| 1 |
#! /bin/sh /usr/share/dpatch/dpatch-run |
| 2 |
## 05_lucenedemo.dpatch by <gregor+debian@comodo.priv.at> |
| 3 |
## |
| 4 |
## All lines beginning with `## DP:' are a description of the patch. |
| 5 |
## DP: Work around missing lucene demo. |
| 6 |
|
| 7 |
@DPATCH@ |
| 8 |
diff -urNad libpdfbox-java~/src/org/pdfbox/searchengine/lucene/IndexFiles.java libpdfbox-java/src/org/pdfbox/searchengine/lucene/IndexFiles.java |
| 9 |
--- libpdfbox-java~/src/org/pdfbox/searchengine/lucene/IndexFiles.java 2007-05-17 22:34:43.000000000 +0200 |
| 10 |
+++ libpdfbox-java/src/org/pdfbox/searchengine/lucene/IndexFiles.java 2007-09-07 15:33:43.000000000 +0200 |
| 11 |
@@ -61,9 +61,9 @@ |
| 12 |
|
| 13 |
import org.apache.lucene.analysis.standard.StandardAnalyzer; |
| 14 |
|
| 15 |
-import org.apache.lucene.demo.HTMLDocument; |
| 16 |
- |
| 17 |
+import org.apache.lucene.ant.HtmlDocument; |
| 18 |
import org.apache.lucene.document.Document; |
| 19 |
+import org.apache.lucene.document.DateField; |
| 20 |
|
| 21 |
import org.apache.lucene.index.IndexReader; |
| 22 |
import org.apache.lucene.index.IndexWriter; |
| 23 |
@@ -207,7 +207,7 @@ |
| 24 |
while (uidIter.term() != null && uidIter.term().field().equals( "uid" ) ) |
| 25 |
{ |
| 26 |
System.out.println("deleting " + |
| 27 |
- HTMLDocument.uid2url(uidIter.term().text())); |
| 28 |
+ uid2url(uidIter.term().text())); |
| 29 |
reader.deleteDocuments(uidIter.term()); |
| 30 |
uidIter.next(); |
| 31 |
} |
| 32 |
@@ -240,7 +240,7 @@ |
| 33 |
{ |
| 34 |
if (uidIter != null) |
| 35 |
{ |
| 36 |
- String uid = HTMLDocument.uid(file); // construct uid for doc |
| 37 |
+ String uid = uid(file); // construct uid for doc |
| 38 |
|
| 39 |
while( uidIter.term() != null && |
| 40 |
uidIter.term().field().equals( "uid" ) && |
| 41 |
@@ -249,7 +249,7 @@ |
| 42 |
if (deleting) |
| 43 |
{ // delete stale docs |
| 44 |
System.out.println("deleting " + |
| 45 |
- HTMLDocument.uid2url(uidIter.term().text())); |
| 46 |
+ uid2url(uidIter.term().text())); |
| 47 |
reader.deleteDocuments(uidIter.term()); |
| 48 |
} |
| 49 |
uidIter.next(); |
| 50 |
@@ -287,7 +287,7 @@ |
| 51 |
path.endsWith(".TXT")) |
| 52 |
{ |
| 53 |
System.out.println( "Indexing Text document: " + file ); |
| 54 |
- doc = HTMLDocument.Document(file); |
| 55 |
+ doc = HtmlDocument.Document(file); |
| 56 |
} |
| 57 |
else if( path.endsWith( ".PDF" ) ) |
| 58 |
{ |
| 59 |
@@ -304,4 +304,27 @@ |
| 60 |
writer.addDocument(doc); |
| 61 |
} |
| 62 |
} |
| 63 |
-} |
| 64 |
\ No newline at end of file |
| 65 |
+ |
| 66 |
+ |
| 67 |
+ /* |
| 68 |
+ * The following 2 methods are taken from the |
| 69 |
+ * org.apache.lucene.demo.HTMLDocument class shipped with |
| 70 |
+ * Lucene 1.4.3. |
| 71 |
+ */ |
| 72 |
+ private static char dirSep = System.getProperty("file.separator").charAt(0); |
| 73 |
+ |
| 74 |
+ private static String uid(File f) { |
| 75 |
+ // Append path and date into a string in such a way that lexicographic |
| 76 |
+ // sorting gives the same results as a walk of the file hierarchy. Thus |
| 77 |
+ // null (\u0000) is used both to separate directory components and to |
| 78 |
+ // separate the path from the date. |
| 79 |
+ return f.getPath().replace(dirSep, '\u0000') + |
| 80 |
+ "\u0000" + |
| 81 |
+ DateField.timeToString(f.lastModified()); |
| 82 |
+ } |
| 83 |
+ |
| 84 |
+ private static String uid2url(String uid) { |
| 85 |
+ String url = uid.replace('\u0000', '/'); // replace nulls with slashes |
| 86 |
+ return url.substring(0, url.lastIndexOf('/')); // remove date from end |
| 87 |
+ } |
| 88 |
+} |
Parent Directory
| 
Revision Log