In order to ease the patch submitting process, *.patch was added as per
Linux kernel's .gitignore. Test cache directory was also added in its
own gitignore (as per kernel's recommendation on .gitignore location).
Since I am adding a test I also took time to fix all flake8 errors in
this test file, since I may add more tests in there and I prefer to work
on a coherent code base.
This adds a new test and verify that the old test has the
"unknown.<extension>" format. As funny as it seems, the "image/jpeg"
Content-Type will generate a ".jpe" extension but I was expecting a more
common ".jpg" extension. This may be a bug, but this is not the subject
of this patch.
When uploading a file using the API its same is always going to be
'unknown.<extension>', which is not good looking for file urls. This
patch proposes a custom header named 'X-File-Name' that allows to
specify a file name using an API client.
This changeset was tested with a local OAuth1 client.
On a fresh clone the celeryd executable is not present. This is because
of the celery version migration to celery 4.x. As per celery's
documentation at http://docs.celeryproject.org/en/latest/whatsnew-4.0.html,
celeryd is no longer shipped in this version.
The admin panel headers are white by default, making them almost
unreadable on the light background used in the Airy theme. Make the
headers dark, like text elsewhere in the theme.
Make the batchaddmedia command look for a 'collection-slug' column, that
for each row may specify the slug of a collection that the media should
be added to. This matches the '--collection-slug' argument of the
addmedia command.
Prior to this commit, using a Unicode character in a configuration
string would result in a `UnicodeDecodeError` being raised. Supporting
Unicode characters is especially useful in user-facing configuration
strings, such as `html_title`.
Commit c0ea2bad changed the expected location column header of csv files
used by the batchaddmedia command from 'media:location' to 'location'.
This change updates the docs to match.
Add a unit test to verify that the change introduced in 07c535c no
longer resulst in a zero divide error when gps data contains no divisor.
This commit includes a test image created by Andrew Browning. Andrew
Browning dedicates this image to the public domain under the CC0 license.
Re-arrange code to check whether a collection has been selected prior
to checking whether an item already exists in the selected collection.
This fixes a regression introduced in 0f3bf8d.
Modify the @get_media_entry_by_id decorator to return media regardless
of processing state. Separately modify all view functions that use the
@get_media_entry_by_id decorator to require that the media be in the
processed state, other than for the media_confirm_delete view. This
allows blog post drafts to be deleted without returning a 404. Further,
it adds the ability to delete unprocessed media in the future, which
would be a nice addition to the user processing panel.
After alembic updates we need to add foundations. The tables to add
foundations to are created as part of dbupdate process. For some reason
Alembic doesn't commit a session in the end. Because the session is not
commited, the tables do not really get created. And because of that,
foundations fail.
The solution is to commit the session after alembic updates.
