Consider health scores in the search API

[jviotti]

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

[augmentcode]

🤖 Augment PR Summary

Summary: This PR updates the search index ranking so schema “health” scores influence result ordering.

Changes:

• Extended SearchEntry with a health field.
• Populated search-index entries with health values from explorer directory metadata.
• Updated make_search() sorting to rank by metadata completeness first, then by higher health, then by path.
• Expanded unit tests to cover health-driven ordering, tie-breaks, and interactions with metadata completeness.

Technical Notes: Health is used as a sort key during index construction; the serialized JSONL payload format returned by search() remains unchanged (path/title/description only).

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

[augmentcode]

Review completed. 2 suggestions posted.

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

[augmentcode]

SearchEntry::health is currently left uninitialized for default-constructed entries (e.g., SearchEntry e;), and make_search() reads it during sorting, which can be UB. Consider giving health a default initializer (e.g., std::uint8_t health{0};) so callers that don’t explicitly set it get deterministic behavior.

Severity: medium

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

[augmentcode]

When converting JSON health to std::uint8_t, out-of-range (or negative) integers will truncate/wrap, which could silently mis-rank results. Consider validating/clamping the integer value before casting so unexpected input can’t distort ordering.

Severity: medium

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

[cubic-dev-ai]

1 issue found across 4 files

Prompt for AI agents (unresolved issues)

Check if these issues are valid — if so, understand the root cause of each and fix them. If appropriate, use sub-agents to investigate and fix each issue separately.


<file name="src/search/include/sourcemeta/one/search.h">

<violation number="1" location="src/search/include/sourcemeta/one/search.h:25">
P1: `health` has no default member initializer, so a default-constructed `SearchEntry` leaves it indeterminate. Reading it in the `make_search()` comparator is undefined behavior. Initialize it: `std::uint8_t health{0};`.</violation>
</file>

_{Reply with feedback, questions, or to request a fix. Tag @cubic-dev-ai to re-run a review.}

[cubic-dev-ai]

P1: health has no default member initializer, so a default-constructed SearchEntry leaves it indeterminate. Reading it in the make_search() comparator is undefined behavior. Initialize it: std::uint8_t health{0};.

Prompt for AI agents

Check if this issue is valid — if so, understand the root cause and fix it. At src/search/include/sourcemeta/one/search.h, line 25:

<comment>`health` has no default member initializer, so a default-constructed `SearchEntry` leaves it indeterminate. Reading it in the `make_search()` comparator is undefined behavior. Initialize it: `std::uint8_t health{0};`.</comment>

<file context>
@@ -22,6 +22,7 @@ struct SearchEntry {
   std::string path;
   std::string title;
   std::string description;
+  std::uint8_t health;
 };
 
</file context>

Suggested change

	std::uint8_t health;
	std::uint8_t health{0};

[github-actions]

Benchmark Index (community)

Details

Benchmark suite	Current: 1b87016	Previous: dd5d964	Ratio
Add one schema (0 existing)	20 ms	25 ms	0.80
Add one schema (100 existing)	25 ms	26 ms	0.96
Add one schema (1000 existing)	83 ms	87 ms	0.95
Add one schema (10000 existing)	718 ms	904 ms	0.79
Update one schema (1 existing)	18 ms	20 ms	0.90
Update one schema (101 existing)	26 ms	27 ms	0.96
Update one schema (1001 existing)	85 ms	90 ms	0.94
Update one schema (10001 existing)	724 ms	766 ms	0.95
Cached rebuild (1 existing)	10 ms	11 ms	0.91
Cached rebuild (101 existing)	13 ms	13 ms	1
Cached rebuild (1001 existing)	37 ms	38 ms	0.97
Cached rebuild (10001 existing)	304 ms	311 ms	0.98
Index 100 schemas	131 ms	150 ms	0.87
Index 1000 schemas	1096 ms	1004 ms	1.09
Index 10000 schemas	13885 ms	14287 ms	0.97

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

[github-actions]

Benchmark Index (enterprise)

Details

Benchmark suite	Current: 1b87016	Previous: dd5d964	Ratio
Add one schema (0 existing)	21 ms	22 ms	0.95
Add one schema (100 existing)	26 ms	27 ms	0.96
Add one schema (1000 existing)	78 ms	86 ms	0.91
Add one schema (10000 existing)	667 ms	867 ms	0.77
Update one schema (1 existing)	20 ms	20 ms	1
Update one schema (101 existing)	26 ms	26 ms	1
Update one schema (1001 existing)	78 ms	87 ms	0.90
Update one schema (10001 existing)	663 ms	707 ms	0.94
Cached rebuild (1 existing)	11 ms	11 ms	1
Cached rebuild (101 existing)	13 ms	14 ms	0.93
Cached rebuild (1001 existing)	35 ms	38 ms	0.92
Cached rebuild (10001 existing)	271 ms	303 ms	0.89
Index 100 schemas	119 ms	131 ms	0.91
Index 1000 schemas	1098 ms	1120 ms	0.98
Index 10000 schemas	13774 ms	13941 ms	0.99

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