svn commit: r467345 - in /lucene/nutch/trunk: CHANGES.txt src/java/org/apache/nutch/parse/ParseUtil.java

[si...@apache.org]

Author: siren
Date: Tue Oct 24 07:28:46 2006
New Revision: 467345

URL: http://svn.apache.org/viewvc?view=rev&rev=467345
Log:
fix for NUTCH-391

Modified:
    lucene/nutch/trunk/CHANGES.txt
    lucene/nutch/trunk/src/java/org/apache/nutch/parse/ParseUtil.java

Modified: lucene/nutch/trunk/CHANGES.txt
URL: http://svn.apache.org/viewvc/lucene/nutch/trunk/CHANGES.txt?view=diff&rev=467345&r1=467344&r2=467345
==============================================================================
--- lucene/nutch/trunk/CHANGES.txt (original)
+++ lucene/nutch/trunk/CHANGES.txt Tue Oct 24 07:28:46 2006
@@ -50,6 +50,9 @@
 
 17. NUTCH-383 - upgrade to Hadoop 0.7.1 and Lucene 2.0.0. (ab)
 
+18. NUTCH-391 - ParseUtil logs file contents to log file when it cannot
+    find parser (siren)
+
   ****************************** WARNING !!! ********************************
   * This upgrade breaks data format compatibility. A tool 'convertdb'       *
   * was added to migrate existing CrawlDb-s to the new format. Segment data *

Modified: lucene/nutch/trunk/src/java/org/apache/nutch/parse/ParseUtil.java
URL: http://svn.apache.org/viewvc/lucene/nutch/trunk/src/java/org/apache/nutch/parse/ParseUtil.java?view=diff&rev=467345&r1=467344&r2=467345
==============================================================================
--- lucene/nutch/trunk/src/java/org/apache/nutch/parse/ParseUtil.java (original)
+++ lucene/nutch/trunk/src/java/org/apache/nutch/parse/ParseUtil.java Tue Oct 24 07:28:46 2006
@@ -68,8 +68,8 @@
       parsers = this.parserFactory.getParsers(content.getContentType(), "");
     } catch (ParserNotFound e) {
       if (LOG.isWarnEnabled()) {
-        LOG.warn("No suitable parser found when trying to parse content " +
-                 content);
+        LOG.warn("No suitable parser found when trying to parse content " + content.getUrl() +
+               " of type " + content.getContentType());
       }
       throw new ParseException(e.getMessage());
     }
@@ -123,8 +123,8 @@
       p = this.parserFactory.getParserById(extId);
     } catch (ParserNotFound e) {
       if (LOG.isWarnEnabled()) {
-        LOG.warn("No suitable parser found when trying to parse content " +
-                 content);
+        LOG.warn("No suitable parser found when trying to parse content " + content.getUrl() +
+            " of type " + content.getContentType());
       }
       throw new ParseException(e.getMessage());
     }