Message181789
| Author |
ncoghlan |
| Recipients |
Arfrever, Julian, Yaroslav.Halchenko, abingham, bfroehle, borja.ruiz, brett.cannon, brian.curtin, chris.jerdonek, eric.araujo, eric.snow, exarkun, ezio.melotti, fperez, hpk, michael.foord, nchauvat, ncoghlan, pitrou, r.david.murray, santoso.wijaya, serhiy.storchaka, spiv |
| Date |
2013年02月10日.11:43:18 |
| SpamBayes Score |
-1.0 |
| Marked as misclassified |
Yes |
| Message-id |
<CADiSq7fA50FMACQrHUPm2U-vgz0jmUmMgEhfeg9byo_mj7Y4nA@mail.gmail.com> |
| In-reply-to |
<1360495605.88.0.971111076027.issue16997@psf.upfronthosting.co.za> |
| Content |
You can use subtests to build parameterized tests, you can't use
parameterized tests to build subtests. The standard library can also
be converted to using subtests *far* more readily than it could be
converted to parameterized tests. There's also the fact that creating
a decent parameterized tests without subtests support requires PEP
422.
If you're telling us we can only have one, then I choose subtests, and
third party test frameworks can layer parameterized tests on top.
However, I think you're making a mistaking by seeing them as
*competing* APIs, rather than seeing subtests as a superior
implementation strategy for the possible later introduction of a
higher level parameterized tests API. |
|
History
|
|---|
| Date |
User |
Action |
Args |
| 2013年02月10日 11:43:19 | ncoghlan | set | recipients:
+ ncoghlan, brett.cannon, spiv, exarkun, pitrou, ezio.melotti, eric.araujo, Arfrever, r.david.murray, michael.foord, brian.curtin, hpk, fperez, chris.jerdonek, Yaroslav.Halchenko, santoso.wijaya, nchauvat, Julian, abingham, eric.snow, serhiy.storchaka, borja.ruiz, bfroehle |
| 2013年02月10日 11:43:19 | ncoghlan | link | issue16997 messages |
| 2013年02月10日 11:43:18 | ncoghlan | create |
|