| msg178679 - (view) |
Author: Brett Cannon (brett.cannon) *  |
Date: 2012-12-31 14:34 |
| Importlib, when checking for PYTHONCASEOK, does not respect -E as it did in Python 3.2 and earlier (http://hg.python.org/cpython/file/0786dfc3b2b4/Python/import.c#l1933). |
|
|
| msg178723 - (view) |
Author: Meador Inge (meador.inge) * |
Date: 2013-01-01 03:21 |
| Is the attached OK? |
|
|
| msg179483 - (view) |
Author: Éric Araujo (eric.araujo) * |
Date: 2013-01-09 18:10 |
| Code patch looks good, but for tests, I prefer to run subprocesses with the options and output I want, instead of changing the expected results depending on the python-running-tests’ options (and thus, having two tests into one, but always running only one). |
|
|
| msg180001 - (view) |
Author: Meador Inge (meador.inge) * |
Date: 2013-01-15 03:25 |
| I agree that it is better to cover both cases in one test irrespective of the interpreter command line options. I will take a look at creating a subprocess. Thanks for the review. |
|
|
| msg186964 - (view) |
Author: Meador Inge (meador.inge) * |
Date: 2013-04-15 00:56 |
| Here is a version which implements the subprocess method for testing suggested by Éric. |
|
|
| msg196876 - (view) |
Author: Roundup Robot (python-dev)  |
Date: 2013-09-03 22:49 |
| New changeset 934e650abc4d by Meador Inge in branch '3.3': Issue #16826: Don't check for PYTHONCASEOK when using -E. http://hg.python.org/cpython/rev/934e650abc4d New changeset ba850a78cbbc by Meador Inge in branch 'default': Issue #16826: Don't check for PYTHONCASEOK when using -E. http://hg.python.org/cpython/rev/ba850a78cbbc |
|
|
| msg196880 - (view) |
Author: Meador Inge (meador.inge) * |
Date: 2013-09-03 23:45 |
| My last commit caused some buildbot failures (http://buildbot.python.org/all/builders/AMD64%20Windows7%20SP1%203.3/builds/1069). I am investigating. |
|
|
| msg196882 - (view) |
Author: Roundup Robot (python-dev) |
Date: 2013-09-04 00:54 |
| New changeset 7801ef4a4ce3 by Meador Inge in branch '3.3': Issue #16826: Revert fix while Windows issues are being worked out. http://hg.python.org/cpython/rev/7801ef4a4ce3 New changeset a1282b67b4cf by Meador Inge in branch 'default': Issue #16826: Revert fix while Windows issues are being worked out. http://hg.python.org/cpython/rev/a1282b67b4cf |
|
|
| msg196928 - (view) |
Author: Meador Inge (meador.inge) * |
Date: 2013-09-04 17:38 |
| Reopening to rework test cases. |
|
|
| msg275077 - (view) |
Author: Eric Snow (eric.snow) * |
Date: 2016-09-08 17:36 |
| What ended up happening with this? |
|
|
| msg275114 - (view) |
Author: Meador Inge (meador.inge) * |
Date: 2016-09-08 18:57 |
| Nothing. I didn't have a Windows box handy to investigate at the time. I do now and can look into this in the next day or so. |
|
|
| msg360917 - (view) |
Author: Brett Cannon (brett.cannon) * |
Date: 2020-01-29 00:33 |
| This is still occurring. Probably need to add `and sys.ignore_environment` to https://github.com/python/cpython/blob/0cd5bff6b7da3118d0c5a88fc2b80f80eb7c3059/Lib/importlib/_bootstrap_external.py#L38. |
|
|
| msg368886 - (view) |
Author: STINNER Victor (vstinner) * |
Date: 2020-05-15 00:47 |
| Issue fixed in bpo-38691: commit fc72ab6913f2b5337ae7fda711f2de846d38f479 Author: idomic <michael.ido@gmail.com> Date: Mon Mar 9 07:57:53 2020 -0400 bpo-38691: importlib ignores PYTHONCASEOK if -E is used (GH-18627) The importlib module now ignores the PYTHONCASEOK environment variable when the -E or -I command line options are being used. |
|
|