Mercurial > rlmcintyre

comparison javadoc-search.pl @ 0:0d795f02a8bb tip

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
initial committ. what was I thinking?
author Robert McIntyre <rlm@mit.edu>
date Mon, 27 Sep 2010 16:57:26 -0400
parents
children
comparison
equal deleted inserted replaced
-1:000000000000 0:0d795f02a8bb
1 #!/usr/bin/perl
2
3 use strict;
4 use vars qw / $javadocRoots $indexFile $resultFrame $linkFrame $gzip /;
5 #####################################################################
6 #
7 # javadoc-search
8 #
9 # http://sourceforge.net/projects/javadoc-search/
10 #
11 # Copyright (C) 2002-2003, Nicholas Sushkin.
12 #
13 # This program is free software; you can redistribute it and/or modify
14 # it under the terms of the GNU General Public License as published by
15 # the Free Software Foundation; either version 2 of the License, or
16 # (at your option) any later version.
17 #
18 # This program is distributed in the hope that it will be useful,
19 # but WITHOUT ANY WARRANTY; without even the implied warranty of
20 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
21 # GNU General Public License for more details.
22 #
23 # You should have received a copy of the GNU General Public License
24 # along with this program; if not, write to the Free Software
25 # Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
26 #
27 #
28 # Instructions are at http://javadoc-search.sourceforge.net/
29 #
30 #
31 #####################################################################
32 #
33 # $Id: javadoc-search.pl,v 1.8 2003/08/19 00:11:09 nsushkin Exp $
34 #
35 #####################################################################
36
37 ### Indexing setup
38
39 # list of subdirectories of docroot to search for javadoc index files
40 # Windows and Apache - prefix with drive letter and use only lowercase letters.
41 # $javadocRoots = [ "c:/program\ files/apache/htdocs/docs" ];
42 # Unix
43 # $javadocRoots = [ "/usr/local/apache/htdocs/docs" ];
44 $javadocRoots = [ "/usr/local/apache/htdocs/docs" ];
45
46 # index file, if ends with .gz will use gzip
47 # Windows
48 # $indexFile = "c:/bin/apache\ group/apache/javadoc-index.txt";
49 # Unix
50 # $indexFile = "/usr/local/apache/htdocs/docs/javadoc-index.txt.gz";
51 $indexFile = "/usr/local/apache/htdocs/docs/javadoc-index.txt.gz";
52
53 ### Display style setup
54
55 # if $resultFrame is not empty, display search results in this frame
56 $resultFrame = "";
57
58 # if $linkFrame is not empty, open javadoc links in this frame
59 $linkFrame = "";
60
61 ### Auxillary programs setup
62
63 # which gzip
64 $gzip = "/usr/bin/gzip";
65
66 # nothing below is customizable
67 #####################################################################
68 use Getopt::Std;
69 use File::Find;
70
71 my @indexFiles;
72 my %links;
73
74 (my $rev = '$Revision: 1.8 $') =~ s/^\$\w+: ([\.\d]+) \$$/$1/;
75
76 sub usage
77 {
78 print STDERR <<EOUSAGE;
79 Javadoc Search $rev
80
81 Usage: $0 [-h] [-r]
82
83 -h prints this help
84 -r builds index
85
86 This is a cgi script to search javadoc for classes, members, and methods
87 using an index.
88 EOUSAGE
89
90 }
91
92 #
93 # Build index part
94 #
95 sub indexWrite
96 {
97 print STDERR "Writing all ", indexGetLinkCount(), " links to $indexFile .. ";
98
99 if ($indexFile =~ /\.gz$/)
100 {
101 open(OUTPUT, "| $gzip -c > \"$indexFile\"") || die "unable to gzip to $indexFile: $!\n";
102 }
103 else
104 {
105 open(OUTPUT, ">" . $indexFile) || die "unable to write to $indexFile: $!\n";
106 }
107
108 my $oldSearchKey = "";
109 foreach (sort { $links{$a} cmp $links{$b} } keys %links)
110 {
111 my $searchKey = $links{$_};
112 if ($searchKey ne $oldSearchKey)
113 {
114 if ($oldSearchKey ne "") { print OUTPUT "\n"; }
115 print OUTPUT $searchKey;
116 }
117 print OUTPUT "\t", $_;
118
119 $oldSearchKey = $searchKey;
120 }
121 print OUTPUT "\n";
122
123 close(OUTPUT);
124
125 print STDERR "done\n";
126 }
127
128 sub indexGetLinkCount
129 {
130 scalar(keys %links);
131 }
132
133 sub indexRebuild
134 {
135 find(\&indexFileFinder, @$javadocRoots);
136
137 foreach (@indexFiles)
138 {
139 parseLinks($$_[0], $$_[1]);
140 }
141 }
142
143 sub indexFileFinder
144 {
145 return unless -f $File::Find::name;
146 return unless /^index-.*html??/;
147
148 push @indexFiles, [$File::Find::name, $File::Find::dir];
149 }
150
151 sub indexAddLink
152 {
153 my $key = shift;
154 my $link = shift;
155
156 $links{$link} = $key;
157 }
158
159 sub parseLinks()
160 {
161 my $fileName = shift;
162 my $dirName = shift;
163
164 print STDERR "Parsing ", $fileName, ": ";
165
166 open(FILE, $fileName);
167
168 while(<FILE>)
169 {
170 if($_ =~ m/<A[^>]+?HREF\s*=\s*[\"\']?([^\'\" >]+?)[ \'\"]?(>| title=)/sig)
171 {
172 parseLink($1, $dirName);
173 }
174 }
175
176 close(FILE);
177
178 print STDERR indexGetLinkCount(), " links\n";
179 }
180
181 sub parseLink()
182 {
183 my $link = shift;
184 my $dirName = shift;
185
186 return if $link =~ qr'^(http|ftp)://'i;
187
188 # print "# link=", $link, "\n";
189
190 my $abslink = ( $link =~ qr'^[\./]*([^\./].*)' ) ? $1 : $link;
191 # print "# abslink= $abslink\n";
192
193 return unless $abslink =~ qr'/';
194
195 my ($file,$anchor) = split '#', $abslink;
196 # print "# file=$file, anchor=$anchor\n";
197
198 my $class = ($file =~ /(.*)\.html??/ ) ? $1 : return;
199 $class =~ tr{\/}{\.};
200
201 # print "# class=", $class, "\n";
202 # print "# dirname=", $dirName, "\n";
203
204 # cancel out ".."
205 if ($link =~ /^\.\.\//)
206 {
207 my @linkSegs = split '/', $link;
208 my @dirSegs = split '/', $dirName;
209
210 while (@linkSegs[0] eq '..' and $#dirSegs >= 0)
211 {
212 shift @linkSegs;
213 pop @dirSegs;
214 }
215 $link = join '/', @linkSegs;
216 $dirName = join '/', @dirSegs;
217 }
218 my $fullLink = ( $link =~ /^\// ) ? $link : $dirName . "/" . $link;
219
220 my $key = $anchor eq "" ? $class : $class . '#' . $anchor;
221 indexAddLink($key, $fullLink);
222 }
223
224 #
225 # CGI part
226 #
227 use CGI qw(:standard);
228
229 my $searchTypes = {'c'=>'class','cm'=>'class#member','cma'=>'class#member(args)'};
230
231 sub topFrameGenerate
232 {
233 Delete('render');
234 param('render', 'query');
235 my $queryFrameUrl = self_url;
236
237 Delete('render');
238 param('render', 'noResults');
239 my $noResultsFrameUrl = self_url;
240
241 return
242 header(),
243 html(head(title('Javadoc search')),
244 frameset({-rows => '150,*'},
245 frame({-name => 'query', -src => $queryFrameUrl}),
246 frame({-name => $resultFrame, -src => $noResultsFrameUrl})));
247 }
248
249 sub noResultsGenerate
250 {
251 return
252 header,
253 start_html('Javadoc search'),
254 h2('No results yet'),
255 end_html,
256 "\n";
257 }
258
259 sub searchFormGenerate
260 {
261 Delete('render');
262
263 return
264 h1('Javadoc search'),
265 $resultFrame ? start_form(-target => $resultFrame, -action=>script_name) : start_form(-action=>script_name),
266 "Search pattern ",
267 textfield(-name=>'pattern', -size=>15),
268 " in ",
269 scrolling_list(-name=>'type',
270 -values=>[sort(keys(%$searchTypes))],
271 -labels=>$searchTypes,
272 -defaults=>['c'],
273 -size=>1),
274 " and display ",
275 scrolling_list(-name=>'results',
276 -values=>['1', '10', '20', '50', '0'],
277 -labels=>{'1'=>'1', '10'=>'10', '20'=>'20', '50'=>'50', '0'=>'All'},
278 -defaults=>['10'],
279 -size=>1),
280 " results ",
281 submit(-name=>'go', -value=>'Go'), submit(-name=>'lucky', -value=>'Go!'),
282 $resultFrame ? hidden(-name => 'render', -default => 'results') : "",
283 end_form,
284 $resultFrame ? "" : hr;
285 }
286
287 sub doSearch
288 {
289 my ($pattern, $type, $results) = @_;
290
291 if (! -r $indexFile)
292 {
293 print strong("Error: index file is not readable");
294 return;
295 }
296
297 if ($indexFile =~ /\.gz$/)
298 {
299 open(INPUT, "$gzip -dc \"$indexFile\" |") || die "unable to un-gzip $indexFile: $!\n";
300 }
301 else
302 {
303 open(INPUT, $indexFile) || die "unable to read $indexFile: $!\n";
304 }
305
306 my $docRoot = $ENV{'DOCUMENT_ROOT'};
307 my @matches;
308
309 INPUTLINE:
310 while(<INPUT>)
311 {
312 chomp;
313 my ($key, @urls) = split (/\t/);
314
315 for ($key)
316 {
317 if (($type eq "c" and /\#/) ||
318 ($type eq "cm" and !(/\#/ and !/\)/)) ||
319 ($type eq "cma" and !(/\#/ and /\)/)))
320 {
321 next INPUTLINE;
322 }
323 }
324
325 if ($key =~ /$pattern/ and ($results == 0 or $#matches < $results) )
326 {
327 push @matches, [ ( $key, map { s/$docRoot//; $_ } @urls ) ];
328 }
329
330 last if $results > 0 and $#matches >= $results;
331
332 }
333 close(INPUT);
334
335 return \@matches;
336 }
337
338 sub displayLuckyMatch
339 {
340 my $matches = shift;
341
342 if ( $#$matches < 0 )
343 {
344 print
345 header,
346 start_html('No matches'),
347 p("No matches"),
348 end_html;
349 }
350 else
351 {
352 print
353 $linkFrame ? header(-target => $linkFrame) : "",
354 redirect('-location'=> $$matches[0][1]);
355 }
356 }
357
358
359 sub displayMatches
360 {
361 my $matches = shift;
362
363 if ( $#$matches < 0 )
364 {
365 print p("No matches");
366 }
367 else
368 {
369 print
370 p(
371 "Searching pattern ", em(escapeHTML(param('pattern'))),
372 " in ", em(escapeHTML($$searchTypes{param('type')}))),
373 h2("Search results"),
374 "<ul>";
375
376 foreach my $match (@$matches)
377 {
378 my $searchKey = escapeHTML(shift @$match);
379 my $cnt = -1;
380 print li( map {
381 a( $linkFrame ? {-href => $_, -target => $linkFrame } : { -href => $_}, ( $cnt++ >= 0 ? "alt".($cnt) : $searchKey) )
382 } @$match );
383
384 }
385
386 print "</ul>";
387 }
388 }
389
390 #
391 # main
392 #
393 use vars qw / $opt_h $opt_r /;
394
395 if (server_software() eq 'cmdline')
396 {
397 getopts("rh");
398
399 if ($opt_h)
400 {
401 usage;
402 exit;
403 }
404
405 if ($opt_r)
406 {
407 indexRebuild;
408 indexWrite;
409 exit;
410 }
411 }
412
413
414 if (param('render'))
415 {
416 $resultFrame = 'results';
417 $linkFrame = '';
418 }
419
420 my $thisFrame = param('render');
421
422 if ($thisFrame eq 'topFrame')
423 {
424 print topFrameGenerate();
425 exit 0;
426 }
427 elsif ($thisFrame eq 'noResults')
428 {
429 print noResultsGenerate();
430 exit 0;
431 }
432
433 unless (param('lucky') and param('pattern'))
434 {
435 print
436 header,
437 start_html('Javadoc search'),
438 defined(param('pattern')) && $resultFrame ? "" : searchFormGenerate();
439 }
440
441 if (param('pattern'))
442 {
443 my $matches = doSearch(param('pattern'), param('type'), param('results'));
444
445 if (param('lucky'))
446 {
447 displayLuckyMatch($matches);
448 }
449 else
450 {
451 displayMatches($matches);
452 }
453 }
454 else
455 {
456 if ($thisFrame ne 'query')
457 {
458 print p('Specify non-blank search pattern');
459 }
460 }
461
462 unless (param('lucky') and param('pattern'))
463 {
464 print end_html;
465 }