- Test functions defined using @given can now be called from other
threads
- Attempting to delete a settings property would previously have
silently done the wrong thing. Now it raises an AttributeError.
- Creating a settings object with a custom database_file parameter was
silently getting ignored and the default was being used instead. Now
it’s not.
We have some things like pos_px stubbed which will fail any test because
of the stub warning - but some tests don't actually need that, it just
happens when e.g. loading something.
So let's not fail tests based on stub warnings, and see how much works
that way.
We've had many checks disabled - these are the ones we actually lose:
L104
Docstrings must use Napoleon, not reStructuredText fields.
L205
__init__.py is not allowed to contain function or class definitions.
L206
Implicit relative imports are not allowed.
L208
Pokémon exception handling is always a mistake. If the intent is
really to catch and ignore exceptions, explicitly name which
exception types to silence.
L209
return, del, raise, assert, print (in python 2, without
print_function) yield, and yield from are statements, not functions,
and as such, do not require parentheses.
L210
Instead of intentionally relying on the side effects of map, filter,
or a comprehension, write an explicit for loop.
L211
Using map or filter with a lambda as the first argument is always
better written as list comprehension or generator expression. An
expression is more readable and extensible, and less importantly,
doesn't incur as much function call overhead.
L212
Using @staticmethod is always wrong.
L301
Files must end with a trailing newline.
L303
noqa is ignored, and as such, # noqa comments should be deleted to
reduce pointless noise.
However, most of those are also checked by pylint (and the rest I don't
really care about), and ebb-lint increases flake8's runtime a lot
(45s -> almost 2min).
Per one of the diff comments on #1597:
> I used to use a tuple for constant things, but nowadays I'd actually
> prefer a list as a tuple is something more heterogeneous (i.e. it
> makes sense to have a `(x, y)` point as a tuple, but a list of points
> would be a list).
> At some point I should probably change it to a list everywhere 😉
This approach is not as weird in Python and still works.
DownloadTarget.OpenDownload has been renamed to OpenFileDownloadTarget,
since OpenDownloadDownloadTarget didn't look as nice.
This parameter replaces the filename and fileobj parameters. This makes
it easier to add more download targets, since only one may be "chosen".
With the OPEN_DOWNLOAD special case added, handling of filename got a
bit ugly, since it may be either None, OPEN_DOWNLOAD or a str with the
file path, and we had to make sure only one target was chosen.
With the new target enum, this handling can be simplified and we
automatically get the guarantee that only one target is chosen.
The test was failing because of two reasons:
First, the old code had filename questions in DownloadManager.get and
DownloadManager.fetch which were almost identical, thus the part in
DownloadManager.get was removed in an earlier commit. All filename
asking is now done by DownloadManager.fetch. The good part is code
deduplication, the bad part is slightly modified behavior: The new code
doesn't wait for a filename to start the download, instead it tries to
fill the buffer immediately. This made the test fail because qute:// has
no registered handler, so in order for the test to pass now, the "no
crash" part is not enough, we also need to expect the "No handler"
error.
Secondly, and a rather rare (race) condition was the handling of errors
in the DownloadItem. If an error occured after the registration of
self.on_reply_error as error handler and before the check
reply.error() != QNetworkReply.NoError
at the end of the function, the error signal would be emitted twice:
Once by _die() (called by on_reply_error), and once by the init_reply
function directly (in the last if block). This lead to duplicated error
messages. This is also explained in a comment in the file (with small
"stack traces").
This way, all temporary downloads will end up in the same directory and
everything is cleaned up at program exit, not when the corresponding
window is closed.
