| 1 |
#!/usr/bin/perl |
| 2 |
# |
| 3 |
# bib2html - make an html file that is equivalent to the bib files, |
| 4 |
# using one of the html bib styles |
| 5 |
# |
| 6 |
# usage: |
| 7 |
$usage = "usage: $0 {alpha|index|long|longp|long-pario|short|short-pario|cv} [-o <outfile>] file.bib...\ |
| 8 |
OR \ |
| 9 |
$0 [ {alpha|index|long|longp|long-pario|short|short-pario|cv} ] [-o <outfile>] -a file.aux"; |
| 10 |
# Output is to bib.html. |
| 11 |
# alpha makes a bibliography like bibstyle{alpha} |
| 12 |
# long makes a bibliography like bibstyle{abstract} |
| 13 |
# short makes a bibliography like bibstyle{alpha}, but with citekey for tags |
| 14 |
# cv lists title first, does not print key, does not sort |
| 15 |
# (useful for listing your papers on your home page) |
| 16 |
# |
| 17 |
# David Kotz 7/94 |
| 18 |
# [email protected] |
| 19 |
|
| 20 |
# updated Dan Gildea 1/2003: |
| 21 |
# . now handles nested \bf \em etc correctly |
| 22 |
# . added some more accents |
| 23 |
# . new -a file.aux usage to use existing .aux file, useful to |
| 24 |
# list only papers cited by a given latex document |
| 25 |
# http://www.cs.rochester.edu/~gildea/ |
| 26 |
|
| 27 |
# updated 2/2004, thanks to Juergen Vollmer |
| 28 |
# a) allowing \emph{..}, \texttt{...} and \textbf{..} |
| 29 |
# b) setting the environment variable BSTINPUTS to directory from $0 |
| 30 |
# c) if run on existing .aux file, bibstyle is now optional on the command line. |
| 31 |
# if not bib style nor specified on command line, uses the html version |
| 32 |
# of bib style from original document. |
| 33 |
|
| 34 |
use File::Basename; |
| 35 |
use File::Spec::Functions; |
| 36 |
|
| 37 |
$tmp = "/tmp/bib2html$$"; |
| 38 |
|
| 39 |
my $cmd_dir = catfile (dirname ($0), |
| 40 |
dirname ((defined readlink($0)) ? readlink($0) : $0)); |
| 41 |
|
| 42 |
$ENV{BSTINPUTS} .= ":$cmd_dir"; |
| 43 |
|
| 44 |
if ( scalar(@ARGV) < 2 ) { |
| 45 |
die "$usage\n"; |
| 46 |
} |
| 47 |
|
| 48 |
|
| 49 |
## pick up the args |
| 50 |
my %Opts; # will hold the option values |
| 51 |
|
| 52 |
if ($ARGV[0] !~ /^-/) { |
| 53 |
$Opts{style} = shift @ARGV; |
| 54 |
} |
| 55 |
while ($ARGV[0] =~ /^-/) { |
| 56 |
$tmpArg = shift @ARGV; |
| 57 |
if ($tmpArg eq "-o" ) { |
| 58 |
$Opts{outfile} = shift @ARGV; |
| 59 |
} elsif ($tmpArg eq "-a" ) { |
| 60 |
$Opts{auxfile} = shift @ARGV; |
| 61 |
} |
| 62 |
} |
| 63 |
|
| 64 |
$Opts{outfile} = "bib.html" if ( !defined ($Opts{outfile}) ); |
| 65 |
|
| 66 |
$SIG{INT} = \&cleanup; |
| 67 |
|
| 68 |
print "Creating $tmp.aux for $files\n"; |
| 69 |
open AUX, ">$tmp.aux"; |
| 70 |
|
| 71 |
if ($Opts{auxfile}) { |
| 72 |
system "ls"; |
| 73 |
open(AUXIN, $Opts{auxfile}) || die "could not open $Opts{auxfile}"; |
| 74 |
while (<AUXIN>) { |
| 75 |
# if style not specified on command line, use html version |
| 76 |
# of style specified in .aux file |
| 77 |
s/^\\bibstyle\{([^\}]*)\}/"\\bibstyle{html-".($Opts{style}||$1)."}"/e; |
| 78 |
print AUX; |
| 79 |
} |
| 80 |
} else { |
| 81 |
if (!defined $Opts{style}) { |
| 82 |
die "bib style not defined"; |
| 83 |
} |
| 84 |
## get all the remaining file names |
| 85 |
my $files = killSuffix ( shift @ARGV ); |
| 86 |
foreach $file ( @ARGV ) { |
| 87 |
$files .= ','; # make sure the list is comma delimited, bibtex needs it |
| 88 |
$files .= killSuffix ($file); |
| 89 |
} |
| 90 |
|
| 91 |
|
| 92 |
## need double slashes here to go around slash interpretation |
| 93 |
## only one is printed |
| 94 |
print AUX <<EOF; |
| 95 |
\\relax |
| 96 |
\\citation{*} |
| 97 |
\\bibstyle{html-$Opts{style}} |
| 98 |
\\bibdata{$files} |
| 99 |
EOF |
| 100 |
} |
| 101 |
close AUX; |
| 102 |
|
| 103 |
unlink <$tmp.{bbl,blg}>; |
| 104 |
|
| 105 |
if (-r "html-split.bst.gz" ) { |
| 106 |
system ( "gunzip html-split.bst.gz" ); |
| 107 |
} |
| 108 |
|
| 109 |
print "bibtex $tmp\n"; |
| 110 |
|
| 111 |
system "bibtex $tmp"; |
| 112 |
|
| 113 |
print "Bibtex done\n\n"; |
| 114 |
|
| 115 |
|
| 116 |
## ------------------------------------------------------------------- |
| 117 |
## the cleanup operation |
| 118 |
open BBL, "$tmp.bbl"; |
| 119 |
open OUT, ">$Opts{outfile}"; |
| 120 |
|
| 121 |
## stack of formatted text types that were started and did not finish on teh same line |
| 122 |
## elements are em, tt or b or i and when a candidate for a format-closing brace is found |
| 123 |
## this stack will be popped to see which format we have to close |
| 124 |
my @formatsToClose; |
| 125 |
|
| 126 |
while ( <BBL> ) { |
| 127 |
|
| 128 |
|
| 129 |
## accent handlers |
| 130 |
## these are quite definite about their replacemnets and so should |
| 131 |
## come before the more generic operations later |
| 132 |
## all braces in the latex constructs are optional |
| 133 |
## info on the html codes can be found at http://www.w3.org/TR/REC-html40/ |
| 134 |
## \\? is to handle latex \'\i which you do to put acute on i without dot |
| 135 |
s/\\ \` (?: \{ )? \\? ([aeiouAEIOU]) (?: \} )? |
| 136 |
/&$1grave;/gx; # grave accent |
| 137 |
s/ \\ \' (?: \{ )? \\? ([aeiouAEIOU]) (?: \} )? |
| 138 |
/&$1acute;/gx; # 'acute accent |
| 139 |
s/ \\ \^ (?: \{ )? \\? ([aeiouAEIOU]) (?: \} )? |
| 140 |
/&$1circ;/gx; # circumflex |
| 141 |
s/ \\ \" (?: \{ )? \\? ([aeiouyAEIOUY]) (?: \} )? |
| 142 |
/&$1uml;/gx; # "umlaut |
| 143 |
s/ \\ \~ (?: \{ )? \\? ([anoANO]) (?: \} )? |
| 144 |
/&$1tilde;/gx; # tilde |
| 145 |
s/ \\ c (?: \{ )? ([cC]) (?: \} )? |
| 146 |
/&$1cedil;/gx; # cedilla |
| 147 |
s/ \\ [v\'] (?: \{ )? ([cC]) (?: \} )? |
| 148 |
/$1/gx; # no html code for czech hook |
| 149 |
s/ \{? \\[s\"]s \}? |
| 150 |
/ß/gx; # german Scharf-S |
| 151 |
|
| 152 |
## several weird symbols |
| 153 |
s/ \\copyright |
| 154 |
/©/gx; |
| 155 |
s/ \\pounds |
| 156 |
/£/gx; |
| 157 |
|
| 158 |
## weirder symbols |
| 159 |
s/ \\ (ae|AE) |
| 160 |
/&$1lig;/gx; |
| 161 |
s/ \\ (o|O) |
| 162 |
/&$1slash;/gx; |
| 163 |
s/ \\ss |
| 164 |
/ß/gx; |
| 165 |
|
| 166 |
## greek letters, case insensitive matching, but upper case in latex and |
| 167 |
## html have the first letter of the english word capitalized |
| 168 |
s/ \\ (?: var )? (alpha|beta|gamma|delta|epsilon|theta|lambda|pi|rho|sigma|omega) |
| 169 |
/&$1;/gxi; |
| 170 |
|
| 171 |
## remove any \/ space-increasing symbols |
| 172 |
s+ ([^\\]) \\ \/ |
| 173 |
+$1+gx; |
| 174 |
|
| 175 |
s+ \\ \ + +gx; |
| 176 |
|
| 177 |
## deal with \cite stuff, change it to a link to a record the same html file |
| 178 |
s+ ([^\\]) \\cite\{ (.*?) \} |
| 179 |
+$1<a href="#$2">$2</a>+xg; |
| 180 |
|
| 181 |
## These rules are to deal with my (DFK) macros |
| 182 |
s/\\ie/i.e./g; |
| 183 |
s/\\eg/e.g./g; |
| 184 |
s/\\etc/etc./g; |
| 185 |
s+\\vs\\+<EM>vs.</EM>+g; |
| 186 |
s/\\usec/usec/g; |
| 187 |
s/\\mbox //g; |
| 188 |
s/\\par / <P> /g; |
| 189 |
s/\\par$/ <P>/g; |
| 190 |
|
| 191 |
|
| 192 |
s/\\\&/\&/g; # ampersand |
| 193 |
|
| 194 |
s/-{2,3}/-/g; # multiple dashes |
| 195 |
|
| 196 |
# a few rules are needed to compensate for BibTeXs way of splitting |
| 197 |
# long words over two lines by sticking a % (TeX comment character) at |
| 198 |
# the end of the line. This works when one word (usually a URL) is |
| 199 |
# split over more than one line. |
| 200 |
## if we have an unescaped % at the end of the line, remove it and the newline and |
| 201 |
## join the next line on |
| 202 |
## example straight out of the camel book, pg 204. amazing |
| 203 |
if ( s/ ([^\\]) \% \n$ /$1/x and $nextline = <BBL> ) { |
| 204 |
$_ .= $nextline; |
| 205 |
redo; # back to the top |
| 206 |
} |
| 207 |
|
| 208 |
# hyphenation characters should be removed |
| 209 |
s+\\-++g; |
| 210 |
|
| 211 |
# tildes - |
| 212 |
# tilde not preceded by \ or / is a nbsp |
| 213 |
# \~{} is ~ (likely in a URL) |
| 214 |
# all other tildes left alone, notably /~ (URL) |
| 215 |
s+ ([^\\/]) ~ +$1 +xg; # normal standalone tilde - nbsp |
| 216 |
s/ \\~ \{\} /~/xg; # \~{} to ~ - do before removing braces below |
| 217 |
|
| 218 |
## -------------------------------------------------------- |
| 219 |
## deal with em and tt and bf text surrounded by braces |
| 220 |
## deal with \emph{..}, \texttt{...} and \textbf{..} |
| 221 |
|
| 222 |
## the fancy groupings around the em etc are because in html bold font is not 'bf' |
| 223 |
## but just b and italic is i not 'it' so we have to pick out only a part of those |
| 224 |
## latex tags |
| 225 |
## final |\w+ is to skip over \rm or any other commands we don't handle |
| 226 |
## also removes braces with no command |
| 227 |
|
| 228 |
while ( /(text(b)f|text(tt)|(em)ph)?([\{\}])/ ) { |
| 229 |
if (defined $1) { |
| 230 |
my $cmd = "$1"; |
| 231 |
my $fmt = $2 || $3 || $4; |
| 232 |
s/\\$cmd\{/<$fmt>/; |
| 233 |
push ( @formatsToClose, $fmt ) |
| 234 |
} elsif ($5 eq '{') { |
| 235 |
## beginning of format |
| 236 |
s/ \{ (?: \\ (?: (em)|(b)f|(tt)|(i)t|\w+ ) )? \s* |
| 237 |
/ ( $format = $1 or $2 or $3 or $4 ) ? "<$format>" : '' /ex ; |
| 238 |
## push the format to be closed onto the stack (may be nothing) |
| 239 |
push ( @formatsToClose, $format ) |
| 240 |
} else { |
| 241 |
## pop format to close from stack |
| 242 |
s+ \s*\} |
| 243 |
+ ( $format = pop (@formatsToClose) ) ? "</$format>" : '' +ex ; |
| 244 |
} |
| 245 |
} |
| 246 |
|
| 247 |
## -------------------------------------------------------- |
| 248 |
|
| 249 |
##retrieve symbols escaped by backslashes |
| 250 |
my $escapedChars = quotemeta ( '#$%&_{}' ); |
| 251 |
s/ ([^\\]) \\ ([$escapedChars]) |
| 252 |
/$1$2/gxo; |
| 253 |
|
| 254 |
print OUT $_; |
| 255 |
} |
| 256 |
|
| 257 |
|
| 258 |
print "\n"; |
| 259 |
print "\noutput is in $Opts{outfile}\n"; |
| 260 |
|
| 261 |
cleanup(); |
| 262 |
|
| 263 |
sub killSuffix { |
| 264 |
$file = shift(); |
| 265 |
( $name, $path ) = fileparse ( $file, '\.[^.]*$' ); # the pattern indicates what a suffix looks like |
| 266 |
return ($path . $name); |
| 267 |
} |
| 268 |
|
| 269 |
sub cleanup { |
| 270 |
unlink ( glob ("$tmp.{aux,bbl,blg}") ); |
| 271 |
} |