1 | # This document processor outputs a document
|
---|
2 | # for mg to process
|
---|
3 |
|
---|
4 |
|
---|
5 | package mgbuildproc;
|
---|
6 |
|
---|
7 | use classify;
|
---|
8 | use doc;
|
---|
9 | use docproc;
|
---|
10 | use util;
|
---|
11 |
|
---|
12 |
|
---|
13 | BEGIN {
|
---|
14 | @ISA = ('docproc');
|
---|
15 | }
|
---|
16 |
|
---|
17 |
|
---|
18 | sub new {
|
---|
19 | my ($class, $collection, $source_dir, $build_dir, $verbosity) = @_;
|
---|
20 | my $self = new docproc ();
|
---|
21 |
|
---|
22 | $self->{'collection'} = $collection;
|
---|
23 | $self->{'source_dir'} = $source_dir;
|
---|
24 | $self->{'build_dir'} = $build_dir;
|
---|
25 | $self->{'verbosity'} = $verbosity;
|
---|
26 | $self->{'classifiers'} = [];
|
---|
27 | $self->{'mode'} = "text";
|
---|
28 | $self->{'index'} = "section:text";
|
---|
29 | $self->{'indexexparr'} = [];
|
---|
30 | $self->{'output_handle'} = "STDOUT";
|
---|
31 | $self->{'num_docs'} = 0;
|
---|
32 | $self->{'num_sections'} = 0;
|
---|
33 | $self->{'num_bytes'} = 0;
|
---|
34 |
|
---|
35 | $self->{'indexing_text'} = 0;
|
---|
36 |
|
---|
37 | return bless $self, $class;
|
---|
38 | }
|
---|
39 |
|
---|
40 | sub reset {
|
---|
41 | my $self = shift (@_);
|
---|
42 |
|
---|
43 | $self->{'num_docs'} = 0;
|
---|
44 | $self->{'num_sections'} = 0;
|
---|
45 | $self->{'num_bytes'} = 0;
|
---|
46 | }
|
---|
47 |
|
---|
48 | sub get_num_docs {
|
---|
49 | my $self = shift (@_);
|
---|
50 |
|
---|
51 | return $self->{'num_docs'};
|
---|
52 | }
|
---|
53 |
|
---|
54 | sub get_num_sections {
|
---|
55 | my $self = shift (@_);
|
---|
56 |
|
---|
57 | return $self->{'num_sections'};
|
---|
58 | }
|
---|
59 |
|
---|
60 | sub get_num_bytes {
|
---|
61 | my $self = shift (@_);
|
---|
62 |
|
---|
63 | return $self->{'num_bytes'};
|
---|
64 | }
|
---|
65 |
|
---|
66 | sub set_output_handle {
|
---|
67 | my $self = shift (@_);
|
---|
68 | my ($handle) = @_;
|
---|
69 |
|
---|
70 | $self->{'output_handle'} = $handle;
|
---|
71 | }
|
---|
72 |
|
---|
73 | sub set_mode {
|
---|
74 | my $self = shift (@_);
|
---|
75 | my ($mode) = @_;
|
---|
76 |
|
---|
77 | $self->{'mode'} = $mode;
|
---|
78 | }
|
---|
79 |
|
---|
80 | sub set_index {
|
---|
81 | my $self = shift (@_);
|
---|
82 | my ($index, $indexexparr) = @_;
|
---|
83 |
|
---|
84 | $self->{'index'} = $index;
|
---|
85 | $self->{'indexexparr'} = $indexexparr if defined $indexexparr;
|
---|
86 | }
|
---|
87 |
|
---|
88 | sub set_classifiers {
|
---|
89 | my $self = shift (@_);
|
---|
90 | my ($classifiers) = @_;
|
---|
91 |
|
---|
92 | $self->{'classifiers'} = $classifiers;
|
---|
93 | }
|
---|
94 |
|
---|
95 | sub set_indexing_text {
|
---|
96 | my $self = shift (@_);
|
---|
97 | my ($indexing_text) = @_;
|
---|
98 |
|
---|
99 | $self->{'indexing_text'} = $indexing_text;
|
---|
100 | }
|
---|
101 |
|
---|
102 | sub process {
|
---|
103 | my $self = shift (@_);
|
---|
104 | my $method = $self->{'mode'};
|
---|
105 |
|
---|
106 | $self->$method(@_);
|
---|
107 | }
|
---|
108 |
|
---|
109 | sub infodb {
|
---|
110 | my $self = shift (@_);
|
---|
111 | my ($doc_obj, $filename) = @_;
|
---|
112 | my $handle = $self->{'output_handle'};
|
---|
113 | # $handle = "main::STDOUT";
|
---|
114 |
|
---|
115 | my $doctype = $doc_obj->get_doc_type();
|
---|
116 |
|
---|
117 | # only output this document if it is one to be indexed
|
---|
118 | return if ($doctype ne "indexed_doc");
|
---|
119 |
|
---|
120 | # this is another document
|
---|
121 | $self->{'num_docs'} += 1 unless ($doctype eq "classification");
|
---|
122 |
|
---|
123 | my $section = $doc_obj->get_top_section ();
|
---|
124 | my $doc_OID = $doc_obj->get_OID();
|
---|
125 | while (defined $section) {
|
---|
126 | # update a few statistics
|
---|
127 | $self->{'num_bytes'} += $doc_obj->get_text_length ($section);
|
---|
128 | $self->{'num_sections'} += 1 unless ($doctype eq "classification");
|
---|
129 |
|
---|
130 | # output the section name
|
---|
131 | if ($section eq "") { print $handle "[$doc_OID]\n"; }
|
---|
132 | else { print $handle "[$doc_OID.$section]\n"; }
|
---|
133 |
|
---|
134 | # output the fact that this document is a document
|
---|
135 | print $handle "<doctype>doc\n";
|
---|
136 |
|
---|
137 | # output whether this node contains text
|
---|
138 | if ($doc_obj->get_text_length($section) > 0) {
|
---|
139 | print $handle "<hastxt>1\n";
|
---|
140 | } else {
|
---|
141 | print $handle "<hastxt>0\n";
|
---|
142 | }
|
---|
143 |
|
---|
144 | # output all the section metadata
|
---|
145 | my $metadata = $doc_obj->get_all_metadata ($section);
|
---|
146 | foreach $pair (@$metadata) {
|
---|
147 | my ($field, $value) = (@$pair);
|
---|
148 |
|
---|
149 | if ($field ne "Identifier" && $field !~ /^gsdl/ &&
|
---|
150 | defined $value && $value ne "") {
|
---|
151 | # escape problematic stuff
|
---|
152 | $value =~ s/\\/\\\\/g;
|
---|
153 | $value =~ s/\n/\\n/g;
|
---|
154 | $value =~ s/\r/\\r/g;
|
---|
155 |
|
---|
156 | print $handle "<$field>$value\n";
|
---|
157 | }
|
---|
158 | }
|
---|
159 |
|
---|
160 | # output archivedir if at top level
|
---|
161 | if ($section eq $doc_obj->get_top_section()) {
|
---|
162 | my ($archivedir) = $filename =~ /^(.*?)(?:\/|\\)[^\/\\]*$/;
|
---|
163 | $archivedir = "" unless defined $archivedir;
|
---|
164 | $archivedir =~ s/^(\/|\\)*//;
|
---|
165 | $archivedir =~ s/(\/|\\)*$//;
|
---|
166 | print $handle "<archivedir>$archivedir\n";
|
---|
167 | }
|
---|
168 |
|
---|
169 | # output a list of children
|
---|
170 | my $children = $doc_obj->get_children ($section);
|
---|
171 | if (scalar(@$children) > 0) {
|
---|
172 | print $handle "<contains>";
|
---|
173 | my $firstchild = 1;
|
---|
174 | foreach $child (@$children) {
|
---|
175 | print $handle ";" unless $firstchild;
|
---|
176 | $firstchild = 0;
|
---|
177 | if ($child =~ /^.*?\.(\d+)$/) {
|
---|
178 | print $handle "\".$1";
|
---|
179 | } else {
|
---|
180 | print $handle "\".$child";
|
---|
181 | }
|
---|
182 | # if ($child eq "") { print $handle "$doc_OID"; }
|
---|
183 | # elsif ($section eq "") { print $handle "$doc_OID.$child"; }
|
---|
184 | # else { print $handle "$doc_OID.$section.$child"; }
|
---|
185 | }
|
---|
186 | print $handle "\n";
|
---|
187 | }
|
---|
188 |
|
---|
189 | # output the matching document number
|
---|
190 | print $handle "<docnum>$self->{'num_sections'}\n";
|
---|
191 |
|
---|
192 | print $handle '-' x 70, "\n";
|
---|
193 |
|
---|
194 |
|
---|
195 | # output a database entry for the document number
|
---|
196 | print $handle "[$self->{'num_sections'}]\n";
|
---|
197 | if ($section eq "") { print $handle "<section>$doc_OID\n"; }
|
---|
198 | else { print $handle "<section>$doc_OID.$section\n"; }
|
---|
199 | print $handle '-' x 70, "\n";
|
---|
200 |
|
---|
201 |
|
---|
202 | $section = $doc_obj->get_next_section($section);
|
---|
203 | }
|
---|
204 |
|
---|
205 | # classify this document
|
---|
206 | &classify::classify_doc ($self->{'classifiers'}, $doc_obj);
|
---|
207 |
|
---|
208 | }
|
---|
209 |
|
---|
210 | sub find_paragraphs {
|
---|
211 | $_[1] =~ s/(<p\b)/\cC$1/gi;
|
---|
212 | }
|
---|
213 |
|
---|
214 | sub filter_text {
|
---|
215 | # $self->filter_text ($field, $new_text);
|
---|
216 | # don't want to do anything for this version, however,
|
---|
217 | # in a particular collection you might want to override
|
---|
218 | # this method to post-process certain fields depending on
|
---|
219 | # the field, or whether we are outputting it for indexing
|
---|
220 | }
|
---|
221 |
|
---|
222 | sub text {
|
---|
223 | my $self = shift (@_);
|
---|
224 | my ($doc_obj) = @_;
|
---|
225 | my $handle = $self->{'output_handle'};
|
---|
226 | my $indexed_doc = 1;
|
---|
227 |
|
---|
228 | # only output this document if it is one to be indexed
|
---|
229 | return if ($doc_obj->get_doc_type() ne "indexed_doc");
|
---|
230 |
|
---|
231 | # see if this document belongs to this subcollection
|
---|
232 | foreach $indexexp (@{$self->{'indexexparr'}}) {
|
---|
233 | $indexed_doc = 0;
|
---|
234 | my ($field, $exp, $options) = split /\//, $indexexp;
|
---|
235 | if (defined ($field) && defined ($exp)) {
|
---|
236 | my ($bool) = $field =~ /^(.)/;
|
---|
237 | $field =~ s/^.// if $bool eq '!';
|
---|
238 | if ($field eq "filename") {
|
---|
239 | $field = $doc_obj->get_source_filename();
|
---|
240 | } else {
|
---|
241 | $field = $doc_obj->get_metadata_element($doc_obj->get_top_section(), $field);
|
---|
242 | }
|
---|
243 | next unless defined $field;
|
---|
244 | if ($bool eq '!') {
|
---|
245 | if ($options =~ /^i$/i) {
|
---|
246 | if ($field !~ /$exp/i) {$indexed_doc = 1; last;}
|
---|
247 | } else {
|
---|
248 | if ($field !~ /$exp/) {$indexed_doc = 1; last;}
|
---|
249 | }
|
---|
250 | } else {
|
---|
251 | if ($options =~ /^i$/i) {
|
---|
252 | if ($field =~ /$exp/i) {$indexed_doc = 1; last;}
|
---|
253 | } else {
|
---|
254 | if ($field =~ /$exp/) {$indexed_doc = 1; last;}
|
---|
255 | }
|
---|
256 | }
|
---|
257 | }
|
---|
258 | }
|
---|
259 |
|
---|
260 | # this is another document
|
---|
261 | $self->{'num_docs'} += 1;
|
---|
262 |
|
---|
263 | # get the parameters for the output
|
---|
264 | my ($level, $fields) = split (/:/, $self->{'index'});
|
---|
265 | $fields =~ s/\ball\b/Title,Creator,text/;
|
---|
266 | $fields =~ s/\btopall\b/topTitle,topCreator,toptext/;
|
---|
267 |
|
---|
268 | my $doc_section = 0; # just for this document
|
---|
269 | my $text = "";
|
---|
270 | my $text_extra = "";
|
---|
271 |
|
---|
272 | # get the text for this document
|
---|
273 | my $section = $doc_obj->get_top_section();
|
---|
274 | while (defined $section) {
|
---|
275 | # update a few statistics
|
---|
276 | $doc_section++;
|
---|
277 | $self->{'num_sections'} += 1;
|
---|
278 |
|
---|
279 | if ($indexed_doc) {
|
---|
280 | $self->{'num_bytes'} += $doc_obj->get_text_length ($section);
|
---|
281 | foreach $field (split (/,/, $fields)) {
|
---|
282 | # only deal with this field if it doesn't start with top or
|
---|
283 | # this is the first section
|
---|
284 | my $real_field = $field;
|
---|
285 | if (!($real_field =~ s/^top//) || ($doc_section == 1)) {
|
---|
286 | my $new_text = "";
|
---|
287 | if ($real_field eq "text") {
|
---|
288 | $new_text = $doc_obj->get_text ($section);
|
---|
289 | $new_text =~ s/[\cB\cC]//g;
|
---|
290 | $self->find_paragraphs($new_text);
|
---|
291 |
|
---|
292 | } else {
|
---|
293 | $new_text = join ("\cC", @{$doc_obj->get_metadata ($section, $real_field)});
|
---|
294 | }
|
---|
295 |
|
---|
296 | # filter the text
|
---|
297 | $self->filter_text ($field, $new_text);
|
---|
298 |
|
---|
299 | if ($self->{'indexing_text'} &&
|
---|
300 | $new_text =~ /[\(\)\{\}]/) {
|
---|
301 | }
|
---|
302 |
|
---|
303 | $text .= "$new_text\cC";
|
---|
304 | }
|
---|
305 | }
|
---|
306 | }
|
---|
307 |
|
---|
308 | if ($level eq "document") { $text_extra .= "\cB"; }
|
---|
309 | else { $text .= "\cB"; }
|
---|
310 |
|
---|
311 | $section = $doc_obj->get_next_section($section);
|
---|
312 | }
|
---|
313 |
|
---|
314 | print $handle "$text$text_extra";
|
---|
315 | }
|
---|
316 |
|
---|
317 | 1;
|
---|
318 |
|
---|