Some fixes for non-ASCII chars in MPD uris #1805
Conversation
Codecov Report
@@ Coverage Diff @@
## release-2.2 #1805 +/- ##
==============================================
+ Coverage 80.99% 81% +<.01%
==============================================
Files 83 83
Lines 6647 6649 +2
==============================================
+ Hits 5384 5386 +2
Misses 1263 1263
Continue to review full report at Codecov.
|
| @@ -49,6 +49,7 @@ def get_distinct(self, field, query=None): | |||
| return self.dummy_get_distinct_result.get(field, set()) | |||
|
|
|||
| def lookup(self, uri): | |||
| uri = Ref.track(uri=uri).uri | |||
There was a problem hiding this comment.
This looked like a no-op to me at first, but this actually encodes the URI as bytes!
There was a problem hiding this comment.
Yes! It's so tempting to just compare two stringy things called uri but you can't.
| @@ -22,6 +22,7 @@ def __init__(self, *args, **kwargs): | |||
| self.response = [] | |||
|
|
|||
| def queue_send(self, data): | |||
| data = data.decode('utf-8') | |||
There was a problem hiding this comment.
data is one small step away from being sent over the network, so it wouldn't be unnatural to keep it as bytes. If we change this to the following, it should work correctly on both Python 2 and 3:
if isinstance(data, compat.text_type):
data = data.encode('utf-8')
lines = (line for line in data.split(b'\n') if line) # Note the explicit 'b' here
...There was a problem hiding this comment.
OK, sounds sensible. Will update.
There was a problem hiding this comment.
Hmm but actually, no. This isn't going over the network and don't we want unicode here so that we can just write unicode in the tests?
|
I've just realised I missed the other URI at https://github.com/mopidy/mopidy/blob/develop/mopidy/mpd/translator.py#L113 |
Related to mopidy#1805.
Related to mopidy#1805.
Fixes #1759
Most of this comes from Track
uris beingbytesbut falling over when formatted withunicodestrings for output / logging. And then I went a bit mad sprinkling non-ASCII throughout the tests.