svn commit: r1071438 - in /lucene/dev/branches/branch_3x: ./ lucene/ solr/ solr/contrib/dataimporthandler/src/main/java/org/apache/solr/handler/dataimport/ solr/contrib/dataimporthandler/src/test/java/org/apache/solr/handler/dataimport/

[yo...@apache.org]

Author: yonik
Date: Wed Feb 16 23:11:59 2011
New Revision: 1071438

URL: http://svn.apache.org/viewvc?rev=1071438&view=rev
Log:
SOLR-1191: fix DIH deltaQyery when pk has prefix, change NPE to better error reporting

Added:
    lucene/dev/branches/branch_3x/solr/contrib/dataimporthandler/src/test/java/org/apache/solr/handler/dataimport/TestSqlEntityProcessorDeltaPrefixedPk.java
      - copied unchanged from r1071435, lucene/dev/trunk/solr/contrib/dataimporthandler/src/test/java/org/apache/solr/handler/dataimport/TestSqlEntityProcessorDeltaPrefixedPk.java
Modified:
    lucene/dev/branches/branch_3x/   (props changed)
    lucene/dev/branches/branch_3x/lucene/   (props changed)
    lucene/dev/branches/branch_3x/solr/   (props changed)
    lucene/dev/branches/branch_3x/solr/CHANGES.txt
    lucene/dev/branches/branch_3x/solr/contrib/dataimporthandler/src/main/java/org/apache/solr/handler/dataimport/DocBuilder.java

Modified: lucene/dev/branches/branch_3x/solr/CHANGES.txt
URL: http://svn.apache.org/viewvc/lucene/dev/branches/branch_3x/solr/CHANGES.txt?rev=1071438&r1=1071437&r2=1071438&view=diff
==============================================================================
--- lucene/dev/branches/branch_3x/solr/CHANGES.txt (original)
+++ lucene/dev/branches/branch_3x/solr/CHANGES.txt Wed Feb 16 23:11:59 2011
@@ -520,6 +520,12 @@ Bug Fixes
 * SOLR-2339: Fix sorting to explicitly generate an error if you
   attempt to sort on a multiValued field. (hossman)
 
+* SOLR-1191: resolve DataImportHandler deltaQuery column against pk when pk
+  has a prefix (e.g. pk="book.id" deltaQuery="select id from ..."). More
+  useful error reporting when no match found (previously failed with a
+  NullPointerException in log and no clear user feedback). (gthb via yonik)
+
+
 Other Changes
 ----------------------
 

Modified: lucene/dev/branches/branch_3x/solr/contrib/dataimporthandler/src/main/java/org/apache/solr/handler/dataimport/DocBuilder.java
URL: http://svn.apache.org/viewvc/lucene/dev/branches/branch_3x/solr/contrib/dataimporthandler/src/main/java/org/apache/solr/handler/dataimport/DocBuilder.java?rev=1071438&r1=1071437&r2=1071438&view=diff
==============================================================================
--- lucene/dev/branches/branch_3x/solr/contrib/dataimporthandler/src/main/java/org/apache/solr/handler/dataimport/DocBuilder.java (original)
+++ lucene/dev/branches/branch_3x/solr/contrib/dataimporthandler/src/main/java/org/apache/solr/handler/dataimport/DocBuilder.java Wed Feb 16 23:11:59 2011
@@ -318,7 +318,11 @@ public class DocBuilder {
       String keyName = root.isDocRoot ? root.getPk() : root.getSchemaPk();
       Object key = map.get(keyName);
       if(key == null) {
-        LOG.warn("no key was available for deleteted pk query. keyName = " + keyName);
+        keyName = findMatchingPkColumn(keyName, map);
+        key = map.get(keyName);
+      }
+      if(key == null) {
+        LOG.warn("no key was available for deleted pk query. keyName = " + keyName);
         continue;
       }
       writer.deleteDoc(key);
@@ -816,6 +820,28 @@ public class DocBuilder {
     return entity.processor = new EntityProcessorWrapper(entityProcessor, this);
   }
 
+  private String findMatchingPkColumn(String pk, Map<String, Object> row) {
+    if (row.containsKey(pk))
+      throw new IllegalArgumentException(
+        String.format("deltaQuery returned a row with null for primary key %s", pk));
+    String resolvedPk = null;
+    for (String columnName : row.keySet()) {
+      if (columnName.endsWith("." + pk) || pk.endsWith("." + columnName)) {
+        if (resolvedPk != null)
+          throw new IllegalArgumentException(
+            String.format(
+              "deltaQuery has more than one column (%s and %s) that might resolve to declared primary key pk='%s'",
+              resolvedPk, columnName, pk));
+        resolvedPk = columnName;
+      }
+    }
+    if (resolvedPk == null)
+      throw new IllegalArgumentException(
+        String.format("deltaQuery has no column to resolve to declared primary key pk='%s'", pk));
+    LOG.info(String.format("Resolving deltaQuery column '%s' to match entity's declared pk '%s'", resolvedPk, pk));
+    return resolvedPk;
+  }
+
   /**
    * <p> Collects unique keys of all Solr documents for whom one or more source tables have been changed since the last
    * indexed time. </p> <p> Note: In our definition, unique key of Solr document is the primary key of the top level
@@ -852,13 +878,20 @@ public class DocBuilder {
     Map<String, Map<String, Object>> deltaSet = new HashMap<String, Map<String, Object>>();
     LOG.info("Running ModifiedRowKey() for Entity: " + entity.name);
     //get the modified rows in this entity
+    String pk = entity.getPk();
     while (true) {
       Map<String, Object> row = entityProcessor.nextModifiedRowKey();
 
       if (row == null)
         break;
 
-      deltaSet.put(row.get(entity.getPk()).toString(), row);
+      Object pkValue = row.get(pk);
+      if (pkValue == null) {
+        pk = findMatchingPkColumn(pk, row);
+        pkValue = row.get(pk);
+      }
+
+      deltaSet.put(pkValue.toString(), row);
       importStatistics.rowsCount.incrementAndGet();
       // check for abort
       if (stop.get())
@@ -873,8 +906,14 @@ public class DocBuilder {
 
       deletedSet.add(row);
       
+      Object pkValue = row.get(pk);
+      if (pkValue == null) {
+        pk = findMatchingPkColumn(pk, row);
+        pkValue = row.get(pk);
+      }
+
       // Remove deleted rows from the delta rows
-      String deletedRowPk = row.get(entity.getPk()).toString();
+      String deletedRowPk = pkValue.toString();
       if (deltaSet.containsKey(deletedRowPk)) {
         deltaSet.remove(deletedRowPk);
       }