Filters/FilterDOC.cs

   1 //
   2 // FilterDOC.cs : Trivial implementation of a MS Word-document filter.
   3 //                This filter uses wv1 library - http://wvware.sourceforge.net/
   4 //
   5 // Author: Veerapuram Varadhan <vvaradhan@novell.com>
   6 //
   7 // Copyright (C) 2004 Novell, Inc.
   8 //
   9
  10 //
  11 // Permission is hereby granted, free of charge, to any person obtaining a
  12 // copy of this software and associated documentation files (the "Software"),
  13 // to deal in the Software without restriction, including without limitation
  14 // the rights to use, copy, modify, merge, publish, distribute, sublicense,
  15 // and/or sell copies of the Software, and to permit persons to whom the
  16 // Software is furnished to do so, subject to the following conditions:
  17 //
  18 // The above copyright notice and this permission notice shall be included in
  19 // all copies or substantial portions of the Software.
  20 //
  21 // THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
  22 // IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
  23 // FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
  24 // AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
  25 // LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
  26 // FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
  27 // DEALINGS IN THE SOFTWARE.
  28 //
  29
  30 using System;
  31 using System.IO;
  32 using System.Runtime.InteropServices;
  33
  34 using Beagle.Util;
  35 using Beagle.Daemon;
  36
  37 using Gsf;
  38
  39 namespace Beagle.Filters {
  40
  41         public class FilterDOC : FilterOle {
  42
  43                 //////////////////////////////////////////////////////////
  44
  45                 private delegate void TextHandlerCallback (IntPtr byteArray, int dataLen,
  46                                                            IntPtr byteHotArray, int hotDataLen,
  47                                                            bool appendStructBrk);
  48
  49                 [DllImport ("libbeagleglue")]
  50                 private static extern int wv1_glue_init_doc_parsing (string fname, TextHandlerCallback callback);
  51
  52                 [DllImport ("libbeagleglue")]
  53                 private static extern int wv1_init ();
  54
  55                 //////////////////////////////////////////////////////////
  56
  57                 static bool wv1_Initted = false;
  58
  59                 public FilterDOC ()
  60                 {
  61                         AddSupportedFlavor (FilterFlavor.NewFromMimeType ("application/msword"));
  62                         AddSupportedFlavor (FilterFlavor.NewFromMimeType ("application/vnd.ms-word"));
  63                         AddSupportedFlavor (FilterFlavor.NewFromMimeType ("application/x-msword"));
  64                         SnippetMode = true;
  65                 }
  66
  67                 private void IndexText (IntPtr byteArray, int dataLen,
  68                                         IntPtr byteHotArray, int hotDataLen,
  69                                         bool appendStructBrk)
  70                 {
  71                         byte[] data = null;
  72                         string str = null;
  73                         string strHot = null;
  74
  75                         try {
  76                                 if (dataLen > 0){
  77                                         data = new byte[dataLen];
  78                                         Marshal.Copy (byteArray, data, 0, dataLen);
  79                                 }
  80
  81                                 if (data != null)
  82                                         str = System.Text.Encoding.UTF8.GetString (data, 0, dataLen);
  83
  84                                 data = null;
  85                                 if (hotDataLen > 0) {
  86                                         data = new byte [hotDataLen];
  87                                         Marshal.Copy (byteHotArray, data, 0, hotDataLen);
  88                                 }
  89                                 if (data != null)
  90                                         strHot = System.Text.Encoding.UTF8.GetString (data, 0, hotDataLen);
  91
  92                                 AppendText (str, strHot);
  93
  94                                 if (appendStructBrk)
  95                                         AppendStructuralBreak ();
  96                         } catch (Exception e) {
  97                                 Logger.Log.Debug ("Exception occurred in Word-Doc filter. {0}", e);
  98                         }
  99                 }
 100
 101                 override protected void OpenStorage (FileInfo info)
 102                 {
 103                         FileName = info.FullName;
 104                 }
 105
 106                 override protected void ExtractMetaData (Gsf.Input sumStream, Gsf.Input docSumStream)
 107                 {
 108                         int count = 0;
 109                         DocProp prop = null;
 110
 111                         if (sumMeta != null) {
 112                                 prop = sumMeta.Lookup ("gsf:word-count");
 113                                 if (prop != null)
 114                                         count = (int) prop.Val;
 115                                 if (count > 0)
 116                                         AddProperty (Beagle.Property.NewUnsearched ("fixme:word-count", count));
 117
 118                                 count = 0;
 119                                 prop = sumMeta.Lookup ("gsf:page-count");
 120                                 if (prop != null)
 121                                         count = (int) prop.Val;
 122                                 if (count > 0)
 123                                         AddProperty (Beagle.Property.NewUnsearched ("fixme:page-count", count));
 124                         }
 125                 }
 126
 127                 override protected void DoPull ()
 128                 {
 129                         int ret;
 130                         TextHandlerCallback textHandler;
 131                         textHandler = new TextHandlerCallback (IndexText);
 132
 133                         if (!wv1_Initted) {
 134                                 wv1_init ();
 135                                 wv1_Initted = true;
 136                         }
 137
 138                         Stopwatch stopwatch = new Stopwatch ();
 139                         stopwatch.Start ();
 140
 141                         ret = wv1_glue_init_doc_parsing (FileName, textHandler);
 142                         if (ret == -2)
 143                                 Logger.Log.Error ("{0} : is password protected", FileName);
 144                         else if (ret == -1)
 145                                 Logger.Log.Error ("{0} : Unable to read", FileName);
 146                         else if (ret == -3)
 147                                 Logger.Log.Error ("Unable to initiate the parser for {0}", FileName);
 148                         stopwatch.Stop ();
 149                         Logger.Log.Info ("Word document extraction done in {0}", stopwatch);
 150                         Finished ();
 151                 }
 152         }
 153 }