dcpomatic/test, branch debug-corruption

Hopefully fix occasional hang in j2k_encoder_deadlock_test.

2024-09-12T23:14:21Z

Previously too many frames were queued for encoding, which AFAICS meant that (if the CPU was busy) we would get to the point where too many frames were in the encoder queue, so that we blocked waiting for it to clear, and then simultaneously too many frames were in the writer queue, which (in this test) would never clear. At this point we would be backed up waiting for Writer::write() to happen in J2KEncoder::encoder_thread() so that the encoder queue could be cleared, but nobody is calling Writer::write().

Bump sleep a little to try and de-flake a test.

2024-09-12T20:49:48Z

Fix error when importing bad subtitle file (#2838).

2024-06-27T18:14:31Z

The subtitle XML refers to a font with no corresponding .

Fix font ID allocation from DCP when there are subs and ccaps using the same IDs (#2822).

2024-06-07T19:09:03Z

Previously we would always use _id as a disambiguated ID for every case.

Fix stream length for DCP content (#2688).

2024-05-21T07:38:03Z

Make sure we don't say some content has audio if none will be emitted.

2024-05-17T13:02:57Z

In the FFmpeg (and shortly DCP) case we don't emit any audio if none is mapped, so we must take this into account in the player.

Work around deadlock when destroying J2KEncoder with a full writer queue (#2784).

2024-05-07T23:33:41Z

This feels like a hack, but I can't think of a nicer way to do it. The interruption disable makes sense because when we destroy encoder threads during a DCP encode (because a remote server goes away, for example) we don't want any frames to be lost due to the encode thread being interrupted between taking the frame off the queue and sending it to the writer. When we're destroying the encoder we don't care about this, but I can't see how you'd differentiate. Maybe the encoder queue could have two lists: to-do and in-progress; the encoder thread atomically moves a frame from to-do to in-progress, but then how do you know when the in-progress ones are orphaned and need to be re-added to the main queue. You could make the writer return saying "no" if the queue is full (rather than blocking and waiting for the queue to empty) but that seems wasteful as then the frame would be re-encoded.

Fix hanging/overlapping dvb subtitles (#2792).

2024-03-31T23:04:43Z

This reverts a change made in 8ca6fd6d97e6d42492afddb655fa85130946853c "Fix doubled subtitles if subtitle stop times are specified." That change breaks the case where a subtitle _does_ have a stop time, but it's wrong (30s from the start time) and we want the next subtitle to clear the previous one. I can't now see how reverting this could cause doubled subtitles, so maybe that problem wlil come back. At least now there's a test for #2792.

Fix failing tests.

2024-03-20T06:41:24Z

Broken in the fix for #2780, when we started to check that referenced audio MXFs have the correct number of channels.

Fix screen name order in KDM emails.

2024-03-07T22:33:03Z