Collect relation-level stats during compression

[erimatnor]

During compression, column min/max stats are collected on a per-segment basis for orderby columns and those that have indexes.

This change uses the same mechanism to collect relation-level min/max stats to be used by chunk skipping. This avoids, in worst case, an extra full table scan to gather these chunk column stats.

For simplicity, stats gathering is enabled for all columns that can support it, even though a column might use neither segment-level stats nor relation-level (chunk column) stats. The overhead of collecting min/max values should be negligible.

Disable-check: force-changelog-file

[codecov]

Codecov Report

Attention: Patch coverage is 89.36170% with 10 lines in your changes missing coverage. Please review.

Project coverage is 82.17%. Comparing base (59f50f2) to head (25d61ff).
Report is 641 commits behind head on main.

Files with missing lines	Patch %	Lines
src/ts_catalog/chunk_column_stats.c	66.66%	1 Missing and 3 partials ⚠️
tsl/src/compression/compression.c	87.50%	0 Missing and 4 partials ⚠️
tsl/src/compression/api.c	75.00%	0 Missing and 1 partial ⚠️
tsl/src/compression/segment_meta.c	97.82%	0 Missing and 1 partial ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #7520      +/-   ##
==========================================
+ Coverage   80.06%   82.17%   +2.11%     
==========================================
  Files         190      230      +40     
  Lines       37181    43183    +6002     
  Branches     9450    10854    +1404     
==========================================
+ Hits        29770    35487    +5717     
- Misses       2997     3369     +372     
+ Partials     4414     4327      -87     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.