From a79fe26ade3a46ddf5649b2318e33ea95e15e0fa Mon Sep 17 00:00:00 2001
From: A Farzat <a@farzat.xyz>
Date: Wed, 8 Oct 2025 09:45:15 +0300
Subject: Add a module to get vid info through scraping

---
 components/extractor/obtain_vid_info.py | 13 +++++++++++++
 1 file changed, 13 insertions(+)
 create mode 100644 components/extractor/obtain_vid_info.py

(limited to 'components')

diff --git a/components/extractor/obtain_vid_info.py b/components/extractor/obtain_vid_info.py
new file mode 100644
index 0000000..bfbe861
--- /dev/null
+++ b/components/extractor/obtain_vid_info.py
@@ -0,0 +1,13 @@
+from urllib.request import urlopen
+
+from bs4 import BeautifulSoup
+from isodate import parse_duration # type: ignore
+
+def obtain_vid_duration(url: str, html: str = '') -> int:
+    html = html or urlopen(url).read().decode('utf-8')
+    soup = BeautifulSoup(html, 'html.parser')
+
+    duration_meta = soup.find('meta', itemprop='duration')
+    assert duration_meta
+    duration = parse_duration(duration_meta['content'])
+    return int(duration.total_seconds())
-- 
cgit v1.2.3-70-g09d2