Before 5218e06904, deleting an item
loaded its state from the database again. Now it stores the state
of that object. markItemPlayed() did not reset the object's playback
position, so when auto-delete was enabled, the position was overwritten again.
GitHub switched their MacOS runners to ARM, which makes the Android emulator fail to start. Since we introduced the CI workflow, GitHub upgraded the Ubuntu runners as well, now supporting hardware acceleration. This means we no longer need MacOS. The Ubuntu runner is also about 2 times faster.
Till 5713b18267 many classes like FeedItem
used to inherit from FeedComponent which provided those two methods.
However since that commit the component no longer exists and now the
classes need to implement it on their own. Without this, ArrayList.remove breaks.
When the position saver ticks while the service is just about to be
stopped, it might happen that we first reset the position and then
set it to the end again. This works around this.
- Run Checkstyle with gradle to make it easier for users
- No longer needs different configuration for new code
- Exclude current violations
- Fix some violations that somehow couldn't be specified in the exclusion file
- Print SpotBugs/Lint/Checkstly violations in GitHub format
- Then the CI run gets annotated on the web UI
- When checking whether there is a subscription, there is no need to create feed objects (plus counters etc). Just the number of episodes is enough.
- Downloads section only needs to load the items it actually displays.
- No need to load FeedMedia, just to load FeedItem including the same FeedMedia afterwards.
- No need to convert columns to Strings and back to Longs.
- No need to join favorites when we are only interested in the list of IDs anyway.
AntennaPod is quite slow with huge feeds. The reason is that we have a bunch of workarounds for misbehaving feeds that also make it slower to work with feeds that do not misbehave.
Changes:
- Only start guessing duplicate episodes when no "proper" match is found
- Only parse non-html as HTML for attributes that really need it
- Do not log failed Long parsing when size is not specified
- Try to parse dates with RFC822 first before falling back to workarounds for other formats
I ran a benchmark with "Stuff you should know" (for which the workarounds are not needed) containing 2k episodes. Includes download of 8MB of feed XML (~5 seconds), debug build.
Before: 44 seconds, after: 13 seconds ==> 3.4 times faster feed refresh
ByteHamster
ByteHamster
flofriday
Fredrik Wallén
Taco
loucasal