For in-memory databases, queries with a "group by" column that is also a hash…

For in-memory databases, queries with a "group by" column that is also a hash index threw a RuntimeException.

h2/src/docsrc/html/changelog.html

@@ -18,7 +18,9 @@ Change Log
 <h1>Change Log</h1>
 
 <h2>Next Version (unreleased)</h2>
-<ul><li>Improved error message for some syntax errors.
+<ul><li>For in-memory databases, queries with a "group by" column that
+    is also a hash index threw a RuntimeException.
+</li><li>Improved error message for some syntax errors.
 </li><li>File system abstraction: if used directly, some file systems did not work correctly
     with spliced byte buffers (the database engine doesn't use those).
 </li><li>To use the MVStore storage engine (which is still work in progress), append

h2/src/main/org/h2/command/dml/Select.java

@@ -250,6 +250,10 @@ public class Select extends Query {
                 if (index.getIndexType().isScan()) {
                     continue;
                 }
+                if (index.getIndexType().isHash()) {
+                    // does not allow scanning entries
+                    continue;
+                }
                 if (isGroupSortedIndex(topTableFilter, index)) {
                     return index;
                 }

h2/src/test/org/h2/test/testScript.sql

@@ -4,6 +4,20 @@
 -- Initial Developer: H2 Group
 --
 --- special grammar and test cases ---------------------------------------------------------------------------------------------
+create table test(x int);
+> ok
+
+create hash index on test(x);
+> ok
+
+select 1 from test group by x;
+> 1
+> -
+> rows: 0
+
+drop table test;
+> ok
+
 call regexp_replace('x', 'x', '\');
 > exception