Os to access a leaf node [$\frac{1}{2}$]. The % number of rows locked in a single leaf node will be reduced % [$\frac{1}{2}$]. % % Low percent fill means that most pointers in a node will not be % allocated when the node is created [$\frac{1}{2}$]. This leaves % little free space in the B-tree for new key values or key values % that change [$\frac{1}{2}$]. % \end{marking} % \end{subsubquestion} % \end{subquestion} % \end{question} % % % % \begin{question}{5}\label{clusterq} % % \begin{subquestion}{2} % Define what is meant by the term ``clustering'' in the context of % physical database design and tuning, and explain why it is beneficial. % % \begin{marking} % Clustering is the grouping of related rows physically near to each % other on disk [1] so that they may be retrieved in as few I/Os as % possible (ideally one)---beneficial because of the slowness of disk % I/O compared to memory I/O [1]. % \end{marking} % \end{subquestion} % % \begin{subquestion}{3}\label{clusterqcode} % Consider the following Oracle8 SQL table definition: % % \begin{lstlisting}[language={[Oracle8]SQL}]{} % create table enrolment % ( student_id char(7), % paper_code char(7), % enrol_year number(4), % % primary key (student_id, paper_code, enrol_year) % ); % \end{lstlisting} % % The \textsf{Enrolment} table is often queried for all rows relating to % a particular student in a particular year (for example, student % 1234567 in 1998, student 9876543 in 2000, etc.). Write appropriate % Oracle8 SQL code that will cluster the rows of the table in such a way % as to support these types of query (you do not need to calculate % cluster sizes). Relevant Oracle8 syntax diagrams are given on the next % page. % \begin{marking} % \begin{lstlisting}[language={[Oracle8]SQL}]{} % create cluster enrolment_cluster (student_id char(7), year number(4)); % % create index enrol_cluster_index on cluster enrolment_cluster; % % create table enrolment % ( student_id char(7), % paper_code char(7), % enrol_year number(4), % % primary key (student_id, paper_code, enrol_year) % ) cluster enrolment_cluster(student_id, enrol_year); % \end{lstlisting} % \end{marking} % [1] each for correct cluster, cluster index and table clustering. % \end{subquestion} % % \end{question} % % \end{examsection} % . % . % . % \end{document} % \end{verbatim} % \normalsize % % \begin{figure} % \fbox{\includegraphics[scale=0.28]{eg1-1}} % \hfill % \fbox{\includegraphics[scale=0.28]{eg1-2}} % % \begin{center} % \fbox{\includegraphics[scale=0.28]{eg1-3}} % \end{center} % % \caption{Output produced by \textsf{ouexam}.} % \label{Fig.Example} % \end{figure} % % \begin{figure} % \fbox{\includegraphics[scale=0.28]{eg2-2}} % \hfill % \fbox{\includegraphics[scale=0.28]{eg2-3}} % % \bigskip % % \fbox{\includegraphics[scale=0.28]{eg2-4}} % \hfill % \fbox{\includegraphics[scale=0.28]{eg2-5}} % % \caption{Output produced by \textsf{ouexam} when % \textsf{markingschedule} is used.} % \label{Fig.Example2} % \end{figure} % % % \StopEventually{} % % \changes{2.0}{2002/01/10}{NJS Revamped and tided up documentation and code.} % % \section{The code} % % \subsection{Preamble} % % \begin{macrocode} \NeedsTeXFormat{LaTeX2e}[1998/06/01] \ProvidesClass{ouexam}[\ouexamshortdate\space % v\ouexamversion\space Otago % University examination paper] % \end{macrocode} % % \begin{macro}{\@unsupported} % \changes{2.0}{2002/01/10}{NJS New \cs{@unsupported} macro.} % This macro is used to handle documents written for older versions of % \textsf{ouexam}. All obsolete macros map to this macro, which prints out an % appropriate error message. It needs to be defined early because it is used % to trap use of the obsolete \textsf{multichoice} class option. % % \begin{macrocode} \def\@unsupported#1#2{\ClassError{ouexam}{% ^^JINCOMPATIBLE (#1)^^J% This document was written for an earlier version of ouexam and is^^J% not compatible with ouexam v\ouexamversion. Please use ouexam v#2 or earlier^^J% to process this document. Version #2 can be downloaded from the^^J% same location as version\space\ouexamversion}} % \end{macrocode} % \end{macro} % % \begin{macro}{multichoice} % \changes{1.1}{1999/04/20}{NJS New \textsf{multichoice} class option.} % \changes{2.0}{2000/09/04}{NJS Removed support for \textsf{multichoice} option.} % The \textsf{multichoice} class option is no longer supported, but rather than % just breaking older documents, we can at least try to present a reasonable % error message. % % \begin{macrocode} \DeclareOption{multichoice}{% \OptionNotUsed\@unsupported{multichoice class option}{1.2}% } % \end{macrocode} % \end{macro} % % \begin{macro}{twoside} % \changes{1.1}{1999/04/20}{NJS Turned off \textsf{twoside} option.} % This document class is based on the \textsf{article} class and accepts any of % the options accepted by \textsf{article}. The \textsf{twoside} option does % not really make sense, however---all Otago examination papers are printed % single-sided anyway, and the format is such that two-sided printing would % look no different. The \textsf{twoside} option is therefore not used in this % class. % % \begin{macrocode} \DeclareOption{twoside}{\OptionNotUsed} % \end{macrocode} % \end{macro} % % \begin{macro}{10pt} % \begin{macro}{11pt} % \changes{2.3}{1999/04/20}{NJS Turned off \textsf{10pt} and \textsf{11pt} % options.} % The Examinations Office specifies a minimum of 12pt for the font size, so the % \textsf{10pt} and \textsf{11pt} options are not used in this class. % % \begin{macrocode} \DeclareOption{10pt}{\OptionNotUsed} \DeclareOption{11pt}{\OptionNotUsed} % \end{macrocode} % \end{macro} % \end{macro} % % \begin{macro}{draft} % \changes{2.0}{2001/05/15}{NJS New \textsf{draft} class option.} % \begin{macro}{\if@draft} % \changes{2.0}{2001/05/15}{NJS New switch for \textsf{draft}.} % The \textsf{draft} option determines whether the examination is to be printed % in draft mode (default off). The \cs{if@draft} switch is used to determine % whether this is in effect (default false). Draft mode has the usual effects % that you would expect for draft mode in the \textsf{article} class, plus % \fbox{DRAFT} is printed in the footer of every page. % % \begin{macrocode} \newif\if@draft \@draftfalse \DeclareOption{draft}{\@drafttrue\PassOptionsToClass{\CurrentOption}{article}} % \end{macrocode} % % Note that if you use the \textsf{graphicx} package and turn on the % \textsf{draft} option in \textsf{ouexam}, all included graphics will % be drawn in draft mode unless you specify the \|draft=false\| option to the % \cs{includegraphics} macro. The amount of effort required to fix what is % a relatively small issue isn't really worth it, so this will not change. % \end{macro} % \end{macro} % % \begin{macro}{markingschedule} % \changes{2.0}{2000/09/05}{NJS New \textsf{markingschedule} class option.} % \begin{macro}{\if@markingschedule} % \changes{2.0}{2000/09/05}{NJS New switch for \textsf{markingschedule}.} % The \textsf{markingschedule} option determines whether marking schedule % information is printed in addition to the questions (default off). The % \cs{if@markingschedule} switch is used to determine whether this is in effect % (default false). % % \begin{macrocode} \newif\if@markingschedule \@markingschedulefalse \DeclareOption{markingschedule}{\@markingscheduletrue} % \end{macrocode} % \end{macro} % \end{macro} % % \begin{macro}{times} % \changes{2.3}{2012/09/03}{NJS New \textsf{times} class option.} % \begin{macro}{\if@times} % \changes{2.3}{2012/09/03}{NJS New switch for \textsf{times}.} % The \textsf{times} option formats the exam paper using 13pt Times. If this % option is not specified, then the exam paper is formatted using 12pt Palatino. % % \begin{macrocode} \newif\if@times \@timesfalse \DeclareOption{times}{\@timestrue} % \end{macrocode} % \end{macro} % \end{macro} % % All other options are passed directly to the \textsf{article} class: % % \begin{macrocode} \DeclareOption{\PassOptionsToClass{\CurrentOption}{article}} \ProcessOptions % \end{macrocode} % % \changes{1.1}{1999/04/20}{NJS Changed default point size to 12pt.} % The defaults for this class are \textsf{onecolumn}, \textsf{oneside}, % \textsf{a4paper} and \textsf{12pt}. The font size is the only one that might % need to change. % % \begin{macrocode} \LoadClass[onecolumn,oneside,a4paper,12pt]{article} % \end{macrocode} % % % \subsection{Required packages} % % \changes{2.0}{2001/05/08}{NJS Removed requirement for \textsf{calc} and % \textsf{ifthen} packages.} % The \textsf{verbatim} package, is used to implement the \textsf{marking} % environment: % % \begin{macrocode} \RequirePackage{verbatim} % \end{macrocode} % % \changes{2.3}{2012/09/04}{NJS Added requirement for \textsf{fontenc}, % \textsf{textcomp} and \textsf{lmodern} packages.} % % Running under vanilla OT1 encoding often causes all sorts of font substitution % warnings, so the \textsf{fontenc} (with \textsf{T1} option), \textsf{textcomp} % and \textsf{lmodern} packages are used to bring \textsf{ouexam}'s font % handling up to more modern standards. % % \begin{macrocode} \RequirePackage[T1]{fontenc} \RequirePackage{textcomp} \RequirePackage{lmodern} % \end{macrocode} % % % \changes{2.3}{2012/09/04}{NJS Added font setup for \textsf{times} class % option.} % \subsection{Font setup} % % The Examinations Office allows for two standard font styles: 12pt Palatino % (preferred) or 13pt Times New Roman. The former is the default; the latter is % only used if the \textsf{times} class option is used, and we need to % explicitly define all the standard sizes. The following is extrapolated from % the definitions in the \textsf{article} class and seems about right: % % \begin{macrocode} \if@times% \usepackage{mathptmx}% \renewcommand{\normalsize}{\fontsize{13}{15.7}\selectfont}% \renewcommand{\tiny}{\fontsize{\@viipt}{\@viiipt}\selectfont}% \renewcommand{\scriptsize}{\fontsize{\@ixpt}{11}\selectfont}% \renewcommand{\footnotesize}{\fontsize{\@xpt}{\@xiipt}\selectfont}% \renewcommand{\small}{\fontsize{\@xipt}{13.6}\selectfont}% \renewcommand{\large}{\fontsize{\@xivpt}{18}\selectfont}% \renewcommand{\Large}{\fontsize{\@xviipt}{22}\selectfont}% \renewcommand{\LARGE}{\fontsize{\@xxpt}{25}\selectfont}% \renewcommand{\huge}{\fontsize{\@xxvpt}{30}\selectfont}% \let\Huge=\huge% \else% \usepackage{mathpazo}% \fi% % \end{macrocode} % % % \subsection{Page setup} % % \begin{macro}{\oddsidemargin} % \begin{macro}{\topmargin} % \begin{macro}{\textwidth} % \begin{macro}{\textheight} % A4 paper is assumed (since this class is intended only for Otago examination % papers, this seems a reasonable assumption). The margins are top and bottom % 2cm, left and right 2.54cm (1in). % % \begin{macrocode} \setlength{\oddsidemargin}{0cm} \setlength{\topmargin}{-0.54cm} \setlength{\textwidth}{15.92cm} \setlength{\textheight}{25.7cm} \advance\textheight by-\headheight \advance\textheight by-\headsep \advance\textheight by-\footskip % \end{macrocode} % \end{macro} % \end{macro} % \end{macro} % \end{macro} % % % \subsection{Page styles}\label{pagestyles} % % \changes{2.0}{2001/05/17}{NJS Added support to page styles for \textsf{draft} and \textsf{markingschedule} class options.} % \changes{2.0}{2002/01/15}{NJS Fixed headers so that the page number is always centered.} % \changes{2.1.1}{2006/08/21}{NJS Updated \textsf{lastpage} style to display ``\textbf{END}''.} % \changes{2.3}{2012/09/03}{NJS Rearranged headers and footers to conform to the latest formatting requirements. Swapped \fbox{MARKING SCHEDULE} and \fbox{DRAFT} between header and footer so that the page number doesn't get overprinted.} % \changes{2.4}{2013/06/06}{NJS Replaced the \texttt{lastpage} page style with direct manipulation of the footer contents, as \cs{pagestyle} and \cs{thispagestyle} don't get applied to automatically generated float pages on the last page of the document.} % %\begin{macro}{\if@lastpagetext} %\begin{macro}{\if@lastpagefloat} % These two switches are used in some slightly convoluted logic to ensure that ``\textbf{END}'' is printed on the last page of the document (see below for details). % \begin{macrocode} \newif\if@lastpagetext \@lastpagetextfalse \newif\if@lastpagefloat \@lastpagefloatfalse % \end{macrocode} % \end{macro} % \end{macro} % % \begin{macro}{\ps@plain} % The \textsf{plain} page style is based on that in the \textsf{article} class, and produces pages with the page number centered at the bottom, the paper number in the top right corner and the text ``\textbf{TURN OVER}'' in the bottom right. If semester information is specified, this is included in parentheses after the paper code, for example, ``COMP 101 (Semester Two)''. \fbox{MARKING SCHEDULE} is printed at top left if the \textsf{markingschedule} class option is used. % \begin{macrocode} \def\ps@plain{% \def\@oddhead{\@markingschedule\hfill\@pnumber% \ifx\@semester\@empty\else\ (\@semester)\fi}% \let\@evenhead\@oddhead% \def\@oddfoot{% % \end{macrocode} % % Note that we have to get a bit clever to ensure that the page number is properly centered in the page footer. The naive solution is just to use \cs{hfill}s, but this will cause the page number to shift around depending on the length of \cs{@pnumber}. The \textsf{fancyhdr} package will do this for us, but it seems a bit of overkill to require this package merely so that we can ensure that the page number remains centered. Instead, we've borrowed the relevant technique from that class, which draws the left, centre and right bits of the footer as three overlapping boxes. You just have to be careful that the value of \cs{@pnumber} isn't so long that it overwrites the page number. % \begin{macrocode} \rlap{\parbox[b]{\columnwidth}{\raggedright\@draft\strut}}\hfill% \parbox[b]{\columnwidth}{\centering\textrm{\thepage}\strut}\hfill% \llap{\parbox[b]{\columnwidth}{\raggedleft\textbf{% % \end{macrocode} % % An additional complication is that only the very last page of the document should have ``\textbf{END}'' displayed in the footer instead of ``\textbf{TURN OVER}''. This turns out to be a lot harder than you might think! Simply applying a page style \cs{AtEndDocument} works fine, \emph{except} when you have unprocessed floats at the end of the document. These are placed on an automatically-generated float page at the end, which \cs{pagestyle} and \cs{thispagestyle} don't affect. That is, a \textsf{lastpage} page style simply won't work in this situation. % % The solution is therefore to not use a page style, but rather check in the \textsf{plain} page style whether or not this is the last page. Again, this isn't as simple as it sounds! The logic goes like this: % % \begin{itemize} % % \item If there are no unprocessed floats, meaning that the last page of the document will be a text page (indicated by \cs{@lastpagetexttrue}), then output ``\textbf{END}''. % % \item If there are unprocessed floats, meaning that the last page of the document will be a float page (indicated by \cs{@lastpagefloattrue}), then: % % \begin{itemize} % % \item If the float page is currently being generated (indicated by \cs{@fcolmadetrue} from \texttt{latex.ltx}), output ``\textbf{END}''. % % \item Otherwise, output ``\textbf{TURN OVER}''. % % \end{itemize} % % \item Otherwise, output ``\textbf{TURN OVER}''. % % \end{itemize} % % Note that this currently fails if more than one float page needs to be generated at the end of the document. \emph{All} such pages will have ``\textbf{END}'' printed at the bottom. % \begin{macrocode} \if@lastpagetext END% \else\if@lastpagefloat\if@fcolmade END% \else TURN OVER\fi% \else TURN OVER\fi% \fi}}}% }% \let\@evenfoot\@oddfoot% } % \end{macrocode} % \end{macro} % % \textsf{plain} is the default page style: % \begin{macrocode} \pagestyle{plain} % \end{macrocode} % % \cs{AtEndDocument}, we need to set the \texttt{lastpage} switches accordingly. The \cs{@deferlist} macro from \texttt{latex.ltx} holds a list of any unprocessed floats, so if this is empty, we know that there's nothing more to do and can just set \cs{@lastpagetexttrue}. If the list still contains unprocessed floats, set \cs{@lastpagefloattrue} then \cs{clearpage} to force the floats out. We also need to decrement the page counter by one, otherwise it ends up being one too many, presumably as a result of the \cs{clearpage}. % % I suspect that this will fail if the document uses \textsf{twocolumn} and has double floats that span both columns, because there is a separate \cs{@dbldeferlist}. Also, as noted above, this currently doesn't account for multiple float pages at the end of the document. % \begin{macrocode} \AtEndDocument{% \ifx\@deferlist\@empty% \@lastpagetexttrue% \else% \@lastpagefloattrue% \clearpage% \addtocounter{page}{-1}% \fi% } % \end{macrocode} % % \begin{macro}{\ps@titlepage} % The \textsf{titlepage} page style is similar to \textsf{plain}, but without the header. % \begin{macrocode} \def\ps@titlepage{% \def\@oddhead{\@markingschedule}% \let\@evenhead\@oddhead% \def\@oddfoot{\@draft\hfill\textbf{TURN OVER}}% \let\@evenfoot\@oddfoot% } % \end{macrocode} % \end{macro} % % % \subsection{Various counters} % % Earlier versions of \textsf{ouexam} used the \textsf{enumerate} % environment to build questions, which meant that special counters were not % required. This version of the class, however, uses a different mechanism for % building questions, so we need to define custom counters for several things. % % \begin{macro}{question} % \begin{macro}{subquestion} % \changes{2.0}{2000/09/11}{NJS New \texttt{subquestion} counter.} % \changes{2.0.1}{2000/04/30}{NJS Removed prefixed question number from % \cs{labelsubquestion} macro so that references now print as just ``(a)'' % instead of ``1(a)''.} % \begin{macro}{subsubquestion} % \changes{2.0}{2000/09/11}{NJS New \texttt{subsubquestion} counter.} % \changes{2.0.1}{2000/04/30}{NJS The \texttt{subsubquestion} was incorrectly % printing as ``(a)'' rather than ``(i)''. Fixed.} % \changes{2.0.1}{2000/04/30}{NJS Removed prefixed question numbers from % \cs{labelsubsubquestion} macro so that references now print as just ``(i)'' % instead of ``1(a)(i)''.} % First, we define three counters for the three possible levels of question: % top-level question, second-level sub-question or third-level sub-sub-question: % \begin{macrocode} \newcounter{question} \newcounter{subquestion}[question] \newcounter{subsubquestion}[subquestion] % \end{macrocode} % The three question levels are numbered as 1., (a) and (i) respectively, so we % also need to redefine the associated \cs{the} and \cs{label} macros % appropriately: % \begin{macrocode} \renewcommand{\thequestion}{\arabic{question}} \newcommand{\labelquestion}{\hfil\arabic{question}.} \renewcommand{\thesubquestion}{(\alph{subquestion})} \newcommand{\labelsubquestion}{\hfil(\alph{subquestion})} \renewcommand{\thesubsubquestion}{(\roman{subsubquestion})} \newcommand{\labelsubsubquestion}{\hfil(\roman{subsubquestion})} % \end{macrocode} % \end{macro} % \end{macro} % \end{macro} % % \begin{macro}{examexpected} % \changes{2.0}{2000/09/11}{NJS New \texttt{examexpected} counter.} % The \|examexpected\| counter stores the total expected number of marks for the % examination, and defaults to 100: % \begin{macrocode} \newcounter{examexpected} \setcounter{examexpected}{100} % \end{macrocode} % \end{macro} % \begin{macro}{examoutof} % \changes{2.0}{2000/09/11}{NJS New \texttt{examoutof} macro.} % To specify that an examination is out of some number of marks other than % 100, use the \cs{examoutof} macro to set the value: % \begin{macrocode} \newcommand{\examoutof}[1]{\setcounter{examexpected}{#1}} % \end{macrocode} % \end{macro} % % \begin{macro}{sectexpected} % \changes{2.0}{2002/01/09}{NJS New \cs{sectexpected} counter.} % \begin{macro}{qexpected} % \changes{2.0}{2000/09/11}{NJS New \cs{qexpected} counter.} % \begin{macro}{subqexpected} % \changes{2.0}{2000/09/11}{NJS New \texttt{subqexpected} counter.} % \begin{macro}{subsubqexpected} % \changes{2.0}{2000/09/11}{NJS New \texttt{subsubqexpected} counter.} % The next four counters perform the same function as the \|examexpected\| counter % for sections, questions, sub-questions and sub-sub-questions respectively. No % macros are required to set these values as they are set automatically by the % question-building environments. % % \begin{macrocode} \newcounter{sectexpected} \newcounter{qexpected} \newcounter{subqexpected} \newcounter{subsubqexpected} % \end{macrocode} % \end{macro} % \end{macro} % \end{macro} % \end{macro} % % \begin{macro}{lastexpected} % \changes{2.0}{2000/09/11}{NJS New \texttt{lastexpected} counter.} % The \|lastexpected\| counter records the expected marks total for the last % question. This will eventually be used to verify that the marks total for a % marking schedule equals the marks total for the corresponding question, but % this hasn't been implemented yet. % % \begin{macrocode} \newcounter{lastexpected} % \end{macrocode} % \end{macro} % % \begin{macro}{examrunning} % \changes{2.0}{2000/09/11}{NJS New \texttt{examrunning} counter.} % \begin{macro}{sectrunning} % \changes{2.0}{2002/01/09}{NJS New \texttt{sectrunning} counter.} % \begin{macro}{qrunning} % \changes{2.0}{2000/09/11}{NJS New \texttt{qrunning} counter.} % \begin{macro}{subqrunning} % \changes{2.0}{2000/09/11}{NJS New \texttt{subqrunning} counter.} % These four counters keep track of the running total of marks for the % current examination, section, question and subquestion respectively. This % is later compared against the corresponding expected total. A running total % is not needed for sub-sub-questions because they do not have sub-parts. The % \|qrunning\| and \|subqrunning\| counters are reset when the associated % question counters are incremented. % % \begin{macrocode} \newcounter{examrunning} \setcounter{examrunning}{0} \newcounter{sectrunning} \newcounter{qrunning}[question] \newcounter{subqrunning}[subquestion] % \end{macrocode} % \end{macro} % \end{macro} % \end{macro} % \end{macro} % % \begin{macro}{hassubs} % \changes{2.0}{2000/09/11}{NJS New \texttt{hassubs} counter.} % \begin{macro}{hassubsubs} % \changes{2.0}{2000/09/11}{NJS New \texttt{hassubsubs} counter.} % These two counters are used to track whether questions and % sub-questions have sub-parts. Counters are used rather than booleans because % counters are set globally and booleans are not (or at least do not appear to % be). Both counters reset when the associated question counters are % incremented. % % \begin{macrocode} \newcounter{hassubs}[question] \newcounter{hassubsubs}[subquestion] % \end{macrocode} % \end{macro} % \end{macro} % % \begin{macro}{qdepth} % \changes{2.3.1}{2013/05/31}{NJS New \texttt{qdepth} counter.} % This counter is incremented every time a \textsf{question}, \textsf{subquestion} or \textsf{subsubquestion} environment is opened, and decremented every time these environments close. It measures the nesting depth of questions, and is used to check that we do not have things like a \texttt{subquestion} not enclosed by a \texttt{question}, or a \texttt{question} nested inside another \texttt{question}. Note that this counter does \emph{not} reset inside any other environment. % % \begin{macrocode} \newcounter{qdepth} \setcounter{qdepth}{0} % \end{macrocode} % \end{macro} % % At the end of the document, check that the running total of marks for the % examination matches what we were expecting: % % \begin{macrocode} \AtEndDocument{% \ifnum\theexamexpected=\theexamrunning% \else\ClassWarning{ouexam}{% actual number of marks for exam (\theexamrunning) does not^^J% match expected number of marks (\theexamexpected)}% \fi% } % \end{macrocode} % % \begin{macro}{xsection} % \changes{2.0}{2002/01/29}{NJS New \texttt{xsection} counter.} % Finally, we set up a counter for the section ``number'' and define it so % that it prints out as an upper case letter rather than a number. We could % just redefine the \|section\| counter, but that interferes with the section % numbering in the documentation, and we can't have that, can we? \|:)\| % % \begin{macrocode} \newcounter{xsection} \setcounter{xsection}{0} \renewcommand{\thexsection}{\Alph{xsection}} % \end{macrocode} % \end{macro} % % % \subsection{Question-building environments and associated items} % % \changes{2.0}{2000/09/11}{NJS \textsf{question} environment no longer a % redefinition of \textsf{enumerate}.} % \changes{2.0}{2001/05/15}{NJS Added penalty handling to ensure correct % placement of marks in output.} % \changes{2.0}{2001/05/15}{NJS Total marks for a question are now printed only % if the question has no sub-questions.} % \begin{environment}{question} % \changes{2.3.1}{2013/05/31}{NJS Check question depth when opening the % environment.} % The \textsf{question} environment specifies a top-level question, and produces % questions numbered in the form ``1.'', ``2.'', etc. It has one mandatory % argument, the number of marks allocated to the question, which is used to % initialise the \|qexpected\| counter\footnote{If it were not for the fact that % you can only refer to environment arguments in the environment preamble, the % \texttt{qexpected} counter would be unnecessary.}. If the argument is left % empty, default to zero for the number of marks: % % \begin{macrocode} \newenvironment{question}[1]{% \def\@nummarks{#1}% \ifx\@nummarks\@empty\setcounter{qexpected}{0}% \else\setcounter{qexpected}{#1}\fi% % \end{macrocode} % % Check that the environment is being opened at the correct depth. The \textsf{question} environment should normally only be opened at depth 0: % \begin{macrocode} \ifnum\theqdepth=0% \else\ClassError{ouexam}{% question environment nested inside another question,^^J% subquestion or subsubquestion environment}% \fi% \addtocounter{qdepth}{1}% % \end{macrocode} % % For each top-level question, we increment the \|question\| counter, which also % resets the \|hassubs\| and \|qrunning\| counters to zero: % \begin{macrocode} \refstepcounter{question}% % \end{macrocode} % The question itself is built using a single-item \textsf{list} environment. % \begin{macrocode} \begin{list}{\labelquestion}{\settowidth{\labelwidth}{88.}}% \item \ignorespaces% % \end{macrocode} % % When the environment closes, the following things happen: % \begin{enumerate} % \item If the question has no sub-questions, the value of the \|qexpected\| % counter is added to \|examrunning\| and \|sectrunning\|, and both \|qrunning\| % and \|lastexpected\| are set to the value of \|qexpected\|. % % \begin{macrocode} }{% \ifnum\thehassubs=0% \addtocounter{examrunning}{\value{qexpected}}% \addtocounter{sectrunning}{\value{qexpected}}% \setcounter{qrunning}{\value{qexpected}}% % \end{macrocode} % % The total number of marks for the question is then printed right-justified % on the line as ``(\emph{m} marks)'' where \emph{m} is the value of the % \|qrunning\| counter. The environment determines whether to print ``mark'' or % ``marks'' automatically, and figures out whether the number of marks will fit % on the last line of the question or needs to be placed on the next line. % The code for handling the line breaking is derived from an example on page % 106 of \emph{The \TeX{}book}: % % \begin{macrocode} \unskip\nobreak\hfil\penalty50\hskip2em\hbox{}\nobreak% \hfil(\theqrunning~\ifnum\theqrunning=1 mark\else marks\fi)% \parfillskip=0pt \finalhyphendemerits=0 \par% % \end{macrocode} % % \item If the question \emph{does} have sub-questions, then \|qrunning\|, % \|examrunning\| and \|sectrunning\| have already been set by the various % sub-environments. The value of \|qrunning\| is then compared with % \|qexpected\|, and a warning is raised if they do not match. The total number % of marks for the question is \emph{not} printed in this case. % % \begin{macrocode} \else% \ifnum\theqrunning=\theqexpected% \else\ClassWarning{ouexam}{% actual mark (\theqrunning) for question % \thequestion\space doesn't match^^J% expected mark (\theqexpected)}% \fi% \fi% % \end{macrocode} % % \item \|lastexpected\| is set to the value of \|qexpected\| so it can be used % in any subsequent \textsf{marking} environment. % % \begin{macrocode} \setcounter{lastexpected}{\value{qexpected}}% \end{list}% % \end{macrocode} % % \item \|qdepth\| is decremented. % % \begin{macrocode} \addtocounter{qdepth}{-1}% } % \end{macrocode} % % \end{enumerate} % \end{environment} % % \begin{environment}{subquestion} % \changes{2.0}{2000/09/11}{NJS \textsf{subquestion} environment no longer a % redefinition of \textsf{enumerate}.} % \changes{2.0}{2001/05/13}{NJS Added penalty handling to ensure correct placement % of marks in output.} % \changes{2.3.1}{2013/05/31}{NJS Check question depth when opening the % environment.} % The \textsf{subquestion} environment is the analogue of \textsf{question} for % sub-questions. On initialisation, it sets \|subqexpected\| to the value of the % mandatory argument (defaulting to zero if the argument is empty), increments % the \|hassubs\| counter so that the enclosing \textsf{question} environment can % react appropriately, increments \|subquestion\| (thus resetting \|hassubsubs\| and % \|subqrunning\| to zero) and opens a single-item list with numbering of the form % ``(a)'', ``(b)'', etc. % % \begin{macrocode} \newenvironment{subquestion}[1]{% \def\@nummarks{#1}% \ifx\@nummarks\@empty\setcounter{subqexpected}{0}% \else\setcounter{subqexpected}{#1}\fi% \setcounter{subqexpected}{#1}% \addtocounter{hassubs}{1}% \refstepcounter{subquestion}% % \end{macrocode} % The \textsf{subquestion} environment should normally only be opened at depth 1: % \begin{macrocode} \ifnum\theqdepth=1% \else\ClassError{ouexam}{% subquestion environment either outside a question^^J% environment, or nested inside another subquestion or subsubquestion^^J% environment}% \fi% \addtocounter{qdepth}{1}% \begin{list}{\labelsubquestion}{\settowidth{\labelwidth}{(m)}}% \item \ignorespaces% }{% % \end{macrocode} % % When the environment closes, if performs similar checks to the % \textsf{question} environment. If the sub-question has no sub-sub-questions, % the value of \|subqexpected\| is added to \|qrunning\| and \|sectrunning\|: % % \begin{macrocode} \ifnum\thehassubsubs=0% \addtocounter{examrunning}{\value{subqexpected}}% \addtocounter{sectrunning}{\value{subqexpected}}% \addtocounter{qrunning}{\value{subqexpected}}% \setcounter{subqrunning}{\value{subqexpected}}% % \end{macrocode} % Then the number of marks for the sub-question are typeset in a similar manner % to the \textsf{question} environment: % \begin{macrocode} \unskip\nobreak\hfil\penalty50\hskip2em\hbox{}\nobreak% \hfil(\thesubqrunning~\ifnum\thesubqrunning=1 mark\else marks\fi)% \parfillskip=0pt \finalhyphendemerits=0 \par% % \end{macrocode} % % If the sub-question \emph{does} have sub-sub-questions, check the running total % against the expected number and raise an error if they don't match. % \begin{macrocode} \else% \ifnum\thesubqrunning=\thesubqexpected% \else\ClassWarning{ouexam}{% actual mark (\thesubqrunning) for question % \thequestion\thesubquestion\space doesn't match^^J% expected mark (\thesubqexpected)}% \fi% \fi% % \end{macrocode} % Finally, set \|lastexpected\| to the value of \|qexpected\| so it can be used % in any subsequent \textsf{marking} environment. % \begin{macrocode} \setcounter{lastexpected}{\value{subqexpected}}% \end{list}% \addtocounter{qdepth}{-1}% } % \end{macrocode} % \end{environment} % % \begin{environment}{subsubquestion} % \changes{2.0}{2000/09/11}{NJS \textsf{subsubquestion} environment no longer a % redefinition of \textsf{enumerate}.} % \changes{2.0}{2001/05/13}{NJS Added penalty handling to ensure correct placement % of marks in output.} % \changes{2.3.1}{2013/05/31}{NJS Check question depth when opening the % environment.} % This is similar to both \textsf{question} and \textsf{subquestion}, and % generates an appropriately numbered sub-sub-question (``(i)'', ``(ii)'', etc.). % \begin{macrocode} \newenvironment{subsubquestion}[1]{% \def\@nummarks{#1}% \ifx\@nummarks\@empty\setcounter{subsubqexpected}{0}% \else\setcounter{subsubqexpected}{#1}\fi% % \end{macrocode} % Increment \|hassubsubs\| so that the enclosing \textsf{subquestion} environment % can react appropriately: % \begin{macrocode} \addtocounter{hassubsubs}{1}% \refstepcounter{subsubquestion}% % \end{macrocode} % The \textsf{subsubquestion} environment should normally only be opened at depth 2: % \begin{macrocode} \ifnum\theqdepth=2% \else\ClassError{ouexam}{% subsubquestion environment either outside a subquestion^^J% environment, or nested inside another subsubquestion environment}% \fi% \addtocounter{qdepth}{1}% \begin{list}{\labelsubsubquestion}{\settowidth{\labelwidth}{(viii)}}% \item \ignorespaces% }{% % \end{macrocode} % % Closing this environment is a bit simpler than the previous two because we % don't need to check whether the sub-sub-question has any sub-components (this % never happens). All we have to do is set the appropriate counters and typeset % the number of marks. % % \begin{macrocode} \addtocounter{examrunning}{\value{subsubqexpected}}% \addtocounter{sectrunning}{\value{subsubqexpected}}% \addtocounter{qrunning}{\value{subsubqexpected}}% \addtocounter{subqrunning}{\value{subsubqexpected}}% \setcounter{lastexpected}{\value{subsubqexpected}}% \unskip\nobreak\hfil\penalty50\hskip2em\hbox{}\nobreak% \hfil(\thesubsubqexpected~\ifnum\thesubsubqexpected=1 mark% \else marks\fi)% \parfillskip=0pt \finalhyphendemerits=0 \par% \end{list}% \addtocounter{qdepth}{-1}% } % \end{macrocode} % \end{environment} % % % \subsection{Draft examination printing} % % \begin{macro}{\@marking} % \changes{2.0}{2001/05/17}{NJS New \cs{@draft} macro.} % Specifying the \textsf{draft} option causes \textsf{ouexam} to print % \fbox{DRAFT} in large letters in the footer of every page. This is drawn by % the \cs{@draft} macro, which is included in the page footer definition of % all the page styles for this class (see % \hyperref[pagestyles]{section~\ref{pagestyles}}). \textbf{Note:} Because % the \cs{@draft} macro is included in the page footer, it will usually % overflow the page boundaries, causing ``\|Overfull \vbox\|'' warnings. Note % the extra set of braces to limit the scope of the \cs{Huge}. % % \begin{macrocode} \if@draft\def\@draft{{\Huge\fbox{DRAFT}}} \else\let\@draft\@empty \fi % \end{macrocode} % \end{macro} % % % \changes{2.2}{2010/04/26}{NJS Added time allocation macros.} % \subsection{Time allocations} % % \begin{macro}{\timeallocation} % The \cs{timeallocation} macro takes a single argument representing the number of minutes suggested time allocation for a question, and outputs it in the form: \textbf{[The suggested time allocation for this question is \emph{t} minutes.]} The argument is not checked and can be anything (numbers, text). The macro will need to be redefined if either units other than minutes or different phrasing are required. % \begin{macrocode} \newcommand{\timeallocation}[1]{% \textbf{[The suggested time allocation for answering this question % is #1 minutes.]}% } % \end{macrocode} % \end{macro} % % \begin{macro}{\readingtime} % The \cs{readingtime} macro takes a single argument representing the number of minutes reading time allocated to some part of the examination, and outputs it in the form: \textbf{[Allow \emph{t} minutes reading time.]} The argument is not checked and can be anything (numbers, text). The macro will need to be redefined if either units other than minutes or different phrasing are required. % \begin{macrocode} \newcommand{\readingtime}[1]{\textbf{[Allow #1 minutes reading time.]}} % \end{macrocode} % \end{macro} % % \textsf{ouexam} currently does not check that any of these time allocations sum to the correct total. If implemented, this would work in a similar way to how marks totals are checked, and would be keyed in some way off \cs{@hours}. % % % \subsection{Marking schedule information} % % \begin{environment}{marking} % \begin{macro}{\@markingschedule} % \changes{1.2}{1999/10/26}{NJS New \textsf{marking} environment for including % answers and/or marking schedule information.} % \changes{2.0}{2000/09/11}{NJS Now prints ``MARKING SCHEDULE'' at the bottom % of every page if the \textsf{markingschedule} class option is specified.} % The \textsf{marking} environment allows the specification of marking % schedule information in the same location as the questions. If the % \textsf{markingschedule} class option is specified, marking schedule % information is printed in italics. In addition, \fbox{MARKING SCHEDULE} is % printed in large letters in the header of every page. This is drawn by the % \cs{@markingschedule} macro, which is included in the page header % definition of all the page styles for this class (see % \hyperref[pagestyles]{section~\ref{pagestyles}}). Note the extra set of % braces to limit the scope of the \cs{Huge}. % % \begin{macrocode} \if@markingschedule \newenvironment{marking}{\itshape}{\normalfont} \def\@markingschedule{{\Huge\fbox{MARKING SCHEDULE}}} \else % \end{macrocode} % % Normally \textsf{marking} just maps to the \textsf{comment} environment from % Rainer Sch\"{o}pf's \textsf{verbatim} package, i.e., marking information is not % printed: % % \begin{macrocode} \let\marking\comment \let\endmarking\endcomment \let\@markingschedule\@empty \fi % \end{macrocode} % \end{macro} % \end{environment} % % % \subsection{Section handling} % % \changes{2.0}{2002/01/09}{NJS new \textsf{examsection} environment.} % \changes{2.0}{2002/01/09}{NJS New \cs{@defsecinst} macro.} % \begin{environment}{examsection} % \begin{macro}{\@defsecinst} % The \textsf{examsection} environment specifies a major section of an % examination paper. Sections are sequentially numbered ``A'', ``B'', etc. The % environment has three arguments: % % \begin{enumerate} % \item The expected number of marks for the section. % % \item The instructions for this section. This defaults to ``ANSWER ALL % QUESTIONS.'' if left blank. To change the default text, redefine the % \cs{@defsecinst} macro. % \begin{macrocode} \def\@defsecinst{ANSWER ALL QUESTIONS.} % \end{macrocode} % % \item A description of the contents of the section which may be left blank. % \end{enumerate} % % \begin{macrocode} \newenvironment{examsection}[3]{% % \end{macrocode} % Every section begins on a new page. % \begin{macrocode} \newpage% % \end{macrocode} % If the argument for the number of marks is left empty, default to zero. % \begin{macrocode} \def\@nummarks{#1}% \ifx\@nummarks\@empty\setcounter{sectexpected}{0}% \else\setcounter{sectexpected}{#1}\fi% \refstepcounter{xsection}% \setcounter{sectrunning}{0}% % \end{macrocode} % The section title is formatted as ``\textbf{\underline{Section A}}'', in % \cs{large} size. % \begin{macrocode} {\large\noindent\textbf{\underline{Section~\thexsection}}}% \\[0.5\baselineskip]% \def\@usersecinst{#2}% \ifx\@usersecinst\@empty\@defsecinst\else\@usersecinst\fi\par% \def\@usersectopic{#3}% \ifx\@usersectopic\@empty\else\par\noindent\@usersectopic\fi \\% }{% % \end{macrocode} % % When the environment closes, the value of \|sectrunning\| is compared against % \|sectexpected\|, and a warning is raised if they do not match. The total number % marks for the section is printed in the form ``\textbf{[SECTION A TOTAL % \emph{m} MARKS]}'' where \emph{m} is the value of \|sectrunning\|. % % \begin{macrocode} \ifnum\thesectrunning=\thesectexpected% \else\ClassWarning{ouexam}{% actual mark (\thesectrunning) for section % \thexsection\space doesn't match^^J% expected mark (\thesectexpected)}% \fi% \bigskip\hfill\textbf{[SECTION \thexsection\ TOTAL % \thesectrunning\ MARKS]}% } % \end{macrocode} % \end{macro} % \end{environment} % % % \subsection{Title page generation} % % \begin{macro}{\examyear} % \begin{macro}{\@year} % \changes{2.0.2}{2002/08/22}{NJS Put back \@eha at end of the \ClassError. Don't know what this does, but the macro crashes without it there. It seems to be required if there's a \cs{protect}.} % The \cs{examyear} macro specifies the year in which the examination is being held. It redefines the \cs{@year} macro which is used in \cs{@maketitlepage}: % \begin{macrocode} \newcommand{\examyear}[1]{\def\@year{#1}} % \end{macrocode} % This macro is required: % \begin{macrocode} \def\@year{% \ClassError{ouexam}{no \protect\examyear\space was specified}\@eha% } % \end{macrocode} % \end{macro} % \end{macro} % % \begin{macro}{\department} % \begin{macro}{\@dept} % \changes{2.0.2}{2002/08/22}{NJS Put back \@eha at end of the \ClassError. Don't know what this does, but the macro crashes without it there. It seems to be required if there's a \cs{protect}.} % The \cs{department} macro specifies the name of the department, and is required. It redefines the \cs{@dept} macro which is used in \cs{@maketitlepage}: % \begin{macrocode} \newcommand{\department}[1]{\def\@dept{#1}} \def\@dept{% \ClassError{ouexam}{no \protect\department\space was specified}\@eha% } % \end{macrocode} % \end{macro} % \end{macro} % % \begin{macro}{\papernumber} % \begin{macro}{\@pnumber} % \changes{2.0.2}{2002/08/22}{NJS Put back \@eha at end of the \ClassError. Don't know what this does, but the macro crashes without it there. It seems to be required if there's a \cs{protect}.} % The \cs{papernumber} macro specifies the paper number, and is required. It redefines the \cs{@pnumber} macro which is used in \cs{@maketitlepage}: % \begin{macrocode} \newcommand{\papernumber}[1]{\def\@pnumber{#1}} \def\@pnumber{% \ClassError{ouexam}{no \protect\papernumber\space was specified}\@eha% } % \end{macrocode} % \end{macro} % \end{macro} % % \begin{macro}{\papertitle} % \begin{macro}{\@ptitle} % \changes{2.0.2}{2002/08/22}{NJS Put back \@eha at end of the \ClassError. Don't know what this does, but the macro crashes without it there. It seems to be required if there's a \cs{protect}.} % The \cs{papertitle} macro specifies the title of the paper, and is required. It redefines the \cs{@ptitle} macro which is used in \cs{@maketitlepage}: % \begin{macrocode} \newcommand{\papertitle}[1]{\def\@ptitle{#1}} \def\@ptitle{% \ClassError{ouexam}{no \protect\papertitle\space was specified}\@eha% } % \end{macrocode} % \end{macro} % \end{macro} % % \begin{macro}{\semester} % \begin{macro}{\@semester} % \changes{2.0.2}{2002/08/22}{NJS Added full-year support to \cs{semester}.} % \changes{2.0}{2002/01/14}{NJS Rewrote \cs{semester} to support summer school % and special exams.} % The \cs{semester} macro specifies which semester the examination is for. % Legal values for the argument are ``1'', ``2'', ``SS'' for summer school, % ``FY'' for full-year or ``SP'' for a special examination---anything else is % ignored. % \begin{macrocode} \def\@ONE{1} \def\@TWO{2} \def\@SS{SS} \def\@FY{FY} \def\@SP{SP} % \end{macrocode} % This macro redefines the \cs{@semester} macro which is used in % \cs{@maketitlepage}: % \begin{macrocode} \newcommand{\semester}[1]{% \def\@sem{#1}% \ifx\@sem\@ONE\def\@semester{Semester One}% \else\ifx\@sem\@TWO\def\@semester{Semester Two}% \else\ifx\@sem\@SS\def\@semester{Summer School}% \else\ifx\@sem\@FY\def\@semester{Full Year}% \else\ifx\@sem\@SP\def\@semester{Special Examination}% \else\ClassWarning{ouexam}{% invalid value `#1' for \protect\semester;^^J% valid values are `1', `2', `SS', `FY' and `SP'. No semester^^J% information will be printed}% \fi\fi\fi\fi\fi% } % \end{macrocode} % \cs{semester} is optional---if you omit it or give it an invalid argument, % \cs{@semester} remains empty. % \begin{macrocode} \let\@semester\@empty % \end{macrocode} % \end{macro} % \end{macro} % % \begin{macro}{\timeallowed} % \begin{macro}{\@hours} % The \cs{timeallowed} macro specifies the length of the examination in hours. % It redefines the \cs{@hours} macro which is used in \cs{@maketitlepage}: % \begin{macrocode} \newcommand{\timeallowed}[1]{\def\@hours{#1}} % \end{macrocode} % \cs{timeallowed} is optional---if you omit it, \cs{@hours} defaults to ``3'': % \begin{macrocode} \def\@hours{3} % \end{macrocode} % \end{macro} % \end{macro} % % \begin{macro}{num.pages} % \changes{1.1}{1999/04/20}{NJS Renamed \texttt{page.last} label to \texttt{num.pages}.} % \changes{2.4}{2013/06/05}{NJS Added check for \textsf{hyperref} being loaded.} % There is no macro to specify the number of pages in the examination because it is done automatically by inserting a \cs{newlabel} that refers to the last page directly into the \|.aux\| file. The \textsf{hyperref} package redefines the \cs{newlabel} macro (in particular it increases the number of arguments), so we need to test whether \textsf{hyperref} is loaded and write the appropriate string to the \|.aux\| file. We borrow a technique from the \textsf{pageslts} package of checking for the existence of \cs{Hy@Warning}. (Ideally we would use \cs{if@packageloaded}, but this can't be used after \cs{AtBeginDocument}, and can be foiled by a package loading \textsf{hyperref} \cs{AtBeginDocument}.) % \begin{macrocode} \AtEndDocument{% \@ifundefined{Hy@Warning}{% hyperref not loaded \immediate\write\@auxout{\string\newlabel{num.pages}{{}{\thepage}}}% }{% hyperref loaded \immediate\write\@auxout{\string\newlabel{num.pages}{{}{\thepage}{}{}{}}}% }% } % \end{macrocode} % This label is then referenced in \cs{@maketitlepage}. % \end{macro} % % \begin{macro}{num.questions} % \changes{1.1}{1999/04/20}{NJS Added \cs{num.questions} label representing the number of questions.} % \changes{2.4}{2013/06/05}{NJS Added check for \textsf{hyperref} being loaded.} % Similarly, there is no macro to specify the number of questions in the examination. This is calculated using the same method as for the number of pages. % \begin{macrocode} \AtEndDocument{% \@ifundefined{Hy@Warning}{% \immediate\write\@auxout{\string\newlabel{num.questions}{{}{\thequestion}}}% }{% \immediate\write\@auxout{\string\newlabel{num.questions}{{}{\thequestion}{}{}{}}}% }% } % \end{macrocode} % \end{macro} % % \begin{macro}{\instructions} % \begin{macro}{\@instructions} % The \cs{instructions} macro specifies instructions on how candidates should % answer questions. It redefines the \cs{@instructions} macro that is used in % \cs{@maketitlepage}: % \begin{macrocode} \def\@instructions{Answer \underline{ALL} questions.} % \end{macrocode} % \cs{instructions} is optional. If you omit it, \cs{@instructions} defaults % to ``Answer \underline{ALL} questions.'' % \begin{macrocode} \newcommand{\instructions}[1]{\def\@instructions{#1}} % \end{macrocode} % \end{macro} % \end{macro} % % \begin{macro}{\@NA} % \changes{2.1.3}{2009/04/28}{NJS Added \cs{@NA} macro.} % The \cs{@NA} macro enables us to easily assign and test for the default value % (``N/A'') for some of the title page macros. (There is certain text that % should not be output when particular macros are allowed to default.) % \begin{macrocode} \def\@NA{N/A} % \end{macrocode} % \end{macro} % % \begin{macro}{\material} % \begin{macro}{\@material} % \changes{2.1.3}{2009/04/28}{NJS Changed default to ``N/A'' for \cs{material}.} % The \cs{material} macro lets specifies any additional material that candidates % are provided in addition to the examination paper itself, and is optional. % It redefines the \cs{@material} macro that is used in \cs{@maketitlepage}: % \begin{macrocode} \let\@material\@NA \newcommand{\material}[1]{\def\@material{#1}} % \end{macrocode} % If you omit it, it defaults to ``N/A''. % \end{macro} % \end{macro} % % \begin{macro}{\allowcalculators} % \begin{macro}{\@calculators} % \changes{2.1}{2004/04/05}{NJS Modified calculator macros to conform to % new University calculator regulations.} % The \cs{allowcalculators} macro specifies whether or not calculators are % allowed in the examination (\cs{permitcalculators} is a synonym). If % \cs{allowcalculators} is omitted, \cs{@calculators} defaults to ``No % calculators are permitted.'': % \begin{macrocode} \def\@calculators{No calculators are permitted.} % \end{macrocode} % Issuing an \cs{allowcalculators[none]} does nothing, as the default is % to disallow calculators anyway. Issuing an \cs{allowcalculators} or % \cs{allowcalculators[any]} redefines \cs{@calculators} to ``No % restriction on the model of calculator to be used, but no device with % communication capability shall be accepted as a calculator.'' Issuing an % \cs{allowcalculators[approved]} redefines \cs{@calculators} to ``Only % calculators on the University of Otago list of approved calculators are % permitted.'': % \begin{macrocode} \def\@CalcNONE{none} \def\@CalcANY{any} \def\@CalcAPPROVED{approved} % \end{macrocode} % \changes{2.1.2}{2008/08/14}{NJS Ensured that the correct default is set % regardless of whether the calculators macro is called.} % \begin{macrocode} \let\@calc\@CalcNONE \newcommand{\allowcalculators}[1][any]{% \def\@calc{#1}% \ifx\@calc\@CalcNONE% \else\ifx\@calc\@CalcANY\def\@calculators{No restriction on the % model of calculator to be used, but no device with communication % capability shall be accepted as a calculator.}% \else\ifx\@calc\@CalcAPPROVED\def\@calculators{Only calculators on % the University of Otago list of approved calculators are permitted.}% \else\ClassWarning{ouexam}{% invalid argument `#1' for^^J% \protect\allowcalculators; valid values are `none', `any' and `approved'.^^J% No calculators will be permitted}% \fi\fi\fi% } % \end{macrocode} % \cs{@calculators} is used in \cs{@maketitlepage}. % \end{macro} % \end{macro} % % \begin{macro}{\permitcalculators} % \changes{2.1}{2004/04/05}{NJS Added \cs{permitcalculators} as a synonym % for \cs{allowcalculators}.} % \cs{permitcalculators} is a synonym for \cs{allowcalculators}. % \begin{macrocode} \let\permitcalculators\allowcalculators % \end{macrocode} % \end{macro} % % \begin{macro}{\copiesof} % \begin{macro}{\@copiesof} % \changes{2.1.3}{2009/04/28}{NJS Changed default to ``N/A'' for \cs{copiesof}.} % The \cs{copiesof} macro specifies any material that candidates are allowed to % bring into the examination, and is optional. It redefines the \cs{@copiesof} % macro that is used in \cs{@maketitlepage}: % \begin{macrocode} \let\@copiesof\@NA \newcommand{\copiesof}[1]{\def\@copiesof{#1}} % \end{macrocode} % \end{macro} % \end{macro} % % \begin{macro}{\otherinstructions} % \begin{macro}{\@otherinst} % \changes{2.1.3}{2009/04/28}{NJS Changed default to ``N/A'' for % \cs{otherinstructions}.} % The \cs{otherinstructions} macro specifies any other instructions not covered % by any of the above, and is optional. It redefines the \cs{@otherinst} % macro that is used in \cs{@maketitlepage}: % \begin{macrocode} \let\@otherinst\@NA \newcommand{\otherinstructions}[1]{\def\@otherinst{#1}} % \end{macrocode} % \end{macro} % \end{macro} % % \begin{macro}{\maketitlepage} % The \cs{maketitlepage} macro is the \textsf{ouexam} analogue of \cs{maketitle}. % It takes all of the information provided by the above macros and generates a % properly formatted examination title page. If you want to design your own title % page styles, redefine \cs{@maketitlepage} (this is not normally recommended % however). % \begin{macrocode} \newcommand{\maketitlepage}{\@maketitlepage} % \end{macrocode} % \end{macro} % % \begin{macro}{\@maketitlepage} % \changes{1.1}{1999/04/20}{NJS Renamed macro from \cs{maketitlepage}.} % \changes{2.4}{2013/06/05}{NJS Added check for \textsf{hyperref} being loaded.} % The \cs{@maketitlepage} macro generates an examination title page that meets the Otago University requirements for final examination papers. You can redefine this if you want to change the format, but normally you would not do this. % \begin{macrocode} \def\@maketitlepage{% \thispagestyle{titlepage}% % \end{macrocode} % The header information for the examination is printed centered at the top % of the page. The department, paper number and paper title (and the semester % information if required) are placed inside a double box. % \begin{macrocode} \begin{center}% {\Large \textbf{UNIVERSITY OF OTAGO EXAMINATIONS \@year}}% \\[\baselineskip]% \fbox{\framebox[\linewidth]{% \begin{tabular}{c}% \\% {\large \@dept} \\% \\% {\large Paper \@pnumber} \\% \\% {\large \@ptitle} \\% \ifx\@semester\@empty\else\@semester \\ \fi% \\% \end{tabular}% }}% \mbox{}\\[\baselineskip]% % \end{macrocode} % Next, the time allowed for completing the examination: % \begin{macrocode} \textbf{(TIME ALLOWED: \@hours\ HOURS)}% \\[\baselineskip]% \end{center}% % \end{macrocode} % The number of pages is automatically calculated, as described earlier. All we need to do is \cs{pageref} the \|num.pages\| label that we wrote into the \|.aux\| file, or \cs{pageref} if \textsf{hyperref} is loaded to inhibit hyperlink generation. % \begin{macrocode} \underline{This examination paper comprises % \@ifundefined{Hy@Warning}{\pageref{num.pages}}{\pageref{num.pages}} pages.% }% \\[\baselineskip]% % \end{macrocode} % Print out the instructions for answering questions: % \begin{macrocode} \underline{Candidates should answer questions as follows:}% \\[\baselineskip]% \hspace{1cm}\begin{minipage}{13.5cm}\@instructions\end{minipage}% \\[\baselineskip]% % \end{macrocode} % Print out any provided material: % \begin{macrocode} \underline{The following material is provided:}% \\[\baselineskip]% \hspace{1cm}\begin{minipage}{13.5cm}\@material\end{minipage}% \\[\baselineskip]% % \end{macrocode} % Print out use of calculators section: % \begin{macrocode} \underline{Use of calculators:} \\[\baselineskip]% \hspace{1cm}\begin{minipage}{13.5cm}\@calculators\end{minipage}% \\[\baselineskip]% % \end{macrocode} % \changes{2.1.3}{2009/04/28}{NJS Fixed vertical spacing.} % \changes{2.1.2}{2008/08/14}{NJS Ensure that the ``subject to inspection'' % line for calculators is only printed when calculators actually are allowed.} % Only print out the ``subject to inspection'' line if calculators actually % are allowed (irrelevant otherwise): % \begin{macrocode} \ifx\@calc\@CalcNONE% \else\mbox{}\hfill(Subject to inspection by the examiners.)% \\[\baselineskip]% \fi% % \end{macrocode} % Print out anything that candidates are allowed to bring: % \begin{macrocode} \underline{Candidates are permitted copies of:}% \\[\baselineskip]% \hspace{1cm}\begin{minipage}{13.5cm}\@copiesof\end{minipage}% \\[\baselineskip]% % \end{macrocode} % \changes{2.1.3}{2009/04/28}{NJS Fixed vertical spacing.} % \changes{2.1.2}{2008/08/14}{NJS Ensure that the ``subject to inspection'' % line for other material is only printed when there actually \emph{is} % other material.} % Only print out the ``subject to inspection'' line if there actually % \emph{is} other material (irrelevant otherwise): % \begin{macrocode} \ifx\@copiesof\@NA% \else\mbox{}\hfill(Subject to inspection by the examiners.)% \\[\baselineskip]% \fi% % \end{macrocode} % \changes{2.1.3}{2009/04/28}{NJS Removed superfluous blank line.} % Print out other instructions, and finish. % \begin{macrocode} \underline{Other Instructions:}% \\[\baselineskip]% \hspace{1cm}\begin{minipage}{13.5cm}\@otherinst\end{minipage}% \newpage% } % \end{macrocode} % \end{macro} % % % \subsection{Obsolete macros and environments} % % \changes{2.0}{2002/01/09}{NJS All \textsf{multichoice} macros and environments % made obsolete.} % \changes{2.0}{2002/01/09}{NJS \cs{markingschedule}, \cs{marks}, \cs{totalmarks} % macros made obsolete.} % \changes{2.0}{2002/01/09}{NJS \textsf{questions}, \textsf{subquestions}, % \textsf{subsubquestions} environments made obsolete.} % The following macros and environments are no longer supported as of version % 2.0. Documents that use these macros can be processed using ouexam v1.2 or % earlier. The obsolete multiple-choice examination support is dealt with % during option processing, so we can ignore the obsolete multiple-choice % macros. % % \begin{environment}{questions} % \begin{environment}{subquestions} % \changes{1.1}{1999/04/20}{NJS Now disabled when \textsf{multichoice} is % active.} % \begin{environment}{subsubquestions} % \changes{1.1}{1999/04/20}{NJS Now disabled when \textsf{multichoice} is % active.} % \begin{macro}{markingschedule} % \changes{1.2}{1999/10/26}{NJS New macro to turn marking information on and % off.} % The \textsf{questions}, \textsf{subquestions} and \textsf{subsubquestions} % environments have been replaced by the \textsf{question}, % \textsf{subquestion} and \textsf{subsubquestion} environments respectively. % \begin{macrocode} \newenvironment{questions}{% \def\item{\@unsupported{questions environment}{1.2}}}{} \newenvironment{subquestions}{% \def\item{\@unsupported{subquestions environment}{1.2}}}{} \newenvironment{subsubquestions}{% \def\item{\@unsupported{subsubquestions environment}{1.2}}}{} % \end{macrocode} % The \cs{markingschedule} macro has been replaced by the % \textsf{markingschedule} class option. We can't map this to the % \cs{@unsupported} macro because \cs{markingschedule} occurs in the document % preamble. Mapping it to \cs{@unsupported} causes \TeX\ to blow up for some % reason. It's safe enough to ignore it. % \begin{macrocode} \let\markingschedule\@empty % \end{macrocode} % \end{macro} % \end{environment} % \end{environment} % \end{environment} % % \changes{2.1}{2004/04/16}{NJS Removed the \texttt{marks} macro entirely % because it conflicts with a macro in e-TeX.} % \changes{1.1}{1999/04/20}{NJS This macro now increments the \texttt{marks} % counter.} % \begin{macro}{\totalmarks} % \changes{1.1}{1999/04/20}{NJS Removed the parameter for the number of marks; % Both these macros have been replaced by automatic calculations within the % question-building environments.} % The \cs{marks} macro has been replaced by the argument to the various % question-building environments. The \cs{totalmarks} macro has been replaced % by automatic calculations within these environments. % \begin{macrocode} \newcommand{\totalmarks}{\@unsupported{totalmarks macro}{1.2}} % \end{macrocode} % \end{macro} % % % % \Finale % % \PrintChanges

Ignore Space Show notes View

ouexam.dtx

\def\ouexamdate{13 June 2013}
\def\ouexamversion{2.4}
\def\ouexamshortdate{2013/06/13}
% \iffalse meta-comment
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%
%% Copyright 1999--2013 Nigel Stanger and University of Otago
%% 
%% You may use this package freely, and also distribute it
%% provided that you don't change it, make any money off
%% it or pretend that you wrote it.
%%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%<*driver>
\documentclass[a4paper]{ltxdoc}
\usepackage{doc}
\DisableCrossrefs
\CodelineNumbered
\RecordChanges

\usepackage{graphicx}
% change hyperref options as required
\usepackage[colorlinks,pdfpagemode=None,urlcolor=blue,
            linkcolor=red,pdfauthor={Nigel Stanger}]{hyperref}

\title{The \textsf{ouexam} document class, v\ouexamversion}
\author{Nigel Stanger\\nstanger@infoscience.otago.ac.nz}
\date{\ouexamdate}

\begin{document}
\maketitle
\DocInput{ouexam.dtx}
\end{document}
%</driver>
%
% \fi
%
%% \CheckSum{896}
%%
%% \CharacterTable
%%  {Upper-case    \A\B\C\D\E\F\G\H\I\J\K\L\M\N\O\P\Q\R\S\T\U\V\W\X\Y\Z
%%   Lower-case    \a\b\c\d\e\f\g\h\i\j\k\l\m\n\o\p\q\r\s\t\u\v\w\x\y\z
%%   Digits        \0\1\2\3\4\5\6\7\8\9
%%   Exclamation   \!     Double quote  \"     Hash (number) \#
%%   Dollar        \$     Percent       \%     Ampersand     \&
%%   Acute accent  \'     Left paren    $     Right paren   $
%%   Asterisk      \*     Plus          \+     Comma         \,
%%   Minus         \-     Point         \.     Solidus       \/
%%   Colon         \:     Semicolon     \;     Less than     \<
%%   Equals        \=     Greater than  \>     Question mark \?
%%   Commercial at \@     Left bracket  \[     Backslash     \\
%%   Right bracket \]     Circumflex    \^     Underscore    \_
%%   Grave accent  \`     Left brace    \{     Vertical bar  \|
%%   Right brace   \}     Tilde         \~}
%%
%
% \MakeShortVerb{\|}
% 
% \changes{1.0}{1999/04/15}{Initial version.}
%
% \begin{abstract}
% This document class allows you to create Otago University final
% examination papers using \LaTeXe\@. It implements the major formatting
% requirements specified by the University, provides useful
% macros to ease the process of building the title page and automatically
% deals with fiddly little issues such as question and section numbering,
% the number of pages in the paper, printing ``\textbf{TURN OVER}'' in the
% bottom right corner of every page except the last one and ensuring that the
% total number of marks for a question adds up to the expected number.
% \end{abstract}
% 
% 
% \tableofcontents
% 
% 
% \section{Overview}
% 
% \changes{2.1}{2004/04/10}{NJS Added requirements and installation instructions.}
% \changes{2.3}{2012/09/04}{NJS Dropped vanilla \LaTeX\ for building docs.}
% \subsection{Requirements and installation}
% 
% The \textsf{ouexam} package requires the \textsf{verbatim},
% \textsf{fontenc}, \textsf{lmodern} and \textsf{textcomp} packages, all
% of which should come standard with most \TeX\ installations. To build
% the documentation and example files, you will need at least version 1.1
% of the \textsf{listings} package, and the \textsf{pdfjam} tool.
% 
% Installation is relatively simple:
% \begin{enumerate}
% 
% 	\item Unpack the distribution archive and \texttt{cd} to the
% 	distribution directory.
% 	
% 	\item \texttt{make}.
% 	
% 	\item \texttt{make install TEXMF\_INSTALL=/path/to/texmf}.
% 	\texttt{/path/to/texmf} should be the root of your preferred
% 	\texttt{texmf} tree (e.g., \texttt{/usr/share/texmf}). You may need
% 	to do this as root depending on which \texttt{texmf} tree you are
% 	installing into. You can also define \texttt{TEXMF\_INSTALL} as an
% 	environment variable then simply type \texttt{make install}.
% 
% \end{enumerate}
% You can, of course, always install manually if you wish.
% 
% 
% \changes{2.0}{2002/01/11}{NJS Added note on backwards compatibility.}
% \subsection{Important note on backwards compatibility}
% 
% Version 2.0 of \textsf{ouexam} introduced several major changes from earlier
% versions. Consequently, documents created using an earlier version of
% \textsf{ouexam} are \emph{incompatible} with version 2.0 or later. Versions 2.0
% and greater of \textsf{ouexam} will detect attempts to process a version 1.x
% document and display an error message.
% 
% The older version (1.2) is still available and can be downloaded from the same
% location as the current version. To process older version documents, just copy
% the version 1.2 |ouexam.cls| file into the same directory as the document file.
% \TeX\ searches the current directory first, so this copy will take precedence
% over any other installed version of \textsf{ouexam}.
% 
% 
% \subsection{Class options}
%
% This document class is based on the \textsf{article} class and accepts
% most of the options accepted by \textsf{article} (the \textsf{twoside},
% \textsf{10pt} and \textsf{11pt} options are ignored). The standard
% options are \textsf{onecolumn}, \textsf{oneside}, \textsf{a4paper} and
% \textsf{12pt}. In addition, \textsf{ouexam} accepts the following
% ``native'' class options:
% \begin{description}
% 	\item[\textsf{draft}] \DescribeMacro{draft} This option has the usual
% 	effects that you would expect for \textsf{draft} mode in the
% 	\textsf{article} class, plus \fbox{DRAFT} is printed in
% 	the footer of every page. Note that if you use the \textsf{graphicx}
% 	package and turn on the \textsf{draft} option in \textsf{ouexam}, all
% 	included graphics will be drawn in draft mode unless you specify the
% 	|draft=false| option to \cs{includegraphics}.
% 	
% 	\item[\textsf{markingschedule}] \DescribeMacro{markingschedule}
% 	Rather than write a separate marking schedule for an examination paper, you
% 	can use the \textsf{marking} environment to embed marking information
% 	within questions (see 
% 	\hyperref[Sec:Questions:Marking]{section~\ref*{Sec:Questions:Marking}}).
% 	By default this information is not printed (for obvious reasons!), but
% 	when it comes time to print a marking schedule for the examination,
% 	including the \textsf{markingschedule} class option in the
% 	\cs{usepackage} command will cause \textsf{ouexam} to print the hidden
% 	content. It will also print \fbox{MARKING SCHEDULE} in the header of
% 	every page.
% 	
% 	\item[\textsf{times}] \DescribeMacro{times}
% 	The Examinations Office prefers all examination papers to be set in 12pt
% 	Palatino. However, if this is not available, they will accept 13pt Times
% 	instead. The \textsf{times} class option enables the latter, but should
% 	only be used if Palatino is not available.
% \end{description}
% 
% 
% \subsection{Required packages}
% 
% This class requires the \textsf{verbatim} package in order to implement
% the marking schedule functionality. It also expects modern \LaTeX\ font
% handling capabilities, and therefore requires the \textsf{fontenc},
% \textsf{textcomp} and \textsf{lmodern} classes.
% 
% 
% \subsection{Page margins}
% 
% This class assumes A4 paper. You will probably get weird results if you try
% to do anything different. The margins are set up as follows: top and bottom
% 2cm (headers and footers are inside this margin), left and right 2.54cm (1in).
% 
% 
% \changes{2.1.1}{2006/08/21}{NJS Updated \textsf{lastpage} style to display
% ``\textbf{END}''.}
% \subsection{Page styles}
% 
% There are three page styles defined in this class:
% \begin{description}
% 	\item[\textsf{plain}] \DescribeMacro{plain} This is a slight modification
% 	of the \textsf{plain} page style from \textsf{article}. It produces pages
% 	that have the page number centered at the bottom, the paper number in the
% 	top right corner and the text ``\textbf{TURN OVER}'' in the bottom right.
% 	This is the default page style.
% 	
% 	\item[\textsf{lastpage}] \DescribeMacro{lastpage} This is similar to
% 	\textsf{plain} but with ``\textbf{END}'' instead of ``\textbf{TURN
% 	OVER}''. It is used for the last page of the examination. You
% 	normally will not have to use this yourself---the class should take
% 	care of it automatically. The class does however occasionally seem
% 	to get confused, so there may times when you have to set the page
% 	style of the last page manually. It will be fairly obvious when you
% 	need to do this---the usual effect is that the last page
% 	has ``\textbf{TURN OVER}'' printed on it when it should not.
% 	
% 	\item[\textsf{titlepage}] \DescribeMacro{titlepage} This is similar to
% 	\textsf{plain} but without the page header, and is used for the title page
% 	of the paper. As with \textsf{lastpage}, you will normally not use this
% 	yourself, because the \cs{maketitlepage} macro handles this automatically
% 	(see \hyperref[Sec:TitlePage]{section~\ref*{Sec:TitlePage}}).
% \end{description}
% 
% 
% \section{Writing an examination paper}
% 
% 
% \changes{2.0.2}{2002/05/11}{NJS Moved the title page subsection to the start
% of this section.}
% \subsection{The title page}
% \label{Sec:TitlePage}
% 
% This class defines a collection of macros that let you fill in the various
% parts of the examination title page. This is analogous to the process you use
% to generate the title of a document in, for example, the \textsf{article} class
% (using \cs{title}, \cs{author}, \cs{date} and \cs{maketitle}). That is, you
% issue the macros described below in the document preamble, then issue a
% \cs{maketitlepage} in the document body.
% 
% The \DescribeMacro{\examyear} \cs{examyear} macro lets you specify the year
% in which the examination is being held, for example, |\examyear{1999}|. This
% macro is mandatory.
% 
% The \DescribeMacro{\department} \cs{department} macro lets you specify the
% name of the department that produced the examination paper, for example,
% |\department{Information| |Science}|. This macro is mandatory.
% 
% The \DescribeMacro{\papernumber} \cs{papernumber} macro lets you specify
% the paper number that the examination is for, for example,
% |\papernumber{COMP 101}|. This macro is mandatory. Note that when you
% specify semester information using \cs{semester}, this is appended to
% the paper number in the page header. For example, if you specify
% |\papernumber{COMP 101}| and |\semester{2}|, the header will contain
% ``COMP 101 (Semester Two)'' rather than just ``COMP 101''.
% 
% The \DescribeMacro{\papertitle} \cs{papertitle} macro lets you specify the
% title of the paper, for example, 
% |\papertitle{Systems Analysis and Design Methods}|. This macro is mandatory.
% 
% \changes{2.0.2}{2002/08/22}{NJS Added `FY' to the \cs{semester} macro
% description.}
% Some papers are offered in more than one semester. The
% \DescribeMacro{\semester} \cs{semester} macro lets you specify which
% semester the examination is for. The argument can be either ``1'' or ``2''
% for semesters 1 and 2, ``SS'' for summer school, ``FY'' for a full-year
% paper or ``SP'' for a special examination, for example, |\semester{2}|.
% Invalid values for the argument cause a warning to be raised. This macro is
% optional---if you omit it, no semester information is generated.
% 
% The \DescribeMacro{\timeallowed} \cs{timeallowed} macro lets you specify the
% length of the examination in hours, for example, |\timeallowed{2}|. This
% macro is optional---if you omit it, it defaults to three (3) hours.
% 
% \changes{2.1}{2004/04/05}{NJS Modified \cs{allowcalculators} to conform
% to the new University calculator regulations.}
% \changes{2.1}{2004/04/05}{NJS Added \cs{permitcalculators} as a synonym
% for \cs{allowcalculators}.}
% If calculators are permitted in the examination, use the
% \DescribeMacro{\allowcalculators}
% \DescribeMacro{\permitcalculators}
% \cs{allowcalculators} macro (\cs{permitcalculators} will also work).
% This macro is optional---if you omit it, a sentence is inserted saying
% that calculators are \emph{not} permitted. \cs{allowcalculators} has a
% single optional argument that specifies the kind of calculators
% permitted. If omitted, it defaults to ``any'' (see below). Otherwise,
% this argument must be one of the following values:
% \begin{description}
% 
% 	\item[\textsf{none}] A sentence is inserted noting that no calculators are
% 	permitted.
% 
% 	\item[\textsf{approved}] A sentence is inserted noting that only approved
% 	calculators are permitted.
% 
% 	\item[\textsf{any}] (default) A sentence is inserted noting that any
% 	calculator that does not have a communication capability is
% 	permitted.
% 
% \end{description}
% 
% The \DescribeMacro{\instructions} \cs{instructions} macro lets you specify
% instructions on how candidates should complete the
% examination, for example, |\instructions{Answer ALL| |questions.}|. This macro
% is optional. Note that the content of the instructions can be just about
% anything. It is up to you to control formatting, such as how you want lines
% broken, etc. This also true of the \cs{material}, \cs{copiesof} and
% \cs{otherinstructions} macros described below.
% 
% The \DescribeMacro{\material} \cs{material} macro lets you specify any
% additional material that candidates are provided in addition to the
% examination paper itself, for example, |\material{SQL schema definition}|.
% This macro is optional.
% 
% The \DescribeMacro{\copiesof} \cs{copiesof} macro lets you specify any
% material that candidates are allowed to bring into the examination, for
% example, |\copiesof{McFadden &| |Hoffer, 5th edition}|. This macro is
% optional.
% 
% The \DescribeMacro{\otherinstructions} \cs{otherinstructions} macro lets you
% specify any other instructions not covered by any of the above, for example,
% |\otherinstructions{No smoking| |allowed.}|. This macro is optional.
% 
% Once you have specified the content of the title page using the above
% macros, you simply issue a \DescribeMacro{\maketitlepage} \cs{maketitlepage}
% to generate the front page of the examination (cf. \cs{maketitle}). The title
% page generated meets University formatting requirements.
% Note that the number of pages in the examination paper is generated
% and inserted into the title page automatically---you do not need to specify
% it manually.
% 
% 
% \subsection{Sections}
% \label{Sec:Sections}
% 
% Examination papers may optionally have multiple sections, ``numbered'' A, B, \ldots. Rather than redefine the existing section macros, this class defines an \DescribeEnv{examsection} \textsf{examsection} environment that generates a new section. The \textsf{examsection} environment has three mandatory arguments:
% \begin{enumerate}
% 	\item The expected number of marks for the question. If you leave this
% 	argument empty it defaults to zero. \textsf{ouexam} keeps a running total
% 	of the \emph{actual} number of marks encountered within a section, which is
% 	compared with the value of this argument when the environment is closed.
% 	An error is raised if the values are not equal. The running total is also
% 	typeset right-justified at the end of the section in the form
% 	``\textbf{[SECTION A TOTAL 20 MARKS]}'' as an additional verification.
% 	
% 	\item Instructions for answering questions in this section (such as
% 	``Answer any TWO questions.''). If you leave this argument empty it
% 	defaults to ``ANSWER ALL QUESTIONS.''.
% 	
% 	\item A description of the topic of the section. This can be left
% 	empty.
% \end{enumerate}
% Every section begins on a new page, and the section title is formatted as
% ``\textbf{\underline{Section A}}'', in \cs{large} size. For example:	\\
% 
% \noindent|\begin{examsection}{25}{}{These questions are remarkably boring.}|	\\
% \hspace*{1cm}\vdots	\\
% |\end{examsection}|	\\
% 
% \noindent will produce the following (assuming that the actual number of
% marks in the question is correct):
% 
% \begin{center}
% 	\fbox{%
% 	\begin{minipage}{0.9\columnwidth}
% 		{\large\noindent\textbf{\underline{Section~A}}}	\\[0.5\baselineskip]
% 		ANSWER ALL QUESTIONS.	\\
% 		These questions are remarkably boring.	\\
% 		\hspace*{2cm}\vdots	\\
% 		\mbox{}\hfill\textbf{[SECTION A TOTAL 25 MARKS]}
% 	\end{minipage}
% 	}
% \end{center}
% 
% 
% \subsection{Questions}
% \label{Sec:Questions}
% 
% \textsf{ouexam} provides three environments for building examination
% questions: \DescribeEnv{question} \textsf{question},
% \DescribeEnv{subquestion} \textsf{subquestion} and \textsf{subsubquestion},
% which correspond to top-level questions, parts of questions, and
% \DescribeEnv{subsubquestion} sub-parts of questions respectively. They
% produce questions that are numbered according to University examination
% formatting requirements. Note that questions are normally numbered
% sequentially throughout the entire paper regardless of any section
% boundaries.
% 
% All three environments have a single mandatory argument which is the
% expected number of marks for the question, part or sub-part. If this
% argument is left empty it will default to zero. This argument works in much
% the same way as the first argument to the \textsf{examsection} environment
% (see \hyperref[Sec:Sections]{section~\ref*{Sec:Sections}}): \textsf{ouexam}
% keeps a running total of the number of marks encountered within each
% question part and sub-part, and compares this total with the expected value
% when the environment closes. Where appropriate, the running total is
% typeset right-justified in the form ``(5 marks)''. For example:
% \begin{verbatim}
% \begin{question}{5}
%   \begin{subquestion}{2}
%     Why is the sky blue?
%   \end{subquestion}
%   \begin{subquestion}{3}
%     Explain in detail how the sky can be made pink.
%   \end{subquestion}
% \end{question}
% \begin{question}{1}
%   Define the term ``floccinaucinihilipilification''.
% \end{question}
% \end{verbatim}
% 
% \noindent will produce the following:
% \begin{center}
% 	\fbox{%
% 	\begin{minipage}{0.9\columnwidth}
% 		\begin{enumerate}
% 			\item
% 			\begin{enumerate}
% 				\item Why is the sky blue? \hfill (2 marks)
% 
% 				\item Explain in detail how the sky can be made pink. \hfill (3
% 				marks)
% 			\end{enumerate}
% 
% 			\item Define the term ``floccinaucinihilipilification''. \hfill (1
% 			mark)
% 		\end{enumerate}
% 	\end{minipage}
% 	}
% \end{center}
% 
% The \textsf{question} environments handle formatting of marks totals
% automatically, including considerations such as:
% \begin{itemize}
% 	\item whether or not to print the number of marks for a question at all
% 	(they should not be printed if the question has sub-parts);
% 	
% 	\item whether to print ``mark'' or ``marks'' (e.g., question 2 above); and
% 	
% 	\item where to position the number of marks relative to the question,
% 	depending on how full the last line of the question is.
% \end{itemize}
% 
% \changes{2.3.1}{2013/05/31}{NJS Enforced proper question environment nesting.}
% The various \textsf{question} environments must be correctly nested in order for the features described above to work correctly, i.e., \textsf{question} at the top level, followed by \textsf{subquestion} and \textsf{subsubquestion}. Any variation to this nesting order will cause an error.
% 
% 
% \changes{2.2}{2010/04/26}{NJS Added time allocation macros.}
% \subsection{Time allocations}
% \label{Sec-Questions-Time}
% 
% If you wish to suggest a time allocation for completing a question, use the \DescribeMacro{\timeallocation} \cs{timeallocation} macro. The only argument is the number of minutes to be allocated for this question. The macro will output the time allocation in the form: \textbf{[The suggested time allocation for this question is \emph{t} minutes.]}
% 
% If you wish to allow a certain amount of time for candidates to read some text, use the \DescribeMacro{\readingtime} \cs{readingtime} macro. The only argument is the number of minutes to be allocated for reading time. The macro will output the reading time in the form: \textbf{[Allow \emph{t} minutes reading time.]}
% 
% Both macros assume the base unit to be minutes; if you wish to change this you will need to redefine the macros. Otherwise, the argument can be almost anything: text, numbers, etc.
% 
% \textsf{ouexam} currently does not check that time allocations sum correctly. This may be implemented in a future version.
%
% 
% \subsection{Marking schedule information}
% \label{Sec:Questions:Marking}
% 
% Rather than writing a separate marking schedule for an examination paper,
% you can use the \DescribeEnv{marking} \textsf{marking} environment to embed
% marking information within questions. By default this information is not
% printed (for obvious reasons!), but when it comes time to print a marking
% schedule for the examination, use the \textsf{markingschedule} class option
% to print the hidden content. The marking information will be set in
% \emph{italics} to differentiate it from the main body of the examination,
% and \fbox{MARKING SCHEDULE} will also be printed in the header of every
% page.
% 
% \changes{2.1}{2004/04/10}{NJS Added driver file tip.}
% A useful tip for writing examination papers is to create two separate
% ``driver'' documents that are set up the document class for the
% ``plain'' and ``marking schedule'' versions of the examination,
% respectively. Each should then include a separate document that contains
% the actual content of the examination paper, for example:
% 
% \begin{verbatim}
% \documentclass[markingschedule]{ouexam}
% 
% \input{exampaper} % contains the actual examination content
% \end{verbatim}
% 
% Doing this enables you to build the ``plain'' and ``marking schedule''
% versions independently, without overwriting each other or having to
% change the original source file. Simply run \LaTeX\ on the appropriate
% driver file to produce the version that you want. See the example files
% that came with the \textsf{ouexam} distribution for an example of this
% approach in action.
% 
% 
% \subsection{Miscellaneous}
% 
% Most examinationss are marked out of 100, and \textsf{ouexam} defaults to
% this. However, if you want an examination that is marked out of some other
% number, for example, 90 marks, you can specify this using the
% \DescribeMacro{\examoutof} \cs{examoutof} macro. Thus, |\examoutof{90}|
% will set the expected number of marks for the examination to 90. This value
% is used by \textsf{ouexam} to verify that the marks for all the questions
% add up to the expected number of marks for the examination.
% 
% 
% \section{Errors and warnings}
% 
% In this section are described the error messages and warning produced by
% \textsf{ouexam}, and the reasons why they occur.
% 
% 
% \subsection{Error messages}
% 
% \noindent\texttt{INCOMPATIBLE (\ldots): this document was written for an
% earlier}	\\
% \texttt{version of ouexam \ldots}	\\
% You are trying to use v2.0 or later of \textsf{ouexam} with a document that was written for \textsf{ouexam} v1.2 or earlier. Versions 2.0 and later of \textsf{ouexam} are \emph{fundamentally incompatible} with earlier versions. The only solutions here are either to revert to \textsf{ouexam} v1.2 or earlier (you can put a copy of the class file in the same directory as the document), or rewrite the document to conform to the current version of \textsf{ouexam}. You should be able to download a copy of v1.2 from the same location that you found the current version.	\\
% 	
% \noindent\texttt{no \cs{examyear} was specified}	\\
% \noindent\texttt{no \cs{department} was specified}	\\
% \noindent\texttt{no \cs{papernumber} was specified}	\\
% \noindent\texttt{no \cs{papertitle} was specified}	\\
% You have not specified one of these macros. All four of these macros are mandatory and must be included in the preamble of any \textsf{ouexam} document.	\\
% 
% 
% \noindent\texttt{question environment nested inside another question, subquestion or	\\*
% \hspace*{1em}subsubquestion environment}	\\
% \noindent\texttt{subquestion environment either outside a question environment, or	\\*
% \hspace*{1em}nested inside another subquestion or subsubquestion environment}	\\
% \noindent\texttt{subsubquestion environment either outside a subquestion environment,	\\*
% \hspace*{1em}or nested inside another subsubquestion environment}	\\
% These errors all mean that you have incorrectly nested your \textsf{question} environments in some way. \textsf{question}s must always appear at the outermost level, \textsf{subquestion}s must always appear inside \textsf{question}s, and \textsf{subsubquestion}s must always appear inside \textsf{subquestion}s. Any other kind of nesting is an error.	\\
% 
% 
% \subsection{Warnings}
% 
% \noindent\texttt{actual number of marks for exam (\ldots) does not match expected	\\*
% \hspace*{1em}number of marks (\ldots)}	\\
% The actual number of marks supplied for the whole examination (calculated by summing the marks for all the questions) is not the same as the expected total for the examination. Adjust the marks for the questions until the totals match.	\\
% 
% \noindent\texttt{actual mark (\ldots) for section \ldots\ does not match expected	\\*
% \hspace*{1em}mark (\ldots)}	\\
% The actual number of marks supplied for a particular section (calculated by summing the marks for all the questions in that section) is not the same as the expected total for that section. Adjust the marks for the questions in the section concerned until the totals match. The section number is given in the warning.	\\
% 
% \noindent\texttt{actual mark (\ldots) for question \ldots\ does not match expected	\\*
% \hspace*{1em}mark (\ldots)}	\\
% The actual number of marks supplied for a particular question (calculated by summing the marks of all its sub-parts) is not the same as the expected total for that question. Adjust the marks for the sub-parts of the question until the totals match. The question number is given in the warning.	\\
% 
% \changes{2.0.2}{2002/08/22}{NJS Added `FY' to \cs{semester} error description.}
% \noindent\texttt{invalid value `\ldots' for \cs{semester}; valid values are `1', `2',	\\*
% \hspace*{1em}`SS', `FY' and `SP'. No semester information will be printed}	\\
% You have provided an invalid value as the argument to the \cs{semester} macro. The allowed values are ``1'' (semester one), ``2'' (semester two), ``SS'' (summer school), ``FY'' (full-year) and ``SP'' (special examination). Any other values will be ignored. Note that these are case-sensitive, for example, `sp' is invalid. 	\\
% 
% \changes{2.1}{2004/04/05}{NJS Added \cs{allowcalculators} error description.}
% \noindent\texttt{invalid value `\ldots' for \cs{allowcalculators}; valid values are	\\*
% \hspace*{1em}`none', `any' and `approved'. No calculators will be permitted}	\\
% You have provided an invalid value as the argument to the \cs{allowcalculators} macro. The allowed values are ``none'' (no calculators permitted), ``any'' (any calculator permitted) and ``approved'' (only approved calculators permitted). Any other values will be ignored, and \textsf{ouexam} will assume that no calculators are permitted. Note that these are case-sensitive, for example, `NONE' is invalid. 	\\
% 
% 
% \section{Example}
% 
% \changes{2.1}{2004/04/10}{NJS Updated example to conform to 2.1 changes.}
% \changes{2.0}{2002/01/25}{NJS Updated example to conform to 2.0 changes.}
% \changes{1.1}{1999/04/20}{NJS Updated example to conform to 1.1 changes.}
% The code given below is for the first three pages of the INFO~321 2001
% final examination. The output produced by this source is shown in
% \hyperref[Fig.Example]{figure~\ref*{Fig.Example} on
% page~\pageref*{Fig.Example}}. Particular points to note in this example
% are:
% \begin{itemize}
% 
% 	\item The number of hours is not specified, so it defaults to three.
% 
% 	\item This examination allows only approved calculators.
% 	
% 	\item The \textsf{listings} package (v1.1 or later) is used to
% 	typeset code listings.
% 
% 	\item The argument of the \cs{instructions} macro can be just about
% 	anything, as can the arguments of the \cs{material}, \cs{copiesof} and
% 	\cs{otherinstructions} macros. In this example, the \cs{instructions}
% 	macro has been omitted, causing \textsf{ouexam} to output the default
%	instructions.
% 
% 	\item Marking schedule information has been embedded within the
% 	questions. \hyperref[Fig.Example2]{Figure~\ref*{Fig.Example2} on
% 	page~\pageref*{Fig.Example2}} shows the output produced by this source
% 	when the \textsf{markingschedule} class option is used.
% 
% \end{itemize}
% The full source for this example may be found in the
% \texttt{example*.tex} files that came with this distribution.
% 
% \small
% \begin{verbatim}
% \documentclass{ouexam}
% 
% \usepackage{graphicx}
% \usepackage{listings}
% 
% \examyear{2001}
% \department{Information Science}
% \papernumber{INFO 321}
% \papertitle{Database Systems}
% \allowcalculators[approved]
% 
% \begin{document}
% 
% \lstloadlanguages{ODL,OQL,[Oracle8]SQL,[Oracle8]PLSQL}
% \lstset{basicstyle=\sffamily, showstringspaces=false, tabsize=2,
%   xleftmargin=1cm, belowskip=0pt, commentstyle=\itshape,
%   numbers=left, numberstyle=\scriptsize, numbersep=5pt,
%   columns=fullflexible,}
% 
% \maketitlepage
% 
% \begin{examsection}{25}{}{Questions in this section (total 25 marks) 
%   relate to physical database design and tuning.}
% 
% 
% 
% \begin{question}{10}
% 
%   A major U.S. mail order firm is having performance problems with queries
%   on the \textsf{Customer} table in its Oracle8 order-processing database.
%   Consider the following information about the \textsf{Customer} table,
%   then answer the questions below:	\\
% 
%   \textsf{\textbf{Customer}(\underline{customer\_no}, name, address, city,
%   state, country, phone, email, status)}
% 
%   \begin{itemize}
%     \item \textsf{customer\_no} is the primary key and because this is an
%     Oracle8 database, a unique B-tree index has been automatically created
%     on this column.
% 
%     \item \textsf{status} is a single character column that holds one of
%     the values `I' (inactive), `B' (bronze), `S' (silver) or `G' (gold).
%     The value of this column is determined by a customer's order history.
% 
%     \item About 10\% of customers are inactive, about 50\% have bronze
%     status, about 35\% have silver status and about 5\% have gold status.
% 
%     \item The average row size is about 200 bytes and there are
%     approximately 5,000,000 rows.
% 
%     \item The \textsf{Customer} table is heavily queried (averaging thirty
%     new queries per minute) by four different groups of employees. Group~1
%     queries only inactive customers, group~2 queries only bronze status
%     customers and groups~3 and~4 both query only silver and gold status
%     customers. Members of any group may query the table at any time and
%     queries often overlap.
% 
%     \item UPDATE, INSERT and DELETE operations on the table are uncommon.
% 
%     \item Very few queries are for individual customers; rather, most
%     queries are based on various combinations of values from
%     \textsf{status}, \textsf{city}, \textsf{state} and \textsf{country}.
%   \end{itemize}\medskip
% 
%   \begin{subquestion}{3}
%     Describe and justify an \emph{index-based} physical tuning solution
%     that will improve the performance of queries on the \textsf{Customer}
%     table. (No code is required.)
%     \begin{marking}
%       Since queries tend to be on combinations of several low-cardinality
%       columns [1] and the table is very large [1], the most effective
%       solution would be to place bitmap indexes on status, city, state and
%       country [1].
% 
%       B-tree indexes on any of the individual columns would not be ideal
%       because of their low cardinality. A possible alternative is a
%       composite B-tree index on all four columns [1 for suggesting this if
%       appropriate], but this is not ideal and would be much larger than
%       the four bitmap indexes [1 for this explanation if appropriate].
%       Hashing is definitely \emph{not} an option, because very few queries
%       are exact match.
%     \end{marking}
%   \end{subquestion}
% 
%   \begin{subquestion}{7}
%     \begin{subsubquestion}{5}\label{nonindex}
%       Assuming that there are two identical fast disks available, describe
%       and justify two alternative \emph{non index-based} physical tuning
%       solutions that will improve the performance of queries on the
%       \textsf{Customer} table. For each solution include details of how
%       the two disks will be used. (No code is required.)
%       \begin{marking}
%         Alternative 1: partition the table across the two disks based on
%         status (four partitions, one for each status value) [1]. This will
%         allow simultaneous parallel access to different parts of the table
%         by the different groups [1]. Since two groups both access silver
%         and gold customers, it makes sense to put the silver customers on
%         disk one and the gold customers on disk two. Bronze customers
%         account for 50\% of the table, so it makes sense to also put them
%         on disk two (as the gold customers are the smallest group), and
%         put the inactive customers on disk one with the silver customers.
%         [1 for any sensible partitioning scheme]
%   
%         Alternative 2: replicate the table across both disks (one replica
%         on each disk) [1]. This will also provide simultaneous parallel
%         access to the data (hard to tell whether this would be better or
%         worse than the partitioning above) [1].
%   
%         Note that any solution must be capable of being implemented in
%         Oracle8.
%       \end{marking}
%     \end{subsubquestion}
% 
%     \begin{subsubquestion}{2}
%       Briefly discuss the relative advantages and/or disadvantages of the
%       two alternative solutions you described in part~(\ref{nonindex})
%       above.
%       \begin{marking}
%         The partitioning solution given above doesn't really help with the
%         bronze customer rows---we would need random horizontal
%         partitioning to really help with this, which Oracle8 doesn't
%         support. More generally, Oracle8 does not migrate rows into the
%         correct partition if a customer's status changes, so the data will
%         need to be reloaded occasionally [1].
% 
%         For replication, we have the usual problem of keeping the replicas
%         synchronised if they are both read/write. This is somewhat
%         mitigated by the fact that there are only two copies [1].
%       \end{marking}
%     \end{subsubquestion}
%   \end{subquestion}
% \end{question}
% 
% \newpage
% \begin{question}{10}
%   There are two generic parameters that affect the structure of a B-tree
%   index.
% 
%   \begin{subquestion}{6}
%     Identify these two parameters and describe what each parameter
%     specifies. Use diagrams to illustrate your answer.
%     \begin{marking}
%       \emph{Node size} [1] specifies the maximum number of pointers in
%       each index node [1]. Suitable diagram [1].
% 
%       \emph{Percentage fill} [1] specifies the number of pointers in each
%       index node that are allocated when a node is created [1]. In effect,
%       it determines the amount of free space left in each node. Suitable
%       diagram [1].
%     \end{marking}
%   \end{subquestion}
%   
%   \begin{subquestion}{4}
%     For each of the two parameters, discuss the practical effect(s) of:
%     \begin{subsubquestion}{2}
%       high values of the parameter; and
%       \begin{marking}
%         Large node size reduces the height of the B-tree, resulting in
%         potentially fewer I/Os to access a leaf node [$\frac{1}{2}$]. It
%         can however increase the number of rows locked in a single leaf
%         node [$\frac{1}{2}$].
%   
%         High percent fill means that most pointers in a node will be
%         allocated when the node is created [$\frac{1}{2}$]. This leaves
%         little free space in the B-tree for new key values or key values
%         that change [$\frac{1}{2}$].
%       \end{marking}
%     \end{subsubquestion}
%     \begin{subsubquestion}{2}
%       low values of the parameter.
%       \begin{marking}
%         Small node size increases the height of the B-tree, resulting in
%         potentially more I/Os to access a leaf node [$\frac{1}{2}$]. The
%         number of rows locked in a single leaf node will be reduced
%         [$\frac{1}{2}$].
%   
%         Low percent fill means that most pointers in a node will not be
%         allocated when the node is created [$\frac{1}{2}$]. This leaves
%         little free space in the B-tree for new key values or key values
%         that change [$\frac{1}{2}$].
%       \end{marking}
%     \end{subsubquestion}
%   \end{subquestion}
% \end{question}
%   
%   
%   
% \begin{question}{5}\label{clusterq}
% 
%   \begin{subquestion}{2}
%     Define what is meant by the term ``clustering'' in the context of
%     physical database design and tuning, and explain why it is beneficial.
% 
%     \begin{marking}
%       Clustering is the grouping of related rows physically near to each
%       other on disk [1] so that they may be retrieved in as few I/Os as
%       possible (ideally one)---beneficial because of the slowness of disk
%       I/O compared to memory I/O [1].
%     \end{marking}
%   \end{subquestion}
%   
%   \begin{subquestion}{3}\label{clusterqcode}
%     Consider the following Oracle8 SQL table definition:
%     
%     \begin{lstlisting}[language={[Oracle8]SQL}]{}
% create table enrolment
% (  student_id char(7),
%   paper_code char(7),
%   enrol_year number(4),
%   
%   primary key (student_id, paper_code, enrol_year)
% );
%     \end{lstlisting}
% 
%     The \textsf{Enrolment} table is often queried for all rows relating to
%     a particular student in a particular year (for example, student
%     1234567 in 1998, student 9876543 in 2000, etc.). Write appropriate
%     Oracle8 SQL code that will cluster the rows of the table in such a way
%     as to support these types of query (you do not need to calculate
%     cluster sizes). Relevant Oracle8 syntax diagrams are given on the next
%     page.
%     \begin{marking}
%       \begin{lstlisting}[language={[Oracle8]SQL}]{}
% create cluster enrolment_cluster (student_id char(7), year number(4));
% 
% create index enrol_cluster_index on cluster enrolment_cluster;
% 
% create table enrolment
% (  student_id char(7),
%   paper_code char(7),
%   enrol_year number(4),
%   
%   primary key (student_id, paper_code, enrol_year)
% ) cluster enrolment_cluster(student_id, enrol_year);
%       \end{lstlisting}
%     \end{marking}
%     [1] each for correct cluster, cluster index and table clustering.
%   \end{subquestion}
% 
% \end{question}
% 
% \end{examsection}
%       .
%       .
%       .
% \end{document}
% \end{verbatim}
% \normalsize
% 
% \begin{figure}
% 	\fbox{\includegraphics[scale=0.28]{eg1-1}}
% 	\hfill
% 	\fbox{\includegraphics[scale=0.28]{eg1-2}}
% 	
% 	\begin{center}
% 		\fbox{\includegraphics[scale=0.28]{eg1-3}}
% 	\end{center}
% 
% 	\caption{Output produced by \textsf{ouexam}.}
% 	\label{Fig.Example}
% \end{figure}
% 
% \begin{figure}
% 	\fbox{\includegraphics[scale=0.28]{eg2-2}}
% 	\hfill
% 	\fbox{\includegraphics[scale=0.28]{eg2-3}}
% 
% 	\bigskip
% 	
% 	\fbox{\includegraphics[scale=0.28]{eg2-4}}
% 	\hfill
% 	\fbox{\includegraphics[scale=0.28]{eg2-5}}
% 
% 	\caption{Output produced by \textsf{ouexam} when
% 			\textsf{markingschedule} is used.}
% 	\label{Fig.Example2}
% \end{figure}
% 
% 
% \StopEventually{}
% 
% \changes{2.0}{2002/01/10}{NJS Revamped and tided up documentation and code.}
% 
% \section{The code}
% 
% \subsection{Preamble}
% 
%    \begin{macrocode}
\NeedsTeXFormat{LaTeX2e}[1998/06/01]
\ProvidesClass{ouexam}[\ouexamshortdate\space %
                           v\ouexamversion\space Otago %
                           University examination paper]
%    \end{macrocode}
% 
% \begin{macro}{\@unsupported}
% \changes{2.0}{2002/01/10}{NJS New \cs{@unsupported} macro.}
% This macro is used to handle documents written for older versions of
% \textsf{ouexam}. All obsolete macros map to this macro, which prints out an
% appropriate error message. It needs to be defined early because it is used
% to trap use of the obsolete \textsf{multichoice} class option.
%
%    \begin{macrocode}
\def\@unsupported#1#2{\ClassError{ouexam}{%
  ^^JINCOMPATIBLE (#1)^^J%
  This document was written for an earlier version of ouexam and is^^J%
  not compatible with ouexam v\ouexamversion. Please use ouexam v#2 or earlier^^J%
  to process this document. Version #2 can be downloaded from the^^J%
  same location as version\space\ouexamversion}}
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{multichoice}
% \changes{1.1}{1999/04/20}{NJS New \textsf{multichoice} class option.}
% \changes{2.0}{2000/09/04}{NJS Removed support for \textsf{multichoice} option.}
% The \textsf{multichoice} class option is no longer supported, but rather than
% just breaking older documents, we can at least try to present a reasonable
% error message.
%
%    \begin{macrocode}
\DeclareOption{multichoice}{%
  \OptionNotUsed\@unsupported{multichoice class option}{1.2}%
}
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{twoside}
% \changes{1.1}{1999/04/20}{NJS Turned off \textsf{twoside} option.}
% This document class is based on the \textsf{article} class and accepts any of
% the options accepted by \textsf{article}. The \textsf{twoside} option does
% not really make sense, however---all Otago examination papers are printed
% single-sided anyway, and the format is such that two-sided printing would
% look no different. The \textsf{twoside} option is therefore not used in this
% class.
%
%    \begin{macrocode}
\DeclareOption{twoside}{\OptionNotUsed}
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{10pt}
% \begin{macro}{11pt}
% \changes{2.3}{1999/04/20}{NJS Turned off \textsf{10pt} and \textsf{11pt}
% options.}
% The Examinations Office specifies a minimum of 12pt for the font size, so the
% \textsf{10pt} and \textsf{11pt} options are not used in this class.
%
%    \begin{macrocode}
\DeclareOption{10pt}{\OptionNotUsed}
\DeclareOption{11pt}{\OptionNotUsed}
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \begin{macro}{draft}
% \changes{2.0}{2001/05/15}{NJS New \textsf{draft} class option.}
% \begin{macro}{\if@draft}
% \changes{2.0}{2001/05/15}{NJS New switch for \textsf{draft}.}
% The \textsf{draft} option determines whether the examination is to be printed
% in draft mode (default off). The \cs{if@draft} switch is used to determine
% whether this is in effect (default false). Draft mode has the usual effects
% that you would expect for draft mode in the \textsf{article} class, plus
% \fbox{DRAFT} is printed in the footer of every page.
%
%    \begin{macrocode}
\newif\if@draft \@draftfalse
\DeclareOption{draft}{\@drafttrue\PassOptionsToClass{\CurrentOption}{article}}
%    \end{macrocode}
% 
% Note that if you use the \textsf{graphicx} package and turn on the
% \textsf{draft} option in \textsf{ouexam}, all included graphics will
% be drawn in draft mode unless you specify the |draft=false| option to the
% \cs{includegraphics} macro. The amount of effort required to fix what is
% a relatively small issue isn't really worth it, so this will not change.
% \end{macro}
% \end{macro}
% 
% \begin{macro}{markingschedule}
% \changes{2.0}{2000/09/05}{NJS New \textsf{markingschedule} class option.}
% \begin{macro}{\if@markingschedule}
% \changes{2.0}{2000/09/05}{NJS New switch for \textsf{markingschedule}.}
% The \textsf{markingschedule} option determines whether marking schedule
% information is printed in addition to the questions (default off). The
% \cs{if@markingschedule} switch is used to determine whether this is in effect
% (default false).
%
%    \begin{macrocode}
\newif\if@markingschedule \@markingschedulefalse
\DeclareOption{markingschedule}{\@markingscheduletrue}
%    \end{macrocode}
% \end{macro}
% \end{macro}
% 
% \begin{macro}{times}
% \changes{2.3}{2012/09/03}{NJS New \textsf{times} class option.}
% \begin{macro}{\if@times}
% \changes{2.3}{2012/09/03}{NJS New switch for \textsf{times}.}
% The \textsf{times} option formats the exam paper using 13pt Times. If this
% option is not specified, then the exam paper is formatted using 12pt Palatino.
%
%    \begin{macrocode}
\newif\if@times \@timesfalse
\DeclareOption{times}{\@timestrue}
%    \end{macrocode}
% \end{macro}
% \end{macro}
% 
% All other options are passed directly to the \textsf{article} class:
%
%    \begin{macrocode}
\DeclareOption*{\PassOptionsToClass{\CurrentOption}{article}}
\ProcessOptions
%    \end{macrocode}
%
% \changes{1.1}{1999/04/20}{NJS Changed default point size to 12pt.}
% The defaults for this class are \textsf{onecolumn}, \textsf{oneside},
% \textsf{a4paper} and \textsf{12pt}. The font size is the only one that might
% need to change.
%
%    \begin{macrocode}
\LoadClass[onecolumn,oneside,a4paper,12pt]{article}
%    \end{macrocode}
% 
% 
% \subsection{Required packages}
% 
% \changes{2.0}{2001/05/08}{NJS Removed requirement for \textsf{calc} and
% \textsf{ifthen} packages.}
% The \textsf{verbatim} package, is used to implement the \textsf{marking}
% environment:
%
%    \begin{macrocode}
\RequirePackage{verbatim}
%    \end{macrocode}
%
% \changes{2.3}{2012/09/04}{NJS Added requirement for \textsf{fontenc},
% \textsf{textcomp} and \textsf{lmodern} packages.}
%
% Running under vanilla OT1 encoding often causes all sorts of font substitution
% warnings, so the \textsf{fontenc} (with \textsf{T1} option), \textsf{textcomp}
% and \textsf{lmodern} packages are used to bring \textsf{ouexam}'s font
% handling up to more modern standards.
%
%    \begin{macrocode}
\RequirePackage[T1]{fontenc}
\RequirePackage{textcomp}
\RequirePackage{lmodern}
%    \end{macrocode}
% 
%
% \changes{2.3}{2012/09/04}{NJS Added font setup for \textsf{times} class
% option.}
% \subsection{Font setup}
% 
% The Examinations Office allows for two standard font styles: 12pt Palatino
% (preferred) or 13pt Times New Roman. The former is the default; the latter is
% only used if the \textsf{times} class option is used, and we need to
% explicitly define all the standard sizes. The following is extrapolated from
% the definitions in the \textsf{article} class and seems about right:
%
%    \begin{macrocode}
\if@times%
  \usepackage{mathptmx}%
  \renewcommand{\normalsize}{\fontsize{13}{15.7}\selectfont}%
  \renewcommand{\tiny}{\fontsize{\@viipt}{\@viiipt}\selectfont}%
  \renewcommand{\scriptsize}{\fontsize{\@ixpt}{11}\selectfont}%
  \renewcommand{\footnotesize}{\fontsize{\@xpt}{\@xiipt}\selectfont}%
  \renewcommand{\small}{\fontsize{\@xipt}{13.6}\selectfont}%
  \renewcommand{\large}{\fontsize{\@xivpt}{18}\selectfont}%
  \renewcommand{\Large}{\fontsize{\@xviipt}{22}\selectfont}%
  \renewcommand{\LARGE}{\fontsize{\@xxpt}{25}\selectfont}%
  \renewcommand{\huge}{\fontsize{\@xxvpt}{30}\selectfont}%
  \let\Huge=\huge%
\else%
  \usepackage{mathpazo}%
\fi%
%    \end{macrocode}
% 
% 
% \subsection{Page setup}
% 
% \begin{macro}{\oddsidemargin}
% \begin{macro}{\topmargin}
% \begin{macro}{\textwidth}
% \begin{macro}{\textheight}
% A4 paper is assumed (since this class is intended only for Otago examination
% papers, this seems a reasonable assumption). The margins are top and bottom
% 2cm, left and right 2.54cm (1in).
%
%    \begin{macrocode}
\setlength{\oddsidemargin}{0cm}
\setlength{\topmargin}{-0.54cm}
\setlength{\textwidth}{15.92cm}
\setlength{\textheight}{25.7cm}
\advance\textheight by-\headheight
\advance\textheight by-\headsep
\advance\textheight by-\footskip
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% 
% 
% \subsection{Page styles}\label{pagestyles}
%
% \changes{2.0}{2001/05/17}{NJS Added support to page styles for \textsf{draft} and \textsf{markingschedule} class options.}
% \changes{2.0}{2002/01/15}{NJS Fixed headers so that the page number is always centered.}
% \changes{2.1.1}{2006/08/21}{NJS Updated \textsf{lastpage} style to display ``\textbf{END}''.}
% \changes{2.3}{2012/09/03}{NJS Rearranged headers and footers to conform to the latest formatting requirements. Swapped \fbox{MARKING SCHEDULE} and \fbox{DRAFT} between header and footer so that the page number doesn't get overprinted.}
% \changes{2.4}{2013/06/06}{NJS Replaced the \texttt{lastpage} page style with direct manipulation of the footer contents, as \cs{pagestyle} and \cs{thispagestyle} don't get applied to automatically generated float pages on the last page of the document.}
% 
%\begin{macro}{\if@lastpagetext}
%\begin{macro}{\if@lastpagefloat}
% These two switches are used in some slightly convoluted logic to ensure that ``\textbf{END}'' is printed on the last page of the document (see below for details).
%    \begin{macrocode}
\newif\if@lastpagetext \@lastpagetextfalse
\newif\if@lastpagefloat \@lastpagefloatfalse
%    \end{macrocode}
% \end{macro}
% \end{macro}
% 
% \begin{macro}{\ps@plain}
% The \textsf{plain} page style is based on that in the \textsf{article} class, and produces pages with the page number centered at the bottom, the paper number in the top right corner and the text ``\textbf{TURN OVER}'' in the bottom right. If semester information is specified, this is included in parentheses after the paper code, for example, ``COMP 101 (Semester Two)''. \fbox{MARKING SCHEDULE} is printed at top left if the \textsf{markingschedule} class option is used.
%    \begin{macrocode}
\def\ps@plain{%
  \def\@oddhead{\@markingschedule\hfill\@pnumber%
      \ifx\@semester\@empty\else\ (\@semester)\fi}%
  \let\@evenhead\@oddhead%
  \def\@oddfoot{%
%    \end{macrocode}
% 
% Note that we have to get a bit clever to ensure that the page number is properly centered in the page footer. The naive solution is just to use \cs{hfill}s, but this will cause the page number to shift around depending on the length of \cs{@pnumber}. The \textsf{fancyhdr} package will do this for us, but it seems a bit of overkill to require this package merely so that we can ensure that the page number remains centered. Instead, we've borrowed the relevant technique from that class, which draws the left, centre and right bits of the footer as three overlapping boxes. You just have to be careful that the value of \cs{@pnumber} isn't so long that it overwrites the page number.
%    \begin{macrocode}
    \rlap{\parbox[b]{\columnwidth}{\raggedright\@draft\strut}}\hfill%
    \parbox[b]{\columnwidth}{\centering\textrm{\thepage}\strut}\hfill%
    \llap{\parbox[b]{\columnwidth}{\raggedleft\textbf{%
%    \end{macrocode}
% 
% An additional complication is that only the very last page of the document should have ``\textbf{END}'' displayed in the footer instead of ``\textbf{TURN OVER}''. This turns out to be a lot harder than you might think! Simply applying a page style \cs{AtEndDocument} works fine, \emph{except} when you have unprocessed floats at the end of the document. These are placed on an automatically-generated float page at the end, which \cs{pagestyle} and \cs{thispagestyle} don't affect. That is, a \textsf{lastpage} page style simply won't work in this situation.
% 
% The solution is therefore to not use a page style, but rather check in the \textsf{plain} page style whether or not this is the last page. Again, this isn't as simple as it sounds! The logic goes like this:
% 
% \begin{itemize}
% 
% 	\item If there are no unprocessed floats, meaning that the last page of the document will be a text page (indicated by \cs{@lastpagetexttrue}), then output ``\textbf{END}''.
% 	
% 	\item If there are unprocessed floats, meaning that the last page of the document will be a float page (indicated by \cs{@lastpagefloattrue}), then:
% 	
% 	\begin{itemize}
% 		
% 		\item If the float page is currently being generated (indicated by \cs{@fcolmadetrue} from \texttt{latex.ltx}), output ``\textbf{END}''.
% 	
% 		\item Otherwise, output ``\textbf{TURN OVER}''.
% 	
% 	\end{itemize}
% 	
% 	\item Otherwise, output ``\textbf{TURN OVER}''.
% 
% \end{itemize}
% 
% Note that this currently fails if more than one float page needs to be generated at the end of the document. \emph{All} such pages will have ``\textbf{END}'' printed at the bottom.
%    \begin{macrocode}
      \if@lastpagetext END%
      \else\if@lastpagefloat\if@fcolmade END%
                            \else TURN OVER\fi%
            \else TURN OVER\fi%
      \fi}}}%
    }%
  \let\@evenfoot\@oddfoot%
}
%    \end{macrocode}
% \end{macro}
%
% \textsf{plain} is the default page style:
%    \begin{macrocode}
\pagestyle{plain}
%    \end{macrocode}
% 
% \cs{AtEndDocument}, we need to set the \texttt{lastpage} switches accordingly. The \cs{@deferlist} macro from \texttt{latex.ltx} holds a list of any unprocessed floats, so if this is empty, we know that there's nothing more to do and can just set \cs{@lastpagetexttrue}. If the list still contains unprocessed floats, set \cs{@lastpagefloattrue} then \cs{clearpage} to force the floats out. We also need to decrement the page counter by one, otherwise it ends up being one too many, presumably as a result of the \cs{clearpage}.
% 
% I suspect that this will fail if the document uses \textsf{twocolumn} and has double floats that span both columns, because there is a separate \cs{@dbldeferlist}. Also, as noted above, this currently doesn't account for multiple float pages at the end of the document.
%    \begin{macrocode}
\AtEndDocument{%
  \ifx\@deferlist\@empty%
    \@lastpagetexttrue%
  \else%
    \@lastpagefloattrue%
  	\clearpage%
  	\addtocounter{page}{-1}%
  \fi%
}
%    \end{macrocode}
% 
% \begin{macro}{\ps@titlepage}
% The \textsf{titlepage} page style is similar to \textsf{plain}, but without the header.
%    \begin{macrocode}
\def\ps@titlepage{%
  \def\@oddhead{\@markingschedule}%
  \let\@evenhead\@oddhead%
  \def\@oddfoot{\@draft\hfill\textbf{TURN OVER}}%
  \let\@evenfoot\@oddfoot%
}
%    \end{macrocode}
% \end{macro}
% 
% 
% \subsection{Various counters}
% 
% Earlier versions of \textsf{ouexam} used the \textsf{enumerate}
% environment to build questions, which meant that special counters were not
% required. This version of the class, however, uses a different mechanism for
% building questions, so we need to define custom counters for several things.
% 
% \begin{macro}{question}
% \begin{macro}{subquestion}
% \changes{2.0}{2000/09/11}{NJS New \texttt{subquestion} counter.}
% \changes{2.0.1}{2000/04/30}{NJS Removed prefixed question number from
% \cs{labelsubquestion} macro so that references now print as just ``(a)''
% instead of ``1(a)''.}
% \begin{macro}{subsubquestion}
% \changes{2.0}{2000/09/11}{NJS New \texttt{subsubquestion} counter.}
% \changes{2.0.1}{2000/04/30}{NJS The \texttt{subsubquestion} was incorrectly
% printing as ``(a)'' rather than ``(i)''. Fixed.}
% \changes{2.0.1}{2000/04/30}{NJS Removed prefixed question numbers from
% \cs{labelsubsubquestion} macro so that references now print as just ``(i)''
% instead of ``1(a)(i)''.}
% First, we define three counters for the three possible levels of question:
% top-level question, second-level sub-question or third-level sub-sub-question:
%    \begin{macrocode}
\newcounter{question}
\newcounter{subquestion}[question]
\newcounter{subsubquestion}[subquestion]
%    \end{macrocode}
% The three question levels are numbered as 1., (a) and (i) respectively, so we
% also need to redefine the associated \cs{the} and \cs{label} macros
% appropriately:
%    \begin{macrocode}
\renewcommand{\thequestion}{\arabic{question}}
\newcommand{\labelquestion}{\hfil\arabic{question}.}

\renewcommand{\thesubquestion}{(\alph{subquestion})}
\newcommand{\labelsubquestion}{\hfil(\alph{subquestion})}

\renewcommand{\thesubsubquestion}{(\roman{subsubquestion})}
\newcommand{\labelsubsubquestion}{\hfil(\roman{subsubquestion})}
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
% 
% \begin{macro}{examexpected}
% \changes{2.0}{2000/09/11}{NJS New \texttt{examexpected} counter.}
% The |examexpected| counter stores the total expected number of marks for the
% examination, and defaults to 100:
%    \begin{macrocode}
\newcounter{examexpected}
\setcounter{examexpected}{100}
%    \end{macrocode}
% \end{macro}
% \begin{macro}{examoutof}
% \changes{2.0}{2000/09/11}{NJS New \texttt{examoutof} macro.}
% To specify that an examination is out of some number of marks other than
% 100, use the \cs{examoutof} macro to set the value:
%    \begin{macrocode}
\newcommand{\examoutof}[1]{\setcounter{examexpected}{#1}}
%    \end{macrocode}
% \end{macro}
% 
% \begin{macro}{sectexpected}
% \changes{2.0}{2002/01/09}{NJS New \cs{sectexpected} counter.}
% \begin{macro}{qexpected}
% \changes{2.0}{2000/09/11}{NJS New \cs{qexpected} counter.}
% \begin{macro}{subqexpected}
% \changes{2.0}{2000/09/11}{NJS New \texttt{subqexpected} counter.}
% \begin{macro}{subsubqexpected}
% \changes{2.0}{2000/09/11}{NJS New \texttt{subsubqexpected} counter.}
% The next four counters perform the same function as the |examexpected| counter
% for sections, questions, sub-questions and sub-sub-questions respectively. No
% macros are required to set these values as they are set automatically by the
% question-building environments.
% 
%    \begin{macrocode}
\newcounter{sectexpected}
\newcounter{qexpected}
\newcounter{subqexpected}
\newcounter{subsubqexpected}
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% 
% \begin{macro}{lastexpected}
% \changes{2.0}{2000/09/11}{NJS New \texttt{lastexpected} counter.}
% The |lastexpected| counter records the expected marks total for the last
% question. This will eventually be used to verify that the marks total for a
% marking schedule equals the marks total for the corresponding question, but
% this hasn't been implemented yet.
% 
%    \begin{macrocode}
\newcounter{lastexpected}
%    \end{macrocode}
% \end{macro}
% 
% \begin{macro}{examrunning}
% \changes{2.0}{2000/09/11}{NJS New \texttt{examrunning} counter.}
% \begin{macro}{sectrunning}
% \changes{2.0}{2002/01/09}{NJS New \texttt{sectrunning} counter.}
% \begin{macro}{qrunning}
% \changes{2.0}{2000/09/11}{NJS New \texttt{qrunning} counter.}
% \begin{macro}{subqrunning}
% \changes{2.0}{2000/09/11}{NJS New \texttt{subqrunning} counter.}
% These four counters keep track of the running total of marks for the
% current examination, section, question and subquestion respectively. This
% is later compared against the corresponding expected total. A running total
% is not needed for sub-sub-questions because they do not have sub-parts. The
% |qrunning| and |subqrunning| counters are reset when the associated
% question counters are incremented.
% 
%    \begin{macrocode}
\newcounter{examrunning}
\setcounter{examrunning}{0}
\newcounter{sectrunning}
\newcounter{qrunning}[question]
\newcounter{subqrunning}[subquestion]
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% 
% \begin{macro}{hassubs}
% \changes{2.0}{2000/09/11}{NJS New \texttt{hassubs} counter.}
% \begin{macro}{hassubsubs}
% \changes{2.0}{2000/09/11}{NJS New \texttt{hassubsubs} counter.}
% These two counters are used to track whether questions and
% sub-questions have sub-parts. Counters are used rather than booleans because
% counters are set globally and booleans are not (or at least do not appear to
% be). Both counters reset when the associated question counters are
% incremented.
% 
%    \begin{macrocode}
\newcounter{hassubs}[question]
\newcounter{hassubsubs}[subquestion]
%    \end{macrocode}
% \end{macro}
% \end{macro}
% 
% \begin{macro}{qdepth}
% \changes{2.3.1}{2013/05/31}{NJS New \texttt{qdepth} counter.}
% This counter is incremented every time a \textsf{question}, \textsf{subquestion} or \textsf{subsubquestion} environment is opened, and decremented every time these environments close. It measures the nesting depth of questions, and is used to check that we do not have things like a \texttt{subquestion} not enclosed by a \texttt{question}, or a \texttt{question} nested inside another \texttt{question}. Note that this counter does \emph{not} reset inside any other environment.
% 
%    \begin{macrocode}
\newcounter{qdepth}
\setcounter{qdepth}{0}
%    \end{macrocode}
% \end{macro}
% 
% At the end of the document, check that the running total of marks for the
% examination matches what we were expecting:
% 
%    \begin{macrocode}
\AtEndDocument{%
  \ifnum\theexamexpected=\theexamrunning%
  \else\ClassWarning{ouexam}{%
    actual number of marks for exam (\theexamrunning) does not^^J%
    match expected number of marks (\theexamexpected)}%
  \fi%
}
%    \end{macrocode}
% 
% \begin{macro}{xsection}
% \changes{2.0}{2002/01/29}{NJS New \texttt{xsection} counter.}
% Finally, we set up a counter for the section ``number'' and define it so
% that it prints out as an upper case letter rather than a number. We could
% just redefine the |section| counter, but that interferes with the section
% numbering in the documentation, and we can't have that, can we? |:)|
% 
%    \begin{macrocode}
\newcounter{xsection}
\setcounter{xsection}{0}
\renewcommand{\thexsection}{\Alph{xsection}}
%    \end{macrocode}
% \end{macro}
%
% 
% \subsection{Question-building environments and associated items}
% 
% \changes{2.0}{2000/09/11}{NJS \textsf{question} environment no longer a
% redefinition of \textsf{enumerate}.}
% \changes{2.0}{2001/05/15}{NJS Added penalty handling to ensure correct
% placement of marks in output.}
% \changes{2.0}{2001/05/15}{NJS Total marks for a question are now printed only
% if the question has no sub-questions.}
% \begin{environment}{question}
% \changes{2.3.1}{2013/05/31}{NJS Check question depth when opening the
% environment.}
% The \textsf{question} environment specifies a top-level question, and produces
% questions numbered in the form ``1.'', ``2.'', etc. It has one mandatory
% argument, the number of marks allocated to the question, which is used to
% initialise the |qexpected| counter\footnote{If it were not for the fact that
% you can only refer to environment arguments in the environment preamble, the
% \texttt{qexpected} counter would be unnecessary.}. If the argument is left
% empty, default to zero for the number of marks:
% 
%    \begin{macrocode}
\newenvironment{question}[1]{%
  \def\@nummarks{#1}%
  \ifx\@nummarks\@empty\setcounter{qexpected}{0}%
  \else\setcounter{qexpected}{#1}\fi%
%    \end{macrocode}
% 
% Check that the environment is being opened at the correct depth. The \textsf{question} environment should normally only be opened at depth 0:
%    \begin{macrocode}
  \ifnum\theqdepth=0%
  \else\ClassError{ouexam}{%
    question environment nested inside another question,^^J%
    subquestion or subsubquestion environment}%
  \fi%
  \addtocounter{qdepth}{1}%
%    \end{macrocode}
% 
% For each top-level question, we increment the |question| counter, which also
% resets the |hassubs| and |qrunning| counters to zero:
%    \begin{macrocode}
  \refstepcounter{question}%
%    \end{macrocode}
% The question itself is built using a single-item \textsf{list} environment.
%    \begin{macrocode}
  \begin{list}{\labelquestion}{\settowidth{\labelwidth}{88.}}%
    \item \ignorespaces%
%    \end{macrocode}
% 
% When the environment closes, the following things happen:
% \begin{enumerate}
% 	\item If the question has no sub-questions, the value of the |qexpected|
% 	counter is added to |examrunning| and |sectrunning|, and both |qrunning|
% 	and |lastexpected| are set to the value of |qexpected|.
% 
%    \begin{macrocode}
}{%
    \ifnum\thehassubs=0%
      \addtocounter{examrunning}{\value{qexpected}}%
      \addtocounter{sectrunning}{\value{qexpected}}%
      \setcounter{qrunning}{\value{qexpected}}%
%    \end{macrocode}
% 
% 	The total number of marks for the question is then printed right-justified
% 	on the line as ``(\emph{m} marks)'' where \emph{m} is the value of the
% 	|qrunning| counter. The environment determines whether to print ``mark'' or
% 	``marks'' automatically, and figures out whether the number of marks will fit
% 	on the last line of the question or needs to be placed on the next line.
% 	The code for handling the line breaking is derived from an example on page
% 	106 of \emph{The \TeX{}book}:
% 
%    \begin{macrocode}
      \unskip\nobreak\hfil\penalty50\hskip2em\hbox{}\nobreak%
      \hfil(\theqrunning~\ifnum\theqrunning=1 mark\else marks\fi)%
      \parfillskip=0pt \finalhyphendemerits=0 \par%
%    \end{macrocode}
% 
% 	\item If the question \emph{does} have sub-questions, then |qrunning|,
% 	|examrunning| and |sectrunning| have already been set by the various
% 	sub-environments. The value of |qrunning| is then compared with
% 	|qexpected|, and a warning is raised if they do not match. The total number
% 	of marks for the question is \emph{not} printed in this case.
% 
%    \begin{macrocode}
    \else%
      \ifnum\theqrunning=\theqexpected%
      \else\ClassWarning{ouexam}{%
        actual mark (\theqrunning) for question %
        \thequestion\space doesn't match^^J%
        expected mark (\theqexpected)}%
      \fi%
    \fi%
%    \end{macrocode}
% 
% 	\item |lastexpected| is set to the value of |qexpected| so it can be used
% 	in any subsequent \textsf{marking} environment.
% 
%    \begin{macrocode}
    \setcounter{lastexpected}{\value{qexpected}}%
  \end{list}%
%    \end{macrocode}
% 
% 	\item |qdepth| is decremented.
% 
%    \begin{macrocode}
  \addtocounter{qdepth}{-1}%
}
%    \end{macrocode}
% 
% \end{enumerate}
% \end{environment}
% 
% \begin{environment}{subquestion}
% \changes{2.0}{2000/09/11}{NJS \textsf{subquestion} environment no longer a
% redefinition of \textsf{enumerate}.}
% \changes{2.0}{2001/05/13}{NJS Added penalty handling to ensure correct placement
% of marks in output.}
% \changes{2.3.1}{2013/05/31}{NJS Check question depth when opening the
% environment.}
% The \textsf{subquestion} environment is the analogue of \textsf{question} for
% sub-questions. On initialisation, it sets |subqexpected| to the value of the
% mandatory argument (defaulting to zero if the argument is empty), increments
% the |hassubs| counter so that the enclosing \textsf{question} environment can
% react appropriately, increments |subquestion| (thus resetting |hassubsubs| and
% |subqrunning| to zero) and opens a single-item list with numbering of the form
% ``(a)'', ``(b)'', etc.
% 
%    \begin{macrocode}
\newenvironment{subquestion}[1]{%
  \def\@nummarks{#1}%
  \ifx\@nummarks\@empty\setcounter{subqexpected}{0}%
  \else\setcounter{subqexpected}{#1}\fi%
  \setcounter{subqexpected}{#1}%
  \addtocounter{hassubs}{1}%
  \refstepcounter{subquestion}%
%    \end{macrocode}
% The \textsf{subquestion} environment should normally only be opened at depth 1:
%    \begin{macrocode}
  \ifnum\theqdepth=1%
  \else\ClassError{ouexam}{%
    subquestion environment either outside a question^^J%
    environment, or nested inside another subquestion or subsubquestion^^J%
    environment}%
  \fi%
  \addtocounter{qdepth}{1}%
  \begin{list}{\labelsubquestion}{\settowidth{\labelwidth}{(m)}}%
    \item \ignorespaces%
}{%
%    \end{macrocode}
% 
% When the environment closes, if performs similar checks to the
% \textsf{question} environment. If the sub-question has no sub-sub-questions,
% the value of |subqexpected| is added to |qrunning| and |sectrunning|:
% 
%    \begin{macrocode}
    \ifnum\thehassubsubs=0%
      \addtocounter{examrunning}{\value{subqexpected}}%
      \addtocounter{sectrunning}{\value{subqexpected}}%
      \addtocounter{qrunning}{\value{subqexpected}}%
      \setcounter{subqrunning}{\value{subqexpected}}%
%    \end{macrocode}
% Then the number of marks for the sub-question are typeset in a similar manner
% to the \textsf{question} environment:
%    \begin{macrocode}
      \unskip\nobreak\hfil\penalty50\hskip2em\hbox{}\nobreak%
      \hfil(\thesubqrunning~\ifnum\thesubqrunning=1 mark\else marks\fi)%
      \parfillskip=0pt \finalhyphendemerits=0 \par%
%    \end{macrocode}
% 
% If the sub-question \emph{does} have sub-sub-questions, check the running total
% against the expected number and raise an error if they don't match.
%    \begin{macrocode}
    \else%
      \ifnum\thesubqrunning=\thesubqexpected%
      \else\ClassWarning{ouexam}{%
        actual mark (\thesubqrunning) for question %
        \thequestion\thesubquestion\space doesn't match^^J%
        expected mark (\thesubqexpected)}%
      \fi%
    \fi%
%    \end{macrocode}
% Finally, set |lastexpected| to the value of |qexpected| so it can be used
% in any subsequent \textsf{marking} environment.
%    \begin{macrocode}
    \setcounter{lastexpected}{\value{subqexpected}}%
  \end{list}%
  \addtocounter{qdepth}{-1}%
}
%    \end{macrocode}
% \end{environment}
% 
% \begin{environment}{subsubquestion}
% \changes{2.0}{2000/09/11}{NJS \textsf{subsubquestion} environment no longer a
% redefinition of \textsf{enumerate}.}
% \changes{2.0}{2001/05/13}{NJS Added penalty handling to ensure correct placement
% of marks in output.}
% \changes{2.3.1}{2013/05/31}{NJS Check question depth when opening the
% environment.}
% This is similar to both \textsf{question} and \textsf{subquestion}, and
% generates an appropriately numbered sub-sub-question (``(i)'', ``(ii)'', etc.).
%    \begin{macrocode}
\newenvironment{subsubquestion}[1]{%
  \def\@nummarks{#1}%
  \ifx\@nummarks\@empty\setcounter{subsubqexpected}{0}%
  \else\setcounter{subsubqexpected}{#1}\fi%
%    \end{macrocode}
% Increment |hassubsubs| so that the enclosing \textsf{subquestion} environment
% can react appropriately:
%    \begin{macrocode}
  \addtocounter{hassubsubs}{1}%
  \refstepcounter{subsubquestion}%
%    \end{macrocode}
% The \textsf{subsubquestion} environment should normally only be opened at depth 2:
%    \begin{macrocode}
  \ifnum\theqdepth=2%
  \else\ClassError{ouexam}{%
    subsubquestion environment either outside a subquestion^^J%
    environment, or nested inside another subsubquestion environment}%
  \fi%
  \addtocounter{qdepth}{1}%
  \begin{list}{\labelsubsubquestion}{\settowidth{\labelwidth}{(viii)}}%
    \item \ignorespaces%
}{%
%    \end{macrocode}
% 
% Closing this environment is a bit simpler than the previous two because we
% don't need to check whether the sub-sub-question has any sub-components (this
% never happens). All we have to do is set the appropriate counters and typeset
% the number of marks.
% 
%    \begin{macrocode}
    \addtocounter{examrunning}{\value{subsubqexpected}}%
    \addtocounter{sectrunning}{\value{subsubqexpected}}%
    \addtocounter{qrunning}{\value{subsubqexpected}}%
    \addtocounter{subqrunning}{\value{subsubqexpected}}%
    \setcounter{lastexpected}{\value{subsubqexpected}}%
    \unskip\nobreak\hfil\penalty50\hskip2em\hbox{}\nobreak%
    \hfil(\thesubsubqexpected~\ifnum\thesubsubqexpected=1 mark%
                              \else marks\fi)%
    \parfillskip=0pt \finalhyphendemerits=0 \par%
  \end{list}%
  \addtocounter{qdepth}{-1}%
}
%    \end{macrocode}
% \end{environment}
% 
% 
% \subsection{Draft examination printing}
% 
% \begin{macro}{\@marking}
% \changes{2.0}{2001/05/17}{NJS New \cs{@draft} macro.}
% Specifying the \textsf{draft} option causes \textsf{ouexam} to print
% \fbox{DRAFT} in large letters in the footer of every page. This is drawn by
% the \cs{@draft} macro, which is included in the page footer definition of
% all the page styles for this class (see
% \hyperref[pagestyles]{section~\ref*{pagestyles}}). \textbf{Note:} Because
% the \cs{@draft} macro is included in the page footer, it will usually
% overflow the page boundaries, causing ``|Overfull \vbox|'' warnings. Note
% the extra set of braces to limit the scope of the \cs{Huge}.
% 
%    \begin{macrocode}
\if@draft\def\@draft{{\Huge\fbox{DRAFT}}}
\else\let\@draft\@empty
\fi
%    \end{macrocode}
% \end{macro}
% 
% 
% \changes{2.2}{2010/04/26}{NJS Added time allocation macros.}
% \subsection{Time allocations}
% 
% \begin{macro}{\timeallocation}
% The \cs{timeallocation} macro takes a single argument representing the number of minutes suggested time allocation for a question, and outputs it in the form: \textbf{[The suggested time allocation for this question is \emph{t} minutes.]} The argument is not checked and can be anything (numbers, text). The macro will need to be redefined if either units other than minutes or different phrasing are required.
%    \begin{macrocode}
\newcommand{\timeallocation}[1]{%
  \textbf{[The suggested time allocation for answering this question %
           is #1 minutes.]}%
}
%    \end{macrocode}
% \end{macro}
% 
% \begin{macro}{\readingtime}
% The \cs{readingtime} macro takes a single argument representing the number of minutes reading time allocated to some part of the examination, and outputs it in the form: \textbf{[Allow \emph{t} minutes reading time.]} The argument is not checked and can be anything (numbers, text). The macro will need to be redefined if either units other than minutes or different phrasing are required.
%    \begin{macrocode}
\newcommand{\readingtime}[1]{\textbf{[Allow #1 minutes reading time.]}}
%    \end{macrocode}
% \end{macro}
% 
% \textsf{ouexam} currently does not check that any of these time allocations sum to the correct total. If implemented, this would work in a similar way to how marks totals are checked, and would be keyed in some way off \cs{@hours}.
% 
% 
% \subsection{Marking schedule information}
% 
% \begin{environment}{marking}
% \begin{macro}{\@markingschedule}
% \changes{1.2}{1999/10/26}{NJS New \textsf{marking} environment for including
% answers and/or marking schedule information.}
% \changes{2.0}{2000/09/11}{NJS Now prints ``MARKING SCHEDULE'' at the bottom
% of every page if the \textsf{markingschedule} class option is specified.}
% The \textsf{marking} environment allows the specification of marking
% schedule information in the same location as the questions. If the
% \textsf{markingschedule} class option is specified, marking schedule
% information is printed in italics. In addition, \fbox{MARKING SCHEDULE} is
% printed in large letters in the header of every page. This is drawn by the
% \cs{@markingschedule} macro, which is included in the page header
% definition of all the page styles for this class (see
% \hyperref[pagestyles]{section~\ref*{pagestyles}}). Note the extra set of
% braces to limit the scope of the \cs{Huge}.
% 
%    \begin{macrocode}
\if@markingschedule
  \newenvironment{marking}{\itshape}{\normalfont}
  \def\@markingschedule{{\Huge\fbox{MARKING SCHEDULE}}}
\else
%    \end{macrocode}
% 
% Normally \textsf{marking} just maps to the \textsf{comment} environment from
% Rainer Sch\"{o}pf's \textsf{verbatim} package, i.e., marking information is not
% printed:
% 
%    \begin{macrocode}
  \let\marking\comment
  \let\endmarking\endcomment
  \let\@markingschedule\@empty
\fi
%    \end{macrocode}
% \end{macro}
% \end{environment}
% 
% 
% \subsection{Section handling}
% 
% \changes{2.0}{2002/01/09}{NJS new \textsf{examsection} environment.}
% \changes{2.0}{2002/01/09}{NJS New \cs{@defsecinst} macro.}
% \begin{environment}{examsection}
% \begin{macro}{\@defsecinst}
% The \textsf{examsection} environment specifies a major section of an
% examination paper. Sections are sequentially numbered ``A'', ``B'', etc. The
% environment has three arguments:
% 
% \begin{enumerate}
% 	\item The expected number of marks for the section.
% 
% 	\item The instructions for this section. This defaults to ``ANSWER ALL
%	QUESTIONS.'' if left blank. To change the default text, redefine the
%	\cs{@defsecinst} macro.
%    \begin{macrocode}
\def\@defsecinst{ANSWER ALL QUESTIONS.}
%    \end{macrocode}
% 
% 	\item A description of the contents of the section which may be left blank.
% \end{enumerate}
% 
%    \begin{macrocode}
\newenvironment{examsection}[3]{%
%    \end{macrocode}
% Every section begins on a new page.
%    \begin{macrocode}
  \newpage%
%    \end{macrocode}
% If the argument for the number of marks is left empty, default to zero.
%    \begin{macrocode}
  \def\@nummarks{#1}%
  \ifx\@nummarks\@empty\setcounter{sectexpected}{0}%
  \else\setcounter{sectexpected}{#1}\fi%
  \refstepcounter{xsection}%
  \setcounter{sectrunning}{0}%
%    \end{macrocode}
% The section title is formatted as ``\textbf{\underline{Section A}}'', in
% \cs{large} size.
%    \begin{macrocode}
  {\large\noindent\textbf{\underline{Section~\thexsection}}}%
  \\[0.5\baselineskip]%
  \def\@usersecinst{#2}%
  \ifx\@usersecinst\@empty\@defsecinst\else\@usersecinst\fi\par%
  \def\@usersectopic{#3}%
  \ifx\@usersectopic\@empty\else\par\noindent\@usersectopic\fi \\%
}{%
%    \end{macrocode}
% 
% When the environment closes, the value of |sectrunning| is compared against
% |sectexpected|, and a warning is raised if they do not match. The total number
% marks for the section is printed in the form ``\textbf{[SECTION A TOTAL
% \emph{m} MARKS]}'' where \emph{m} is the value of |sectrunning|.
% 
%    \begin{macrocode}
  \ifnum\thesectrunning=\thesectexpected%
  \else\ClassWarning{ouexam}{%
    actual mark (\thesectrunning) for section %
    \thexsection\space doesn't match^^J%
    expected mark (\thesectexpected)}%
  \fi%
  \bigskip\hfill\textbf{[SECTION \thexsection\ TOTAL %
                              \thesectrunning\ MARKS]}%
}
%    \end{macrocode}
% \end{macro}
% \end{environment}
% 
% 
% \subsection{Title page generation}
% 
% \begin{macro}{\examyear}
% \begin{macro}{\@year}
% \changes{2.0.2}{2002/08/22}{NJS Put back \@eha at end of the \ClassError. Don't know what this does, but the macro crashes without it there. It seems to be required if there's a \cs{protect}.}
% The \cs{examyear} macro specifies the year in which the examination is being held. It redefines the \cs{@year} macro which is used in \cs{@maketitlepage}:
%    \begin{macrocode}
\newcommand{\examyear}[1]{\def\@year{#1}}
%    \end{macrocode}
% This macro is required:
%    \begin{macrocode}
\def\@year{%
  \ClassError{ouexam}{no \protect\examyear\space was specified}\@eha%
}
%    \end{macrocode}
% \end{macro}
% \end{macro}
% 
% \begin{macro}{\department}
% \begin{macro}{\@dept}
% \changes{2.0.2}{2002/08/22}{NJS Put back \@eha at end of the \ClassError. Don't know what this does, but the macro crashes without it there. It seems to be required if there's a \cs{protect}.}
% The \cs{department} macro specifies the name of the department, and is required. It redefines the \cs{@dept} macro which is used in \cs{@maketitlepage}:
%    \begin{macrocode}
\newcommand{\department}[1]{\def\@dept{#1}}
\def\@dept{%
  \ClassError{ouexam}{no \protect\department\space was specified}\@eha%
}
%    \end{macrocode}
% \end{macro}
% \end{macro}
% 
% \begin{macro}{\papernumber}
% \begin{macro}{\@pnumber}
% \changes{2.0.2}{2002/08/22}{NJS Put back \@eha at end of the \ClassError. Don't know what this does, but the macro crashes without it there. It seems to be required if there's a \cs{protect}.}
% The \cs{papernumber} macro specifies the paper number, and is required. It redefines the \cs{@pnumber} macro which is used in \cs{@maketitlepage}:
%    \begin{macrocode}
\newcommand{\papernumber}[1]{\def\@pnumber{#1}}
\def\@pnumber{%
  \ClassError{ouexam}{no \protect\papernumber\space was specified}\@eha%
}
%    \end{macrocode}
% \end{macro}
% \end{macro}
% 
% \begin{macro}{\papertitle}
% \begin{macro}{\@ptitle}
% \changes{2.0.2}{2002/08/22}{NJS Put back \@eha at end of the \ClassError. Don't know what this does, but the macro crashes without it there. It seems to be required if there's a \cs{protect}.}
% The \cs{papertitle} macro specifies the title of the paper, and is required. It redefines the \cs{@ptitle} macro which is used in \cs{@maketitlepage}:
%    \begin{macrocode}
\newcommand{\papertitle}[1]{\def\@ptitle{#1}}
\def\@ptitle{%
  \ClassError{ouexam}{no \protect\papertitle\space was specified}\@eha%
}
%    \end{macrocode}
% \end{macro}
% \end{macro}
% 
% \begin{macro}{\semester}
% \begin{macro}{\@semester}
% \changes{2.0.2}{2002/08/22}{NJS Added full-year support to \cs{semester}.}
% \changes{2.0}{2002/01/14}{NJS Rewrote \cs{semester} to support summer school
% and special exams.}
% The \cs{semester} macro specifies which semester the examination is for.
% Legal values for the argument are ``1'', ``2'', ``SS'' for summer school,
% ``FY'' for full-year or ``SP'' for a special examination---anything else is
% ignored.
%    \begin{macrocode}
\def\@ONE{1}
\def\@TWO{2}
\def\@SS{SS}
\def\@FY{FY}
\def\@SP{SP}
%    \end{macrocode}
% This macro redefines the \cs{@semester} macro which is used in
% \cs{@maketitlepage}:
%    \begin{macrocode}
\newcommand{\semester}[1]{%
  \def\@sem{#1}%
  \ifx\@sem\@ONE\def\@semester{Semester One}%
  \else\ifx\@sem\@TWO\def\@semester{Semester Two}%
  \else\ifx\@sem\@SS\def\@semester{Summer School}%
  \else\ifx\@sem\@FY\def\@semester{Full Year}%
  \else\ifx\@sem\@SP\def\@semester{Special Examination}%
  \else\ClassWarning{ouexam}{%
    invalid value `#1' for \protect\semester;^^J%
    valid values are `1', `2', `SS', `FY' and `SP'. No semester^^J%
    information will be printed}%
  \fi\fi\fi\fi\fi%
}
%    \end{macrocode}
% \cs{semester} is optional---if you omit it or give it an invalid argument,
% \cs{@semester} remains empty.
%    \begin{macrocode}
\let\@semester\@empty
%    \end{macrocode}
% \end{macro}
% \end{macro}
% 
% \begin{macro}{\timeallowed}
% \begin{macro}{\@hours}
% The \cs{timeallowed} macro specifies the length of the examination in hours.
% It redefines the \cs{@hours} macro which is used in \cs{@maketitlepage}:
%    \begin{macrocode}
\newcommand{\timeallowed}[1]{\def\@hours{#1}}
%    \end{macrocode}
% \cs{timeallowed} is optional---if you omit it, \cs{@hours} defaults to ``3'':
%    \begin{macrocode}
\def\@hours{3}
%    \end{macrocode}
% \end{macro}
% \end{macro}
% 
% \begin{macro}{num.pages}
% \changes{1.1}{1999/04/20}{NJS Renamed \texttt{page.last} label to \texttt{num.pages}.}
% \changes{2.4}{2013/06/05}{NJS Added check for \textsf{hyperref} being loaded.}
% There is no macro to specify the number of pages in the examination because it is done automatically by inserting a \cs{newlabel} that refers to the last page directly into the |.aux| file. The \textsf{hyperref} package redefines the \cs{newlabel} macro (in particular it increases the number of arguments), so we need to test whether \textsf{hyperref} is loaded and write the appropriate string to the |.aux| file. We borrow a technique from the \textsf{pageslts} package of checking for the existence of \cs{Hy@Warning}. (Ideally we would use \cs{if@packageloaded}, but this can't be used after \cs{AtBeginDocument}, and can be foiled by a package loading \textsf{hyperref} \cs{AtBeginDocument}.)
%    \begin{macrocode}
\AtEndDocument{%
  \@ifundefined{Hy@Warning}{% hyperref not loaded
    \immediate\write\@auxout{\string\newlabel{num.pages}{{}{\thepage}}}%
  }{% hyperref loaded
    \immediate\write\@auxout{\string\newlabel{num.pages}{{}{\thepage}{}{}{}}}%
  }%
}
%    \end{macrocode}
% This label is then referenced in \cs{@maketitlepage}.
% \end{macro}
% 
% \begin{macro}{num.questions}
% \changes{1.1}{1999/04/20}{NJS Added \cs{num.questions} label representing the number of questions.}
% \changes{2.4}{2013/06/05}{NJS Added check for \textsf{hyperref} being loaded.}
% Similarly, there is no macro to specify the number of questions in the examination. This is calculated using the same method as for the number of pages.
%    \begin{macrocode}
\AtEndDocument{%
  \@ifundefined{Hy@Warning}{%
    \immediate\write\@auxout{\string\newlabel{num.questions}{{}{\thequestion}}}%
  }{%
    \immediate\write\@auxout{\string\newlabel{num.questions}{{}{\thequestion}{}{}{}}}%
  }%
}
%    \end{macrocode}
% \end{macro}
% 
% \begin{macro}{\instructions}
% \begin{macro}{\@instructions}
% The \cs{instructions} macro specifies instructions on how candidates should
% answer questions. It redefines the \cs{@instructions} macro that is used in
% \cs{@maketitlepage}:
%    \begin{macrocode}
\def\@instructions{Answer \underline{ALL} questions.}
%    \end{macrocode}
% \cs{instructions} is optional. If you omit it, \cs{@instructions} defaults
% to ``Answer \underline{ALL} questions.'' 
%    \begin{macrocode}
\newcommand{\instructions}[1]{\def\@instructions{#1}}
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \begin{macro}{\@NA}
% \changes{2.1.3}{2009/04/28}{NJS Added \cs{@NA} macro.}
% The \cs{@NA} macro enables us to easily assign and test for the default value
% (``N/A'') for some of the title page macros. (There is certain text that
% should not be output when particular macros are allowed to default.)
%    \begin{macrocode}
\def\@NA{N/A}
%    \end{macrocode}
% \end{macro}
% 
% \begin{macro}{\material}
% \begin{macro}{\@material}
% \changes{2.1.3}{2009/04/28}{NJS Changed default to ``N/A'' for \cs{material}.}
% The \cs{material} macro lets specifies any additional material that candidates
% are provided in addition to the examination paper itself, and is optional.
% It redefines the \cs{@material} macro that is used in \cs{@maketitlepage}:
%    \begin{macrocode}
\let\@material\@NA
\newcommand{\material}[1]{\def\@material{#1}}
%    \end{macrocode}
% If you omit it, it defaults to ``N/A''.
% \end{macro}
% \end{macro}
% 
% \begin{macro}{\allowcalculators}
% \begin{macro}{\@calculators}
% \changes{2.1}{2004/04/05}{NJS Modified calculator macros to conform to
% new University calculator regulations.}
% The \cs{allowcalculators} macro specifies whether or not calculators are
% allowed in the examination (\cs{permitcalculators} is a synonym). If
% \cs{allowcalculators} is omitted, \cs{@calculators} defaults to ``No
% calculators are permitted.'':
%    \begin{macrocode}
\def\@calculators{No calculators are permitted.}
%    \end{macrocode}
% Issuing an \cs{allowcalculators[none]} does nothing, as the default is
% to disallow calculators anyway. Issuing an \cs{allowcalculators} or
% \cs{allowcalculators[any]} redefines \cs{@calculators} to ``No
% restriction on the model of calculator to be used, but no device with
% communication capability shall be accepted as a calculator.'' Issuing an
% \cs{allowcalculators[approved]} redefines \cs{@calculators} to ``Only
% calculators on the University of Otago list of approved calculators are
% permitted.'':
%    \begin{macrocode}
\def\@CalcNONE{none}
\def\@CalcANY{any}
\def\@CalcAPPROVED{approved}
%    \end{macrocode}
% \changes{2.1.2}{2008/08/14}{NJS Ensured that the correct default is set
% regardless of whether the calculators macro is called.}
%    \begin{macrocode}
\let\@calc\@CalcNONE
\newcommand{\allowcalculators}[1][any]{%
  \def\@calc{#1}%
  \ifx\@calc\@CalcNONE%
  \else\ifx\@calc\@CalcANY\def\@calculators{No restriction on the %
    model of calculator to be used, but no device with communication %
    capability shall be accepted as a calculator.}%
  \else\ifx\@calc\@CalcAPPROVED\def\@calculators{Only calculators on %
    the University of Otago list of approved calculators are permitted.}%
  \else\ClassWarning{ouexam}{%
    invalid argument `#1' for^^J%
    \protect\allowcalculators; valid values are `none', `any' and `approved'.^^J%
    No calculators will be permitted}%
  \fi\fi\fi%
}
%    \end{macrocode}
% \cs{@calculators} is used in \cs{@maketitlepage}.
% \end{macro}
% \end{macro}
% 
% \begin{macro}{\permitcalculators}
% \changes{2.1}{2004/04/05}{NJS Added \cs{permitcalculators} as a synonym
% for \cs{allowcalculators}.}
% \cs{permitcalculators} is a synonym for \cs{allowcalculators}.
%    \begin{macrocode}
\let\permitcalculators\allowcalculators
%    \end{macrocode}
% \end{macro}
% 
% \begin{macro}{\copiesof}
% \begin{macro}{\@copiesof}
% \changes{2.1.3}{2009/04/28}{NJS Changed default to ``N/A'' for \cs{copiesof}.}
% The \cs{copiesof} macro specifies any material that candidates are allowed to
% bring into the examination, and is optional. It redefines the \cs{@copiesof}
% macro that is used in \cs{@maketitlepage}:
%    \begin{macrocode}
\let\@copiesof\@NA
\newcommand{\copiesof}[1]{\def\@copiesof{#1}}
%    \end{macrocode}
% \end{macro}
% \end{macro}
% 
% \begin{macro}{\otherinstructions}
% \begin{macro}{\@otherinst}
% \changes{2.1.3}{2009/04/28}{NJS Changed default to ``N/A'' for
% \cs{otherinstructions}.}
% The \cs{otherinstructions} macro specifies any other instructions not covered
% by any of the above, and is optional. It redefines the \cs{@otherinst}
% macro that is used in \cs{@maketitlepage}:
%    \begin{macrocode}
\let\@otherinst\@NA
\newcommand{\otherinstructions}[1]{\def\@otherinst{#1}}
%    \end{macrocode}
% \end{macro}
% \end{macro}
% 
% \begin{macro}{\maketitlepage}
% The \cs{maketitlepage} macro is the \textsf{ouexam} analogue of \cs{maketitle}.
% It takes all of the information provided by the above macros and generates a
% properly formatted examination title page. If you want to design your own title
% page styles, redefine \cs{@maketitlepage} (this is not normally recommended
% however).
%    \begin{macrocode}
\newcommand{\maketitlepage}{\@maketitlepage}
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\@maketitlepage}
% \changes{1.1}{1999/04/20}{NJS Renamed macro from \cs{maketitlepage}.}
% \changes{2.4}{2013/06/05}{NJS Added check for \textsf{hyperref} being loaded.}
% The \cs{@maketitlepage} macro generates an examination title page that meets the Otago University requirements for final examination papers. You can redefine this if you want to change the format, but normally you would not do this.
%    \begin{macrocode}
\def\@maketitlepage{%
  \thispagestyle{titlepage}%
%    \end{macrocode}
% The header information for the examination is printed centered at the top
% of the page. The department, paper number and paper title (and the semester
% information if required) are placed inside a double box.
%    \begin{macrocode}
  \begin{center}%
    {\Large \textbf{UNIVERSITY OF OTAGO EXAMINATIONS \@year}}%
    \\[\baselineskip]%
    \fbox{\framebox[\linewidth]{%
      \begin{tabular}{c}%
        \\%
        {\large \@dept} \\%
        \\%
        {\large Paper \@pnumber} \\%
        \\%
        {\large \@ptitle} \\%
        \ifx\@semester\@empty\else\@semester \\ \fi%
        \\%
      \end{tabular}%
    }}%
    \mbox{}\\[\baselineskip]%
%    \end{macrocode}
% Next, the time allowed for completing the examination:
%    \begin{macrocode}
    \textbf{(TIME ALLOWED: \@hours\ HOURS)}%
    \\[\baselineskip]%
  \end{center}%
%    \end{macrocode}
% The number of pages is automatically calculated, as described earlier. All we need to do is \cs{pageref} the |num.pages| label that we wrote into the |.aux| file, or \cs{pageref*} if \textsf{hyperref} is loaded to inhibit hyperlink generation.
%    \begin{macrocode}
  \underline{This examination paper comprises %
    \@ifundefined{Hy@Warning}{\pageref{num.pages}}{\pageref*{num.pages}} pages.%
  }%
  \\[\baselineskip]%
%    \end{macrocode}
% Print out the instructions for answering questions:
%    \begin{macrocode}
  \underline{Candidates should answer questions as follows:}%
  \\[\baselineskip]%
  \hspace*{1cm}\begin{minipage}{13.5cm}\@instructions\end{minipage}%
  \\[\baselineskip]%
%    \end{macrocode}
% Print out any provided material:
%    \begin{macrocode}
  \underline{The following material is provided:}%
  \\[\baselineskip]%
  \hspace*{1cm}\begin{minipage}{13.5cm}\@material\end{minipage}%
  \\[\baselineskip]%
%    \end{macrocode}
% Print out use of calculators section:
%    \begin{macrocode}
  \underline{Use of calculators:} \\[\baselineskip]%
  \hspace*{1cm}\begin{minipage}{13.5cm}\@calculators\end{minipage}%
  \\[\baselineskip]%
%    \end{macrocode}
% \changes{2.1.3}{2009/04/28}{NJS Fixed vertical spacing.}
% \changes{2.1.2}{2008/08/14}{NJS Ensure that the ``subject to inspection''
% line for calculators is only printed when calculators actually are allowed.}
% Only print out the ``subject to inspection'' line if calculators actually
% are allowed (irrelevant otherwise):
%    \begin{macrocode}
  \ifx\@calc\@CalcNONE%
  \else\mbox{}\hfill(Subject to inspection by the examiners.)%
    \\[\baselineskip]%
  \fi%
%    \end{macrocode}
% Print out anything that candidates are allowed to bring:
%    \begin{macrocode}
  \underline{Candidates are permitted copies of:}%
  \\[\baselineskip]%
  \hspace*{1cm}\begin{minipage}{13.5cm}\@copiesof\end{minipage}%
  \\[\baselineskip]%
%    \end{macrocode}
% \changes{2.1.3}{2009/04/28}{NJS Fixed vertical spacing.}
% \changes{2.1.2}{2008/08/14}{NJS Ensure that the ``subject to inspection''
% line for other material is only printed when there actually \emph{is}
% other material.}
% Only print out the ``subject to inspection'' line if there actually
% \emph{is} other material (irrelevant otherwise):
%    \begin{macrocode}
  \ifx\@copiesof\@NA%
  \else\mbox{}\hfill(Subject to inspection by the examiners.)%
    \\[\baselineskip]%
  \fi%
%    \end{macrocode}
% \changes{2.1.3}{2009/04/28}{NJS Removed superfluous blank line.}
% Print out other instructions, and finish.
%    \begin{macrocode}
  \underline{Other Instructions:}%
  \\[\baselineskip]%
  \hspace*{1cm}\begin{minipage}{13.5cm}\@otherinst\end{minipage}%
 
  \newpage%
}
%    \end{macrocode}
% \end{macro}
%
%
% \subsection{Obsolete macros and environments}
% 
% \changes{2.0}{2002/01/09}{NJS All \textsf{multichoice} macros and environments
% made obsolete.}
% \changes{2.0}{2002/01/09}{NJS \cs{markingschedule}, \cs{marks}, \cs{totalmarks}
% macros made obsolete.}
% \changes{2.0}{2002/01/09}{NJS \textsf{questions}, \textsf{subquestions},
% \textsf{subsubquestions} environments made obsolete.}
% The following macros and environments are no longer supported as of version
% 2.0. Documents that use these macros can be processed using ouexam v1.2 or
% earlier. The obsolete multiple-choice examination support is dealt with
% during option processing, so we can ignore the obsolete multiple-choice
% macros.
% 
% \begin{environment}{questions}
% \begin{environment}{subquestions}
% \changes{1.1}{1999/04/20}{NJS Now disabled when \textsf{multichoice} is
% active.}
% \begin{environment}{subsubquestions}
% \changes{1.1}{1999/04/20}{NJS Now disabled when \textsf{multichoice} is
% active.}
% \begin{macro}{markingschedule}
% \changes{1.2}{1999/10/26}{NJS New macro to turn marking information on and
% off.}
% The \textsf{questions}, \textsf{subquestions} and \textsf{subsubquestions}
% environments have been replaced by the \textsf{question},
% \textsf{subquestion} and \textsf{subsubquestion} environments respectively.
%   \begin{macrocode}
\newenvironment{questions}{%
  \def\item{\@unsupported{questions environment}{1.2}}}{}
\newenvironment{subquestions}{%
  \def\item{\@unsupported{subquestions environment}{1.2}}}{}
\newenvironment{subsubquestions}{%
  \def\item{\@unsupported{subsubquestions environment}{1.2}}}{}
%    \end{macrocode}
% The \cs{markingschedule} macro has been replaced by the
% \textsf{markingschedule} class option. We can't map this to the
% \cs{@unsupported} macro because \cs{markingschedule} occurs in the document
% preamble. Mapping it to \cs{@unsupported} causes \TeX\ to blow up for some
% reason. It's safe enough to ignore it.
%   \begin{macrocode}
\let\markingschedule\@empty
%    \end{macrocode}
% \end{macro}
% \end{environment}
% \end{environment}
% \end{environment}
% 
% \changes{2.1}{2004/04/16}{NJS Removed the \texttt{marks} macro entirely
% because it conflicts with a macro in e-TeX.}
% \changes{1.1}{1999/04/20}{NJS This macro now increments the \texttt{marks}
% counter.}
% \begin{macro}{\totalmarks}
% \changes{1.1}{1999/04/20}{NJS Removed the parameter for the number of marks;
% Both these macros have been replaced by automatic calculations within the
% question-building environments.}
% The \cs{marks} macro has been replaced by the argument to the various
% question-building environments. The \cs{totalmarks} macro has been replaced
% by automatic calculations within these environments.
%    \begin{macrocode}
\newcommand{\totalmarks}{\@unsupported{totalmarks macro}{1.2}}
%    \end{macrocode}
% \end{macro}
% 
%
% 
% \Finale
% 
% \PrintChanges

\def\ouexamdate{4 September 2012}
\def\ouexamversion{2.3}
\def\ouexamshortdate{2012/09/04}
% \iffalse meta-comment
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%
%% File: $Id$
%% Copyright 1999--2006 Nigel Stanger and University of Otago
%% 
%% You may use this package freely, and also distribute it
%% provided that you don't change it, make any money off
%% it or pretend that you wrote it.
%%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%<*driver>
\documentclass[a4paper]{ltxdoc}
\usepackage{doc}
\DisableCrossrefs
\CodelineNumbered
\RecordChanges

\usepackage{graphicx}
% change hyperref options as required
\usepackage[colorlinks,pdfpagemode=None,urlcolor=blue,
            linkcolor=red,pdfauthor={Nigel Stanger}]{hyperref}

\title{The \textsf{ouexam} document class, v\ouexamversion}
\author{Nigel Stanger\\nstanger@infoscience.otago.ac.nz}
\date{\ouexamdate}

\renewcommand{\thesubquestion}{(\alph{subquestion})}
\newcommand{\labelsubquestion}{\hfil(\alph{subquestion})}

Ignore Space Show notes View ouexam.ins
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% %% %% Copyright (c) 1999--2013 Nigel Stanger and University of Otago %% %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% \def\batchfile{ouexam.ins} \input docstrip \generateFile{ouexam.cls}{f}{\from{ouexam.dtx}{}} \Msg{***************************************************************} \Msg{The file ouexam.cls has been created. Move it to a directory in } \Msg{your TeX search path. You can now type `make doc' to produce the } \Msg{documentation. } \Msg{*************************************************************} %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% %% %% File: $Id$ %% Copyright (c) 1999--2002 Nigel Stanger and University of Otago %% %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% \def\batchfile{ouexam.ins} \input docstrip \generateFile{ouexam.cls}{f}{\from{ouexam.dtx}{}} \Msg{*************************************************************} \Msg{The file ouexam.cls has been created. Move it to a directory in } \Msg{your TeX search path. You can now type `make doc' to produce the } \Msg{documentation. } \Msg{***************************************************************}

Ignore Space Show notes View

ouexam.ins

Show line notes below