exostring/new_vk_parser
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat: ingest site videos through raw pipeline

Browse Source
This commit is contained in:
Your Name
2026-08-10 22:35:33 +05:00
parent 7889d4d376
commit faded0132b
9 changed files with 231 additions and 42 deletions
Download Patch File Download Diff File Expand all files Collapse all files
site_parser_worker/test_parser.py
+3 -1
View File
@@ -4,12 +4,14 @@ from app import parse_rss
def test_parse_rss() -> None:
xml = """<rss><channel><item><guid>1</guid><title>Title</title>
<link>https://example.test/1</link>
<description>&lt;p&gt;Body&lt;/p&gt;&lt;img src="https://example.test/1.jpg"&gt;</description>
<description>&lt;p&gt;Body&lt;/p&gt;&lt;img src="/1.jpg"&gt;
&lt;iframe src="https://www.youtube.com/embed/dQw4w9WgXcQ?feature=oembed"&gt;&lt;/iframe&gt;</description>
<pubDate>Mon, 10 Aug 2026 10:00:00 +0000</pubDate></item></channel></rss>"""
item = parse_rss(xml, 20)[0]
assert item.external_id == "1"
assert item.text == "Body"
assert item.media[0]["url"] == "https://example.test/1.jpg"
assert item.media[1] == {"type": "video", "url": "https://www.youtube.com/watch?v=dQw4w9WgXcQ"}
if __name__ == "__main__":