Author: lewismc
Date: Mon Jan 13 13:21:26 2014
New Revision: 1557707
URL: http://svn.apache.org/r1557707
Log:
NUTCH-1672 Inlinks are added twice in DbUpdateReducer
Modified:
nutch/branches/2.x/CHANGES.txt
nutch/branches/2.x/src/java/org/apache/nutch/crawl/DbUpdateReducer.java
Modified: nutch/branches/2.x/CHANGES.txt
URL:
http://svn.apache.org/viewvc/nutch/branches/2.x/CHANGES.txt?rev=1557707&r1=1557706&r2=1557707&view=diff
==============================================================================
--- nutch/branches/2.x/CHANGES.txt (original)
+++ nutch/branches/2.x/CHANGES.txt Mon Jan 13 13:21:26 2014
@@ -2,6 +2,8 @@ Nutch Change Log
Current Development
+* NUTCH-1672 Inlinks are added twice in DbUpdateReducer (Tien Nguyen Manh via
lewismc)
+
* NUTCH-1667 Updatedb always ignore batchId (Tien Nguyen Manh via lewismc)
* NUTCH-1695 NutchDocument.toString() (markus via lewismc)
Modified:
nutch/branches/2.x/src/java/org/apache/nutch/crawl/DbUpdateReducer.java
URL:
http://svn.apache.org/viewvc/nutch/branches/2.x/src/java/org/apache/nutch/crawl/DbUpdateReducer.java?rev=1557707&r1=1557706&r2=1557707&view=diff
==============================================================================
--- nutch/branches/2.x/src/java/org/apache/nutch/crawl/DbUpdateReducer.java
(original)
+++ nutch/branches/2.x/src/java/org/apache/nutch/crawl/DbUpdateReducer.java Mon
Jan 13 13:21:26 2014
@@ -159,10 +159,7 @@ extends GoraReducer<UrlWithScore, NutchW
if (page.getInlinks() != null) {
page.getInlinks().clear();
}
- for (ScoreDatum inlink : inlinkedScoreData) {
- page.putToInlinks(new Utf8(inlink.getUrl()), new
Utf8(inlink.getAnchor()));
- }
-
+
// Distance calculation.
// Retrieve smallest distance from all inlinks distances
// Calculate new distance for current page: smallest inlink distance plus
1.