root / tmp / org.txm.tigersearch.rcp / groovy / org / txm / importer / tigersearch / tigersearchLoader.groovy @ 911
History | View | Annotate | Download (3.2 kB)
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 |
// $LastChangedDate: 2015-06-03 15:04:53 +0200 (mer., 03 juin 2015) $
|
23 |
// $LastChangedRevision: 2984 $
|
24 |
// $LastChangedBy: mdecorde $
|
25 |
//
|
26 |
package org.txm.importer.tigersearch;
|
27 |
|
28 |
import javax.xml.stream.XMLStreamReader; |
29 |
|
30 |
import org.txm.sw.RemoveTag; |
31 |
import org.txm.importer.ApplyXsl2; |
32 |
import org.txm.importer.ValidateXml; |
33 |
import org.txm.objects.*; |
34 |
import org.txm.tokenizer.TokenizerClasses; |
35 |
import org.txm.utils.*; |
36 |
import org.txm.*; |
37 |
import org.txm.scripts.teitxm.*; |
38 |
import org.txm.utils.i18n.*; |
39 |
import org.txm.metadatas.*; |
40 |
import javax.xml.stream.*; |
41 |
import org.w3c.dom.Element |
42 |
import org.txm.utils.xml.DomUtils; |
43 |
import org.txm.importer.xtz.* |
44 |
|
45 |
String userDir = System.getProperty("user.home"); |
46 |
|
47 |
def MONITOR;
|
48 |
boolean debug = org.txm.utils.logger.Log.isPrintingErrors();
|
49 |
BaseParameters params; |
50 |
try {params = paramsBinding;MONITOR=monitor} catch (Exception) |
51 |
{ println "DEV MODE";//exception means we debug |
52 |
debug = true
|
53 |
params = new BaseParameters(new File(userDir, "xml/roland/import.xml")) |
54 |
params.load() |
55 |
if (!org.txm.Toolbox.isInitialized()) {
|
56 |
|
57 |
TokenizerClasses.loadFromNode(params.getTokenizerElement(params.getCorpusElement())); |
58 |
Toolbox.setParam(Toolbox.INSTALL_DIR,new File("/usr/lib/TXM")); |
59 |
//Toolbox.setParam(Toolbox.INSTALL_DIR,new File("C:\\Program Files\\TXM"));//For Windows
|
60 |
Toolbox.setParam(Toolbox.TREETAGGER_INSTALL_PATH,new File(userDir,"treetagger")); |
61 |
//Toolbox.setParam(Toolbox.TREETAGGER_INSTALL_PATH,new File("C:\\Program Files\\treetagger"));//for Windows
|
62 |
Toolbox.setParam(Toolbox.TREETAGGER_MODELS_PATH,new File(userDir,"treetagger/models")); |
63 |
Toolbox.setParam(Toolbox.METADATA_ENCODING, "UTF-8");
|
64 |
Toolbox.setParam(Toolbox.METADATA_COLSEPARATOR, ",");
|
65 |
Toolbox.setParam(Toolbox.METADATA_TXTSEPARATOR, "\"");
|
66 |
//Toolbox.setParam(Toolbox.TREETAGGER_MODELS_PATH,new File("C:\\Program Files\\treetagger\\models"));//for Windows
|
67 |
Toolbox.setParam(Toolbox.USER_TXM_HOME, new File(System.getProperty("user.home"), "TXM")); |
68 |
} |
69 |
} |
70 |
|
71 |
if (params == null) { println "no parameters. Aborting"; return; } |
72 |
|
73 |
//params.getKeyValueParameters().put(ImportKeys.CLEAN, "false")
|
74 |
//params.getKeyValueParameters().put(ImportKeys.MULTITHREAD, "false")
|
75 |
//params.getKeyValueParameters().put(ImportKeys.DEBUG, "false")
|
76 |
//params.getKeyValueParameters().put(ImportKeys.UPDATECORPUS, "false")
|
77 |
|
78 |
TSImport i = new TSImport(params);
|
79 |
i.process(); |
80 |
readyToLoad = i.isSuccessful |