fix(search): down-weight the project name in ranking — completes #720 (#748)

The per-word path fix (#745) brought the backend to parity but not above:
the project name still gave the lexically-matching stack a residual dir
match + an FTS class-name match, so a backend query that included the
project name still ranked the frontend at/above the backend.

Derive the project name from go.mod module / package.json name / repo dir,
and treat a query word matching it as non-discriminative: drop it from path
relevance and from codegraph_explore's PascalCase type-disambiguation bias
(reporter's suggestions #1/#2) — unless it's the only query word, so a bare
project-name search still scores.

Narrow by construction: the down-weighting fires ONLY when a query word
matches the derived project name (≥5 chars), so every query that doesn't
name the project is byte-identical. On the reporter's repro the backend
controllers now top a backend question that includes the project name;
queries without it, bare project-name queries, and normal symbol queries
are unchanged. Query-time only (no re-index).

Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This commit is contained in:
Colby Mchenry
2026-06-08 22:43:54 -04:00
committed by GitHub
co-authored by Claude Opus 4.8
parent afec1282e1
commit 75ae1e8bd9
6 changed files with 154 additions and 10 deletions
+42 -1
View File
@@ -16,7 +16,7 @@ import * as path from 'path';
import * as os from 'os';
import CodeGraph from '../src/index';
import { LOW_CONFIDENCE_MARKER } from '../src/context';
import { isDistinctiveIdentifier, scorePathRelevance } from '../src/search/query-utils';
import { isDistinctiveIdentifier, scorePathRelevance, deriveProjectNameTokens } from '../src/search/query-utils';
describe('isDistinctiveIdentifier', () => {
it('treats plain dictionary words as non-distinctive', () => {
@@ -64,6 +64,47 @@ describe('scorePathRelevance per-word scoring (#720)', () => {
});
});
// The project name is context, not a discriminator: dropping it from path
// scoring stops every file under a `<ProjectName>…/` tree from winning on the
// name alone, so the rest of the query decides the ranking (#720).
describe('project-name down-weighting in path relevance (#720)', () => {
it('derives the project name from go.mod / package.json, skipping short names', () => {
const dir = fs.mkdtempSync(path.join(os.tmpdir(), 'codegraph-projname-'));
try {
fs.writeFileSync(path.join(dir, 'go.mod'), 'module example.com/SuperBizAgent\n\ngo 1.21\n');
fs.writeFileSync(path.join(dir, 'package.json'), JSON.stringify({ name: '@acme/superbizagent-web' }));
const tokens = deriveProjectNameTokens(dir);
expect(tokens.has('superbizagent')).toBe(true);
expect(tokens.has('superbizagentweb')).toBe(true);
} finally {
fs.rmSync(dir, { recursive: true, force: true });
}
});
it('drops a project-name query word from path scoring when other words remain', () => {
const proj = new Set(['superbizagent']);
// Without the project name dropped, the frontend path wins on it (+5).
// With it dropped, only "backend" is left — and it doesn't match this path.
const withDrop = scorePathRelevance('SuperBizAgentFrontend/app.js', 'SuperBizAgent backend', proj);
const noDrop = scorePathRelevance('SuperBizAgentFrontend/app.js', 'SuperBizAgent backend');
expect(withDrop).toBeLessThan(noDrop);
expect(withDrop).toBe(0);
});
it('keeps the project-name word when it is the ONLY query word (bare query still scores)', () => {
const proj = new Set(['superbizagent']);
expect(scorePathRelevance('SuperBizAgentFrontend/app.js', 'SuperBizAgent', proj)).toBe(5);
});
it('does not affect a query that omits the project name', () => {
const proj = new Set(['superbizagent']);
const path0 = 'internal/controller/chat/chat.go';
expect(scorePathRelevance(path0, 'controller chat', proj)).toBe(
scorePathRelevance(path0, 'controller chat')
);
});
});
describe('Context ranking — common-word precision & confidence', () => {
let testDir: string;
let cg: CodeGraph;