 |
Chunk consists of its name, data-size and data itself. But there is a header which isn't a chunk. It's just 12 bytes: "RIFF", <size-of-the-following-data>, "WAVE".
So there is an explicit way to parse the chunk-organized file:
- 1. Read the header to verify that file size equals to a value of an integer+8, consisted by bytes from #5 to #8. If false then file is corrupted. Quit parsing. Else continue.
- 2. Go to the next chunk (first starts from byte #13; following chunks' location found using chunk's data-size value).
- 3. Read chunk's name, data-size and data.
- 4. Parse chunk's data if needed.
- 5. Goto point #2 if not EOF else finish parsing.
Now, there are plenty of chunks: bext, fmt, data, fact, wavl, slnt, cue, plst, list, labl, note, ltxt, smpl. And it's not my goal to describe their content (which can easily be googled). I just wanted to demonstrate how to walk through the chunk-based structure.
Thanks for stopping by.
 |
