- make sure negative row indexes are based on the size of the

number of columns we're actually reporting on
- add more tests for negative row index
- changelog/migration

2 files changed, 27 insertions, 0 deletions

diff --git a/doc/build/changelog/changelog_11.rst b/doc/build/changelog/changelog_11.rst
index 2a0e609b6..18d6d1f74 100644
--- a/doc/build/changelog/changelog_11.rst
+++ b/doc/build/changelog/changelog_11.rst
@@ -22,6 +22,18 @@
     :version: 1.1.0b1
 
     .. change::
+        :tags: feature, sql
+        :pullreq: github:231
+
+        Negative integer indexes are now accommodated by rows
+        returned from a :class:`.ResultProxy`.  Pull request courtesy
+        Emanuele Gaifas.
+
+        .. seealso::
+
+            :ref:`change_gh_231`
+
+    .. change::
         :tags: feature, sqlite
         :tickets: 3629
 
diff --git a/doc/build/changelog/migration_11.rst b/doc/build/changelog/migration_11.rst
index 64ed2d9e2..6f0da3780 100644
--- a/doc/build/changelog/migration_11.rst
+++ b/doc/build/changelog/migration_11.rst
@@ -1179,6 +1179,21 @@ string values::
 
 :ticket:`3292`
 
+.. _change_gh_231:
+
+Negative integer indexes accommodated by Core result rows
+---------------------------------------------------------
+
+The :class:`.RowProxy` object now accomodates single negative integer indexes
+like a regular Python sequence, both in the pure Python and C-extension
+version.  Previously, negative values would only work in slices::
+
+    >>> from sqlalchemy import create_engine
+    >>> e = create_engine("sqlite://")
+    >>> row = e.execute("select 1, 2, 3").first()
+    >>> row[-1], row[-2], row[1], row[-2:2]
+    3 2 2 (2,)
+
 .. _change_3095:
 
 The ``Enum`` type now does in-Python validation of values