beagled/LuceneFileQueryable.cs

   1 //
   2 // LuceneFileQueryable.cs
   3 //
   4 // Copyright (C) 2005 Novell, Inc.
   5 //
   6
   7 //
   8 // Permission is hereby granted, free of charge, to any person obtaining a
   9 // copy of this software and associated documentation files (the "Software"),
  10 // to deal in the Software without restriction, including without limitation
  11 // the rights to use, copy, modify, merge, publish, distribute, sublicense,
  12 // and/or sell copies of the Software, and to permit persons to whom the
  13 // Software is furnished to do so, subject to the following conditions:
  14 //
  15 // The above copyright notice and this permission notice shall be included in
  16 // all copies or substantial portions of the Software.
  17 //
  18 // THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
  19 // IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
  20 // FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
  21 // AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
  22 // LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
  23 // FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
  24 // DEALINGS IN THE SOFTWARE.
  25 //
  26
  27 using System;
  28 using System.Collections;
  29 using System.IO;
  30
  31 using Beagle.Util;
  32
  33 namespace Beagle.Daemon {
  34
  35         /**
  36          * This queryable just takes the LuceneQueryable and adds some sane
  37          * default behavior for indexing files containing multiple indexables.
  38          * Suitable for feedfiles or mbox style mail files.
  39          * Use this only if the backend generates multiple indexables from one
  40          * physical source file; _do not_ use this if multiple indexables
  41          * are created from one indexable while filtering (aka child indexables).
  42          */
  43         public abstract class LuceneFileQueryable : LuceneQueryable {
  44
  45                 public LuceneFileQueryable (string index_name, int minor_version, bool disable_locking) :
  46                         base (index_name, minor_version, disable_locking)
  47                 { }
  48
  49                 public LuceneFileQueryable (string index_name) : this (index_name, -1, false) { }
  50
  51                 public LuceneFileQueryable (string index_name, bool disable_locking) : this (index_name, -1, disable_locking) { }
  52
  53                 public LuceneFileQueryable (string index_name, int minor_version) : this (index_name, minor_version, false) { }
  54
  55                 ///////////////////////////////////////////////////////////////////////////
  56
  57                 private Hashtable file_references_count = new Hashtable ();
  58
  59                 private void IncrementReferenceCount (string path)
  60                 {
  61                         if (! file_references_count.Contains (path)) {
  62                                 file_references_count [path] = 1;
  63                                 return;
  64                         }
  65
  66                         file_references_count [path] = (int) file_references_count [path] + 1;
  67                 }
  68
  69                 // returns
  70                 // true : reference left
  71                 // false: no more reference left
  72                 private bool DecrementReferenceCount (string path)
  73                 {
  74                         if (! file_references_count.Contains (path))
  75                                 throw new Exception ("Shared file is not referenced: " + path);
  76
  77                         int reference_count = (int) file_references_count [path] - 1;
  78
  79                         if (reference_count == 0) {
  80                                 file_references_count.Remove (path);
  81                                 return false;
  82                         }
  83
  84                         file_references_count [path] = reference_count;
  85                         return true;
  86                 }
  87
  88                 ///////////////////////////////////////////////////////////////////////////
  89
  90                 private class CachedFileInfo {
  91                         public Uri Uri;
  92                         public string Path;
  93                         public DateTime Mtime;
  94                         public bool Shared = false;
  95                 }
  96
  97                 private Hashtable file_info_cache = UriFu.NewHashtable ();
  98
  99                 override protected bool PreAddIndexableHook (Indexable indexable)
 100                 {
 101                         // None of this applies for Removes
 102                         if (indexable.Type == IndexableType.Remove)
 103                                 return true;
 104
 105                         CachedFileInfo info = (CachedFileInfo) file_info_cache [indexable.Uri];
 106
 107                         if (info == null)
 108                                 info = new CachedFileInfo ();
 109
 110                         info.Uri = indexable.Uri;
 111
 112                         if (indexable.Uri.IsFile && indexable.IsNonTransient)
 113                                 info.Path = indexable.Uri.LocalPath;
 114                         else if (indexable.ContentUri.IsFile && indexable.IsNonTransient)
 115                                 info.Path = indexable.ContentUri.LocalPath;
 116                         else if (indexable.ParentUri != null && indexable.ParentUri.IsFile) {
 117                                 info.Path = indexable.ParentUri.LocalPath;
 118                                 info.Shared = true;
 119                                 IncrementReferenceCount (info.Path);
 120                         }
 121
 122                         // The path could be null in certain cases:
 123                         //    * The indexable is a non-file URI and no
 124                         //      parent URI is set.
 125                         //    * The indexable is a child indexable and the
 126                         //      parent URI is not a file URI.
 127                         if (info.Path == null)
 128                                 return true;
 129
 130                         try {
 131                                 info.Mtime = FileSystem.GetLastWriteTimeUtc (info.Path);
 132                         } catch (FileNotFoundException ex) {
 133                                 // If we can't get an mtime for the file, it must
 134                                 // have disappeared out from under us.  In that case,
 135                                 // don't bother adding anything.
 136                                 return false;
 137                         }
 138
 139                         file_info_cache [info.Uri] = info;
 140
 141                         return true;
 142                 }
 143
 144                 override protected void PostAddHook (Indexable indexable, IndexerAddedReceipt receipt)
 145                 {
 146                         // Retrieve our cached info about the file.
 147                         CachedFileInfo info;
 148                         info = file_info_cache [receipt.Uri] as CachedFileInfo;
 149                         if (info == null)
 150                                 return;
 151
 152                         // Yeah, this is ghetto. If it's a file that's shared across multiple
 153                         // indexables, only tag it with when the last indexable has been indexed.
 154                         if (info.Shared && DecrementReferenceCount (info.Path))
 155                                 return;
 156
 157                         // Since we know that the file has been successfully
 158                         // indexed, update the file attributes accordingly.
 159                         // Don't set filter information on a file if multiple
 160                         // indexables has been created from it.
 161                         FileAttributes attr;
 162                         attr = FileAttributesStore.ReadOrCreate (info.Path);
 163
 164                         attr.LastWriteTime = info.Mtime;
 165
 166                         // Don't set filter information on a file if multiple indexables has been
 167                         // created from it.
 168                         if (! info.Shared) {
 169                                 attr.FilterName = receipt.FilterName;
 170                                 attr.FilterVersion = receipt.FilterVersion;
 171                         }
 172
 173                         if (! FileAttributesStore.Write (attr))
 174                                 Logger.Log.Warn ("Couldn't write attributes for {0}", info.Path);
 175
 176                         file_info_cache.Remove (info.Uri);
 177                 }
 178
 179                 override protected void PostRemoveHook (Indexable indexable, IndexerRemovedReceipt receipt)
 180                 {
 181                         file_info_cache.Remove (indexable.Uri);
 182                 }
 183
 184                 override protected bool HitIsValid (Uri uri)
 185                 {
 186                         // Do the right thing if the Uri is a file.
 187                         // If the file Uri we need is the ContentUri, this won't work.
 188                         if (! uri.IsFile)
 189                                 return true;
 190
 191                         try {
 192                                 return FileSystem.Exists (uri.LocalPath);
 193                         } catch (Exception e) {
 194                                 Logger.Log.Warn ("Exception executing HitIsValid on {0}", uri.LocalPath);
 195                                 return false;
 196                         }
 197                 }
 198
 199                 ///////////////////////////////////////////////////////////////////////////
 200
 201                 // Convenience functions
 202
 203                 public bool IsUpToDate (string path, Filter filter)
 204                 {
 205                         return FileAttributesStore.IsUpToDate (path, filter);
 206                 }
 207
 208                 public bool IsUpToDate (string path)
 209                 {
 210                         return FileAttributesStore.IsUpToDate (path);
 211                 }
 212
 213         }
 214
 215 }