Feature request: Add support for ALP (Adaptive Lossless floating-Point Compression)

**Describe the bug**

Sorry. This is not a bug. This is a feature request.

ALP (Adaptive Lossless floating-Point Compression) may be a good codec for Blosc. It's a lossless floating-point compression algorithm for double/float. It seems that it's a high speed and high compression algorithm:

* Implementation: https://github.com/cwida/ALP
* Paper: https://dl.acm.org/doi/pdf/10.1145/3626717
* DuckDB integration: https://duckdb.org/2024/02/13/announcing-duckdb-0100.html#adaptive-lossless-floating-point-compression-alp

Are you interesting in this algorithm?

If we add support for ALP, we may do it as a codec like ZFP not a filter.

**To Reproduce**

Not available.

**Expected behavior**

Not available.

**Logs**

Not available.

**System information:**
- OS: Not available
- Compiler: Not available
- Version: Not available

**Additional context**

Not available.

No source file or test is named. Start by reviewing the linked ALP implementation and paper, then compare how the existing ZFP codec is integrated; done means c-blosc2 supports ALP as a codec for double and float values rather than as a filter.

Written by the indexing model from the issue text.