Author r.david.murray
Recipients bethard, ezio.melotti, georg.brandl, louielu, r.david.murray, serhiy.storchaka, vstinner, wolma
Date 2017-04-20.14:56:42
SpamBayes Score -1.0
Marked as misclassified Yes
Message-id <1492700202.56.0.1224984439.issue18971@psf.upfronthosting.co.za>
In-reply-to
Content
Well, part of the point of the tests is to make sure the module's behavior doesn't change when doing the conversion to argparse.  But "behavior changing" doesn't extend to the exact format of the output...that can change, as long as what the module *does* given specific cli argument combinations stays the same.  Ideally, you write the tests so that it is not sensitive to the exact output, but realistically some tests may need to be different between the optparse case and the argparse case to get good coverage.

Note that even with good tests, you can still miss behavior changes.  We tried to do a good job on the tests for compileall, for example, and we still introduced a bug or two when we converted to argparse.
History
Date User Action Args
2017-04-20 14:56:42r.david.murraysetrecipients: + r.david.murray, georg.brandl, bethard, vstinner, ezio.melotti, serhiy.storchaka, wolma, louielu
2017-04-20 14:56:42r.david.murraysetmessageid: <1492700202.56.0.1224984439.issue18971@psf.upfronthosting.co.za>
2017-04-20 14:56:42r.david.murraylinkissue18971 messages
2017-04-20 14:56:42r.david.murraycreate