Hermes Agent
201 lines
7.2 KiB
Python
201 lines
7.2 KiB
Python
#!/usr/bin/env python3
|
|
"""Scrape YouTube trending videos for Germany using InnerTube API with curl_cffi."""
|
|
|
|
import json
|
|
import re
|
|
import sys
|
|
from curl_cffi import requests
|
|
|
|
def get_api_key(session):
|
|
"""Extract the InnerTube API key from YouTube's homepage."""
|
|
r = session.get("https://www.youtube.com/", impersonate="chrome120")
|
|
if r.status_code != 200:
|
|
print(f"ERROR: Got status {r.status_code} from YouTube homepage", file=sys.stderr)
|
|
return None
|
|
match = re.search(r'"INNERTUBE_API_KEY"\s*:\s*"([^"]+)"', r.text)
|
|
if match:
|
|
return match.group(1)
|
|
# Try alternate pattern
|
|
match = re.search(r'"innertubeApiKey"\s*:\s*"([^"]+)"', r.text)
|
|
if match:
|
|
return match.group(1)
|
|
return None
|
|
|
|
def get_trending_videos(session, api_key):
|
|
"""Fetch trending videos via InnerTube API."""
|
|
url = f"https://www.youtube.com/youtubei/v1/browse?key={api_key}"
|
|
headers = {
|
|
"Content-Type": "application/json",
|
|
"Accept-Language": "de-DE,de;q=0.9",
|
|
"X-YouTube-Client-Name": "1",
|
|
"X-YouTube-Client-Version": "2.20260609.07.00",
|
|
"Origin": "https://www.youtube.com",
|
|
"Referer": "https://www.youtube.com/feed/trending"
|
|
}
|
|
body = {
|
|
"context": {
|
|
"client": {
|
|
"hl": "de",
|
|
"gl": "DE",
|
|
"clientName": "WEB",
|
|
"clientVersion": "2.20260609.07.00",
|
|
"utcOffsetMinutes": 120
|
|
}
|
|
},
|
|
"browseId": "FEtrending"
|
|
}
|
|
|
|
r = session.post(url, headers=headers, json=body, impersonate="chrome120")
|
|
if r.status_code != 200:
|
|
print(f"ERROR: Got status {r.status_code} from trending API", file=sys.stderr)
|
|
print(f"Response: {r.text[:500]}", file=sys.stderr)
|
|
return None
|
|
return r.json()
|
|
|
|
def extract_videos(data):
|
|
"""Extract video information from the browse response."""
|
|
videos = []
|
|
|
|
def find_video_renderers(obj):
|
|
if isinstance(obj, dict):
|
|
if 'videoRenderer' in obj:
|
|
vr = obj['videoRenderer']
|
|
video_id = vr.get('videoId', 'N/A')
|
|
title_runs = vr.get('title', {}).get('runs', [])
|
|
title = title_runs[0].get('text', 'N/A') if title_runs else 'N/A'
|
|
|
|
owner_runs = vr.get('ownerText', {}).get('runs', [])
|
|
channel = owner_runs[0].get('text', 'N/A') if owner_runs else 'N/A'
|
|
|
|
views_text = vr.get('viewCountText', {})
|
|
views = views_text.get('simpleText', 'N/A')
|
|
if views == 'N/A':
|
|
views_runs = views_text.get('runs', [])
|
|
views = ''.join(r.get('text', '') for r in views_runs) if views_runs else 'N/A'
|
|
|
|
desc_snippet = vr.get('descriptionSnippet', {})
|
|
desc_runs = desc_snippet.get('runs', [])
|
|
desc = ''.join(r.get('text', '') for r in desc_runs) if desc_runs else ''
|
|
|
|
videos.append({
|
|
'videoId': video_id,
|
|
'title': title,
|
|
'channel': channel,
|
|
'views': views,
|
|
'description_snippet': desc
|
|
})
|
|
|
|
for key, value in obj.items():
|
|
if key != 'videoRenderer':
|
|
find_video_renderers(value)
|
|
|
|
elif isinstance(obj, list):
|
|
for item in obj:
|
|
find_video_renderers(item)
|
|
|
|
find_video_renderers(data)
|
|
return videos
|
|
|
|
def get_video_description(session, api_key, video_id):
|
|
"""Fetch full video description via InnerTube player API."""
|
|
url = f"https://www.youtube.com/youtubei/v1/player?key={api_key}"
|
|
headers = {
|
|
"Content-Type": "application/json",
|
|
"Accept-Language": "de-DE,de;q=0.9",
|
|
"X-YouTube-Client-Name": "1",
|
|
"X-YouTube-Client-Version": "2.20260609.07.00",
|
|
"Origin": "https://www.youtube.com",
|
|
"Referer": f"https://www.youtube.com/watch?v={video_id}"
|
|
}
|
|
body = {
|
|
"context": {
|
|
"client": {
|
|
"hl": "de",
|
|
"gl": "DE",
|
|
"clientName": "WEB",
|
|
"clientVersion": "2.20260609.07.00",
|
|
"utcOffsetMinutes": 120
|
|
}
|
|
},
|
|
"videoId": video_id,
|
|
"racyCheckOk": True
|
|
}
|
|
|
|
r = session.post(url, headers=headers, json=body, impersonate="chrome120")
|
|
if r.status_code != 200:
|
|
print(f" ERROR: Got status {r.status_code} for video {video_id}", file=sys.stderr)
|
|
return None
|
|
|
|
try:
|
|
data = r.json()
|
|
except:
|
|
return None
|
|
|
|
# Extract description from videoDetails
|
|
video_details = data.get('videoDetails', {})
|
|
short_desc = video_details.get('shortDescription', '')
|
|
|
|
# Also try microformat
|
|
microformat = data.get('microformat', {}).get('playerMicroformatRenderer', {})
|
|
full_desc = microformat.get('description', {}).get('simpleText', '')
|
|
if not full_desc:
|
|
desc_runs = microformat.get('description', {}).get('runs', [])
|
|
full_desc = ''.join(r.get('text', '') for r in desc_runs)
|
|
|
|
return full_desc or short_desc
|
|
|
|
def main():
|
|
session = requests.Session()
|
|
session.headers.update({
|
|
"User-Agent": "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36"
|
|
})
|
|
|
|
print("Step 1: Getting API key...", file=sys.stderr)
|
|
api_key = get_api_key(session)
|
|
if not api_key:
|
|
print("ERROR: Could not get API key", file=sys.stderr)
|
|
sys.exit(1)
|
|
print(f"Got API key: {api_key[:15]}...", file=sys.stderr)
|
|
|
|
print("Step 2: Fetching trending videos...", file=sys.stderr)
|
|
data = get_trending_videos(session, api_key)
|
|
if not data:
|
|
print("ERROR: Could not fetch trending data", file=sys.stderr)
|
|
sys.exit(1)
|
|
|
|
videos = extract_videos(data)
|
|
print(f"Found {len(videos)} videos", file=sys.stderr)
|
|
|
|
# Print all found videos for debugging
|
|
for i, v in enumerate(videos[:10]):
|
|
print(f" {i+1}. [{v['videoId']}] {v['title'][:60]} - {v['channel']} ({v['views']})", file=sys.stderr)
|
|
|
|
top5 = videos[:5]
|
|
|
|
print("\nStep 3: Fetching full descriptions...", file=sys.stderr)
|
|
results = []
|
|
for i, v in enumerate(top5):
|
|
print(f" Fetching description for video {i+1}: {v['title'][:50]}...", file=sys.stderr)
|
|
full_text = get_video_description(session, api_key, v['videoId'])
|
|
|
|
desc_de = v.get('description_snippet', '')
|
|
if not desc_de and full_text:
|
|
sentences = re.split(r'(?<=[.!?])\s+', full_text)
|
|
desc_de = ' '.join(sentences[:3])
|
|
|
|
results.append({
|
|
"title": v['title'],
|
|
"channel": v['channel'],
|
|
"views": v['views'],
|
|
"description_de": desc_de[:300] if desc_de else "Keine Beschreibung verfügbar.",
|
|
"source_url": f"https://www.youtube.com/watch?v={v['videoId']}",
|
|
"full_text": full_text or "N/A",
|
|
"tags": ["YOUTUBE", "TRENDING", "DEUTSCHLAND"]
|
|
})
|
|
|
|
print("\n=== RESULTS ===", file=sys.stderr)
|
|
print(json.dumps(results, ensure_ascii=False, indent=2))
|
|
|
|
if __name__ == "__main__":
|
|
main()
|