Merge lp:~vila/bzr/688072-skip-sphinx-failures into lp:bzr
| Status: | Merged | ||||
|---|---|---|---|---|---|
| Approved by: | Vincent Ladeuil on 2010-12-14 | ||||
| Approved revision: | 5565 | ||||
| Merged at revision: | 5568 | ||||
| Proposed branch: | lp:~vila/bzr/688072-skip-sphinx-failures | ||||
| Merge into: | lp:bzr | ||||
| Diff against target: |
62 lines (+13/-0) 4 files modified
bzrlib/tests/doc_generate/__init__.py (+5/-0) bzrlib/tests/doc_generate/builders/test_texinfo.py (+2/-0) bzrlib/tests/doc_generate/writers/test_texinfo.py (+3/-0) doc/en/release-notes/bzr-2.3.txt (+3/-0) |
||||
| To merge this branch: | bzr merge lp:~vila/bzr/688072-skip-sphinx-failures | ||||
| Related bugs: |
|
| Reviewer | Review Type | Date Requested | Status |
|---|---|---|---|
| bzr-core | 2010-12-09 | Pending | |
|
Review via email:
|
|||
Commit Message
Check sphinx version for tests that need it.
Description of the Change
python-sphinx has reached 1.0 (and beyond) but we don't yet have
it widely available in most of our setups.
Lately, some tests have been failing on babune, in fact the last
failures to turn the OSX 10.6 slave blue are due to sphinx-1.0.4.
Since I won't investigate and fix the compatibility with the
various sphinx versions (maverick is at 0.6.6), I'd prefer to
just skip them in the mean time (2 tests only are failing out of 18).
| John A Meinel (jameinel) wrote : | # |
- 5565. By Vincent Ladeuil on 2010-12-09
-
Nice jma's catch, we need ints.
| Vincent Ladeuil (vila) wrote : | # |
> + return tuple(features.
> +
>
>
> ^- I'm pretty sure that isn't enough, since '12' < '6'.
Doh ! Thanks, fixed.
> I'm fine with this, though not supporting maverick is a bit odd.
(0, 6, 6) < (1,0) so we support maverick. Keep in mind though, that's the texinfo version needs work for all the intra and extra links anyway.
| Vincent Ladeuil (vila) wrote : | # |
sent to pqm by email

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
On 12/9/2010 9:09 AM, Vincent Ladeuil wrote: /bugs.launchpad .net/bugs/ 688072
> Vincent Ladeuil has proposed merging lp:~vila/bzr/688072-skip-sphinx-failures into lp:bzr.
>
> Requested reviews:
> bzr-core (bzr-core)
> Related bugs:
> #688072 tests failures with sphinx >= 1.0
> https:/
>
>
> python-sphinx has reached 1.0 (and beyond) but we don't yet have
> it widely available in most of our setups.
>
> Lately, some tests have been failing on babune, in fact the last
> failures to turn the OSX 10.6 slave blue are due to sphinx-1.0.4.
>
> Since I won't investigate and fix the compatibility with the
> various sphinx versions (maverick is at 0.6.6), I'd prefer to
> just skip them in the mean time (2 tests only are failing out of 18).
>
+ def sphinx_ version( self): sphinx. module. __version_ _.split( '.'))
+ # Convert to a tuple to avoid traps in string comparison
+ # ( '1.12' < '1.6' but (1, 12) > (1, 6) )
+ return tuple(features.
+
^- I'm pretty sure that isn't enough, since '12' < '6'.
You need to map(int( features. sphinx. module. __version_ _.split( '.'))
I'm fine with this, though not supporting maverick is a bit odd.
John
=:->
-----BEGIN PGP SIGNATURE----- enigmail. mozdev. org/
BMW4ACgkQJdeBCY SNAANP/ QCdFomY6sDeVSLe vxHMBbLtGf4d fHG/HIIldztsLpq bV
Version: GnuPG v1.4.9 (Cygwin)
Comment: Using GnuPG with Mozilla - http://
iEYEARECAAYFAk0
8dQAoMyRoI9H3M5
=Ro+8
-----END PGP SIGNATURE-----