releases/mcap-cli/v0.0.36

Include metadata in mcap merge (#958)

Includes metadata records from input files in mcap merge via a new read
option. This required a breaking change to read options to avoid a
dependency cycle: since I need to supply a callback option to apply to
metadata records, the readopts package required awareness of "mcap"
while "mcap" required awareness of readopts for configuration.

To address this I have moved readopts.go under the mcap package. Users
who upgrade the library will need to swap out the package name if they
are using any options.

releases/rust/v0.9.0

Update the Crate to be wasm compatible.

releases/mcap-cli/v0.0.35

Add progress bar for `mcap convert` (#973)

### Public-Facing Changes

This adds a progress bar to show percentage completion when running `mcap convert ...`

releases/mcap-cli/v0.0.34

CLI: Properly support --name parameter on add attachment (#953)

Previously this parameter did not function. Not sure if it never
functioned or if we lost the support at some point.

releases/mcap-cli/v0.0.33

Go: Allocate less in indexed message chunk loading (#950)

Prior to this commit, when reading a chunk into memory we would read it
(internally, in io.ReadAll) into a resizing buffer and need to resize it
several times up to final size. Since we know the final size in advance
we can preallocate and save those allocations/copies.

releases/rust/v0.7.1

Update the Writer tracking of message bounds to include log_time 0 as a valid time for bounds.

releases/mcap-cli/v0.0.32

mcap filter: don't die on schema id 0 (#921)

### Public-Facing Changes

CLI commands `filter`, `compress`, `decompress`, and `recover` no longer fail on schemaless channels (schema id = 0).

### Description

Don't fail when encountering schema ID 0.

Fixes FG-4182

See also:
- https://github.com/foxglove/mcap/issues/829
- https://github.com/foxglove/mcap/issues/847
- https://github.com/foxglove/mcap/issues/579

releases/mcap-cli/v0.0.31

Fix mcap conversion for bzip2 bags (#890)

* Fix mcap conversion for bzip2 bags

Prior to this commit, the bzip2 decompression was broken on bags with
more than one chunk. The chunk reader was getting set to a basic reader
on the underlying chunk data, rather than a bz2 decompressor (on chunks
after the first).

This fixes that and also does a minor refactor to remove the concept of
a resettable reader. This was only applicable to the lz4 decompressor
previously (out of three - lz4, none, and bz2) and existed to fake
generic support for resetting readers when underlying readers didn't
support that. Rather than doing that, special case lz4 and use the other
two reader types directly.

* Add test with chunked bz2 bag

releases/mcap-cli/v0.0.30

go/mcap: do not write empty message indexes (#873)

### Public-Facing Changes

Go MCAP writer no longer produces message index records for channels not present in a chunk. This helps indexed readers skip chunks that do not contain any messages on a given channel.

### Description
Fixes #862

releases/mcap-cli/v0.0.29

CLI: info: handle durations > 2**63 (#844)

* CLI: info: handle durations > 2**63

* make a function

* correct comment