|
1 |
// Copyright © 2010-2013 ENS de Lyon.
|
|
2 |
// Copyright © 2007-2010 ENS de Lyon, CNRS, INRP, University of
|
|
3 |
// Lyon 2, University of Franche-Comté, University of Nice
|
|
4 |
// Sophia Antipolis, University of Paris 3.
|
|
5 |
//
|
|
6 |
// The TXM platform is free software: you can redistribute it
|
|
7 |
// and/or modify it under the terms of the GNU General Public
|
|
8 |
// License as published by the Free Software Foundation,
|
|
9 |
// either version 2 of the License, or (at your option) any
|
|
10 |
// later version.
|
|
11 |
//
|
|
12 |
// The TXM platform is distributed in the hope that it will be
|
|
13 |
// useful, but WITHOUT ANY WARRANTY; without even the implied
|
|
14 |
// warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR
|
|
15 |
// PURPOSE. See the GNU General Public License for more
|
|
16 |
// details.
|
|
17 |
//
|
|
18 |
// You should have received a copy of the GNU General
|
|
19 |
// Public License along with the TXM platform. If not, see
|
|
20 |
// http://www.gnu.org/licenses.
|
|
21 |
//
|
|
22 |
//
|
|
23 |
//
|
|
24 |
// $LastChangedDate: 2017-04-06 09:11:32 +0200 (jeu. 06 avril 2017) $
|
|
25 |
// $LastChangedRevision: 3425 $
|
|
26 |
// $LastChangedBy: mdecorde $
|
|
27 |
//
|
|
28 |
package org.txm.tokenizer
|
|
29 |
|
|
30 |
import java.io.FileWriter;
|
|
31 |
import java.util.regex.Pattern;
|
|
32 |
|
|
33 |
import org.txm.utils.xml.DomUtils;
|
|
34 |
import org.w3c.dom.Document;
|
|
35 |
import org.w3c.dom.Element
|
|
36 |
import org.w3c.dom.Node;
|
|
37 |
import org.w3c.dom.NodeList;
|
|
38 |
|
|
39 |
import javax.xml.stream.*;
|
|
40 |
import java.net.URL;
|
|
41 |
|
|
42 |
// TODO: Auto-generated Javadoc
|
|
43 |
/**
|
|
44 |
* The Class TokenizerClasses.
|
|
45 |
*/
|
|
46 |
class TokenizerClasses {
|
|
47 |
|
|
48 |
//// ROOTS ? ////
|
|
49 |
public static boolean debug = false;
|
|
50 |
/** The tag_all. */
|
|
51 |
public static String tag_all = "<[A-Za-z][^>]+>";
|
|
52 |
|
|
53 |
/** The Weblex enclitics. */
|
|
54 |
public static String enclitics = "je|m[eo]i|tu|t[eo]i|il|lui|luy|ilz|ils|no?u?s|vo?u?s|on|ce|ci|là|elles?"; // FRO
|
|
55 |
public static String encliticsFR = "je|tu|il|elle|on|nous|vous|ils|elles|toi|moi|en|y|t|leur|lui|le|la|les"; // FR
|
|
56 |
|
|
57 |
/** The TT enclitics. */
|
|
58 |
public static String FClitic_en = "'(s|re|ve|d|m|em|ll)|n['‘’]t";
|
|
59 |
public static String PClitic_fr = '[dcjlmnstDCJLNMST][\'‘’]|[Qq]u[\'‘’]|[Jj]usqu[\'‘’]|[Ll]orsqu[\'‘’]|[Pp]uisqu[\'‘’]|[Qq]uoiqu[\'‘’]';
|
|
60 |
public static String FClitic_fr = '-t-elles?|-t-ils?|-t-on|-ce|-elles?|-ils?|-je|-la|-les?|-leur|-lui|-mêmes?|-m[\'‘’]|-moi|-nous|-on|-toi|-tu|-t[\'‘’]|-vous|-en|-y|-ci|-là';
|
|
61 |
public static String PClitic_it = '[dD][ae]ll[\'‘’]|[nN]ell[\'‘’]|[Aa]ll[\'‘’]|[lLDd][\'‘’]|[Ss]ull[\'‘’]|[Qq]uest[\'‘’]|[Uu]n[\'‘’]|[Ss]enz[\'‘’]|[Tt]utt[\'‘’]';
|
|
62 |
public static String FClitic_gl = '-la|-las|-lo|-los|-nos';
|
|
63 |
|
|
64 |
/** The div_tags. */
|
|
65 |
public static String div_tags = "TEI|text|front|body|div|div1|div2|div3|div4|div5|div6|back|head|trailer|p|ab|sp|speaker|list|notice|bibl|opener|dateline";
|
|
66 |
|
|
67 |
/** The q_tags. */
|
|
68 |
public static String q_tags = "q|quote|said|item|stage|cit|label|heraldry";
|
|
69 |
|
|
70 |
/** The extraword_tags. */
|
|
71 |
public static String extraword1_tags = "expan|pb|lb|milestone|gap|note|s|locus|title|ref|hi|witDetail";
|
|
72 |
|
|
73 |
/** The corr_tags_no_seg. */
|
|
74 |
public static String corr_tags_no_seg = "expan|unclear|choice|corr|sic|reg|orig|foreign|hi|title|name|supplied|subst|add|del|damage|date|idno|surplus";
|
|
75 |
|
|
76 |
/** The word_tags. */
|
|
77 |
public static String word_tags = "w";
|
|
78 |
|
|
79 |
/** The intraword_tags. */
|
|
80 |
public static String intraword_tags = "c|ex|caesura";
|
|
81 |
|
|
82 |
/** The punct_quotes. */
|
|
83 |
public static String punct_quotes = "'‘’’"
|
|
84 |
|
|
85 |
/** The punct_strong1. */
|
|
86 |
public static String punct_strong1 = ".!?";
|
|
87 |
|
|
88 |
/** The punct_strong2. */
|
|
89 |
public static String punct_strong2 = "\\.\\.|\\.\\.\\.|…|\\|";
|
|
90 |
|
|
91 |
/** The punct_paren_open1. */
|
|
92 |
public static String punct_paren_open1 = "«";
|
|
93 |
|
|
94 |
/** The punct_paren_open2. */
|
|
95 |
public static String punct_paren_open2 = "``|\\(|\\[|\\{";
|
|
96 |
|
|
97 |
/** The punct_paren_close1. */
|
|
98 |
public static String punct_paren_close1 = "»";
|
|
99 |
|
|
100 |
/** The punct_paren_close2. */
|
|
101 |
public static String punct_paren_close2 = "''|‘‘|’’|\\)|\\]|\\}";
|
|
102 |
|
|
103 |
/** The punct_weak. */
|
|
104 |
public static String punct_weak = "\\-,;∼ˆ·*:\"“”\\+±=/\\≤≥<>\\—ν√μ•@≈→αγ∞≡σ&%|#¼½Θĩ†φθΓ§ẽə∈";
|
|
105 |
|
|
106 |
public static String entity = "&[^;]+;";
|
|
107 |
|
|
108 |
public static String seg_tags = "seg"
|
|
109 |
|
|
110 |
//// COMBINAISONS ////
|
|
111 |
|
|
112 |
/** The corr_tags. */
|
|
113 |
public static String corr_tags = "$corr_tags_no_seg|$seg_tags";
|
|
114 |
/** The extraword_tags. */
|
|
115 |
public static String extraword_tags = "$div_tags|$q_tags|$extraword1_tags";
|
|
116 |
|
|
117 |
/** The punct_strong. */
|
|
118 |
public static String punct_strong = "[$punct_strong1]+|$punct_strong2";
|
|
119 |
|
|
120 |
/** The punct_paren_open. */
|
|
121 |
public static String punct_paren_open = "$punct_paren_open1|$punct_paren_open2";
|
|
122 |
|
|
123 |
/** The punct_paren_close. */
|
|
124 |
public static String punct_paren_close = "$punct_paren_close1|$punct_paren_close2";
|
|
125 |
|
|
126 |
/** The punct_paren. */
|
|
127 |
public static String punct_paren = "$punct_paren_open|$punct_paren_close";
|
|
128 |
|
|
129 |
/** The punct_all. */
|
|
130 |
public static String punct_all = "$punct_strong|$punct_paren|[$punct_weak]";
|
|
131 |
|
|
132 |
/** The word_chars. */
|
|
133 |
public static String word_chars = "[^ $punct_quotes$punct_strong1$punct_paren_open1$punct_paren_close1$punct_weak]|$entity";
|
|
134 |
|
|
135 |
public static def tests = [
|
|
136 |
new TTest(/\A(.*-)(je|m[eo]i|tu|t[eo]i|lui|luy|ilz|ils?|no?u?s|vo?u?s|on|leurs|ce|ci|là|elles?)(.*)\Z/),
|
|
137 |
new TTest(/\A(.*)(-)()\Z/, "pon"),
|
|
138 |
//new TTest(/\A(.*)(Bande dessinée|eau de vie|machine à écrire|Moyen Âge|petit pois|poule d'eau|Avoir l'air|se rendre compte|faire semblant|prendre froid|s'en aller|Comme il faut|bon marché|bon enfant|en retard|en colère|à la mode|de bonne humeur|hors de propos|de travers|en vacances|de standing|à l'abandon|sans défense|pieds nus|Tout à fait|d'ores et déjà|de temps en temps|tout de suite|de gré ou de force|ne pas|ne jamais|ne plus|sans doute|on ne peut plus|Quelque chose|ce dernier|n'importe quoi|N'importe quel|beaucoup de|plus d'un|peu de|un tas de|un groupe de|je ne sais quel|De sorte que|bien que|parce que|c'est pourquoi|de même que|depuis que|Quant à|au-dessus de|par rapport à|à côté de|grâce à|hors de|le long de|Par ma barbe !|Mon dieu !|Eh bien !|Bonne nuit !|Pas de panique !)(.*)\Z/),
|
|
139 |
new TTest(/\A([Aa]ujourd'hui)(.*)/,0,1,2),
|
|
140 |
|
|
141 |
// BFM ONLY
|
|
142 |
// new TTest(/\A()([.·][^ .·]+[.·])(.*)\Z/, "num"),
|
|
143 |
|
|
144 |
// TXM REFMAN ONLY
|
|
145 |
// new TTest(/\A(.*)($tag_all)(.*)\Z/, "tag"),
|
|
146 |
// new TTest(/\A(.*)([0-9][0-9][\/][0-9][0-9][\/][0-9][0-9])(.*)\Z/, "date", 1,2,3),
|
|
147 |
// new TTest(/\A(.*)([0-9][0-9]:[0-9][0-9]:[0-9][0-9])(.*)\Z/, "time", 1,2,3),
|
|
148 |
// new TTest(/\A([^0-9]*)([0-9]+(\.[0-9]+[a-z]+))(.*)\Z/, "version", 1,2,4),
|
|
149 |
// new TTest(/\A([^0-9]*)([0-9]+(\.[0-9]+)(\.[0-9]+)+)(\.[0-9]+[a-z]+)?(.*)\Z/, "version", 1,2,6),
|
|
150 |
// new TTest(/\A()([a-z]++:\/\/)(.*)\Z/, "protocol", 1,2,3),
|
|
151 |
// new TTest(/\A()([0-9]+([.,][0-9]+)?)()\Z/, "num", 1,2,4),
|
|
152 |
// new TTest(/\A()(([A-Z]:)?([\/\\][^\/\\]+)+)()\Z/, "path", 1,2,5),
|
|
153 |
|
|
154 |
new TTest(/\A([\p{Ps}|\p{Pe}|\p{Pi}|\p{Pf}|\p{Po}|\p{S}]*)(\p{L}++['‘’])(.*)\Z/, "w", 1, 2, 3),
|
|
155 |
new TTest(/\A(.*)(\.\.\.)(.*)\Z/, "pon"),
|
|
156 |
new TTest(/\A(.*)(\p{Ps}|\p{Pe}|\p{Pi}|\p{Pf}|\p{Po}|\p{S})(.*)\Z/, "pon")
|
|
157 |
|
|
158 |
];
|
|
159 |
//tag_all = "<[^>]+>";
|
|
160 |
//enclitics = "je|m[eo]i|tu|t[eo]i|il|lui|luy|ilz|ils|no?u?s|vo?u?s|on|ce|ci|là|elles?"; // FRO
|
|
161 |
//encliticsFR = "je|tu|il|elle|on|nous|vous|ils|elles|toi|moi|en|y|t|leur|lui|le|la|les"; // FR
|
|
162 |
//div_tags = "TEI|text|front|body|div|div1|div2|div3|div4|div5|div6|back|head|trailer|p|ab|sp|speaker|list|notice|bibl";
|
|
163 |
//q_tags = "q|quote|item|stage";
|
|
164 |
//extraword_tags = "$div_tags|$q_tags|expan|pb|lb|milestone|gap|note|s|locus|title|ref|hi|witDetail";
|
|
165 |
//corr_tags_no_seg = "expan|unclear|choice|corr|sic|reg|orig|foreign|hi|title|name|supplied|subst|add|del|damage|date|idno";
|
|
166 |
//corr_tags = "$corr_tags_no_seg|seg";
|
|
167 |
//word_tags = "w|abbr|num";
|
|
168 |
//intraword_tags = "c|ex";
|
|
169 |
|
|
170 |
public static String whitespaces = "[\\p{Z}\\p{C}]+";
|
|
171 |
public static String regElision = "['‘’]" // obsolete
|
|
172 |
public static String regPunct = "[\\p{Ps}\\p{Pe}\\p{Pi}\\p{Pf}\\p{Po}\\p{S}]"
|
|
173 |
|
|
174 |
//punct_strong1 = ".!?";
|
|
175 |
//punct_strong2 = "\\.\\.|\\.\\.\\.|…|\\|";
|
|
176 |
//punct_strong = "[$punct_strong1]|$punct_strong2";
|
|
177 |
//punct_paren_open1 = "«";
|
|
178 |
//punct_paren_open2 = "<<|``|\\(|\\[|\\{";
|
|
179 |
//punct_paren_open = "$punct_paren_open1|$punct_paren_open2";
|
|
180 |
//punct_paren_close1 = "»";
|
|
181 |
//punct_paren_close2 = ">>|''|‘‘|’’|\\)|\\]|\\}";
|
|
182 |
//punct_paren_close = "$punct_paren_close1|$punct_paren_close2";
|
|
183 |
//punct_paren = "$punct_paren_open|$punct_paren_close";
|
|
184 |
//punct_weak = ",;∼ˆ·*:\"`'“”\\+±=/\\-≤≥<>\\—_ν√μ•@≈→αγ∞≡σ&¼½Θĩ†φθΓ§ẽə∈";
|
|
185 |
//punct_all = "$punct_strong|$punct_paren|[$punct_weak]";
|
|
186 |
//word_chars = "[^ $punct_strong1$punct_paren_open1$punct_paren_close1$punct_weak]|&[^;]+;";
|
|
187 |
|
|
188 |
public static reset() {
|
|
189 |
tag_all = "<[^>]+>";
|
|
190 |
enclitics = "je|m[eo]i|tu|t[eo]i|il|lui|luy|ilz|ils|no?u?s|vo?u?s|on|ce|ci|là|elles?"; // FRO
|
|
191 |
encliticsFR = "je|tu|il|elle|on|nous|vous|ils|elles|toi|moi|en|y|t|leur|lui|le|la|les"; // FR
|
|
192 |
div_tags = "TEI|text|front|body|div|div1|div2|div3|div4|div5|div6|back|head|trailer|p|ab|sp|speaker|list|notice|bibl";
|
|
193 |
q_tags = "q|quote|item|stage|cit";
|
|
194 |
extraword1_tags = "expan|pb|lb|milestone|gap|note|s|locus|title|ref|hi|witDetail";
|
|
195 |
corr_tags_no_seg = "expan|unclear|choice|corr|sic|reg|orig|foreign|hi|title|name|supplied|subst|add|del|damage|date|idno|surplus";
|
|
196 |
word_tags = "w";
|
|
197 |
intraword_tags = "c|ex|caesura";
|
|
198 |
punct_quotes = "'‘’’"
|
|
199 |
punct_strong1 = ".!?";
|
|
200 |
punct_strong2 = "\\.\\.|\\.\\.\\.|…|\\|";
|
|
201 |
punct_paren_open1 = "«";
|
|
202 |
punct_paren_open2 = "``|\\(|\\[|\\{";
|
|
203 |
punct_paren_close1 = "»";
|
|
204 |
punct_paren_close2 = "''|‘‘|’’|\\)|\\]|\\}";
|
|
205 |
punct_weak = "\\-,;∼ˆ·*:\"“”\\+±=/\\-≤≥<>\\—ν√μ•@≈→αγ∞≡σ&¼½Θĩ†φθΓ§ẽə∈";
|
|
206 |
entity = "&[^;]+;";
|
|
207 |
seg_tags = "seg"
|
|
208 |
|
|
209 |
whitespaces = "[\\p{Z}\\p{C}]+";
|
|
210 |
regElision = "['‘’]"
|
|
211 |
regPunct = "[\\p{Ps}\\p{Pe}\\p{Pi}\\p{Pf}\\p{Po}\\p{S}]"
|
|
212 |
|
|
213 |
recombine();
|
|
214 |
}
|
|
215 |
|
|
216 |
public static recombine() {
|
|
217 |
corr_tags = "$corr_tags_no_seg|$seg_tags";
|
|
218 |
extraword_tags = "$div_tags|$q_tags|$extraword1_tags";
|
|
219 |
punct_strong = "[$punct_strong1]|$punct_strong2";
|
|
220 |
punct_paren_open = "$punct_paren_open1|$punct_paren_open2";
|
|
221 |
punct_paren_close = "$punct_paren_close1|$punct_paren_close2";
|
|
222 |
punct_paren = "$punct_paren_open|$punct_paren_close";
|
|
223 |
punct_all = "$punct_strong|$punct_paren|[$punct_weak]";
|
|
224 |
word_chars = "[^ $punct_quotes$punct_strong1$punct_paren_open1$punct_paren_close1$punct_weak]|$entity";
|
|
225 |
}
|
|
226 |
|
|
227 |
/**
|
|
228 |
* To dom.
|
|
229 |
*
|
|
230 |
* @param doc the doc
|
|
231 |
* @param parent the parent
|
|
232 |
* @return the java.lang. object
|
|
233 |
*/
|
|
234 |
public static toDom(Document doc, Node parent) {
|
|
235 |
Element tokenizer = doc.createElement("tokenizer");
|
|
236 |
tokenizer.setAttribute("onlyThoseTests", "false")
|
|
237 |
parent.appendChild(tokenizer)
|
|
238 |
|
|
239 |
//String tag_all = "<[^>]+>";
|
|
240 |
Element p = doc.createElement("param");
|
|
241 |
p.setAttribute("key","tag_all");
|
|
242 |
p.setTextContent(tag_all);
|
|
243 |
tokenizer.appendChild(p)
|
|
244 |
//String enclitics = "je|m[eo]i|tu|t[eo]i|il|lui|luy|ilz|ils|no?u?s|vo?u?s|on|ce|ci|là|elles?"; // FRO
|
|
245 |
p = doc.createElement("param");
|
|
246 |
p.setAttribute("key","enclitics");
|
|
247 |
p.setTextContent(enclitics);
|
|
248 |
tokenizer.appendChild(p)
|
|
249 |
//String encliticsFR = "je|tu|il|elle|on|nous|vous|ils|elles|toi|moi|en|y|t|leur|lui|le|la|les"; // FR
|
|
250 |
p = doc.createElement("param");
|
|
251 |
p.setAttribute("key","encliticsFR");
|
|
252 |
p.setTextContent(encliticsFR);
|
|
253 |
tokenizer.appendChild(p)
|
|
254 |
//String div_tags = "TEI|text|front|body|div|div1|div2|div3|div4|div5|div6|back|head|trailer|p|ab|sp|speaker|list|notice|bibl";
|
|
255 |
p = doc.createElement("param");
|
|
256 |
p.setAttribute("key","div_tags");
|
|
257 |
p.setTextContent(div_tags);
|
|
258 |
tokenizer.appendChild(p)
|
|
259 |
//String q_tags = "q|quote|item|stage";
|
|
260 |
p = doc.createElement("param");
|
|
261 |
p.setAttribute("key","q_tags");
|
|
262 |
p.setTextContent(q_tags);
|
|
263 |
tokenizer.appendChild(p)
|
|
264 |
//String extraword_tags = "$div_tags|$q_tags|expan|pb|lb|milestone|gap|note|s|locus|title|ref|hi|witDetail";
|
|
265 |
p = doc.createElement("param");
|
|
266 |
p.setAttribute("key","extraword_tags");
|
|
267 |
p.setTextContent(extraword_tags);
|
|
268 |
tokenizer.appendChild(p)
|
|
269 |
//String corr_tags_no_seg = "expan|unclear|choice|corr|sic|reg|orig|foreign|hi|title|name|supplied|subst|add|del|damage|date|idno";
|
|
270 |
p = doc.createElement("param");
|
|
271 |
p.setAttribute("key","enclitics");
|
|
272 |
p.setTextContent(enclitics);
|
|
273 |
tokenizer.appendChild(p)
|
|
274 |
//String corr_tags = "$corr_tags_no_seg|seg";
|
|
275 |
p = doc.createElement("param");
|
|
276 |
p.setAttribute("key","corr_tags");
|
|
277 |
p.setTextContent(corr_tags);
|
|
278 |
tokenizer.appendChild(p)
|
|
279 |
//String word_tags = "w|abbr|num";
|
|
280 |
p = doc.createElement("param");
|
|
281 |
p.setAttribute("key","word_tags");
|
|
282 |
p.setTextContent(word_tags);
|
|
283 |
tokenizer.appendChild(p)
|
|
284 |
//String intraword_tags = "c|ex";
|
|
285 |
p = doc.createElement("param");
|
|
286 |
p.setAttribute("key","intraword_tags");
|
|
287 |
p.setTextContent(intraword_tags);
|
|
288 |
tokenizer.appendChild(p)
|
|
289 |
|
|
290 |
//String punct_strong1 = ".!?";
|
|
291 |
p = doc.createElement("param");
|
|
292 |
p.setAttribute("key","punct_strong1");
|
|
293 |
p.setTextContent(punct_strong1);
|
|
294 |
tokenizer.appendChild(p)
|
|
295 |
//String punct_strong2 = "\\.\\.|\\.\\.\\.|…|\\|";
|
|
296 |
p = doc.createElement("param");
|
|
297 |
p.setAttribute("key","punct_strong2");
|
|
298 |
p.setTextContent(punct_strong2);
|
|
299 |
tokenizer.appendChild(p)
|
|
300 |
//String punct_strong = "[$punct_strong1]|$punct_strong2";
|
|
301 |
p = doc.createElement("param");
|
|
302 |
p.setAttribute("key","punct_strong");
|
|
303 |
p.setTextContent(punct_strong);
|
|
304 |
tokenizer.appendChild(p)
|
|
305 |
//String punct_paren_open1 = "«";
|
|
306 |
p = doc.createElement("param");
|
|
307 |
p.setAttribute("key","punct_paren_open1");
|
|
308 |
p.setTextContent(punct_paren_open1);
|
|
309 |
tokenizer.appendChild(p)
|
|
310 |
//String punct_paren_open2 = "<<|``|\\(|\\[|\\{";
|
|
311 |
p = doc.createElement("param");
|
|
312 |
p.setAttribute("key","punct_paren_open2");
|
|
313 |
p.setTextContent(punct_paren_open2);
|
|
314 |
tokenizer.appendChild(p)
|
|
315 |
//String punct_paren_open = "$punct_paren_open1|$punct_paren_open2";
|
|
316 |
p = doc.createElement("param");
|
|
317 |
p.setAttribute("key","punct_paren_open");
|
|
318 |
p.setTextContent(punct_paren_open);
|
|
319 |
tokenizer.appendChild(p)
|
|
320 |
//String punct_paren_close1 = "»";
|
|
321 |
p = doc.createElement("param");
|
|
322 |
p.setAttribute("key","punct_paren_close1");
|
|
323 |
p.setTextContent(punct_paren_close1);
|
|
324 |
tokenizer.appendChild(p)
|
|
325 |
//String punct_paren_close2 = ">>|''|‘‘|’’|\\)|\\]|\\}";
|
|
326 |
p = doc.createElement("param");
|
|
327 |
p.setAttribute("key","punct_paren_close2");
|
|
328 |
p.setTextContent(punct_paren_close2);
|
|
329 |
tokenizer.appendChild(p)
|
|
330 |
//String punct_paren_close = "$punct_paren_close1|$punct_paren_close2";
|
|
331 |
p = doc.createElement("param");
|
|
332 |
p.setAttribute("key","punct_paren_close");
|
|
333 |
p.setTextContent(punct_paren_close);
|
|
334 |
tokenizer.appendChild(p)
|
|
335 |
//String punct_paren = "$punct_paren_open|$punct_paren_close";
|
|
336 |
p = doc.createElement("param");
|
|
337 |
p.setAttribute("key","punct_paren");
|
|
338 |
p.setTextContent(punct_paren);
|
|
339 |
tokenizer.appendChild(p)
|
|
340 |
//String punct_weak = ",;∼ˆ·*:\"`'“”\\+±=/\\-≤≥<>\\—_ν√μ•@≈→αγ∞≡σ&¼½Θĩ†φθΓ§ẽə∈";
|
|
341 |
p = doc.createElement("param");
|
|
342 |
p.setAttribute("key","punct_weak");
|
|
343 |
p.setTextContent(punct_weak);
|
|
344 |
tokenizer.appendChild(p)
|
|
345 |
//String punct_all = "$punct_strong|$punct_paren|[$punct_weak]";
|
|
346 |
p = doc.createElement("param");
|
|
347 |
p.setAttribute("key","punct_all");
|
|
348 |
p.setTextContent(punct_all);
|
|
349 |
tokenizer.appendChild(p)
|
|
350 |
//String word_chars = "[^ $punct_strong1$punct_paren_open1$punct_paren_close1$punct_weak]|&[^;]+;";
|
|
351 |
p = doc.createElement("param");
|
|
352 |
p.setAttribute("key","word_chars");
|
|
353 |
p.setTextContent(word_chars);
|
|
354 |
tokenizer.appendChild(p)
|
|
355 |
//whitespaces = "[\\p{Z}\\p{C}]+";
|
|
356 |
p = doc.createElement("param");
|
|
357 |
p.setAttribute("key","whitespaces");
|
|
358 |
p.setTextContent(whitespaces);
|
|
359 |
tokenizer.appendChild(p)
|
|
360 |
//regElision = "['‘’]"
|
|
361 |
p = doc.createElement("param");
|
|
362 |
p.setAttribute("key","regElision");
|
|
363 |
p.setTextContent(regElision);
|
|
364 |
tokenizer.appendChild(p)
|
|
365 |
//regPunct = "[\\p{Ps}\\p{Pe}\\p{Pi}\\p{Pf}\\p{Po}\\p{S}]"
|
|
366 |
p = doc.createElement("param");
|
|
367 |
p.setAttribute("key","regPunct");
|
|
368 |
p.setTextContent(regPunct);
|
|
369 |
tokenizer.appendChild(p)
|
|
370 |
}
|
|
371 |
|
|
372 |
/**
|
|
373 |
* Dump.
|
|
374 |
*/
|
|
375 |
public static void dump() {
|
|
376 |
println "BRUT"
|
|
377 |
println "tag_all = $tag_all"
|
|
378 |
println "enclitics = $enclitics"
|
|
379 |
println "encliticsFR = $encliticsFR"
|
|
380 |
println "div_tags = $div_tags"
|
|
381 |
println "q_tags = $q_tags"
|
|
382 |
println "extraword1_tags = $extraword1_tags"
|
|
383 |
println "corr_tags_no_seg = $corr_tags_no_seg"
|
|
384 |
println "word_tags = $word_tags"
|
|
385 |
println "intraword_tags = $intraword_tags"
|
|
386 |
println "punct_quotes = $punct_quotes"
|
|
387 |
println "punct_strong1 = $punct_strong1"
|
|
388 |
println "punct_strong2 = $punct_strong2"
|
|
389 |
println "punct_paren_open1 = $punct_paren_open1"
|
|
390 |
println "punct_paren_open2 = $punct_paren_open2"
|
|
391 |
println "punct_paren_close1 = $punct_paren_close1"
|
|
392 |
println "punct_paren_close2 = $punct_paren_close2"
|
|
393 |
println "punct_weak = $punct_weak"
|
|
394 |
println "entity = $entity"
|
|
395 |
println "seg_tags = $seg_tags"
|
|
396 |
println "COMBINED"
|
|
397 |
println "corr_tags = $corr_tags"
|
|
398 |
println "extraword_tags = $extraword_tags"
|
|
399 |
println "punct_strong = $punct_strong"
|
|
400 |
println "punct_paren_open = $punct_paren_open"
|
|
401 |
println "punct_paren_close = $punct_paren_close"
|
|
402 |
println "punct_paren = $punct_paren"
|
|
403 |
println "punct_all = $punct_all"
|
|
404 |
println "word_chars = $word_chars"
|
|
405 |
println "whitespaces = $whitespaces"
|
|
406 |
println "regElision = $regElision"
|
|
407 |
println "regPunct = $regPunct"
|
|
408 |
println "TESTS"
|
|
409 |
for(def test : tests)
|
|
410 |
println " $test"
|
|
411 |
}
|
|
412 |
|
|
413 |
public static boolean isValid() {
|
|
414 |
try { Pattern.compile(tag_all) }
|
|
415 |
catch (Exception e) { println "tag_all=${tag_all}: "+e;return false; }
|
|
416 |
|
|
417 |
try { Pattern.compile(enclitics) }
|
|
418 |
catch (Exception e) { println "enclitics=${enclitics}: "+e;return false; }
|
|
419 |
|
|
420 |
try { Pattern.compile(encliticsFR) }
|
|
421 |
catch (Exception e) { println "encliticsFR=${encliticsFR}: "+e;return false; }
|
|
422 |
|
|
423 |
try { Pattern.compile(div_tags) }
|
|
424 |
catch (Exception e) { println "div_tags=${div_tags}: "+e;return false; }
|
|
425 |
|
|
426 |
try { Pattern.compile(q_tags) }
|
|
427 |
catch (Exception e) { println "q_tags=${q_tags}: "+e;return false; }
|
|
428 |
|
|
429 |
try { Pattern.compile(extraword1_tags) }
|
|
430 |
catch (Exception e) { println "extraword1_tags=${extraword1_tags}: "+e;return false; }
|
|
431 |
|
|
432 |
try { Pattern.compile(corr_tags_no_seg) }
|
|
433 |
catch (Exception e) { println "corr_tags_no_seg=${corr_tags_no_seg}: "+e;return false; }
|
|
434 |
|
|
435 |
try { Pattern.compile(word_tags) }
|
|
436 |
catch (Exception e) { println "word_tags=${word_tags}: "+e;return false; }
|
|
437 |
|
|
438 |
try { Pattern.compile(intraword_tags) }
|
|
439 |
catch (Exception e) { println "intraword_tags=${intraword_tags}: "+e;return false; }
|
|
440 |
|
|
441 |
try { Pattern.compile(punct_quotes) }
|
|
442 |
catch (Exception e) { println "punct_quotes=${punct_quotes}: "+e;return false; }
|
|
443 |
|
|
444 |
try { Pattern.compile(punct_strong1) }
|
|
445 |
catch (Exception e) { println "punct_strong1=${punct_strong1}: "+e;return false; }
|
|
446 |
|
|
447 |
try { Pattern.compile(punct_strong2) }
|
|
448 |
catch (Exception e) { println "punct_strong2=${punct_strong2}: "+e;return false; }
|
|
449 |
|
|
450 |
try { Pattern.compile(punct_paren_open1) }
|
|
451 |
catch (Exception e) { println "punct_paren_open1=${punct_paren_open1}: "+e;return false; }
|
|
452 |
|
|
453 |
try { Pattern.compile(punct_paren_open2) }
|
|
454 |
catch (Exception e) { println "punct_paren_open2=${punct_paren_open2}: "+e;return false; }
|
|
455 |
|
|
456 |
try { Pattern.compile(punct_paren_close1) }
|
|
457 |
catch (Exception e) { println "punct_paren_close1=${punct_paren_close1}: "+e;return false; }
|
|
458 |
|
|
459 |
try { Pattern.compile(punct_paren_close2) }
|
|
460 |
catch (Exception e) { println "punct_paren_close2=${punct_paren_close2}: "+e;return false; }
|
|
461 |
|
|
462 |
try { Pattern.compile(punct_weak) }
|
|
463 |
catch (Exception e) { println "punct_weak=${punct_weak}: "+e;return false; }
|
|
464 |
|
|
465 |
try { Pattern.compile(entity) }
|
|
466 |
catch (Exception e) { println "entity=${entity}: "+e;return false; }
|
|
467 |
|
|
468 |
try { Pattern.compile(seg_tags) }
|
|
469 |
catch (Exception e) { println "seg_tags=${seg_tags}: "+e;return false; }
|
|
470 |
|
|
471 |
try { Pattern.compile(corr_tags) }
|
|
472 |
catch (Exception e) { println "corr_tags: "+e;return false; }
|
|
473 |
|
|
474 |
try { Pattern.compile(extraword_tags) }
|
|
475 |
catch (Exception e) { println "extraword_tags: "+e;return false; }
|
|
476 |
|
|
477 |
if (punct_strong != null)
|
|
478 |
try { Pattern.compile(punct_strong) }
|
|
479 |
catch (Exception e) { println "punct_strong: "+e;return false; }
|
|
480 |
|
|
481 |
try { Pattern.compile(punct_paren_open) }
|
|
482 |
catch (Exception e) { println "punct_paren_open: "+e;return false; }
|
|
483 |
|
|
484 |
try { Pattern.compile(punct_paren_close) }
|
|
485 |
catch (Exception e) { println "punct_paren_close: "+e;return false; }
|
|
486 |
|
|
487 |
try { Pattern.compile(punct_paren) }
|
|
488 |
catch (Exception e) { println "punct_paren: "+e;return false; }
|
|
489 |
|
|
490 |
try { Pattern.compile(punct_all) }
|
|
491 |
catch (Exception e) { println "punct_all: "+e;return false; }
|
|
492 |
|
|
493 |
try { Pattern.compile(word_chars) }
|
|
494 |
catch (Exception e) { println "word_chars: "+e;return false; }
|
|
495 |
|
|
496 |
if (regPunct != null)
|
|
497 |
try { Pattern.compile(regPunct) }
|
|
498 |
catch (Exception e) { println "regPunct=${regPunct}: "+e;return false; }
|
|
499 |
|
|
500 |
if (regElision != null)
|
|
501 |
try { Pattern.compile(regElision) }
|
|
502 |
catch (Exception e) { println "regElision=${regElision}: "+e;return false; }
|
|
503 |
|
|
504 |
if (whitespaces != null)
|
|
505 |
try { Pattern.compile(whitespaces) }
|
|
506 |
catch (Exception e) { println "whitespaces=${whitespaces}: "+e;return false; }
|
|
507 |
|
|
508 |
for(TTest test : tests) {
|
|
509 |
try { Pattern.compile(test.regex) }
|
|
510 |
catch (Exception e) { println test.regex+": "+e;return false; }
|
|
511 |
}
|
|
512 |
return true;
|
|
513 |
}
|
|
514 |
|
|
515 |
public static boolean loadFromNode(Element tokenizerElement) {
|
|
516 |
//load params
|
|
517 |
String tmp_strong_punct;
|
|
518 |
NodeList params = tokenizerElement.getElementsByTagName("param");
|
|
519 |
for (int i = 0 ; i < params.getLength() ; i++) {
|
|
520 |
Element param = params.item(i);
|
|
521 |
String key = param.getAttribute("name")
|
|
522 |
String value = param.getAttribute("value")
|
|
523 |
if (value == null || value.length() == 0) value = param.getTextContent()
|
|
524 |
if (value.length() == 0) value = null;
|
|
525 |
|
|
526 |
if (debug) println (" Tokenizer parametrized with $key=$value.")
|
|
527 |
if (key == "tag_all")
|
|
528 |
tag_all = value
|
|
529 |
else if (key == "enclitics")
|
|
530 |
enclitics = value
|
|
531 |
else if (key == "encliticsFR")
|
|
532 |
encliticsFR = value
|
|
533 |
else if (key == "div_tags")
|
|
534 |
div_tags = value
|
|
535 |
else if (key == "q_tags")
|
|
536 |
q_tags = value
|
|
537 |
else if (key == "extraword1_tags")
|
|
538 |
extraword1_tags = value
|
|
539 |
else if (key == "corr_tags_no_seg")
|
|
540 |
corr_tags_no_seg = value
|
|
541 |
else if (key == "word_tags")
|
|
542 |
word_tags = value
|
|
543 |
else if (key == "intraword_tags")
|
|
544 |
intraword_tags = value
|
|
545 |
else if (key == "punct_quotes")
|
|
546 |
punct_quotes = value
|
|
547 |
else if (key == "punct_strong1")
|
|
548 |
punct_strong1 = value
|
|
549 |
else if (key == "punct_strong2")
|
|
550 |
punct_strong2 = value
|
|
551 |
else if (key == "punct_paren_open1")
|
|
552 |
punct_paren_open1 = value
|
|
553 |
else if (key == "punct_paren_open2")
|
|
554 |
punct_paren_open2 = value
|
|
555 |
else if (key == "punct_paren_close1")
|
|
556 |
punct_paren_close1 = value
|
|
557 |
else if (key == "punct_paren_close2")
|
|
558 |
punct_paren_close2 = value
|
|
559 |
else if (key == "punct_weak")
|
|
560 |
punct_weak = value
|
|
561 |
else if (key == "entity")
|
|
562 |
entity = value
|
|
563 |
else if (key == "seg_tags")
|
|
564 |
seg_tags = value
|
|
565 |
else if (key == "regPunct")
|
|
566 |
regPunct = value
|
|
567 |
else if (key == "regElision")
|
|
568 |
regElision = value
|
|
569 |
else if (key == "whitespaces")
|
|
570 |
whitespaces = value
|
|
571 |
else if (key == "punct_strong") // this is temporary
|
|
572 |
tmp_strong_punct = value // this is temporary
|
|
573 |
else
|
|
574 |
println "MISSING TOKENIZER KEY: "+key
|
|
575 |
}
|
|
576 |
//recombine
|
|
577 |
recombine();
|
|
578 |
if (tmp_strong_punct != null) punct_strong = tmp_strong_punct; // this is temporary
|
|
579 |
|
|
580 |
String shouldResetTests = tokenizerElement.getAttribute("onlyThoseTests")
|
|
581 |
if ("true".equals(shouldResetTests)) {
|
|
582 |
println "Warning: tokenizer only using import parameters tests"
|
|
583 |
tests = []
|
|
584 |
}
|
|
585 |
|
|
586 |
NodeList testsList = tokenizerElement.getElementsByTagName("test");
|
|
587 |
if (testsList.getLength() > 0) {
|
|
588 |
//println "Add "+testsList.getLength()+" tests to the tokenizer"
|
|
589 |
for (int i = 0 ; i < testsList.getLength() ; i++) {
|
|
590 |
Element test = testsList.item(i);
|
|
591 |
tests.add(i, new TTest(test.getTextContent(), test.getAttribute("type"),
|
|
592 |
Integer.parseInt(test.getAttribute("before")),
|
|
593 |
Integer.parseInt(test.getAttribute("hit")),
|
|
594 |
Integer.parseInt(test.getAttribute("after")))
|
|
595 |
)
|
|
596 |
}
|
|
597 |
//println "Tests: "+tests
|
|
598 |
}
|
|
599 |
|
|
600 |
if (isValid())
|
|
601 |
return true
|
|
602 |
else {
|
|
603 |
reset();
|
|
604 |
recombine();
|
|
605 |
return false;
|
|
606 |
}
|
|
607 |
}
|
|
608 |
|
|
609 |
/**
|
|
610 |
* The main method.
|
|
611 |
*
|
|
612 |
* @param args the arguments
|
|
613 |
*/
|
|
614 |
public static void main(String[] args) {
|
|
615 |
|
|
616 |
Document doc = DomUtils.load(new File("/home/mdecorde/xml/discours/import.xml"));
|
|
617 |
Node corpora = doc.getDocumentElement().getElementsByTagName("corpora").item(0);
|
|
618 |
Node corpus = ((Element)corpora).getElementsByTagName("corpus").item(0)
|
|
619 |
Node tokenizer = ((Element)corpus).getElementsByTagName("tokenizer").item(0)
|
|
620 |
println "OK? "+ TokenizerClasses.loadFromNode(tokenizer)
|
|
621 |
}
|
|
622 |
}
|