Fix string_to_array() to correctly handle the case where there are

overlapping possible matches for the separator string, such as
string_to_array('123xx456xxx789', 'xx').
Also, revise the logic of replace(), split_part(), and string_to_array()
to avoid O(N^2) work from redundant searches and conversions to pg_wchar
format when there are N matches to the separator string.
Backpatched the full patch as far as 8.0.  7.4 also has the bug, but the
code has diverged a lot, so I just went for a quick-and-dirty fix of the
bug itself in that branch.

diff --git a/src/backend/utils/adt/varlena.c b/src/backend/utils/adt/varlena.c
index 81f8726c3b4957bc326bced5fc18f51284fd93b3..37de64b46a07a53310324e0d27c42271457908e5 100644 (file)
--- a/src/backend/utils/adt/varlena.c
+++ b/src/backend/utils/adt/varlena.c
@@ -8,7 +8,7 @@
  *
  *
  * IDENTIFICATION
- *       $Header: /cvsroot/pgsql/src/backend/utils/adt/varlena.c,v 1.106.2.6 2006/05/21 20:06:43 tgl Exp $
+ *       $Header: /cvsroot/pgsql/src/backend/utils/adt/varlena.c,v 1.106.2.7 2006/10/07 00:12:12 tgl Exp $
  *
  *-------------------------------------------------------------------------
  */
@@ -779,6 +779,8 @@ text_position(Datum str, Datum search_str, int matchnum)
                                        pos = p + 1;
                                        break;
                                }
+                               p1 += len2 - 1;
+                               p += len2 - 1;
                        }
                        p1++;
                }
@@ -809,6 +811,8 @@ text_position(Datum str, Datum search_str, int matchnum)
                                        pos = p + 1;
                                        break;
                                }
+                               p1 += len2 - 1;
+                               p += len2 - 1;
                        }
                        p1++;
                }