view release on metacpan or  search on metacpan

lib/Biblio/bp/lib/bp-cs-tex.pl  view on Meta::CPAN

package bp_cs_tex;

######

$bib'charsets{'tex', 'i_name'} = 'tex';

$bib'charsets{'tex', 'tocanon'}   = "bp_cs_tex'tocanon";
$bib'charsets{'tex', 'fromcanon'} = "bp_cs_tex'fromcanon";

$bib'charsets{'tex', 'toesc'}   = "[\$\\\\]";
# XXXXX We have so many characters to protect, should we even bother?
$bib'charsets{'tex', 'fromesc'} = "[\\#\$\%\&{}_\|><\^~\200-\377]|${bib'cs_ext}|${bib'cs_meta}";

######

$cs_init = 0;

# package variables for anyone to use
$mine = '';
$unicode = '';
$can = '';

######

sub init_cs {

# Thorn and eth are really nasty since they don't exist in the standard TeX
# fonts.  This is what I came up with in r2b to fake it.  Fortunately they
# aren't used often.  Get the cmoer fonts if you want to do them right.
# My eth is pretty nice, but the thorn leaves a little to be desired.

%charmap = (
'00A1', "!'",
'00A2', '\leavevmode\hbox{\rm\rlap/c}',
'00A3', '{\pounds}',
'00A4', '$\spadesuit$',
'00A5', '\leavevmode\hbox{\rm\rlap=Y}',
'00A6', '\leavevmode
         \hbox{\hskip.4ex\hbox{\ooalign{\vrule width.2ex height.5ex depth.4ex\crcr
         \hfil\raise.8ex\hbox{\vrule width.2ex height.9ex depth0ex}\hfil}}}',
'00A7', '\S ',
'00A8', '{\"{ }}',
'00A9', '\leavevmode\hbox{\raise.6em\hbox{\copyright}}',
'00AA', '${}^{\b{\scriptsize a}}$',
'00AB', '$\scriptscriptstyle\ll$',
'00AC', '$\neg$',
'00AE', '\leavevmode\hbox{\raise.6em\hbox{\ooalign{{\mathhexbox20D}\crcr
         \hfil\raise.07ex\hbox{r}\hfil}}}',
'00AF', '{\={ }}',
'00B0', '${}^\circ$',
'00B1', '$\pm$',
'00B2', '${}^2$',
'00B3', '${}^3$',
'00B4', '{\'{ }}',
'00B5', '$\mu$',
'00B6', '\P ',
'00B7', '$\cdot$',
'00B8', '{\c{ }}',
'00B9', '${}^1$',
'00BA', '${}^{\b{\scriptsize o}}$',
'00BB', '$\scriptscriptstyle\gg$',
'00BC', '$1\over4$',
'00BD', '$1\over2$',
'00BE', '$3\over4$',
'00BF', '?`',
'00C0', '{\`A}',
'00C1', q-{\'A}-,
'00C2', '{\^A}',
'00C3', '{\~A}',
'00C4', '{\"A}',
'00C5', '{\AA}',
'00C6', '{\AE}',
'00C7', '{\c{C}}',
'00C8', '{\`E}',
'00C9', q-{\'E}-,
'00CA', '{\^E}',
'00CB', '{\"E}',
'00CC', '{\`I}',
'00CD', q-{\'I}-,
'00CE', '{\^I}',
'00CF', '{\"I}',
'00D0', '\leavevmode\hbox{\ooalign{{D}\crcr
         \hskip.2ex\raise.25ex\hbox{-}\hfil}}',
'00D1', '{\~N}',
'00D2', '{\`O}',
'00D3', q-{\'O}-,
'00D4', '{\^O}',
'00D5', '{\~O}',
'00D6', '{\"O}',
'00D7', '$\times$',
'00D8', '{\O}',
'00D9', '{\`U}',
'00DA', q-{\'U}-,
'00DB', '{\^U}',
'00DC', '{\"U}',
'00DD', q-{\'Y}-,
'00DE', '\leavevmode\hbox{I\hskip-.6ex\raise.5ex\hbox{$\scriptscriptstyle\supset$}}',
'00DF', '{\ss}',
'00E0', '{\`a}',
'00E1', q-{\'a}-,
'00E2', '{\^a}',
'00E3', '{\~a}',
'00E4', '{\"a}',
'00E5', '{\aa}',
'00E6', '{\ae}',
'00E7', '{\c{c}}',
'00E8', '{\`e}',
'00E9', q-{\'e}-,
'00EA', '{\^e}',
'00EB', '{\"e}',
'00EC', '{\`i}',
'00ED', q-{\'i}-,
'00EE', '{\^i}',
'00EF', '{\"i}',
'00F0', '\leavevmode\hbox{\ooalign{$\partial$\crcr\hskip.8ex\raise.7ex\hbox{-}\hfil}}',
'00F1', '{\~n}',
'00F2', '{\`o}',
'00F3', q-{\'o}-,
'00F4', '{\^o}',
'00F5', '{\~o}',
'00F6', '{\"o}',
'00F7', '$\div$',
'00F8', '{\o}',
'00F9', '{\`u}',
'00FA', q-{\'u}-,
'00FB', '{\^u}',
'00FC', '{\"u}',
'00FD', q-{\'y}-,
'00FE', '\leavevmode\hbox{{\lower.3ex\hbox{\large l}}\hskip-.52ex o}',
'00FF', '{\"y}',
'0107', q-{\'c}-,
'010C', '{\vC}',
'010D', '{\vc}',
'0159', '{\vr}',
'015F', '{\c{s}}',
'0160', '{\vS}',
'0161', '{\vs}',
'017A', q-{\'z}-,
'017E', '{\vz}',
# XXXXX
# Should these be surrounded by $ (math mode)?
# Also, what to do with \mu, which is listed twice?
'03B1', '\alpha',
'03B2', '\beta',
'03B3', '\gamma',
'03B4', '\delta',
'03B5', '\epsilon',
'03B6', '\zeta',
'03B7', '\eta',
'03B8', '\theta',
'03B9', '\iota',
'03BA', '\kappa',
'03BB', '\lambda',
'03BC', '\mu',
'03BD', '\nu',
'03BE', '\xi',
'03C0', '\pi',
'03C1', '\rho',
'03C2', '\varsigma',
'03C3', '\sigma',
'03C4', '\tau',
'03C5', '\upsilon',
'03C6', '\phi',
'03C7', '\chi',
'03C8', '\psi',
'03C9', '\omega',
'2007', '$\:$',
'2009', '$\,$',
'201C', '``',
'201D', '\'\'',
);

# This mapping is only used in the from section.  We'll do these by hand
# in the to mapping.
%charmap2 = (
'00A0', '~',
'00AD', '-',
'2002', '\ ',
'2003', '\ \ ',
'2014', '---',
'03BF', 'o',
);

# Blah.  TeX has such a non-uniform way of handling characters that this is
# really slow.  I'm going to try some optimizations for the tocanon code
# since that will be heavily used.  It makes this stuff less uniform though.
# Remember that we don't have a full TeX parser, or even a partial one.

# Build up a search string to do the reverse map.
$cmap_to_eval = '';
$cmap_from8_eval = '';
$cmap_to_eval_1 = '';
$cmap_to_eval_2 = '';
%rmap = ();
%accent = ();

# Step 1: Build a reverse map
while (($unicode, $mine) = each %charmap) {
  $rmap{$mine} = $unicode;
}
# Step 2: walk through the keys in sorted order
local($mineE);
foreach $mine (sort keys %rmap) {
  $can = &bib'unicode_to_canon( $rmap{$mine} );
  $mineE = $mine;