1 |
#!/usr/bin/perl -w |
2 |
|
3 |
=head1 NAME |
4 |
|
5 |
cpanest - generate an Hyper Estraier index for CPAN |
6 |
|
7 |
=head1 SYNOPSIS |
8 |
|
9 |
B<cpanest> |
10 |
[B<-clean>] [B<-noclean>] |
11 |
[B<-cpan> I<url or directory>] |
12 |
[B<-node> I<node_uri>] |
13 |
[B<-force>] [B<-noforce>] |
14 |
[B<-keep> I<directory>] |
15 |
[B<-match> I<regexp>] |
16 |
[B<-test> I<level>] |
17 |
[B<-trust_mtime>] [B<-notrust_mtime>] |
18 |
|
19 |
=head1 DESCRIPTION |
20 |
|
21 |
This is a port of C<cpanwait> from L<WAIT> perl search engine to node API of |
22 |
Hyper Estraier. |
23 |
|
24 |
All the hard work was done by Ulrich Pfeifer who wrote all parsers and |
25 |
formatters. I just added support for Hyper Estraier back-end after. |
26 |
|
27 |
B<This documentation is somewhat incomplete and off-the-sync with code.> |
28 |
|
29 |
=head1 OPTIONS |
30 |
|
31 |
=over 5 |
32 |
|
33 |
=item B<-clean> / B<-noclean> |
34 |
|
35 |
Clean the table befor indexing. Default is B<off>. |
36 |
|
37 |
=item B<-cpan> I<url or directory> |
38 |
|
39 |
Default directory or URL for indexing. If an URL is given, there |
40 |
currently must be a file F<indices/find-ls.gz> relative to it which |
41 |
contains the output of C<find . -ls | gzip>. |
42 |
Default is F<ftp://ftp.rz.ruhr-uni-bochum.de/pub/CPAN>. |
43 |
|
44 |
|
45 |
=item B<-node> I<http://localhost:1978/node/cpan> |
46 |
|
47 |
Specify node URI |
48 |
|
49 |
=item B<-force> |
50 |
|
51 |
Force reindexing, even if B<cpan> thinks files are up to date. |
52 |
Default is B<off> |
53 |
|
54 |
=item B<-keep> I<directory> |
55 |
|
56 |
If fetching from a remote server, keep files in I<directory>. Default is |
57 |
F</app/unido-i06/src/share/lang/perl/96a/CPAN/sources>. |
58 |
|
59 |
=item B<-match> I<regexp> |
60 |
|
61 |
Limit to patches matching I<regexp>. Default is F<authors/id/>. |
62 |
|
63 |
=item B<-test> I<level> |
64 |
|
65 |
Set test level, were B<0> means normal operation, B<1> means, don't |
66 |
really index and B<2> means, don't even get archives and examine them. |
67 |
|
68 |
=item B<-trust_mtime> / B<-notrust_mtime> |
69 |
|
70 |
If B<on>, the files mtimes are used to decide, which version of an |
71 |
archive is the newest. If b<off>, the version extracted is used |
72 |
(beware, there are far more version numbering schemes than B<cpan> can |
73 |
parse). |
74 |
|
75 |
=back |
76 |
|
77 |
=head1 AUTHORS |
78 |
|
79 |
Ulrich Pfeifer E<lt>F<pfeifer@ls6.informatik.uni-dortumund.de>E<gt> |
80 |
|
81 |
Dobrica Pavlinusic E<lt>F<dpavlin@rot13.org>E<gt> |
82 |
|
83 |
=head1 COPYRIGHT |
84 |
|
85 |
Copyright (c) 1996-1997, Ulrich Pfeifer |
86 |
|
87 |
Copyright (c) 2005, Dobrica Pavlinusic |
88 |
|
89 |
=cut |
90 |
|
91 |
use strict; |
92 |
|
93 |
use File::Path; |
94 |
use Getopt::Long; |
95 |
use File::Find; |
96 |
use File::Basename; |
97 |
use IO::File; |
98 |
use IO::Zlib; |
99 |
use POSIX qw/strftime/; |
100 |
|
101 |
use lib '/data/wait/lib'; |
102 |
|
103 |
use WAIT::Parse::Base; |
104 |
use WAIT::Parse::Pod; |
105 |
use WAIT::Document::Tar; |
106 |
|
107 |
sub fname($); |
108 |
|
109 |
# maximum number of archives to index (set to -1 for unlimited) |
110 |
my $max = -1; |
111 |
|
112 |
my %OPT = ( |
113 |
node => 'http://localhost:1978/node/cpan', |
114 |
clean => 0, |
115 |
remove => [], |
116 |
force => 0, |
117 |
# cpan => '/usr/src/perl/CPAN/sources', |
118 |
cpan => '/rest/cpan/CPAN/', |
119 |
trust_mtime => 1, |
120 |
match => 'authors/id/', |
121 |
test => 0, |
122 |
# cpan => 'ftp://ftp.rz.ruhr-uni-bochum.de/pub/CPAN', |
123 |
# cpan => 'ftp://ftp.uni-hamburg.de:/pub/soft/lang/perl/CPAN', |
124 |
keep => '/tmp/CPAN/', |
125 |
); |
126 |
|
127 |
GetOptions(\%OPT, |
128 |
'node=s', |
129 |
'cpan=s', |
130 |
'keep=s', |
131 |
'match=s', |
132 |
'clean!', |
133 |
'test=i', # test level 0: normal |
134 |
# 1: don't change db |
135 |
# 2: don't look at archives even |
136 |
|
137 |
'remove=s@', |
138 |
'force!', # force indexing even if seen |
139 |
'trust_mtime!', # use mtime instead of version number |
140 |
'max=i', |
141 |
'debug!', |
142 |
) || die "Usage: ...\n"; |
143 |
|
144 |
if ($OPT{max}) { |
145 |
$max = $OPT{max}; |
146 |
print STDERR "processing just first $max modules\n"; |
147 |
} |
148 |
|
149 |
# FIXME |
150 |
#clean_node( |
151 |
# node => $OPT{node}, |
152 |
# ) if $OPT{clean}; |
153 |
|
154 |
my $tb = new HyperEstraier::WAIT::Table( |
155 |
uri => $OPT{node}, |
156 |
attr => ['docid', 'headline', 'source', 'size', 'parent', 'version'], |
157 |
key => 'docid', |
158 |
invindex => [ qw/name synopsis bugs description text environment example author/ ], |
159 |
debug => $OPT{debug}, |
160 |
) or die "Could not open node '$OPT{node}'"; |
161 |
|
162 |
my $layout= new WAIT::Parse::Pod; |
163 |
|
164 |
# Map e.g. '.../latest' to 'perl'. Used in wanted(). Effects version |
165 |
# considerations. Value *must* match common prefix. Aliasing should be |
166 |
# used if CPAN contains serveral distributions with different name but |
167 |
# same root directory. |
168 |
# We still have a problem if there are different root directories! |
169 |
|
170 |
my %ALIAS = (# tar name real (root) name |
171 |
'Games-Scrabble' => 'Games', |
172 |
'HTML-ParseBrowser' => 'HTML', |
173 |
'iodbc_ext' => 'iodbc-ext-0.1', |
174 |
'sol-inst' => 'Solaris', |
175 |
'WebService-Validator-CSS-223C' => 'WebService-Validator-CSS-W3C-0.02', |
176 |
'MPEG-ID3212Tag' => 'MPEG-ID3v2Tag-0.36', |
177 |
'WebService-GoogleHack' => 'WebService', |
178 |
'Db-Mediasurface-ReadConfig' => 'ReadConfig', |
179 |
'Tie-Array-RestrictUpdates' => 'Tie', |
180 |
'HTML-Lister' => 'HTML', |
181 |
'Net-253950-AsyncZ' => 'Net-Z3950-AsyncZ-0.08', |
182 |
'ChildExit_0' => 'ChildExit-0.1', |
183 |
'Tie-TieConstant' => 'TieConstant.pm', |
184 |
'Crypt-OpenSSL-23509' => 'Crypt-OpenSSL-X509-0.2', |
185 |
'subclustv' => 'blib', |
186 |
'finance-yahooquote' => 'Finance-YahooQuote-0.20', |
187 |
'HPUX-FS' => 'FS', |
188 |
'Business-DE-Konto' => 'Business', |
189 |
'Digest-MD5-124p' => 'Digest-MD5-M4p-0.01', |
190 |
'AKDB_Okewo_de' => 'AKDB', |
191 |
'ExtUtils-0577' => 'ExtUtils-F77-1.14', |
192 |
'LispFmt' => 'Lisp::Fmt-0.00', |
193 |
'Acme-Stegano' => 'Acme', |
194 |
'Acme-RTB' => 'Acme', |
195 |
'WWW-Search-PRWire' => 'work', |
196 |
'Video-Capture-214l' => 'Video-Capture-V4l-0.224', |
197 |
'Tie-DirHandle' => 'Tie', |
198 |
'DB2' => 'DBD-DB2-0.71a', |
199 |
'Tie-Scalar-RestrictUpdates' => 'Tie', |
200 |
'Math-MVPoly' => 'MVPoly', |
201 |
'PlugIn' => 'PlugIn.pm', |
202 |
'Lingua-ID-Nums2Words' => 'Nums2Words-0.01', |
203 |
'chronos-1.' => 'Chronos', |
204 |
'jp_beta' => 'jperl_beta_r1', |
205 |
'Bundle-223C-Validator' => 'Bundle-W3C-Validator-0.6.5', |
206 |
'Text-199' => 'Text-T9-1.0', |
207 |
'Games-Literati' => 'Games', |
208 |
'VMS-IndexedFile' => 'VMS', |
209 |
'authen-rbac' => 'Authen', |
210 |
'Graphics-EPS' => 'EPS.pm', |
211 |
'new.spirit-2.' => 'new.spirit', |
212 |
'Tk-MListbox' => 'MListbox-1.11', |
213 |
'DBD-SQLrelay' => 'SQLRelay.pm', |
214 |
'Tie-RDBM-Cached' => 'RDBM', |
215 |
'PDL_IO_HDF' => 'HDF', |
216 |
'HPUX-LVM' => 'LVM', |
217 |
'Parse-Nibbler' => 'Parse', |
218 |
'Digest-Perl-MD4' => 'MD4', |
219 |
'Crypt-Imail' => 'Imail', |
220 |
'ubertext' => 'Text-UberText-0.95', |
221 |
'MP3-123U' => 'M3U', |
222 |
'Qmail-Control' => 'Qmail', |
223 |
'T-LXS' => 'Text-LevenshteinXS-0.02', |
224 |
'HTML-Paginator' => 'HTML', |
225 |
'swig' => 'SWIG1.1p5', |
226 |
'MIDI-Realtime' => 'MIDI', |
227 |
'sparky-public' => 'Sparky-Public-1.06', |
228 |
'Chemistry-MolecularMass' => 'Chemistry', |
229 |
'Net-253950-SimpleServer' => 'Net-Z3950-SimpleServer-0.08', |
230 |
'NewsClipper-OpenSource' => 'NewsClipper-1.32-OpenSource', |
231 |
'Win32API-Resources' => 'Resources.pm', |
232 |
'Unicode-Collate-Standard-2131_1' => 'Unicode-Collate-Standard-V3_1_1-0.1', |
233 |
'Net-026Term' => 'Net-C6Term-0.11', |
234 |
'BitArray1' => 'BitArray', |
235 |
'Audio-Radio-214L' => 'Audio-Radio-V4L-0.01', |
236 |
'Devel-AutoProfiler' => 'Devel', |
237 |
'Brasil-Checar-CGC' => 'Brasil', |
238 |
'AI-NeuralNet-SOM' => 'SOM.pm', |
239 |
'Net-BitTorrent-File-fix' => 'Net-BitTorrent-File-1.01', |
240 |
'VMS-FindFile' => 'VMS', |
241 |
'LoadHtml.' => 'README', |
242 |
'Time-Compare' => 'Time', |
243 |
'ShiftJIS-230213-MapUTF' => 'ShiftJIS-X0213-MapUTF-0.21', |
244 |
'Image-WMF' => 'Image', |
245 |
'sdf-2.0.eta' => 'sdf-2.001beta1', |
246 |
'Math-Expr-LATEST' => 'Math-Expr-0.4', |
247 |
'MP3-Player-PktConcert' => 'MP3', |
248 |
'Apache-OWA' => 'OWA', |
249 |
'Audio-Gramofile' => 'Audio', |
250 |
'DBIx-Copy' => 'Copy', |
251 |
'P4-024' => 'P4-C4-2.021', |
252 |
'Disassemble-2386' => 'Disassemble-X86-0.13', |
253 |
'Proc-Swarm' => 'Swarm-0.5', |
254 |
'Smil' => 'perlysmil', |
255 |
'Net-SSH-2232Perl' => 'Net-SSH-W32Perl-0.05', |
256 |
'Win32-SerialPort' => 'SerialPort-0.19', |
257 |
'Lingua-ID-Words2Nums' => 'Words2Nums-0.01', |
258 |
'Parse-Text' => 'Text', |
259 |
'DBIx-HTMLView-LATEST' => 'DBIx-HTMLView-0.9', |
260 |
'Apache-NNTPGateway' => 'NNTPGateway-0.9', |
261 |
'XPathToXML' => 'XPathToXML.pm', |
262 |
'XML-WMM-ASX' => 'XML', |
263 |
'CGISession' => 'CGI', |
264 |
'Net-SMS-142' => 'Net-SMS-O2-0.019', |
265 |
'Search-253950' => 'Search-Z3950-0.05', |
266 |
'Date-Christmas' => 'Christmas', |
267 |
'Win32-InternetExplorer-Window' => 'Win32', |
268 |
'Apache-WAP-MailPeek' => 'MailPeek', |
269 |
'Statistics-Table-F' => 'Statistics', |
270 |
'BerkeleyDB_Locks' => 'BerkeleyDB-Locks-0_2', |
271 |
'HookPrePostCall' => 'PrePostCall-1.2', |
272 |
'Oak-AAS-Service-DBI_13_PAM' => 'Oak-AAS-Service-DBI_N_PAM-1.8', |
273 |
'Math-Vector' => 'Vector.pm', |
274 |
'Audio-124pDecrypt' => 'Audio-M4pDecrypt-0.04', |
275 |
'libao-perl_0.03' => 'libao-perl-0.03', |
276 |
'CGI-EZForm' => 'EZForm', |
277 |
'Data-Locations-fixed' => 'Data-Locations-5.2-fixed', |
278 |
'HTML-Template-Filter-Dreamweaver' => 'Dreamweaver', |
279 |
'LineByLine' => 'LineByLine.pm', |
280 |
'Geo-0400' => 'Geo-E00-0.05', |
281 |
'WebService-Validator-HTML-223C' => 'WebService-Validator-HTML-W3C-0.03', |
282 |
'DateTime-Format-223CDTF' => 'DateTime-Format-W3CDTF-0.04', |
283 |
'DBD_SQLFLEX' => 'DBD-Sqlflex', |
284 |
'Text-Number' => 'Number', |
285 |
'DBIx-DataLookup' => 'DBIx', |
286 |
'MP3-ID3211Tag' => 'MP3-ID3v1Tag-1.11', |
287 |
'Text-Striphigh' => 'Striphigh-0.02', |
288 |
'Tie-SortHash' => 'SortHash', |
289 |
'Apache-AccessAbuse' => 'AccessAbuse', |
290 |
'MP3-123U-Parser' => 'MP3-M3U-Parser', |
291 |
'Net-253950' => 'Net-Z3950-0.44', |
292 |
'Net-RBLClient' => 'RBLCLient-0.2', |
293 |
'CGI-EasyCGI' => 'CGI', |
294 |
'http-handle' => 'HTTP::Handle', |
295 |
'JPEG-Comment' => 'JPEG', |
296 |
'router-lg' => 'Router', |
297 |
'Db-Mediasurface' => 'Mediasurface', |
298 |
'Text-BarGraph' => 'bargraph', |
299 |
'TL' => 'Text-Levenshtein-0.04', |
300 |
'Config-Vars' => 'Config-0.01', |
301 |
'Tie-PerfectHash' => 'Tie', |
302 |
'DNS-TinyDNS' => 'DNS', |
303 |
'DesignPattern-Factory' => 'Factory', |
304 |
'WWW-01_Rail' => 'WWW-B_Rail-0.01', |
305 |
'Win32-Exchange' => 'blib', |
306 |
'Math-RPN' => 'Math', |
307 |
'Db-Mediasurface-Cache' => 'Cache', |
308 |
'perl_archie.' => 'Archie.pm', |
309 |
'Acme-PGPSign' => 'Acme', |
310 |
'HTML-Widget-sideBar' => 'HTML-Widget-SideBar-1.00', |
311 |
'log' => 'Games', |
312 |
'File-List' => 'File', |
313 |
'Schedule-Cronchik' => 'Schedule', |
314 |
'Curses-Devkit' => 'Cdk', |
315 |
'Pod-PalmDoc' => 'Pod', |
316 |
'Easy-WML' => 'Easy WML 0.1', |
317 |
'Interval.' => 'Date', |
318 |
'Brasil-Checar-CPF' => 'Brasil', |
319 |
'Apache-WAP-AutoIndex' => 'AutoIndex', |
320 |
|
321 |
'SOM.pm' => 'SOM.pm', |
322 |
'PlugIn.pm' => 'PlugIn.pm', |
323 |
'XPathToXML.pm' => 'XPathToXML.pm', |
324 |
'Vector.pm' => 'Vector.pm', |
325 |
'LineByLine.pm' => 'LineByLine.pm', |
326 |
'Archie.pm' => 'Archie.pm', |
327 |
'TieConstant.pm' => 'TieConstant.pm', |
328 |
'EPS.pm' => 'EPS.pm', |
329 |
'SQLRelay.pm' => 'SQLRelay.pm', |
330 |
'Resources.pm' => 'Resources.pm', |
331 |
'README' => 'README', |
332 |
|
333 |
); |
334 |
my %NEW_ALIAS; # found in this pass |
335 |
|
336 |
# Map module names to pathes. Generated by wanted() doing alisaing. |
337 |
my %ARCHIVE; |
338 |
|
339 |
# Map module names to latest version. Generated by wanted() |
340 |
my %VERSION; |
341 |
|
342 |
|
343 |
# Mapping for modules with common root not matching modules name that |
344 |
# are not aliased. This is just for prefix stripping and not strictly |
345 |
# necessary. Takes effect after version considerations. |
346 |
my %TR = (# tar name root to strip |
347 |
'Net_SSLeay.pm' => 'SSLeay/', |
348 |
'EventDrivenServer' => 'Server/', |
349 |
'bio_lib.pl.' => '', |
350 |
'AlarmCall' => 'Sys/', |
351 |
'Cdk-ext' => 'Cdk/', |
352 |
'Sx' => '\d.\d/', |
353 |
'DumpStack' => 'Devel/', |
354 |
'StatisticsDescriptive'=> 'Statistics/', |
355 |
'Term-Gnuplot' => 'Gnuplot/', |
356 |
'iodbc_ext' => 'iodbc-ext-\d.\d/', |
357 |
'UNIVERSAL' => '', |
358 |
'Term-Query' => 'Query/', |
359 |
'SelfStubber' => 'Devel/', |
360 |
'CallerItem' => 'Devel/', |
361 |
); |
362 |
|
363 |
my $LWP; |
364 |
|
365 |
# FIXME |
366 |
my $DIR = '/rest/estseek/cpan/'; |
367 |
my $DATA = $DIR . '/data'; |
368 |
|
369 |
|
370 |
if (@{$OPT{remove}}) { |
371 |
my $pod; |
372 |
for $pod (@{$OPT{remove}}) { |
373 |
unless (-e $pod) { |
374 |
$pod = "$DIR/$pod"; |
375 |
} |
376 |
index_pod(file => $pod, remove => 1) if -f $pod; |
377 |
unlink $pod or warn "Could not unlink '$pod': $!\n"; |
378 |
} |
379 |
exit; |
380 |
} |
381 |
|
382 |
# Now get the beef |
383 |
if ($OPT{cpan} =~ /^(http|ftp):/) { |
384 |
$LWP = 1; |
385 |
require LWP::Simple; |
386 |
LWP::Simple->import(); |
387 |
|
388 |
mkpath($DATA,1,0755) or |
389 |
die "Could not generate '$DATA/': $!" |
390 |
unless -d $DATA; |
391 |
|
392 |
if (! -f "$DATA/find-ls.gz" or -M "$DATA/find-ls.gz" > 0.5) { |
393 |
my $status = mirror("$OPT{cpan}/indices/find-ls.gz", "$DATA/find-ls.gz"); |
394 |
if ($status != &RC_OK and $status != &RC_NOT_MODIFIED) { |
395 |
# we could use Net:FTP here ... |
396 |
die "Was unable to mirror '$OPT{cpan}/indices/find-ls.gz'\n"; |
397 |
} |
398 |
} |
399 |
my $fh = new IO::File "gzip -cd $DATA/find-ls.gz |"; |
400 |
die "Could not open 'gzip -cd $DATA/find-ls.gz': !$\n" unless $fh; |
401 |
|
402 |
my $line; |
403 |
while (defined ($line = <$fh>)) { |
404 |
chomp($line); |
405 |
my ($mon, $mday, $time, $file, $is_link) = (split ' ', $line)[7..11]; |
406 |
|
407 |
next if defined $is_link; |
408 |
my $mtime = mtime($mon, $mday, $time); |
409 |
|
410 |
$file =~ s:^\./::; |
411 |
($_) = fileparse($file); |
412 |
$File::Find::name = $file; |
413 |
wanted($mtime); |
414 |
} |
415 |
} else { |
416 |
find(sub {&wanted((stat($_))[9])}, $OPT{cpan}); |
417 |
} |
418 |
|
419 |
ARCHIVE: |
420 |
for my $tar (sort keys %ARCHIVE) { |
421 |
next if $OPT{match} and $ARCHIVE{$tar} !~ /$OPT{match}/o; |
422 |
my $base = (split /\//, $ARCHIVE{$tar})[-1]; |
423 |
my $parent; |
424 |
|
425 |
my %attr; |
426 |
|
427 |
# logging |
428 |
if ($OPT{trust_mtime}) { |
429 |
$attr{'@mdate'} = strftime('%Y-%m-%dT%H:%M:%S+00:00', gmtime($VERSION{$tar})); |
430 |
$parent->{'@mdate'} = $attr{'@mdate'}; |
431 |
printf "%-20s %10s %s\t", $tar, $attr{'@mdate'}, $base; |
432 |
} else { |
433 |
$attr{'version'} = $VERSION{$tar}; |
434 |
printf "%-20s %10.5f %s\t", $tar, $attr{'version'}, $base; |
435 |
} |
436 |
|
437 |
# Remember the archive |
438 |
# We should have an extra table for the tar file data ... |
439 |
if (!$OPT{force} and $tb->have(docid => $base)) { |
440 |
print "skipping\n"; |
441 |
next ARCHIVE; |
442 |
} else { |
443 |
$parent->{_id} = $tb->insert(docid => $base, |
444 |
headline => $ARCHIVE{$tar}, |
445 |
%attr |
446 |
) unless $OPT{test}; |
447 |
print "indexing\n"; |
448 |
} |
449 |
|
450 |
next ARCHIVE if $OPT{test} > 1; |
451 |
|
452 |
my $TAR = myget($tar); |
453 |
|
454 |
next ARCHIVE unless $TAR; # not able to fetch it |
455 |
|
456 |
my %tar; |
457 |
tie (%tar, |
458 |
'WAIT::Document::Tar', |
459 |
sub { $_[0] =~ /\.(pm|pod|PL)$/i or $_[0] =~ /readme/i}, |
460 |
#sub { $_[0] !~ m:/$: }, |
461 |
$TAR) |
462 |
or warn "Could not tie '$TAR'\n"; |
463 |
|
464 |
my $sloppy; |
465 |
my ($key, $val); |
466 |
|
467 |
FILE: |
468 |
while (($key, $val) = each %tar) { |
469 |
my $file = fname($key); |
470 |
|
471 |
# don't index directories |
472 |
next if $file =~ /\/$/; |
473 |
|
474 |
# is it a POD file? |
475 |
next FILE unless $file =~ /readme/i or $val =~ /\n=head/; |
476 |
|
477 |
# remove directory prefix |
478 |
unless ($sloppy # no common root |
479 |
or $file =~ s:^\Q$tar\E[^/]*/:: # common root, maybe alias |
480 |
or ($TR{$tar} # common root, not aliased |
481 |
and $file =~ s:^\Q$TR{$tar}\E::) |
482 |
) { |
483 |
# try to determine an alias |
484 |
warn "Bad directory prefix: '$file'\n"; |
485 |
my ($prefix) = split /\//, $file; |
486 |
|
487 |
while ($key = (tied %tar)->NEXTKEY) { |
488 |
my $file = fname($key); |
489 |
|
490 |
next if $file =~ /\/$/; |
491 |
unless ($file =~ m:^$prefix/: or $file eq $prefix) { |
492 |
warn "Archive contains different prefixes: $prefix,$file\n"; |
493 |
$prefix = ''; |
494 |
last; |
495 |
} |
496 |
} |
497 |
if ($prefix) { |
498 |
print "Please alias '$tar' to '$prefix' next time!\n"; |
499 |
print "See alias table later.\n"; |
500 |
$NEW_ALIAS{$tar} = $prefix; |
501 |
$tb->delete_by_key($parent->{_id}); |
502 |
next ARCHIVE; |
503 |
} else { |
504 |
print "Assuming that tar file name $tar is a valid prefix\n"; |
505 |
$sloppy = 1; |
506 |
|
507 |
# We may reset too much here! But that this is not exact |
508 |
# science anyway. Maybe we should ignore using 'next ARCHIVE'. |
509 |
|
510 |
$key = (tied %tar)->FIRSTKEY; |
511 |
redo FILE; |
512 |
} |
513 |
} |
514 |
|
515 |
# remove /lib prefix |
516 |
$file =~ s:^lib/::; |
517 |
|
518 |
# generate new path |
519 |
my $path = "$DATA/$tar/$file"; |
520 |
|
521 |
my ($sbase, $sdir) = fileparse($path); |
522 |
my $fh; |
523 |
|
524 |
unless ($OPT{test}) { |
525 |
if (-f $path) { |
526 |
index_pod(file => $path, remove => 1); |
527 |
unlink $path or warn "Could not unlink '$path' $!\n"; |
528 |
} elsif (!-d $sdir) { |
529 |
mkpath($sdir,1,0755) or die "Could not mkpath($sdir): $!\n"; |
530 |
} |
531 |
# $fh = new IO::File "> $path"; |
532 |
$fh = new IO::Zlib "$path.gz","wb"; |
533 |
die "Could not write '$path': $!\n" unless $fh; |
534 |
} |
535 |
|
536 |
if ($file =~ /readme|install/i) { # make READMEs verbatim pods |
537 |
$val =~ s/\n/\n /g; |
538 |
$val = "=head1 NAME\n\n$tar $file\n\n=head1 DESCRIPTION\n\n $val" |
539 |
unless $val =~ /^=head/m; |
540 |
} else { # remove non-pod stuff |
541 |
my $nval = $val; $val = ''; |
542 |
my $cutting = 1; |
543 |
|
544 |
for (split /\n/, $nval) { |
545 |
if (/^=cut|!NO!SUBS!/) { |
546 |
$cutting = 1; |
547 |
} elsif ($cutting and /^=head/) { |
548 |
$cutting = 0; |
549 |
} |
550 |
unless ($cutting) { |
551 |
$val .= $_ . "\n"; |
552 |
} |
553 |
} |
554 |
} |
555 |
unless ($OPT{test}) { |
556 |
$fh->print($val); |
557 |
index_pod(file => $path, parent => $parent, |
558 |
text => $val, source => $ARCHIVE{$tar}, |
559 |
); |
560 |
} |
561 |
} |
562 |
|
563 |
if ($LWP and !$OPT{keep}) { |
564 |
unlink $TAR or warn |
565 |
"Could not unlink '$TAR': $!\n"; |
566 |
} |
567 |
} |
568 |
|
569 |
if (%NEW_ALIAS) { |
570 |
print "\%ALIAS = (\n"; |
571 |
for (keys %NEW_ALIAS) { |
572 |
print "\t'$_'\t=> '$NEW_ALIAS{$_}',\n"; |
573 |
} |
574 |
print "\t);\n"; |
575 |
} |
576 |
|
577 |
exit; |
578 |
|
579 |
sub fname ($) { |
580 |
my $key = shift; |
581 |
my ($ntar, $file) = split $;, $key; |
582 |
|
583 |
# remove leading './' - shudder |
584 |
$file =~ s/^\.\///; |
585 |
|
586 |
return($file); |
587 |
} |
588 |
|
589 |
sub myget { |
590 |
my $tar = shift; |
591 |
my $TAR; |
592 |
|
593 |
if ($LWP) { # fetch the archive |
594 |
if ($OPT{keep}) { |
595 |
$TAR = "$OPT{keep}/$ARCHIVE{$tar}"; |
596 |
print "Keeping in '$TAR'\n" unless -e $TAR; |
597 |
my ($base, $path) = fileparse($TAR); |
598 |
unless (-d $path) { |
599 |
mkpath($path,1,0755) or |
600 |
die "Could not mkpath($path)\n"; |
601 |
} |
602 |
} else { |
603 |
$TAR = "/tmp/$tar.tar.gz"; |
604 |
} |
605 |
unless (-e $TAR) { # lwp mirror seems to fetch ftp: in any case? |
606 |
print "Fetching $OPT{cpan}/$ARCHIVE{$tar}\n"; |
607 |
my $status = mirror("$OPT{cpan}/$ARCHIVE{$tar}", $TAR); |
608 |
if ($status != &RC_OK and $status != &RC_NOT_MODIFIED) { |
609 |
warn "Was unable to mirror '$ARCHIVE{$tar}, skipping'\n"; |
610 |
return; |
611 |
} |
612 |
} |
613 |
} else { |
614 |
$TAR = $ARCHIVE{$tar}; |
615 |
} |
616 |
$TAR; |
617 |
} |
618 |
|
619 |
sub index_pod { |
620 |
my %parm = @_; |
621 |
my $did = $parm{file}; |
622 |
my $rel_did = $did; |
623 |
my $abs_did = $did; |
624 |
|
625 |
if ($rel_did =~ s:$DIR/::) { |
626 |
$abs_did = "$DIR/$rel_did"; |
627 |
} |
628 |
|
629 |
undef $did; |
630 |
|
631 |
# check for both variants |
632 |
if ($tb->have('docid' => $rel_did)) { |
633 |
$did = $rel_did; |
634 |
} elsif ($tb->have('docid' => $abs_did)) { |
635 |
$did = $abs_did; |
636 |
} |
637 |
if ($did) { # have it version |
638 |
if (!$parm{remove} and !$OPT{force}) { |
639 |
warn "duplicate: $did\n"; |
640 |
return; |
641 |
} |
642 |
} else { # not seen yet |
643 |
$did = $rel_did; |
644 |
if ($parm{remove}) { |
645 |
print "missing: $did\n"; |
646 |
return; |
647 |
} |
648 |
} |
649 |
|
650 |
$parm{'text'} ||= WAIT::Document::Find->FETCH($abs_did); |
651 |
|
652 |
unless (defined $parm{'text'}) { |
653 |
print "unavailable: $did\n"; |
654 |
return; |
655 |
} |
656 |
|
657 |
my $record = $layout->split($parm{'text'}); |
658 |
|
659 |
if (! $record) { |
660 |
print "empty pod: $did\n"; |
661 |
return; |
662 |
} |
663 |
|
664 |
$record->{size} = length($parm{'text'}); |
665 |
my $headline = $record->{name} || $did; |
666 |
|
667 |
# additional fields for Hyper Estraier |
668 |
$record->{'@mdate'} = $parm{'mdate'} if ($parm{'mdate'}); |
669 |
|
670 |
$headline =~ s/^$DATA//o; # $did |
671 |
$headline =~ s/\s+/ /g; $headline =~ s/^\s+//; |
672 |
|
673 |
printf "%s %s\n", ($parm{remove})?'-':'+', substr($headline,0,70); |
674 |
if ($parm{remove}) { |
675 |
$tb->delete('docid' => $did, |
676 |
headline => $headline, |
677 |
%{$record}); |
678 |
} else { |
679 |
foreach (keys %{$parm{parent}}) { |
680 |
next if (/^_/); |
681 |
$record->{$_} = $parm{parent}->{$_} if ($parm{parent}->{$_}); |
682 |
} |
683 |
$tb->insert('docid' => $did, |
684 |
headline => $headline, |
685 |
source => $parm{source}, |
686 |
parent => $parm{parent}->{_id}, |
687 |
%{$record}); |
688 |
} |
689 |
} |
690 |
|
691 |
# This *must* remove the version in *any* case. It should compute a |
692 |
# resonable version number - but usually mtimes should be used. |
693 |
sub version { |
694 |
local ($_) = @_; |
695 |
|
696 |
# remove alpha/beta postfix |
697 |
s/([-_\d])(a|b|alpha|beta|src)$/$1/; |
698 |
|
699 |
# jperl1.3@4.019.tar.gz |
700 |
s/@\d.\d+//; |
701 |
|
702 |
# oraperl-v2.4-gk.tar.gz |
703 |
s/-v(\d)/$1/; |
704 |
|
705 |
# lettered versions - shudder |
706 |
s/([-_\d\.])([a-z])([\d\._])/sprintf "$1%02d$3", ord(lc $2) - ord('a') /ei; |
707 |
s/([-_\d\.])([a-z])$/sprintf "$1%02d", ord(lc $2) - ord('a') /ei; |
708 |
|
709 |
# thanks libwww-5b12 ;-) |
710 |
s/(\d+)b/($1-1).'.'/e; |
711 |
s/(\d+)a/($1-2).'.'/e; |
712 |
|
713 |
# replace '-pre' by '0.' |
714 |
s/-pre([\.\d])/-0.$1/; |
715 |
s/\.\././g; |
716 |
s/(\d)_(\d)/$1$2/g; |
717 |
|
718 |
# chop '[-.]' and thelike |
719 |
s/\W$//; |
720 |
|
721 |
# ram's versions Storable-0.4@p |
722 |
s/\@/./; |
723 |
|
724 |
if (s/[-_]?(\d+)\.(0\d+)\.(\d+)$//) { |
725 |
return($_, $1 + "0.$2" + $3 / 1000000); |
726 |
} elsif (s/[-_]?(\d+)\.(\d+)\.(\d+)$//) { |
727 |
return($_, $1 + $2/1000 + $3 / 1000000); |
728 |
} elsif (s/[-_]?(\d+\.[\d_]+)$//) { |
729 |
return($_, $1); |
730 |
} elsif (s/[-_]?([\d_]+)$//) { |
731 |
return($_, $1); |
732 |
} elsif (s/-(\d+.\d+)-/-/) { # perl-4.019-ref-guide |
733 |
return($_, $1); |
734 |
} else { |
735 |
if ($_ =~ /\d/) { # smells like an unknown scheme |
736 |
warn "Odd version Numbering: '$File::Find::name'\n"; |
737 |
return($_, undef); |
738 |
} else { # assume version 0 |
739 |
warn "No version Numbering: '$File::Find::name'\n"; |
740 |
return($_, 0); |
741 |
} |
742 |
|
743 |
} |
744 |
} |
745 |
|
746 |
sub wanted { |
747 |
my $mtime = shift; # called by parse_file_ls(); |
748 |
|
749 |
return if (! $max); |
750 |
$max--; |
751 |
|
752 |
return unless /^(.*)\.(tar\.(gz|Z)|tgz)$/; |
753 |
|
754 |
my ($archive, $version) = version($1); |
755 |
|
756 |
unless (defined $version) { |
757 |
warn "Skipping $1\n"; |
758 |
return; |
759 |
} |
760 |
|
761 |
# Check for file alias |
762 |
$archive = $ALIAS{$archive} if $ALIAS{$archive}; |
763 |
|
764 |
# Check for path alias. |
765 |
if ($File::Find::name =~ m(/CPAN/(?:source/)?(.*\Q$archive\E))) { |
766 |
if ($ALIAS{$1}) { |
767 |
$archive = $ALIAS{$1}; |
768 |
} |
769 |
} |
770 |
|
771 |
if ($OPT{trust_mtime}) { |
772 |
$version = $mtime; |
773 |
} else { |
774 |
$version =~ s/(\d)_/$1/; |
775 |
$version ||= $mtime; # mtime |
776 |
} |
777 |
|
778 |
if (!exists $ARCHIVE{$archive} |
779 |
or $VERSION{$archive} < $version) { |
780 |
$ARCHIVE{$archive} = $File::Find::name; |
781 |
$VERSION{$archive} = $version; |
782 |
} |
783 |
} |
784 |
|
785 |
my %MON; |
786 |
my $YEAR; |
787 |
|
788 |
BEGIN { |
789 |
my $i = 1; |
790 |
for (qw(Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec)) { |
791 |
$MON{$_} = $i++; |
792 |
} |
793 |
$YEAR = (localtime(time))[5]; |
794 |
} |
795 |
|
796 |
# We could/should use Date::GetDate here |
797 |
use Time::Local; |
798 |
sub mtime { |
799 |
my ($mon, $mday, $time) = @_; |
800 |
my ($hour, $min, $year, $monn) = (0,0); |
801 |
|
802 |
if ($time =~ /(\d+):(\d+)/) { |
803 |
($hour, $min) = ($1, $2); |
804 |
$year = $YEAR; |
805 |
} else { |
806 |
$year = $time; |
807 |
} |
808 |
$monn = $MON{$mon} || $MON{ucfirst lc $mon} || warn "Unknown month: '$mon'"; |
809 |
my $guess = timelocal(0,$min,$hour,$mday,$monn-1,$year); |
810 |
if ($guess > time) { |
811 |
$guess = timelocal(0,$min,$hour,$mday,$monn-1,$year-1); |
812 |
} |
813 |
$guess; |
814 |
} |
815 |
|
816 |
package HyperEstraier::WAIT::Table; |
817 |
|
818 |
use HyperEstraier; |
819 |
use Text::Iconv; |
820 |
|
821 |
=head1 NAME |
822 |
|
823 |
HyperEstraier::WAIT::Table |
824 |
|
825 |
=head1 DESCRIPTION |
826 |
|
827 |
This is a mode that emulates C<WAIT::Table> functionality somewhat. |
828 |
|
829 |
There are some limitations and only one key attribute is supported (and used |
830 |
for C<@uri>). |
831 |
|
832 |
=head2 Porting from WAIT to this module. |
833 |
|
834 |
Since only one key is supported (and used as C<@uri> attribute), |
835 |
use first parametar of C<keyset> as C<key>. |
836 |
|
837 |
Full text index is specified as C<invindex>, but you need just name of fields. |
838 |
|
839 |
You will probably need to add |
840 |
|
841 |
use WAIT::Parse::Base; |
842 |
|
843 |
to your code after you remove C<WAIT::Config> and C<WAIT::Database>. |
844 |
|
845 |
=head1 METHODS |
846 |
|
847 |
=head2 new |
848 |
|
849 |
my $tb = new HyperEstraier::WAIT::Table( |
850 |
uri => 'http://localhost:1978/node/cpan', |
851 |
attr => qw/docid headline source size parent/, |
852 |
key => 'docid', |
853 |
invindex => qw/name synopsis bugs description text environment example author/, |
854 |
); |
855 |
|
856 |
=cut |
857 |
|
858 |
sub new { |
859 |
my $class = shift; |
860 |
my $self = {@_}; |
861 |
bless($self, $class); |
862 |
|
863 |
foreach my $p (qw/uri attr key invindex/) { |
864 |
die "need $p" unless ($self->{$p}); |
865 |
} |
866 |
|
867 |
$self->{'iso2utf'} = Text::Iconv->new('ISO-8859-1','UTF-8'); |
868 |
|
869 |
my $node = HyperEstraier::Node->new($self->{'uri'}); |
870 |
$node->set_auth('admin', 'admin'); |
871 |
|
872 |
$self->{'node'} = $node; |
873 |
|
874 |
$self ? return $self : return undef; |
875 |
} |
876 |
|
877 |
=head2 have |
878 |
|
879 |
if ( $tb->have(docid => $something) ) ... |
880 |
|
881 |
=cut |
882 |
|
883 |
sub have { |
884 |
my $self = shift; |
885 |
my $args = {@_}; |
886 |
my $key = $self->{'key'} || die "no key in object"; |
887 |
my $key_v = $args->{$key} || die "no key $key in data"; |
888 |
|
889 |
my $id = $self->{'node'}->uri_to_id('file://' . $key_v); |
890 |
|
891 |
return ($id == -1 ? undef : $id); |
892 |
} |
893 |
|
894 |
=head2 insert |
895 |
|
896 |
my $key = $tb->insert( |
897 |
docid => $base, |
898 |
headline => 'Something', |
899 |
... |
900 |
); |
901 |
|
902 |
=cut |
903 |
|
904 |
sub insert { |
905 |
my $self = shift; |
906 |
my $args = {@_}; |
907 |
|
908 |
my $uri = 'file://'; |
909 |
$uri .= $args->{'docid'} or die "no docid"; |
910 |
|
911 |
my $doc = HyperEstraier::Document->new; |
912 |
|
913 |
$doc->add_attr('@uri', $uri); |
914 |
$doc->add_attr('@title', $args->{'headline'}) if ($args->{'headline'}); |
915 |
$doc->add_attr('@size', $args->{'size'}) if ($args->{'size'}); |
916 |
|
917 |
my @attr = $self->{'attr'} || die "no attr in object"; |
918 |
my @invindex = $self->{'invindex'} || die "no invindex in object"; |
919 |
|
920 |
foreach my $attr (keys %{$args}) { |
921 |
if (grep(/^$attr$/, @{ $self->{'attr'} }) or $attr =~ m/^@/o) { |
922 |
$doc->add_attr($attr, $args->{$attr}); |
923 |
} |
924 |
if (grep(/^$attr$/, @{ $self->{'invindex'} })) { |
925 |
$doc->add_text($args->{$attr}); |
926 |
} |
927 |
} |
928 |
|
929 |
print STDERR $doc->dump_draft if ($self->{'debug'}); |
930 |
|
931 |
my $id; |
932 |
unless ($self->{'node'}->put_doc($doc)) { |
933 |
printf STDERR "ERROR: %d\n", $self->{'node'}->status; |
934 |
} else { |
935 |
$id = $self->{'node'}->uri_to_id( $uri ); |
936 |
if ($id != -1) { |
937 |
print STDERR "id: $id\n" if ($self->{'debug'}) |
938 |
} else { |
939 |
print STDERR "ERROR: can't find id for newly insrted document $uri\n"; |
940 |
} |
941 |
} |
942 |
|
943 |
return $id; |
944 |
} |
945 |
|
946 |
=head2 delete_by_key |
947 |
|
948 |
$tb->delete_by_key($key); |
949 |
|
950 |
=cut |
951 |
|
952 |
sub delete_by_key { |
953 |
my $self = shift; |
954 |
my $key_v = shift || die "no key?"; |
955 |
|
956 |
unless ($self->{'node'}->out_doc_by_uri( 'file://' . $key_v )) { |
957 |
print STDERR "WARNING: can't delete document $key_v\n"; |
958 |
} |
959 |
} |
960 |
|
961 |
=head2 delete |
962 |
|
963 |
$tb->delete( docid => $did, ... ); |
964 |
|
965 |
=cut |
966 |
|
967 |
sub delete { |
968 |
my $self = shift; |
969 |
my $args = {@_}; |
970 |
|
971 |
my $key = $self->{'key'} || die "no key in object"; |
972 |
|
973 |
die "no $key in data" unless (my $key_v = $args->{$key}); |
974 |
|
975 |
$self->delete_by_key($key_v); |
976 |
|
977 |
} |