Attributes of SQLAlchemy objects get "lost". So "cache"
them locally in the code.
This is really the simple explanation for some scarry
sqlalchemy details.
The code often needs to know some fields of the test user
even after doing some sql and stuff. The solultion is to
reload it and properly detach it from its Session. That way
all its fields are available and the whole thing is not
connected to a session. It feels like a normal object.
This merge involved moving the new FilenameBuilder class to
processing/__init__.py, and putting the comment deletion tests back into
test_submission.py using the refactored functions.
- Changed config files of test configs to use SQL
- Updated celery initialization tools, factored them to be able to
use the "big instance" application stuff
- Sniffing now goes through the old extension-based filter before
doing it the bitsniffing way.
- Refractored get_media_type_and_manager(filename).
- Removed ogg extension from video accepted extensions, audio
will take care of that.
- Added custom audio player, still WIP,but working.
- Added test for sniffing. This only tests for the
mediagoblin.media_types.image type, as that is the only
one enabled from start.
This test helps verify that bug #261 is actually fixed.
In order to test that all the processed images are smaller, I needed to add
an image that's bigger than processing.MEDIUM_SIZE, hence bigblue.png.
All the data posts in these tests had a lot of common code. Putting all
that into a function makes it easier to write more tests (which I'll be
doing in a bit) and see what's really being tested.
The actual code is just a simple for loop; there might be a better
implementation but this is a fine start. I also extended test_delete to
check this too.
* Added extlib/freesound/audioprocessing.py
* config_spec
* Added create_spectrogram setting
* Added media:medium and media:thumb max_{width,height} settings
* Added sniffing logic to
- audio.processing:sniff_handler
- video.processing:sniff_handler
* Changed audio.processing:sniff_handler logic
* Added audio thumbnailing functionality to audio.processing
(works only with create_spectrogram enabled)
* Refractored contexts in audio.processing
* Added audio.transcoders:AudioThumbnailer
Used for creating spectrograms and spectrogram thumbnails -
Wadsworth's Constant, we meet again :)
* audio.transcoders:AudioTranscoder
- Added mux_string kwarg
- Delete self.pipeline on self.halt()
* Changed str.format formatting in image.processing:sniff_handler
Had {1} without an {0}, changed to {0}
* Refractored VideoTranscoder to use transcode() for transcoding instead
of __init__()
* Added discover() method to video.transcoders:VideoTranscoder
* Added spectrogram display to media_displays/audio.html
* Updated test_submission to reflect changes in media plugin delegation
* cwebber/celerysql:
Adjust unit tests to match new celery/kombu sqlalchemy setup
"database" is not the sqlalchemy kombu transport... should be "sqlalchemy"
Celery and kombu databases should also be .gitignore'd
kombu-sqlalchemy a requirement in order for kombu sqlalchemy transport to work
Move mediagoblin dbs out of user_dev for race condition directory-creation reasons.
Give kombu its own db. Responding to Elrond "sqlite will lock all the time!" :)
Apparently an absolute path is three slashes after sqlite:. Thx elrond.
Should be all that's needed to switch celery/kombu settings to sqlalchemy
The test checks for a pixel value after rotation (good
idea!). But the value seems to be a bit different on some
platforms, so use a list of seen values.
Not the perfect solution, but it works.
