github/node-DeepResearch
1
0
Fork 0
mirror of https://github.com/jina-ai/node-DeepResearch.git synced 2026-03-22 07:29:35 +08:00
Code Issues Packages Projects Releases Wiki Activity

feat: improve url ranking, fix eval bugs

Browse Source
This commit is contained in:
Han Xiao
2025-03-06 15:24:03 +08:00
parent d9bfc2fd1f
commit 7bb5b18f2e
2 changed files with 10 additions and 11 deletions
Download Patch File Download Diff File Expand all files Collapse all files

10
src/agent.ts
View File

@@ -554,14 +554,10 @@ But then you realized you have asked them before. You decided to to think out of
try { try {
let siteQuery = query let siteQuery = query
if (allURLs && Object.keys(allURLs).length > 0) {
// rerank urls
weightedURLs = rankURLs(getUnvisitedURLs(allURLs, visitedURLs), {
question: currentQuestion
}, context);
}
const topHosts = Object.entries(countUrlParts(weightedURLs).hostnameCount).sort((a, b) => b[1] - a[1]); const topHosts = Object.entries(countUrlParts(
Object.entries(allURLs).map(([, result]) => result)
).hostnameCount).sort((a, b) => b[1] - a[1]);
console.log(topHosts) console.log(topHosts)
if (topHosts.length > 0 && Math.random() < 0.6 && !query.includes('site:')) { if (topHosts.length > 0 && Math.random() < 0.6 && !query.includes('site:')) {
// explore-exploit // explore-exploit

11
src/utils/text-tools.ts
View File

@@ -2,8 +2,8 @@ import {AnswerAction} from "../types";
import i18nJSON from './i18n.json'; import i18nJSON from './i18n.json';
export function buildMdFromAnswer(answer: AnswerAction) { export function buildMdFromAnswer(answer: AnswerAction) {
// Standard footnote regex - updated to handle both [^1] and [1^] formats // Standard footnote regex - updated to handle [^1], [1^], and [1] formats
const footnoteRegex = /\[(\^(\d+)|(\d+)\^)]/g; const footnoteRegex = /\[(\^(\d+)|(\d+)\^|(\d+))]/g;
// New regex to catch grouped footnotes like [^1, ^2, ^3] or [^1,^2,^3] // New regex to catch grouped footnotes like [^1, ^2, ^3] or [^1,^2,^3]
const groupedFootnoteRegex = /\[\^(\d+)(?:,\s*\^(\d+))+]/g; const groupedFootnoteRegex = /\[\^(\d+)(?:,\s*\^(\d+))+]/g;
@@ -35,8 +35,10 @@ export function buildMdFromAnswer(answer: AnswerAction) {
.replace(footnoteRegex, ''); .replace(footnoteRegex, '');
} }
// Normalize footnotes first (convert [1^] to [^1] format) // Normalize footnotes first (convert [1^] to [^1] format and [1] to [^1] format)
let processedAnswer = answer.answer.replace(/\[(\d+)\^]/g, (_, num) => `[^${num}]`); let processedAnswer = answer.answer
.replace(/\[(\d+)\^]/g, (_, num) => `[^${num}]`)
.replace(/\[(\d+)]/g, (_, num) => `[^${num}]`);
// Fix grouped footnotes // Fix grouped footnotes
processedAnswer = processedAnswer.replace(groupedFootnoteRegex, (match) => { processedAnswer = processedAnswer.replace(groupedFootnoteRegex, (match) => {
@@ -133,6 +135,7 @@ ${correctedAnswer}
${formatReferences(answer.references)} ${formatReferences(answer.references)}
`.trim(); `.trim();
} }
export const removeExtraLineBreaks = (text: string) => { export const removeExtraLineBreaks = (text: string) => {
return text.replace(/\n{2,}/gm, '\n\n'); return text.replace(/\n{2,}/gm, '\n\n');
} }