We were still using webob's Response objects for template rendering.
Transition to werkzeug's Response object. One caveat was that it
seemed to have used the default mimetype "text/plain" for all pages,
so we override the default Response class, setting the default mime
type to "text/html".
Signed-off-by: Sebastian Spaeth <Sebastian@SSpaeth.de>
In order to move away from webob with its redirect(location=...) we
need to provide a redirect function that allows to directly specify
the URL rather than the urlgen parameters that we now use.
Extend our MG.tools:redirect helper so we can pass in the direct URL
via the optional "location" keyword.
This commit does not switch over any redirect consumers yet.
Signed-off-by: Sebastian Spaeth <Sebastian@SSpaeth.de>
We were refering to model._id in most of the code base as this is
what Mongo uses. However, each use of _id required a) fixup of queries:
e.g. what we did in our find() and find_one() functions moving all
'_id' to 'id'. It also required using AliasFields to make the ._id
attribute available. This all means lots of superfluous fixing and
transitioning in a SQL world.
It will also not work in the long run. Much newer code already refers
to the objects by model.id (e.g. in the oauth plugin), which will break
with Mongo. So let's be honest, rip out the _id mongoism and live with
.id as the one canonical way to address objects.
This commit modifies all users and providers of model._id to use
model.id instead. This patch works with or without Mongo removed first,
but will break Mongo usage (even more than before)
I have not bothered to fixup db.mongo.* and db.sql.convert
(which converts from Mongo to SQL)
Signed-off-by: Sebastian Spaeth <Sebastian@SSpaeth.de>
empty find() queries would not work anymore with the simplified .find
compatability code, so remove these and use proper sqlalchemy in the
tests.
The storage test failed because my virtualenv environment ran
mediagoblin/local/mediagoblin/tests/test_storage.py and somehow decided
the 2 classes are different objects. Just test against the full class name.
Signed-off-by: Sebastian Spaeth <Sebastian@SSpaeth.de>
De-noisify the transcoding log and db updates. Previously we would log
and save the progress percentage every second, even if it had not changed
at all. Save progress:oercentage in the Transcoder and only log/update
when the percentage has actually changed.
Signed-off-by: Sebastian Spaeth <Sebastian@SSpaeth.de>
We accidentally used the fake translation mechanism here which will not
actually translate anything.
Signed-off-by: Sebastian Spaeth <Sebastian@SSpaeth.de>
The oauth plugin used timedelta.total_seconds which was introduced
in python 2.7 only. To preserve backwards compatability, we simply
calculate the time difference in seconds manually.
I considered monkeypatching total_seconds to the timedelta object,
but it is a built-in type written in C (I believe) and modifying
attributes failed horribly. Switch this to use total_seconds once we
require python 2.7 as minimum version.
Signed-off-by: Sebastian Spaeth <Sebastian@SSpaeth.de>
This button used to be a css style <a href=...> </a> (note:
No contents for the <a>). Using this approach has various
drawbacks. Most notably:
- Not clickable in text mode browsers
- Possibly getting marked as a hidden (spam) link
- No alt attribute
So replaced with a real <img/>.
I have no idea what to put in the alt attribute.
No point in translating <p>{{ var }}</p>. Really.
This does not hurt the string freeze, as it removes a
translateable string. So any translations of this string
are just well ... usefuless afterwards.
We had """<a href="abc.dat"> abc.dat</a>""" (note the
apace). And this space is being rendered as a link by
browsers. This looks strange, really.
So fix the spacing.
Pass in unicode not (binary) strings where sqlite expects unicode
values to prevent the test suite from (correctly) complaining about
errors.
I now pass the full suite without any complaints.
Signed-off-by: Sebastian Spaeth <Sebastian@SSpaeth.de>
These tests output noisy sql complaints about receiving non-unicode
for an unicode field. This was ... well ... because we were handing
in non-unicode usernames and passwords.
Prefixing usernames/passwords with u'' makes the testsuite less noisy
and verbose.
Signed-off-by: Sebastian Spaeth <Sebastian@SSpaeth.de>
CSRF tests apparently passed with earlier versions of webtest, but
failed with the latest webtest (1.4.0) package. It choked on passing
a "key=value; " cookie as it split at the semicolon and failed to
find additional values or something like that. Removing the semicolon
makes this test pass.
Signed-off-by: Sebastian Spaeth <Sebastian@SSpaeth.de>
The reason for wanting to give extra information to the user (this is
a very special case migration) is good, but we don't have a nice
"official" way to capture and present that information during tests,
so removing this.
_fix_query_dict modifies its argument in place. Ensure that the
argument passed in has a local name and will be passed into the
subsequent filter_by call.
We were fetching the user collection gallery by slug only, so if two users
had the same collection slug, we would not have been sure which one we'd get.
Fix this by explicitly only fetching the specific user's collections. Also
switch over the view function to make use of the new active_user_from_url
decorator that fetches the User() object for us.
Signed-off-by: Sebastian Spaeth <Sebastian@SSpaeth.de>
This switches the user gallery page over to use the new decorator, and
cleans up the queries to be in proper sqlalchemy format rather than the
old mongo format.
Signed-off-by: Sebastian Spaeth <Sebastian@SSpaeth.de>
