Package de.anomic.data.ymark

Examples of de.anomic.data.ymark.YMarkMetadata

205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
        } catch (RowSpaceExceededException e) {
          Log.logException(e);
        }
          } else if (isWordCount || isMetadata || isURLdb || isCrawlStart) {
            try {
                  final YMarkMetadata meta = new YMarkMetadata(new DigestURI(post.get(ROOT).substring(2)), sb.indexSegments);
              meta.loadDocument(sb.loader);
              if(isWordCount)  {
                final TreeMap<String,Word> words = meta.getWordCounts();
              final ArrayList<String> topwords = new ArrayList<String>(words.descendingKeySet());
              for(int i = 0; i < 20 && i < topwords.size(); i++) {
                String word = topwords.get(i);
                int occur = words.get(word).occurrences();
                prop.put("folders_"+count+"_foldername","<small><b>"+word+":</b> [" + occur + "]</small>");
                  putProp(count, "meta");
                  count++;
              }
              count--;
              prop.put("folders_"+count+"_comma", "");
              count++;
                  prop.put("folders", count);
              } else if(isMetadata) {
                count = putMeta(count, meta.loadMetadata());
              } else if(isURLdb) {
            count = putMeta(count, meta.getMetadata());
              } else if(isCrawlStart) {
                Log.logInfo("YMark", "I am looking for CrawlStart: "+post.get(ROOT).substring(2));
                final YMarkCrawlStart crawlStart = new YMarkCrawlStart(sb.tables, post.get(ROOT).substring(2));
                final Iterator<String> iter = crawlStart.keySet().iterator();
                String key;
View Full Code Here
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
        } catch (final RowSpaceExceededException e) {
          Log.logException(e);
        }
          } else if (isAutoTagger || isMetadata || isURLdb || isCrawlStart) {
            try {
                  final YMarkMetadata meta = new YMarkMetadata(new DigestURI(post.get(ROOT).substring(2)), sb.indexSegments);
              final Document document = meta.loadDocument(sb.loader);
              final TreeMap<String, YMarkTag> tags = sb.tables.bookmarks.getTags(bmk_user);
              if(isAutoTagger)  {
                prop.put("folders_"+count+"_foldername","<small><b>meta-"+YMarkMetadata.METADATA.KEYWORDS.name().toLowerCase()+":</b> " + meta.loadMetadata().get(YMarkMetadata.METADATA.KEYWORDS) + "</small>");
                putProp(count, "meta");
                count++;
            prop.put("folders_"+count+"_foldername","<small><b>with preference: </b>"+YMarkAutoTagger.autoTag(document, 4, tags)+"</small>");
              putProp(count, "meta");
              count++;
            prop.put("folders_"+count+"_foldername","<small><b>without preference: </b>"+YMarkAutoTagger.autoTag(document, 4, new  TreeMap<String, YMarkTag>())+"</small>");
              putProp(count, "meta");
              count++;
                  prop.put("folders", count);
              } else if(isMetadata) {
                count = putMeta(count, meta.loadMetadata());
              } else if(isURLdb) {
            count = putMeta(count, meta.getMetadata());
              } else if(isCrawlStart) {
                Log.logInfo("YMark", "I am looking for CrawlStart: "+post.get(ROOT).substring(2));
                final YMarkCrawlStart crawlStart = new YMarkCrawlStart(sb.tables, post.get(ROOT).substring(2));
                final Iterator<String> iter = crawlStart.keySet().iterator();
                String key;
View Full Code Here
45
46
47
48
49
50
51
52
53
54
55
56
57
      if (!hasProtocol) {
          url=YMarkTables.PROTOCOLS.HTTP.protocol(url);
      }

          try {
        final YMarkMetadata meta = new YMarkMetadata(new DigestURI(url), sb.indexSegments);
        final Document document = meta.loadDocument(sb.loader);
        final EnumMap<YMarkMetadata.METADATA, String> metadata = meta.loadMetadata();

        prop.putXML("title", metadata.get(YMarkMetadata.METADATA.TITLE));
        prop.putXML("desc", metadata.get(YMarkMetadata.METADATA.DESCRIPTION));
        prop.put("keywords", putTags(document.dc_subject(','), "keywords"));
        prop.put("autotags", putTags(YMarkAutoTagger.autoTag(document, 5, sb.tables.bookmarks.getTags(bmk_user)), "autotags"));
View Full Code Here
TOP

Related Classes of de.anomic.data.ymark.YMarkMetadata

  • de.anomic.crawler.retrieval.Response
  • get_metadata
  • get_treeview
  • net.yacy.document.Condenser
  • net.yacy.kelondro.data.meta.DigestURI
  • net.yacy.kelondro.data.meta.URIMetadataRow
    • 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.