ResourceClass.normalize_name produces different results for py2 vs. py3

[Eric Fried]

Due to a py3 quirk (read: bug they decided not to fix) [1], .upper() works differently in py2 vs. py3 for the sharp S ('ß'). In py2, it stays the same (like all other characters not in the a-z range); in py3 it becomes the two-character string 'SS'. This means that, as written, ResourceClass.normalize_name('ß') will yield 'CUSTOM__' in py2, but 'CUSTOM_SS' in py3.

[1] https://bugs.python.org/issue4610

[OpenStack Infra (hudson-openstack)]

Related fix proposed to branch: master
Review: https://review.openstack.org/560092

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: master
Review: https://review.openstack.org/560093

[melanie witt (melwitt)]

FWIW, reading through the bug in [1], it looks like it was the opposite -- that:

  >>> x='ß'
  >>> print(x, x.upper())
  ß ß

was considered a bug [2] that was later fixed [3] to behave like:

  >>> 'ß'.upper()
  'SS'

in a later python3 version.

[1] https://bugs.python.org/issue4610
[2] https://bugs.python.org/msg77417
[3] https://bugs.python.org/msg191740

[Eric Fried (efried)]

Shrug, I can see it both ways. Nevertheless, it results in some busted behavior in environments where both py2 and py3 need to be supported.

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.openstack.org/560092
Committed: https://git.openstack.org/cgit/openstack/nova/commit/?id=443980399b03282c37e64e7a529be4960689b34d
Submitter: Zuul
Branch: master

commit 443980399b03282c37e64e7a529be4960689b34d
Author: Eric Fried <email address hidden>
Date: Tue Apr 10 12:17:17 2018 -0500

    Test case: ResourceClass.normalize_name with ß

    Demonstrates the cited bug by invoking ResourceClass.normalize_name on a
    string containing sharp S ('ß'). The result is different for py2 and
    py3.

    Change-Id: I632995c1a234f26f05ef0b5675e0e22b9866b03d
    Related-Bug: #1762789

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.openstack.org/560093
Committed: https://git.openstack.org/cgit/openstack/nova/commit/?id=7fd2de85b632575315df0dce2287b3a7b44492f1
Submitter: Zuul
Branch: master

commit 7fd2de85b632575315df0dce2287b3a7b44492f1
Author: Eric Fried <email address hidden>
Date: Tue Apr 10 12:24:40 2018 -0500

    Make ResourceClass.normalize_name handle sharp S

    Fixes ResourceClass.normalize_name to produce the same result in py2 and
    py3 even when sharp S ('ß') is part of the input.

    We do this by waiting until after replacing non-alphanumerics to upcase.

    Change-Id: I431fa29d36d0d633374973fc25168344042d6c1a
    Closes-Bug: #1762789

[OpenStack Infra (hudson-openstack)]

This issue was fixed in the openstack/nova 18.0.0.0b1 development milestone.