Package org.apache.poi.hdgf.extractor

Examples of org.apache.poi.hdgf.extractor.VisioTextExtractor

70
71
72
73
74
75
76
77
78
79
80
81
82
     
        try {
            String contents = "";
            SummaryInformation summary = null;
            try {
                final VisioTextExtractor extractor = new VisioTextExtractor(source);
              contents = extractor.getText();
                summary = extractor.getSummaryInformation();
            } catch (Exception e) {
              Log.logWarning("vsdParser", e.getMessage());
            }

            String author = null;
View Full Code Here
114
115
116
117
118
119
120
121
122
123
124
125
126
                setType(metadata, "application/vnd.ms-excel");
                Locale locale = context.get(Locale.class, Locale.getDefault());
                new ExcelExtractor().parse(filesystem, xhtml, locale);
            } else if ("VisioDocument".equals(name)) {
                setType(metadata, "application/vnd.visio");
                VisioTextExtractor extractor =
                    new VisioTextExtractor(filesystem);
                for (String text : extractor.getAllText()) {
                    xhtml.element("p", text);
                }
            } else if (!outlookExtracted && name.startsWith("__substg1.0_")) {
                // TODO: Cleaner mechanism for detecting Outlook
                outlookExtracted = true;
View Full Code Here
125
126
127
128
129
130
131
132
133
134
      }
      if(entry.getName().equals("PowerPoint Document")) {
        return new PowerPointExtractor(poifsDir, fs);
      }
      if(entry.getName().equals("VisioDocument")) {
        return new VisioTextExtractor(poifsDir, fs);
      }
    }
    throw new IllegalArgumentException("No supported documents found in the OLE2 stream");
  }
View Full Code Here
122
123
124
125
126
127
128
129
130
131
132
133
134
                   setType(metadata, "application/vnd.ms-excel");
                   Locale locale = context.get(Locale.class, Locale.getDefault());
                   new ExcelExtractor().parse(filesystem, xhtml, locale);
               } else if ("VisioDocument".equals(name)) {
                   setType(metadata, "application/vnd.visio");
                   VisioTextExtractor extractor =
                       new VisioTextExtractor(filesystem);
                   for (String text : extractor.getAllText()) {
                       xhtml.element("p", text);
                   }
               } else if (!outlookExtracted && name.startsWith("__substg1.0_")) {
                   // TODO: Cleaner mechanism for detecting Outlook
                   outlookExtracted = true;
View Full Code Here
208
209
210
211
212
213
214
215
216
217
218
219
220
           break;
        case PROJECT:
           // We currently can't do anything beyond the metadata
           break;
        case VISIO:
           VisioTextExtractor visioTextExtractor =
              new VisioTextExtractor(root);
           for (String text : visioTextExtractor.getAllText()) {
              xhtml.element("p", text);
           }
           break;
        case OUTLOOK:
           OutlookExtractor extractor =
View Full Code Here
134
135
136
137
138
139
140
141
142
143
      }
      if(entry.getName().equals("PowerPoint Document")) {
        return new PowerPointExtractor(poifsDir, fs);
      }
      if(entry.getName().equals("VisioDocument")) {
        return new VisioTextExtractor(poifsDir, fs);
      }
    }
    throw new IllegalArgumentException("No supported documents found in the OLE2 stream");
  }
View Full Code Here
112
113
114
115
116
117
118
119
120
121
122
123
124
            } else if ("Workbook".equals(name)) {
                setType(metadata, "application/vnd.ms-excel");
                new ExcelExtractor().parse(filesystem, xhtml);
            } else if ("VisioDocument".equals(name)) {
                setType(metadata, "application/vnd.visio");
                VisioTextExtractor extractor =
                    new VisioTextExtractor(filesystem);
                for (String text : extractor.getAllText()) {
                    xhtml.element("p", text);
                }
            } else if (!outlookExtracted && name.startsWith("__substg1.0_")) {
                // TODO: Cleaner mechanism for detecting Outlook
                outlookExtracted = true;
View Full Code Here
194
195
196
197
198
199
200
201
202
203
204
205
206
                case WORKBOOK:
                    Locale locale = context.get(Locale.class, Locale.getDefault());
                    new ExcelExtractor(context).parse(filesystem, xhtml, locale);
                    break;
                case VISIO:
                    VisioTextExtractor visioTextExtractor =
                        new VisioTextExtractor(filesystem);
                    for (String text : visioTextExtractor.getAllText()) {
                        xhtml.element("p", text);
                    }
                    break;
                case OUTLOOK:
                    if (!outlookExtracted) {
View Full Code Here
195
196
197
198
199
200
201
202
203
204
205
206
207
           break;
        case PROJECT:
           // We currently can't do anything beyond the metadata
           break;
        case VISIO:
           VisioTextExtractor visioTextExtractor =
              new VisioTextExtractor(root);
           for (String text : visioTextExtractor.getAllText()) {
              xhtml.element("p", text);
           }
           break;
        case OUTLOOK:
           OutlookExtractor extractor =
View Full Code Here
231
232
233
234
235
236
237
238
239
240
241
        if (poifsDir.hasEntry("PowerPoint Document")) {
            return new PowerPointExtractor(poifsDir);
        }

        if (poifsDir.hasEntry("VisioDocument")) {
            return new VisioTextExtractor(poifsDir);
        }

        if (poifsDir.hasEntry("Quill")) {
            return new PublisherTextExtractor(poifsDir);
        }
View Full Code Here
TOP

Related Classes of org.apache.poi.hdgf.extractor.VisioTextExtractor

  • net.yacy.document.parser.vsdParser
  • org.apache.poi.extractor.ExtractorFactory
  • org.apache.poi.hdgf.chunks.Chunk
  • org.apache.poi.hdgf.chunks.Chunk.Command
  • org.apache.poi.hdgf.HDGFDiagram
  • org.apache.poi.hdgf.streams.ChunkStream
  • org.apache.poi.hdgf.streams.PointerContainingStream
  • org.apache.poi.hdgf.TestHDGFCore
  • org.apache.poi.poifs.filesystem.POIFSFileSystem
  • org.apache.tika.parser.microsoft.OfficeParser
    • 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.