OLD | NEW |
| (Empty) |
1 #!/usr/bin/perl | |
2 # convert TeX (Patgen) hyphenation patterns to Libhnj format | |
3 # (A utility for finding substring embeddings in patterns) | |
4 # usage: substrings.pl inputfile outputfile [encoding] | |
5 | |
6 # Libhnj is dual licensed under LGPL and MPL. Boilerplate for both | |
7 # licenses follows. | |
8 | |
9 # LibHnj - a library for high quality hyphenation and justification | |
10 # Copyright (C) 1998 Raph Levien, | |
11 # (C) 2001 ALTLinux, Moscow (http://www.alt-linux.org), | |
12 # (C) 2001 Peter Novodvorsky (nidd@cs.msu.su) | |
13 # (C) 2006, 2007, 2008, 2010 László Németh (nemeth at OOo) | |
14 # | |
15 # This library is free software; you can redistribute it and/or | |
16 # modify it under the terms of the GNU Library General Public | |
17 # License as published by the Free Software Foundation; either | |
18 # version 2 of the License, or (at your option) any later version. | |
19 # | |
20 # This library is distributed in the hope that it will be useful, | |
21 # but WITHOUT ANY WARRANTY; without even the implied warranty of | |
22 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU | |
23 # Library General Public License for more details. | |
24 # | |
25 # You should have received a copy of the GNU Library General Public | |
26 # License along with this library; if not, write to the | |
27 # Free Software Foundation, Inc., 59 Temple Place - Suite 330, | |
28 # Boston, MA 02111-1307 USA. | |
29 | |
30 # The contents of this file are subject to the Mozilla Public License | |
31 # Version 1.0 (the "MPL"); you may not use this file except in | |
32 # compliance with the MPL. You may obtain a copy of the MPL at | |
33 # http://www.mozilla.org/MPL/ | |
34 # | |
35 # Software distributed under the MPL is distributed on an "AS IS" basis, | |
36 # WITHOUT WARRANTY OF ANY KIND, either express or implied. See the MPL | |
37 # for the specific language governing rights and limitations under the | |
38 # MPL. | |
39 | |
40 if (!defined $ARGV[1]) { | |
41 print "" . | |
42 "substrings.pl - convert TeX (Patgen) hyphenation patterns to Libhnj format\n" .
| |
43 "(A utility for finding substring embeddings in patterns)\n" . | |
44 "usage: substrings.pl infile outfile [encoding [lefthyphenmin [righthyphenmin]]]
\n"; | |
45 exit 1; | |
46 } | |
47 $fn = $ARGV[0]; | |
48 if (!-e $fn) { $fn = "hyphen.us"; } | |
49 open HYPH, $fn; | |
50 open OUT, ">$ARGV[1]"; | |
51 $encoding = $ARGV[2]; | |
52 $lhmin = $ARGV[3]; | |
53 $rhmin = $ARGV[4]; | |
54 if (defined $encoding) { print OUT "$encoding\n"; } | |
55 if (defined $lhmin) { print OUT "LEFTHYPHENMIN $lhmin\n"; } | |
56 if (defined $rhmin) { print OUT "RIGHTHYPHENMIN $rhmin\n"; } | |
57 | |
58 while (<HYPH>) | |
59 { | |
60 $pat =~ s/%.*$//g; | |
61 if (/^\%/) { | |
62 #comment, ignore | |
63 } elsif (/^(.+)\/([^,]+),([0-9]+),([0-9]+)$/) { | |
64 $origpat = $1; | |
65 $pat = $1; | |
66 $repl = $2; | |
67 $beg = $3; | |
68 $len = $4; | |
69 $pat =~ s/\d//g; | |
70 if ($origpat eq $pat) { | |
71 print "error - missing hyphenation point: $_"; | |
72 exit 1; | |
73 } | |
74 push @patlist, $pat; | |
75 $pattab{$pat} = $origpat; | |
76 $repltab{$pat} = $repl; | |
77 $replbeg{$pat} = $beg - 1; | |
78 $repllen{$pat} = $len; | |
79 } elsif (/^(.+)\/(.+)$/) { | |
80 $origpat = $1; | |
81 $pat = $1; | |
82 $repl = $2; | |
83 $pat =~ s/\d//g; | |
84 if ($origpat eq $pat) { | |
85 print "error - missing hyphenation point: $_"; | |
86 exit 1; | |
87 } | |
88 push @patlist, $pat; | |
89 $pattab{$pat} = $origpat; | |
90 $repltab{$pat} = $repl; | |
91 $replbeg{$pat} = 0; | |
92 $repllen{$pat} = enclen($pat); | |
93 } elsif (/^(.+)$/) { | |
94 $origpat = $1; | |
95 $pat = $1; | |
96 $pat =~ s/\d//g; | |
97 push @patlist, $pat; | |
98 $pattab{$pat} = $origpat; | |
99 } | |
100 } | |
101 | |
102 foreach $pat (@patlist) { | |
103 $patsize = length $pat; | |
104 for $i (0..$patsize - 1) { | |
105 for $j (1..$patsize - $i) { | |
106 $subpat = substr ($pat, $i, $j); | |
107 if (defined $pattab{$subpat}) { | |
108 print "$pattab{$subpat} is embedded in $pattab{$pat}\n"; | |
109 $newpat = substr $pat, 0, $i + $j; | |
110 if (!defined $newpattab{$newpat}) { | |
111 $newpattab{$newpat} = | |
112 substr ($pat, 0, $i).$pattab{$subpat}; | |
113 $ss = substr $pat, 0, $i; | |
114 print "$ss+$pattab{$subpat}\n"; | |
115 push @newpatlist, $newpat; | |
116 if (defined $repltab{$subpat}) { | |
117 $begcorr = (($pat =~ /^[.]/) && !($subpat =~ /^[.]/)) ?
1 : 0; | |
118 $newrepltab{$newpat} = $repltab{$subpat}; | |
119 $newreplbeg{$newpat} = $replbeg{$subpat} + enclen($ss) -
$begcorr; | |
120 $newrepllen{$newpat} = $repllen{$subpat}; | |
121 } | |
122 } else { | |
123 $tmp = $newpattab{$newpat}; | |
124 $newpattab{$newpat} = | |
125 combine ($newpattab{$newpat}, $pattab{$subpat}); | |
126 print "$tmp + $pattab{$subpat} -> $newpattab{$newpat}\n"; | |
127 } | |
128 } | |
129 } | |
130 } | |
131 } | |
132 | |
133 foreach $pat (@newpatlist) { | |
134 if (defined $newrepltab{$pat}) { | |
135 print OUT $newpattab{$pat}."/".$newrepltab{$pat}.",".($newreplbeg{$pat}+
1).",".$newrepllen{$pat}."\n"; | |
136 } else { | |
137 print OUT $newpattab{$pat}."\n"; | |
138 } | |
139 } | |
140 | |
141 #convert 'n1im' to 0n1i0m0 expresed as a list | |
142 sub expand { | |
143 my ($pat) = @_; | |
144 my $last = '.'; | |
145 my @exp = (); | |
146 | |
147 foreach $c (split (//, $pat)) { | |
148 if ($last =~ /[\D]/ && $c =~ /[\D]/) { | |
149 push @exp, 0; | |
150 } | |
151 push @exp, $c; | |
152 $last = $c; | |
153 } | |
154 if ($last =~ /[\D]/) { | |
155 push @exp, 0; | |
156 } | |
157 return @exp; | |
158 } | |
159 | |
160 # Combine two patterns, i.e. .ad4der + a2d becomes .a2d4der | |
161 # The second pattern needs to be a substring of the first (modulo digits) | |
162 sub combine { | |
163 my @exp = expand shift; | |
164 my @subexp = expand shift; | |
165 my $pat1, $pat2; | |
166 my $i; | |
167 | |
168 $pat1 = join ('', map { $_ =~ /\d/ ? () : $_ } @exp); | |
169 $pat2 = join ('', map { $_ =~ /\d/ ? () : $_ } @subexp); | |
170 | |
171 $begcorr = ($pat1 =~ /^[.]/) ? 1 : 0; | |
172 | |
173 for $i (0..length ($pat1) - length ($pat2)) { | |
174 if (substr ($pat1, $i, length $pat2) eq $subpat) { | |
175 for ($j = 0; $j < @subexp; $j += 2) { | |
176 if ($subexp[$j] > $exp[2 * $i + $j]) { | |
177 $exp[2 * $i + $j] = $subexp[$j]; | |
178 if (defined $newrepltab{$pat2} && !defined $newrepltab{$pat1
}) { | |
179 $ss = substr ($pat1, 0, $i); | |
180 $newrepltab{$pat1} = $newrepltab{$pat2}; | |
181 $newreplbeg{$pat1} = $newreplbeg{$pat2} + enclen($ss) -
$begcorr; | |
182 $newrepllen{$pat1} = $newrepllen{$pat2}; | |
183 } | |
184 } | |
185 } | |
186 print ("$pat1 includes $pat2 at pos $i\n"); | |
187 } | |
188 } | |
189 return join ('', map { $_ eq '0' ? () : $_ } @exp); | |
190 } | |
191 | |
192 # 8 bit or UTF-8 character length (calculating right start position for discreti
onary hyphenation) | |
193 sub enclen { | |
194 my $nonchar = 0; | |
195 my $len = length($_[0]); | |
196 if ($encoding eq "UTF-8") { | |
197 # length of an UTF-8 string equals to the count of the characters not st
arted with '10' bits | |
198 for ($i = 0; $i < $len; $i++) { | |
199 if ((ord(substr($_[0], $i, 1)) >> 6) == 2) { $nonchar++; } | |
200 } | |
201 } | |
202 return $len - $nonchar; | |
203 } | |
OLD | NEW |