#!/usr/bin/perl
#
# ========================================================================
# @perl-file{
# author = "Alan Jeffrey",
# version = "0.19",
# date = "15 January 1996",
# time = "12:26:55 GMT",
# filename = "texfaq2html",
# address = "School of Cognitive and Computing Sciences
# University of Sussex
# Brighton BN1 9QH
# UK",
# FAX = "+44 1273 671320"
# email = "alanje@cogs.sussex.ac.uk",
# codetable = "ISO/ASCII",
# keywords = "LaTeX FAQ HTML",
# supported = "yes",
# abstract = "This perl script converts the UKTUG TeX FAQ
# LaTeX source document into HTML, on the fly."
# package = "stands alone",
# dependencies = "faqbody.tex, newfaq.aux, dirctan.tex,
# filectan,tex",
# }
# ========================================================================
# A script to provide a searchable WWW interface to the the UKTUG TeX
# FAQ file.
#
# The script takes parameters in the form of the QUERY_STRING environment
# variable.
# Copyright 1994 Alan Jeffrey
# 15 Sep 1994, v0.01: Created script.
#
# 16 Sep 1994, v0.02: Added the ability to download more then one
# question at once. Added some extra filters.
#
# 23 Sep 1994, v0.03: Updated to new markup syntax.
#
# 5 Oct 1994, v0.04: Updated to new markup syntax.
#
# 7 Oct 1994, v0.05: Added --, ---, \CTANref and \CTANlabel.
#
# 26 Oct 1994, v0.06: Added some more markup. Read CTAN definitions
# from separate files. Read the Qrefs from the aux file.
#
# 28 Oct 1994, v0.07: Added \@ and the optional argument to \Qref.
# Changed the banner.
#
# 8 Nov 1994, v0.08: Added footnoteenv. Changed the \Qref syntax to
# match Robin's. Added \ldots and \large. Added \ProgName. Added
# \&. Moved \cs before \Qref so you can have \cs's inside \Qref's.
# Renamed newfaq.tex to faqbody.tex.
#
# 25 Nov 1994, v0.09: Added \dots, $e$, \ISBN, \#, \ ,
# \begin{comment}...\end{comment}.
# \cs now takes its argument between |...|.
#
# 2 Dec 1994, v0.10: Added \protect and proglist.
#
# 6 Dec 1994, v0.11: Allow \fileversion and \filedate on the same
# line.
#
# 9 Mar 1995, v0.12: Made relative paths explicit for the new httpd.
#
# 15 Sep 1995, v0.13: Fixed a bug where I'd assumed that html-ignored
# material wasn't nested (in fact the document has a footnote
# inside an htmlignore).
#
# 23 Nov 1995, v0.14: Added \MP{}.
#
# 28 Nov 1995, v0.15: Added \TUGboat{} and \Package|...|.
#
# 2 Jan 1996, v0.16: Added \AllTeX{}, \acro{...} and \nothtml{...}.
#
# 15 Jan 1996, v0.17: Added \twee.
#
# 10 Jun 1997, v0.17a: for operation at Cambridge
#
# 18 Aug 1997, v0.18: alternative sites and formats
#
# 8 Oct 1997, v0.19: perl5 udpate, changed format of all file-name variables
# so don't contain directory name, so that reports
# of problems don't constitute a security hole
#
# 14 Sep 2000: version to run with betas of the faq
require "sanitize.pl";
# The site-specific stuff:
$home = "$ENV{FAQ_HOME}" || "/home/rf/tex/faq";
$faqhtml = $home . "/html";
$texfaq = "faqbody.tex";
$auxfaq = "newfaq.aux";
$ctandir = "dirctan.tex";
$ctanfiles = "filectan.tex";
$archive_list = "archive.list";
# defaults; these have to be allowed by the $archive_list file
$default_archive = "cam.ctan.org";
$default_format = "tar.gz";
# pro tem -- can we set these differently?
$arch = $default_archive;
$fmt = $default_format;
# table of symbols we believe in
%SymbolChar = (
92 => "\\",
123 => "\{",
125 => "\}"
);
# This script produces HTML, but we're sending the stuff to file
# so we needn't tell anyone
#
# print ("Content-type: text/html\n\n");
# By default, we convert LaTeX to HTML.
$converting = 1;
$ignoring = 0;
$sectioning = 0;
# two things used in conversion of \item[ ]
$itemset = "";
$enditemset = "";
# Get the list of CTAN directories:
open (CTANDIR, "$home/$ctandir")
|| &oh_dear ("Couldn't open $ctandir");
while (<CTANDIR>) {
if ( /\\CTANdirectory\{([^\}]*)\}\{([^\}]*)\}/ ) {
$ctanref{$1} = "$2.$fmt";
} elsif ( /\\CTANdirectory\*\{([^\}]*)\}\{([^\}]*)\}/ ) {
$ctanref{$1} = "$2/";
}
}
# Get the list of CTAN files
open (CTANFILES, "$home/$ctanfiles")
|| &oh_dear ("Couldn't open $ctanfiles");
while (<CTANFILES>) {
if ( /\\CTANfile\{([^\}]*)\}\{([^\}]*)\}/ ) {
$ctanref{$1} = "$2";
}
}
# Get the list of allowable archives
open (ARCHIVE_LIST, "$home/$archive_list")
|| &oh_dear ("Couldn't open $archive_list");
while (<ARCHIVE_LIST>) {
chop;
($archive, $root_dir) = split(/\s+/, $_, 2);
$archive_root{$archive} = $root_dir;
}
unless ( $root = $archive_root{$default_archive} ) {
&oh_dear("Archive $default_archive isn't in my list") }
# Get the Qrefs:
open (AUXFAQ, "$home/$auxfaq")
|| &oh_dear ("Couldn't open $auxfaq");
while (<AUXFAQ>) {
if (/\\newlabel\{([^\}]*)\}\{\{([^\}]*)\}/) {
$qref{$1} = "FAQ$2.html";
}
}
# $labels to hold a (list of) label names for these question(s)
$labels =~ s/^\+//;
# Open the FAQ file:
open (TEXFAQ, "$home/$texfaq")
|| &oh_dear ("Couldn't open $texfaq");
# Run through to the introduction, grabbing useful info.
while (<TEXFAQ>) {
last if /\\section\{Introduction\}/;
$fileversion=$1 if /\\def\\faqfileversion\{([^\}]*)\}/;
$filedate=$1 if /\\def\\faqfiledate\{([^\}]*)\}/;
}
# open main html file
open (FAQ, ">$faqhtml/index.html") ||
oh_dear ("couldn't open output file");
printf FAQ ("
<html><head>
<title>TeX Frequently Asked Questions</title>
</head><body>
<h1 align=\"center\">Welcome to the UK List of<br>
TeX Frequently Asked Questions</h1>
");
# Blast out the introduction until we get to the first section.
print FAQ "<h2>Introduction</h2>\n";
while (<TEXFAQ>) {
last if /\\section\{([^\}]*)\}/;
&sanitize_line;
print FAQ;
}
# start the index of questions ($_ still left over from last read from file)
/\\section\{([^\}]*)\}/;
$_ = $1;
&sanitize_line;
print FAQ "<h3>$_</h3><ul>\n";
# first pass, we have a file open by default
$faq_file_open = 1;
while ( $faq_file_open ) {
# no question yet open
$qfile_open = 0;
while (<TEXFAQ>) {
if ($converting) {
if ( /^\s*\\section\{(.*)\}\s*$/ ) {
$_=$1;
&sanitize_line;
print FAQ "</ul><h3>$_</h3><ul>\n";
$_="";
} elsif ( /^\s*\\subsection\{(.*)\}\s*$/ ) {
$_=$1;
&sanitize_line;
print FAQ "</ul><h4>$_</h4><ul>\n";
$_="";
}
}
&sanitize_line;
if (/^\s*\\Question(\[.*\])?\{(.*)\}\s*$/) {
$qnum++;
$entry = "<li><a href=\"FAQ$qnum.html\">$2</a>\n";
print FAQ $entry;
if ($qfile_open) {
print QFILE "</body>\n";
close QFILE;
}
open (QFILE, ">$faqhtml/FAQ$qnum.html") ||
oh_dear ("couldn't open file for question $qnum");
$qfile_open=1;
printf QFILE ("<head>
<title>UK TeX FAQ -- question $qnum</title>
</head><body>
<h3>$2</h3>
");
} else {
if ($qfile_open) { print QFILE; }
}
}
if ($qfile_open) {
print QFILE "</body>\n";
close QFILE;
}
# now: are there any other file names in ARGV?
if ( $#ARGV >= 0 ) {
close TEXFAQ;
$next_file = $ARGV[0];
open (TEXFAQ, "$home/$next_file") ||
oh_dear ("Couldn't open $home/$next_file");
shift;
} else {
$faq_file_open = 0;
}
}
# finish off the index.html file listing all questions
print FAQ "</ul>\n";
# report number of questions processed
print "$qnum questions processed\n";
# An error report:
sub oh_dear {
print $_[0], "\n";
die $_;
}
# Print a tail.
printf FAQ ("
<hr><address>
Maintenance of the
<a href=\"http://www.tex.ac.uk/faq\">TeX FAQ</a>
is coordinated by Robin Fairbairns.<p>
");
$qref_noans=$qref{"Q-noans"};
$qref_newans=$qref{"Q-newans"};
printf FAQ ("
Comments, suggestions, or error reports? -- see
\"<a href=\"%s\">Improving the FAQ</a>\" or
\"<a href=\"%s\">Extending the FAQ</a>\".
<p>
This is FAQ version $fileversion, last modified on $filedate.
</address>
</body></html>
", $qref_noans, $qref_newans);
# That's it!