Package com.nexwave.nquindexer

Examples of com.nexwave.nquindexer.SaxHTMLIndex

191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
    // ------------------------------------------

    // Retrieve the clean-up properties for indexing
    retrieveCleanUpProps();

    SaxHTMLIndex spe = new SaxHTMLIndex(cleanUpStrings, cleanUpChars); // use clean-up props files

    System.setProperty("org.xml.sax.driver", "org.ccil.cowan.tagsoup.Parser");
    System.setProperty("javax.xml.parsers.SAXParserFactory", "org.ccil.cowan.tagsoup.jaxp.SAXFactoryImpl");

    if (spe.init(tempDico) == 0) {
      //create a html file description list
      ArrayList filesDescription = new ArrayList();

      String indexerLanguage = getProperty("webhelpIndexerLanguage");
      indexerLanguage = ((indexerLanguage == null) ? "en" : indexerLanguage);
      //TODO: change this when updating webhelpindexer in order to use the new WriteJSFiles.WriteIndex method
        if (getLog().isDebugEnabled())
          getLog().debug("Indexer language is: " + indexerLanguage);

      // parse each html files
      for (int f = 0; f < htmlFiles.size(); f++) {
        File ftemp = (File) htmlFiles.get(f);

        if (getLog().isDebugEnabled())
          getLog().debug("Parsing html file: " + ftemp.getAbsolutePath());

        //tempMap.put(key, value);
        //The HTML file information are added in the list of FileInfoObject
        DocFileInfo docFileInfoTemp = new DocFileInfo(spe.runExtractData(ftemp, indexerLanguage, stemming));

        ftemp = docFileInfoTemp.getFullpath();

        String stemp = ftemp.toString();
        int i = stemp.indexOf(targetBaseDir.getAbsolutePath());
View Full Code Here
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
    // ------------------------------------------

    // Retrieve the clean-up properties for indexing
    retrieveCleanUpProps();

    SaxHTMLIndex spe = new SaxHTMLIndex(cleanUpStrings, cleanUpChars); // use clean-up props files

    if (spe.init(tempDico) == 0) {
      //create a html file description list
      ArrayList filesDescription = new ArrayList();

      String indexerLanguage = getProperty("webhelpIndexerLanguage");
      indexerLanguage = ((indexerLanguage == null) ? "en" : indexerLanguage);
      //TODO: change this when updating webhelpindexer in order to use the new WriteJSFiles.WriteIndex method
      IndexerTask.indexerLanguage = indexerLanguage;
        if (getLog().isDebugEnabled())
          getLog().debug("Indexer language is: " + indexerLanguage);

      // parse each html files
      for (int f = 0; f < htmlFiles.size(); f++) {
        File ftemp = (File) htmlFiles.get(f);

        if (getLog().isDebugEnabled())
          getLog().debug("Parsing html file: " + ftemp.getAbsolutePath());

        //tempMap.put(key, value);
        //The HTML file information are added in the list of FileInfoObject
        DocFileInfo docFileInfoTemp = new DocFileInfo(spe.runExtractData(ftemp, indexerLanguage));

        ftemp = docFileInfoTemp.getFullpath();

        String stemp = ftemp.toString();
        int i = stemp.indexOf(targetBaseDir.getAbsolutePath());
View Full Code Here
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
    // ------------------------------------------

    // Retrieve the clean-up properties for indexing
    retrieveCleanUpProps();

    SaxHTMLIndex spe = new SaxHTMLIndex(cleanUpStrings, cleanUpChars); // use clean-up props files

    if (spe.init(tempDico) == 0) {
      //create a html file description list
      ArrayList filesDescription = new ArrayList();

      // parse each html files
      for (int f = 0; f < htmlFiles.size(); f++) {
        File ftemp = (File) htmlFiles.get(f);

        if (getLog().isDebugEnabled())
          getLog().debug("Parsing html file: " + ftemp.getAbsolutePath());

        //tempMap.put(key, value);
        //The HTML file information are added in the list of FileInfoObject
        String indexerLanguage = getProperty("webhelpIndexerLanguage");
        DocFileInfo docFileInfoTemp = new DocFileInfo(spe.runExtractData(ftemp, (indexerLanguage == null) ? "en"
            : indexerLanguage));

        ftemp = docFileInfoTemp.getFullpath();

        String stemp = ftemp.toString();
View Full Code Here
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
    // ------------------------------------------

    // Retrieve the clean-up properties for indexing
    retrieveCleanUpProps();

    SaxHTMLIndex spe = new SaxHTMLIndex(cleanUpStrings, cleanUpChars); // use clean-up props files

    if (spe.init(tempDico) == 0) {
      //create a html file description list
      ArrayList filesDescription = new ArrayList();

      // parse each html files
      for (int f = 0; f < htmlFiles.size(); f++) {
        File ftemp = (File) htmlFiles.get(f);

        if (getLog().isDebugEnabled())
          getLog().debug("Parsing html file: " + ftemp.getAbsolutePath());

        //tempMap.put(key, value);
        //The HTML file information are added in the list of FileInfoObject
        String      indexerLanguage = getProperty("webhelpIndexerLanguage");
        DocFileInfo docFileInfoTemp =
          new DocFileInfo(spe.runExtractData(ftemp,
                                             (indexerLanguage == null) ? "en" : indexerLanguage));

        ftemp = docFileInfoTemp.getFullpath();

        String stemp = ftemp.toString();
View Full Code Here
TOP

Related Classes of com.nexwave.nquindexer.SaxHTMLIndex

  • com.agilejava.docbkx.maven.AbstractWebhelpMojo
    • Copyright © 2018 www.massapicom. All rights reserved.
    All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.