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

fix: url datetime guessing

Browse Source
This commit is contained in:
Han Xiao
2025-03-07 14:32:47 +08:00
parent 8b836431af
commit 1e097a9ecc
2 changed files with 7 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
src/agent.ts
View File

@@ -693,9 +693,10 @@ You decided to think out of the box or cut from a completely different angle.
const r: SearchSnippet = {
title: link[0],
url: normalizeUrl(link[1]),
description: link[0]
description: link[0],
}
addToAllURLs(r, allURLs);
// in-page link has lower initial weight comparing to search links
addToAllURLs(r, allURLs, 0.1);
})
return {url, result: response};

8
src/utils/url-tools.ts
View File

@@ -223,7 +223,7 @@ export const rankURLs = (urlItems: SearchSnippet[], options: any = {}, trackers:
const {hostname, path} = extractUrlParts(item.url);
// Base weight from original
const freq = item.weight || 1.0; // Default to 1 if weight is missing
const freq = item.weight || 0; // Default to 1 if weight is missing
// Hostname boost (normalized by total URLs)
const hostnameFreq = normalizeCount(hostnameCount[hostname] || 0, totalUrls);
@@ -264,12 +264,12 @@ export const rankURLs = (urlItems: SearchSnippet[], options: any = {}, trackers:
}).sort((a, b) => b.finalScore - a.finalScore);
};
export const addToAllURLs = (r: SearchSnippet, allURLs: Record<string, SearchSnippet>) => {
export const addToAllURLs = (r: SearchSnippet, allURLs: Record<string, SearchSnippet>, weightDelta = 1) => {
if (!allURLs[r.url]) {
allURLs[r.url] = r;
allURLs[r.url].weight = 1;
allURLs[r.url].weight = weightDelta;
} else {
(allURLs[r.url].weight as number)++;
(allURLs[r.url].weight as number)+= weightDelta;
const curDesc = allURLs[r.url].description;
allURLs[r.url].description = smartMergeStrings(curDesc, r.description);
}