Automatically ignore configuration files in collection paths

[jviotti]

Signed-off-by: Juan Cruz Viotti jv@jviotti.com

[github-actions]

Benchmark Index (community)

Details

Benchmark suite	Current: a0472d1	Previous: 947760f	Ratio
Add one schema (0 existing)	21 ms	21 ms	1
Add one schema (100 existing)	26 ms	29 ms	0.90
Add one schema (1000 existing)	88 ms	74 ms	1.19
Add one schema (10000 existing)	725 ms	622 ms	1.17
Update one schema (1 existing)	18 ms	18 ms	1
Update one schema (101 existing)	33 ms	25 ms	1.32
Update one schema (1001 existing)	90 ms	76 ms	1.18
Update one schema (10001 existing)	746 ms	643 ms	1.16
Cached rebuild (1 existing)	11 ms	10 ms	1.10
Cached rebuild (101 existing)	13 ms	12 ms	1.08
Cached rebuild (1001 existing)	38 ms	27 ms	1.41
Cached rebuild (10001 existing)	298 ms	194 ms	1.54
Index 100 schemas	125 ms	147 ms	0.85
Index 1000 schemas	1013 ms	1175 ms	0.86
Index 10000 schemas	14105 ms	14868 ms	0.95

This comment was automatically generated by workflow using github-action-benchmark.

[github-actions]

Benchmark Index (enterprise)

Details

Benchmark suite	Current: a0472d1	Previous: 947760f	Ratio
Add one schema (0 existing)	22 ms	22 ms	1
Add one schema (100 existing)	28 ms	27 ms	1.04
Add one schema (1000 existing)	83 ms	75 ms	1.11
Add one schema (10000 existing)	697 ms	793 ms	0.88
Update one schema (1 existing)	22 ms	21 ms	1.05
Update one schema (101 existing)	28 ms	26 ms	1.08
Update one schema (1001 existing)	88 ms	77 ms	1.14
Update one schema (10001 existing)	668 ms	608 ms	1.10
Cached rebuild (1 existing)	12 ms	12 ms	1
Cached rebuild (101 existing)	14 ms	13 ms	1.08
Cached rebuild (1001 existing)	36 ms	28 ms	1.29
Cached rebuild (10001 existing)	276 ms	198 ms	1.39
Index 100 schemas	143 ms	124 ms	1.15
Index 1000 schemas	1014 ms	1048 ms	0.97
Index 10000 schemas	14656 ms	14118 ms	1.04

This comment was automatically generated by workflow using github-action-benchmark.

[augmentcode]

🤖 Augment PR Summary

Summary: This PR prevents JSON/YAML configuration files from being mistakenly treated as schemas when a collection path points at (or includes) the configuration directory.

Changes:

• Extended Configuration::read with an overload that collects all configuration-related file paths (root config + extends/include chain).
• Tracked referenced configuration files during dereferencing and returned the parsed configuration as before.
• Updated the indexer to pass a set to Configuration::read and skip those files during recursive collection scanning.
• Added CLI tests to ensure collection paths like "." and "" no longer fail due to self-detection.
• Added a regression test ensuring extended/included config JSON files are skipped while real schema files are still indexed.

Technical Notes: Uses std::filesystem::weakly_canonical(...).native() strings for path identity across the read + index phases.

_{🤖 Was this summary useful? React with 👍 or 👎}

[augmentcode]

Review completed. 1 suggestion posted.

Comment augment review to trigger a new review at any time.

[augmentcode]

std::filesystem::weakly_canonical(entry.path()) can throw (e.g., permission issues or odd filesystem states) and abort the whole indexing run; consider using an error_code overload or otherwise handling failures so a single problematic file doesn’t stop directory scanning.

Severity: medium

_{🤖 Was this useful? React with 👍 or 👎, or 🚀 if it prevented an incident/outage.}

[cubic-dev-ai]

No issues found across 9 files