637681 - bybug should return test run counts for all days

Status: RESOLVED FIXED

(Tree Management Graveyard :: OrangeFactor, defect)

Description

[Mark Côté [:mcote]]

Forked from bug 636735.

bybug doesn't return any data for dates on which the bug didn't occur, if you pass a bug to it.  So for bug details graphs, the test run count shows as 0 for dates on which the bug wasn't observed, which is obviously wrong.  See e.g.,
http://localhost:8080/?display=Bug&bugid=633566&startday=2011-02-01&endday=2011-02-23#

There are multiple possible ways to fix this; one would be to change the bybug
API to include testrun data and an orangecount of 0 for dates on which the bug
doesn't exist; I'm not sure if this would have any effects on other code.

Comment 1

[Jonathan Griffin (:jgriffin)]

Attachment: woo_server.py patch

Now, calling count or bybug will return testrun count for every day in the date range, even if there are no relevant oranges.

Comment 2

[Mark Côté [:mcote]]

Comment on attachment 517584 [details] [diff] [review]
woo_server.py patch

Looks good.  There's still an issue on the front end, though, which I'll look into.

Comment 3

[Jonathan Griffin (:jgriffin)]

Pushed as http://hg.mozilla.org/automation/orangefactor/rev/b14c261e8883

Comment 4

[Mark Côté [:mcote]]

Attachment: front-end patch

Okay this patch displays test-run counts of 0.  I did some further simplications:
- there's no point in having a test-run count by bug--they should all be the same
- there's no point in using the date from each orange object, since we already categorized them by date

The only problem left here is that the we seem to have different ideas of what the end date means from view to view.  For example, the OF view does not include the last day in the graph, but the Bug Details view does.  We should sort that out and make it consistent.

Comment 5

[Mark Côté [:mcote]]

Pushed as http://hg.mozilla.org/automation/orangefactor/rev/643b0d368217

I will resolve this; we can deal with the end-date inconsistencies elsewhere.

Comment 6

[(no longer active)]

Dummy question: what is the update schedule from the code base on the production web server?

Comment 7

[Mark Côté [:mcote]]

Pretty arbitrary. :)  But I deployed these changes earlier today.  You might have to refresh to ensure you aren't using cached copies.