GitHub-Proxy/jina_reader
1
0
Fork 0
mirror of https://git.mirrors.martin98.com/https://github.com/jina-ai/reader.git synced 2025-08-19 04:55:54 +08:00
Code Issues Packages Projects Releases Wiki Activity

fix: img with srcset only

Browse Source
This commit is contained in:
Yanlong Wang 2024-11-18 16:37:42 +08:00
parent 6fa8ce309e
commit 1f4620deef
No known key found for this signature in database
GPG Key ID: C0A623C0BADF9F37
1 changed files with 4 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
backend/functions/src/services/snapshot-formatter.ts
View File

@ -230,6 +230,10 @@ export class SnapshotFormatter extends AsyncService {
return alt ? `(Image ${++imgIdx}: ${alt})` : '';
}
let linkPreferredSrc = (node.getAttribute('src') || '').trim();
const maybeSrcSet: string = (node.getAttribute('srcset') || '').trim();
if (!linkPreferredSrc && maybeSrcSet) {
linkPreferredSrc = maybeSrcSet.split(',').map((x) => x.trim()).filter(Boolean)[0];
}
if (!linkPreferredSrc || linkPreferredSrc.startsWith('data:')) {
const dataSrc = (node.getAttribute('data-src') || '').trim();
if (dataSrc && !dataSrc.startsWith('data:')) {