| Age | Commit message (Collapse) | Author |
|
|
|
|
|
Python docs say, "it's usually not necessary to invoke these methods directly."
|
|
|
|
|
|
|
|
|
|
|
|
|
|
It's slightly more compatible than the git protocol.
|
|
|
|
Follow up to a0d166306fbdc41f49e6fadf4ec84b17eb147daa.
|
|
|
|
|
|
queryset cache.
Thanks Luke Benstead for the report Simon Charettes for the fix.
|
|
|
|
|
|
|
|
|
|
Regression in b52c73008a9d67e9ddbb841872dc15cdd3d6ee01.
Thanks Tim Graham for the review.
|
|
|
|
Refs #26709
|
|
|
|
|
|
|
|
- Fixed Oracle backend due to cx_Oracle 5.3 change in the
Cursor.description behavior i.e. "Use None instead of 0 for items in
the Cursor.description attribute that do not have any validity.".
- Used cx_Oracle.Object.size() instead of len().
Thanks Tim Graham for the review.
|
|
Using JSON as non-UTF-8 content is controversial, and the RFC 7159
discourages it.
Thanks Tim Graham for the review.
|
|
|
|
This reverts commit d895fc9ac01db3d3420aa7c943949fe17b3ce028 since the
test is problematic as described in the ticket.
|
|
|
|
positive integer fields.
|
|
|
|
|
|
|
|
|
|
|
|
A JSON Content-Type can contain further content, like charset for example.
|
|
|
|
This reverts commit 4cffa9a1ffb37d4de7e99a851a9ed87b3c02d229.
|
|
Thanks Andrew Brown for the test case.
|
|
Thanks Aymeric Augustin and Carl Meyer.
|
|
|
|
Thanks Tim Graham for the review.
|
|
|
|
|
|
|
|
non-namespaced URLs.
|
|
|
|
'pk'.
|
|
|