PATCH: SQL Bug on summarizers.py (PostgreSQL as the backend)

[Pedro Algarvio, aka, s0undt3ch <ufs@…>]

I seem to have hit an SQL bug while running bitten(trunk) on trac 0.10.3 with PostgreSQL as the backend.

ERROR: column "item_lines.value" must appear in the GROUP BY clause or be used in an aggregate function
SQL state: 42803

Here's the patch that at least makes it work again.

bitten/trac_ext/summarizers.py

@@ -78 +78 @@
         cursor = db.cursor()
         cursor.execute("""
 SELECT item_name.value AS unit, item_file.value AS file,
-       item_lines.value AS loc, item_percentage.value AS cov
+       max(item_lines.value) AS loc, max(item_percentage.value) AS cov
 FROM bitten_report AS report
  LEFT OUTER JOIN bitten_report_item AS item_name
   ON (item_name.report=report.id AND item_name.name='name')
@@ -93 +93 @@
       item_percentage.item=item_name.item AND
       item_percentage.name='percentage')
 WHERE category='coverage' AND build=%s AND step=%s
-GROUP BY file, unit ORDER BY unit""", (build.id, step.name))
+GROUP BY file, item_name.value ORDER BY item_name.value""", (build.id, step.name))
 
         data = []
         total_loc, total_cov = 0, 0

[cmlenz]

Fixed in r445.