An example of how to use this to make a C version of `seek-bzip-table` would be great

[hippietrail]

I've been using seek-bzip in its old C form and slightly more recent JavaScript form for decades.

I'm doing a new project and found this repo thinking it would be perfect to make a new fast seek-bzip-table but I find my skills too lacking to make use of this code directly in my own.

Here's a link to the seek-bzip version of the code that outputs a table that can be used to index a bz2 ready for random-access reading.

---

Oh it took me a long time to find it because this repo is so dense:

ibzip2 -L

[mxmlnkn]

Glad that ibzip2 -L works for you. So, this is exactly what you wanted? On reading your original question, I thought that you wanted to use some kind of C API to use it as a library. This would require some work, which I am not sure is worth it, because it is at heart a C++ library, not C.

I guess the ibzip2 -L use case could be added here and also the help output could be added to the repository for easier search.

[hippietrail]

Glad that ibzip2 -L works for you. So, this is exactly what you wanted? On reading your original question, I thought that you wanted to use some kind of C API to use it as a library. This would require some work, which I am not sure is worth it, because it is at heart a C++ library, not C.

The end goal will be to use it from Rust. So for any FFI C is the interface language.
But I'm not asking your to make a C interface, I'm trying to figure that out myself.

I guess the ibzip2 -L use case could be added here and also the help output could be added to the repository for easier search.

The two things I need are of course, building an index and decompressing individual blocks. I just expected the former to be the easy part.

Ideally I'd love to decompress and index in a single pass, streaming both outputs to different places. Would that be doable with indexed_bzip2?

[mxmlnkn]

Ideally I'd love to decompress and index in a single pass, streaming both outputs to different places. Would that be doable with indexed_bzip2?

Yes, ibzip2 -L file.index -d -o file file.bz2.

[hippietrail]

Ideally I'd love to decompress and index in a single pass, streaming both outputs to different places. Would that be doable with indexed_bzip2?

Yes, ibzip2 -L file.index -d -o file file.bz2.

I'm far from a wizard at *nix shell. Is it already possible to pipe both outputs rather than save them to files?
Also, I'd really like a streaming equivalent to -L that output the offsets as they are found rather than after completion. I expect that would be a trivial change?