78f4fd9d3f
- This fixes the issue when one NAL unit has multiple subsamples, which is allowed in spec and could happen when the clear_bytes exceeds 2^16 bytes. - The new algorithm updates the subsample entries in following steps: 1) When NaluReader parses a new NAL unit, the function splits subsamples and align them with the NAL unit. It also finds the range of subsamples for that NAL unit. 2) Drop or update subsamples of that NAL unit, based on the type. Other inserted bytes are treated as a new all-clear subsample. 3) Merge those subsamples if possible. - This addresses #237. Change-Id: I2cab47cd00d90e29ec52b90d4fde057c632193ac |
||
|---|---|---|
| .. | ||
| base | ||
| chunking | ||
| codecs | ||
| crypto | ||
| demuxer | ||
| event | ||
| file | ||
| formats | ||
| test | ||
| trick_play | ||