fix snippet containig number with comma as desmo point http://mantis.tokeek.de/view.php?id=344

to keep it as one word (by altering the split regex)
- added sniipet test case with number
- regex for word split to match multiple splitcars

2 files changed, 16 insertions, 2 deletions

diff --git a/source/net/yacy/search/snippet/TextSnippet.java b/source/net/yacy/search/snippet/TextSnippet.java
index d864d7c21..b10fbd44b 100644
--- a/source/net/yacy/search/snippet/TextSnippet.java
+++ b/source/net/yacy/search/snippet/TextSnippet.java
@@ -75,9 +75,11 @@ public class TextSnippet implements Comparable<TextSnippet>, Comparator<TextSnip
             Pattern.compile(".+[^\\p{L}\\p{N}]\\Z");

     /**

      * <code>\\A[\\p{L}\\p{N}]+[^\\p{L}\\p{N}].+\\Z</code>

+     * updated to <code>\\A([\\p{L}\\p{N}]+[^\\p{L}\\p{N}].+)([\\p{N}]+[.,][\\p{N}])+\\Z</code>

+     * to detect words with none alphanumeric chars (1) allow comma/dot surrounded by number (2)

      */

     private static final Pattern p3 =

-            Pattern.compile("\\A[\\p{L}\\p{N}]+[^\\p{L}\\p{N}].+\\Z");

+            Pattern.compile("\\A([\\p{L}\\p{N}]+[^\\p{L}\\p{N}].+)([\\p{N}]+[.,][\\p{N}])+\\Z");

     /**

      * <code>[^\\p{L}\\p{N}]</code>

      */

@@ -390,7 +392,7 @@ public class TextSnippet implements Comparable<TextSnippet>, Comparator<TextSnip
         return this.resultStatus;

     }

 

-    private static final Pattern SPLIT_PATTERN = Pattern.compile(" |-");

+    private static final Pattern SPLIT_PATTERN = Pattern.compile("[ |-]+");

 

     /**

      * Marks all words in current line which have the same

diff --git a/test/net/yacy/search/snippet/TextSnippetTest.java b/test/net/yacy/search/snippet/TextSnippetTest.java
index bd92903c6..a038b08fb 100644
--- a/test/net/yacy/search/snippet/TextSnippetTest.java
+++ b/test/net/yacy/search/snippet/TextSnippetTest.java
@@ -152,5 +152,17 @@ public class TextSnippetTest {
         System.out.println("testDescriptionline: snippet=" + sniptxt);

         assertFalse ("HTML code not allowed in snippet text",sniptxt.contains("<pre>")); // display text not to include unwanted html

         assertTrue ("Query word not marked", sniptxt.contains("<b>test</b>")); // query word to be marked

+

+        // test text with some numbers (english/german format)

+        rawtestline = "Test Version 1.83 calculates pi to 3,14 always";

+        ts = new TextSnippet(

+            url.hash(),

+            rawtestline,

+            false, // isMarked,

+            TextSnippet.ResultClass.SOURCE_METADATA, "");

+        sniptxt = ts.descriptionline(qg);

+        System.out.println("testDescriptionline: (with numbers) snippet="+sniptxt);

+        assertTrue ("number (.) broken up",sniptxt.contains("1.83"));

+        assertTrue ("number (,) broken up",sniptxt.contains("3,14"));

     }

 }