Skip to content

perf: hoist uniqueQueryTokens out of per-result loop in rankAnswerEvidence #77

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
BigSimmo merged 1 commit into from
Jun 27, 2026
Merged

perf: hoist uniqueQueryTokens out of per-result loop in rankAnswerEvidence #77

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
19 changes: 9 additions & 10 deletions src/lib/answer-ranking.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -105,13 +105,12 @@ function tokenCoverage(tokens: string[], haystack: string) {
return hits.length / tokens.length;
}

function phraseScore(query: string, haystack: string) {
const queryTokens = uniqueQueryTokens(query);
if (queryTokens.length < 2 || !haystack) return 0;
function phraseScore(tokens: string[], haystack: string) {
if (tokens.length < 2 || !haystack) return 0;
const phrases: string[] = [];
for (let size = Math.min(4, queryTokens.length); size >= 2; size -= 1) {
for (let index = 0; index <= queryTokens.length - size; index += 1) {
phrases.push(queryTokens.slice(index, index + size).join(" "));
for (let size = Math.min(4, tokens.length); size >= 2; size -= 1) {
for (let index = 0; index <= tokens.length - size; index += 1) {
phrases.push(tokens.slice(index, index + size).join(" "));
}
}
const hits = phrases.filter((phrase) => haystack.includes(phrase)).length;
Expand Down Expand Up @@ -168,17 +167,16 @@ function sourceQualityScore(result: SearchResult) {
return score;
}

function answerEvidenceScore(query: string, result: SearchResult, queryClass: RagQueryClass) {
function answerEvidenceScore(tokens: string[], result: SearchResult, queryClass: RagQueryClass) {
const texts = resultTexts(result);
const tokens = uniqueQueryTokens(query);
const base = Math.min(1, result.hybrid_score ?? result.similarity ?? 0) * 0.28;
const contentCoverage = tokenCoverage(tokens, texts.content);
const titleCoverage = tokenCoverage(tokens, texts.title);
const sectionCoverage = tokenCoverage(tokens, texts.section);
const metadataCoverage = tokenCoverage(tokens, texts.metadata);
const combinedCoverage = tokenCoverage(tokens, texts.combined);
const adjacentCoverage = tokenCoverage(tokens, texts.adjacent);
const phraseCoverage = phraseScore(query, texts.combined);
const phraseCoverage = phraseScore(tokens, texts.combined);
const directnessScore =
contentCoverage * 0.34 + titleCoverage * 0.12 + sectionCoverage * 0.1 + metadataCoverage * 0.08;
const weakOverlapPenalty = combinedCoverage < 0.2 ? -0.18 : combinedCoverage < 0.34 ? -0.07 : 0;
Expand Down Expand Up @@ -243,11 +241,12 @@ export function rankAnswerEvidence(
results: SearchResult[],
queryClass: RagQueryClass = classifyRagQuery(query).queryClass,
) {
const tokens = uniqueQueryTokens(query);
const ranked = results
.map((result, index) => ({
result,
index,
score: answerEvidenceScore(query, result, queryClass),
score: answerEvidenceScore(tokens, result, queryClass),
}))
.sort(
(a, b) =>
Expand Down