/[hyperestraier_wrappers]/trunk/perl/scripts/est-spider
This is repository of my old source code which isn't updated any more. Go to git.rot13.org for current projects!
ViewVC logotype

Annotation of /trunk/perl/scripts/est-spider

Parent Directory Parent Directory | Revision Log Revision Log


Revision 27 - (hide annotations)
Sat Sep 17 23:07:52 2005 UTC (17 years, 2 months ago) by dpavlin
File size: 4959 byte(s)
use File::MMagic::XS to perform better checking on input formats
1 dpavlin 5 #!/usr/bin/perl -w
2     use strict;
3     use File::Find;
4     use Getopt::Long;
5     use File::Which;
6     use HyperEstraier;
7     use Text::Iconv;
8 dpavlin 27 #use File::MMagic;
9     use File::MMagic::XS qw/:compat/;
10 dpavlin 5
11 dpavlin 18 # do we use Node API?
12     my $node_url;
13    
14 dpavlin 5 my $collection; # name which will be inserted
15     my $path_add; # add additional info in path
16     my $verbose;
17     my $exclude;
18    
19     #$verbose = 1;
20    
21     my $result = GetOptions(
22     "collection=s" => \$collection,
23     "path=s" => \$path_add,
24     "verbose!" => \$verbose,
25     "debug!" => \$verbose,
26     "exclude=s" => \$exclude,
27 dpavlin 18 "node=s" => \$node_url,
28 dpavlin 5 );
29    
30     my $dir = shift @ARGV || die "usage: $0 [dir]";
31    
32     #my $basedir = $0;
33     #$basedir =~ s,/[^/]+$,/,;
34     #require "$basedir/filter.pm";
35    
36     my $pdftotext = which('pdftotext');
37    
38 dpavlin 27 #my $mm = new File::MMagic('/usr/share/misc/file/magic');
39     my $mm = new File::MMagic::XS();
40    
41 dpavlin 5 my $iconv = new Text::Iconv('iso-8859-2', 'utf-8');
42    
43     select(STDERR); $|=1;
44     select(STDOUT); $|=1;
45    
46     print STDERR "using $pdftotext to convert pdf into html\n" if ($pdftotext && $verbose);
47    
48 dpavlin 18 my $db;
49     if ($node_url) {
50     $db = HyperEstraier::Node->new($node_url);
51     $db->set_auth('admin', 'admin');
52     } else {
53     # open the database
54     $db = HyperEstraier::Database->new();
55     $db->open('/tmp/casket', $HyperEstraier::Database::DBWRITER | $HyperEstraier::Database::DBCREAT);
56 dpavlin 5
57 dpavlin 18 sub signal {
58     my($sig) = @_;
59     print "\nCaught a SIG$sig--syncing database and shutting down\n";
60     $db->sync();
61     exit(0);
62     }
63 dpavlin 5
64 dpavlin 18 $SIG{'INT'} = \&signal;
65     $SIG{'QUIT'} = \&signal;
66 dpavlin 5 }
67    
68     find({ wanted => \&file,
69     follow => 1,
70 dpavlin 26 follow_skip => 2,
71     no_chdir => 1,
72 dpavlin 5 }, $dir);
73    
74 dpavlin 18 unless ($node_url) {
75     print "--- sync\n";
76     $db->sync();
77 dpavlin 5
78 dpavlin 18 print "--- optimize...\n";
79     $db->optimize(0);
80     }
81 dpavlin 5 exit;
82    
83     sub dump_contents($$$$) {
84     my ($db,$contents,$mtime,$path) = @_;
85    
86     return unless ($contents); # don't die on empty files
87    
88     if ($exclude && $path =~ m/$exclude/i) {
89     print STDERR "skip: $path\n" if ($verbose);
90     return;
91     }
92    
93     use bytes;
94     my $size = length $contents;
95    
96     print STDERR " [$size]" if ($verbose);
97    
98     # create a document object
99     my $doc = HyperEstraier::Document->new;
100    
101     my $title = $1 if ($contents =~ m#<title>(.+)</title>#is);
102    
103 dpavlin 27 # chop long titles to 100 chars
104     $title = substr($title, 0, 100) . '...' if ($title && length($title) > 100);
105     # use path if no title is found
106     $title ||= $path;
107    
108 dpavlin 5 # add attributes to the document object
109     $doc->add_attr('@uri', "file:///$path");
110 dpavlin 27 $doc->add_attr('@title', $iconv->convert($title));
111 dpavlin 5 $doc->add_attr('@size', $size);
112     $doc->add_attr('@mtime', $mtime);
113    
114     # html2text
115     $contents =~ s#<[^>]+/*>##gs;
116     $contents =~ s#\s\s+# #gs;
117    
118     $doc->add_text($iconv->convert($contents));
119    
120     # print $doc->dump_draft if ($verbose);
121    
122     # register the document object to the database
123 dpavlin 18 if ($node_url) {
124     $db->put_doc($doc);
125     } else {
126     $db->put_doc($doc, $HyperEstraier::Database::PDCLEAN);
127     }
128 dpavlin 5
129     }
130    
131     sub file {
132    
133     my $path = $_;
134     my $contents;
135    
136 dpavlin 27 return if (-l $path || $path =~ m#/.svn#);
137 dpavlin 5
138 dpavlin 27 my $type = $mm->checktype_filename($path);
139     $type =~ s/\s+/ /gs;
140 dpavlin 5
141 dpavlin 27 print STDERR "# $path $type\n" if ($verbose);
142    
143     if ($pdftotext && -f $path && $type =~ m/pdf/i) {
144    
145 dpavlin 5 print STDERR "$path {converting}" if ($verbose);
146    
147     open(F,"$pdftotext -htmlmeta \"$path\" - |") || die "can't open $pdftotext with '$path'";
148     my $html;
149     while(<F>) {
150     # XXX why pdftotext barks if I try to use this is beyond me.
151     #$contents .= $_;
152    
153     $html .= $_;
154     }
155     close(F);
156    
157     return if (! $html);
158    
159     my $file_only = $path;
160     $file_only =~ s/^.*\/([^\/]+)$/$1/g;
161    
162     my ($pre_html,$pages,$post_html) = ('<html><head><title>$path :: page ##page_nr##</title></head><body><pre>',$html,'</pre></body></html>');
163    
164     ($pre_html,$pages,$post_html) = ($1,$2,$3) if ($html =~ m/^(<html>.+?<pre>)(.+)(<\/pre>.+?)$/si);
165    
166     if ($collection) {
167     $pre_html =~ s/<title>(.+?)<\/title>/<title>$collection :: page ##page_nr##<\/title>/si;
168     } else {
169     $pre_html =~ s/<title>(.+?)<\/title>/<title>$1 :: page ##page_nr##<\/title>/si ||
170     $pre_html =~ s/<title><\/title>/<title>$file_only :: page ##page_nr##<\/title>/si;
171     }
172    
173     my $page_nr = 1;
174     foreach my $page (split(/\f/s,$pages)) {
175     print STDERR " $page_nr" if ($verbose);
176     my $pre_tmp = $pre_html;
177     $pre_tmp =~ s/##page_nr##/$page_nr<\/title>/s;
178     dump_contents($db, $pre_tmp . $page . $post_html,time(), $path) if ($page !~ m/^\s*$/s);
179     $page_nr++;
180     }
181    
182     } else {
183    
184 dpavlin 27 # return if (! -f $path || ! m/\.(html*|php|pl|txt|info|log|text)$/i);
185     return if (! -f $path || (
186     $type !~ m/html/ ||
187     ($type !~ m#text/plain# && m/\.(php|pl|txt|info|log|text)$/)
188     ));
189 dpavlin 5
190     # skip index files
191     return if (m/index_[a-z]\.html*/i || m/index_symbol\.html*/i);
192    
193     open(F,"$path") || die "can't open file: $path";
194 dpavlin 27 print STDERR "$path ($type)" if ($verbose);
195 dpavlin 5 while(<F>) {
196     $contents .= "$_";
197     }
198     $contents .= "\n\n";
199    
200     #$contents = filter($contents,$collection);
201    
202     # add optional components to path
203     $path .= " $path_add" if ($path_add);
204    
205     dump_contents($db, $contents,time(), $path);
206     }
207    
208     print STDERR "\n" if ($verbose);
209     # die "zero size content in '$path'" if (! $contents);
210    
211     }
212    

Properties

Name Value
svn:executable *

  ViewVC Help
Powered by ViewVC 1.1.26