chunk index cache: use cache/chunks.<HASH>, see #8503

[ThomasWaldmann]

this should fix the cache coherency issue when running (and finishing normally) multiple backups in parallel. if another backup then starts at any time, it will merge all already cached chunk indexes before beginning to process files.

Improvements:

• doesn't need a separate file for the hash
• multiple borg running in parallel don't overwrite the other's chunk indexes
• when loading the chunk index from the cache, just merge all we have there
• we can later write multiple partial chunkindexes to the cache

[codecov[bot]]

Codecov Report

Attention: Patch coverage is 86.41975% with 11 lines in your changes missing coverage. Please review.

Project coverage is 81.80%. Comparing base (38ba3d8) to head (56493fc). Report is 9 commits behind head on master.

Files with missing lines	Patch %	Lines
src/borg/cache.py	85.52%	7 Missing and 4 partials :warning:

Additional details and impacted files

```diff @@ Coverage Diff @@ ## master #8531 +/- ## ========================================== + Coverage 81.75% 81.80% +0.04% ========================================== Files 74 74 Lines 13240 13281 +41 Branches 1941 1955 +14 ========================================== + Hits 10824 10864 +40 - Misses 1753 1754 +1 Partials 663 663 ```

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.