Files
last30days-skill/scripts/lib/youtube_yt.py
T
Matt Van Horn 0591f55f0e feat(quality): YouTube relevance scoring and cross-source linking
YouTube videos now get real relevance scores based on token overlap
between the search query and video title (was hardcoded at 0.7).
Uses ratio overlap with stopword removal, floored at 0.1.

Cross-source linking annotates items that discuss the same story
across different platforms (e.g., Reddit + HN + X). Items get
bidirectional cross_refs displayed as [xref: R3, HN5] in compact
output so Claude can triangulate multi-platform coverage.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-02-25 10:46:58 -08:00

408 lines
12 KiB
Python

"""YouTube search and transcript extraction via yt-dlp for /last30days v2.1.
Uses yt-dlp (https://github.com/yt-dlp/yt-dlp) for both YouTube search and
transcript extraction. No API keys needed — just have yt-dlp installed.
Inspired by Peter Steinberger's toolchain approach (yt-dlp + summarize CLI).
"""
import json
import math
import os
import re
import signal
import shutil
import subprocess
import sys
import tempfile
from concurrent.futures import ThreadPoolExecutor, as_completed
from pathlib import Path
from typing import Any, Dict, List, Optional, Set, Tuple
# Depth configurations: how many videos to search / transcribe
DEPTH_CONFIG = {
"quick": 10,
"default": 20,
"deep": 40,
}
TRANSCRIPT_LIMITS = {
"quick": 3,
"default": 5,
"deep": 8,
}
# Max words to keep from each transcript
TRANSCRIPT_MAX_WORDS = 500
# Stopwords for relevance computation (common English words that dilute token overlap)
STOPWORDS = frozenset({
'the', 'a', 'an', 'to', 'for', 'how', 'is', 'in', 'of', 'on',
'and', 'with', 'from', 'by', 'at', 'this', 'that', 'it', 'my',
'your', 'i', 'me', 'we', 'you', 'what', 'are', 'do', 'can',
'its', 'be', 'or', 'not', 'no', 'so', 'if', 'but', 'about',
'all', 'just', 'get', 'has', 'have', 'was', 'will',
})
def _tokenize(text: str) -> Set[str]:
"""Lowercase, strip punctuation, remove stopwords, drop single-char tokens."""
words = re.sub(r'[^\w\s]', ' ', text.lower()).split()
return {w for w in words if w not in STOPWORDS and len(w) > 1}
def _compute_relevance(query: str, title: str) -> float:
"""Compute relevance as ratio of query tokens found in title.
Uses ratio overlap (intersection / query_length) so short queries
score higher when fully represented in the title. Floors at 0.1.
"""
q_tokens = _tokenize(query)
t_tokens = _tokenize(title)
if not q_tokens:
return 0.5 # Neutral fallback for empty/stopword-only queries
overlap = len(q_tokens & t_tokens)
ratio = overlap / len(q_tokens)
return max(0.1, min(1.0, ratio))
def _log(msg: str):
"""Log to stderr."""
sys.stderr.write(f"[YouTube] {msg}\n")
sys.stderr.flush()
def is_ytdlp_installed() -> bool:
"""Check if yt-dlp is available in PATH."""
return shutil.which("yt-dlp") is not None
def _extract_core_subject(topic: str) -> str:
"""Extract core subject from verbose query for YouTube search.
Strips meta/research words to keep only the core product/concept name,
similar to bird_x.py's approach.
"""
text = topic.lower().strip()
# Strip multi-word prefixes
prefixes = [
'what are the best', 'what is the best', 'what are the latest',
'what are people saying about', 'what do people think about',
'how do i use', 'how to use', 'how to',
'what are', 'what is', 'tips for', 'best practices for',
]
for p in prefixes:
if text.startswith(p + ' '):
text = text[len(p):].strip()
# Strip individual noise words
# NOTE: 'tips', 'tricks', 'tutorial', 'guide', 'review', 'reviews'
# are intentionally KEPT — they're YouTube content types that improve search
noise = {
'best', 'top', 'good', 'great', 'awesome', 'killer',
'latest', 'new', 'news', 'update', 'updates',
'trending', 'hottest', 'popular', 'viral',
'practices', 'features',
'recommendations', 'advice',
'prompt', 'prompts', 'prompting',
'methods', 'strategies', 'approaches',
}
words = text.split()
filtered = [w for w in words if w not in noise]
result = ' '.join(filtered) if filtered else text
return result.rstrip('?!.')
def search_youtube(
topic: str,
from_date: str,
to_date: str,
depth: str = "default",
) -> Dict[str, Any]:
"""Search YouTube via yt-dlp. No API key needed.
Args:
topic: Search topic
from_date: Start date (YYYY-MM-DD)
to_date: End date (YYYY-MM-DD)
depth: 'quick', 'default', or 'deep'
Returns:
Dict with 'items' list of video metadata dicts.
"""
if not is_ytdlp_installed():
return {"items": [], "error": "yt-dlp not installed"}
count = DEPTH_CONFIG.get(depth, DEPTH_CONFIG["default"])
core_topic = _extract_core_subject(topic)
_log(f"Searching YouTube for '{core_topic}' (since {from_date}, count={count})")
# yt-dlp search with full metadata (no --flat-playlist so dates are real).
# No --dateafter — we filter by date in Python with a soft fallback,
# because YouTube search returns relevance-sorted results and strict date
# filtering returns 0 for evergreen topics like "thumbnail tips".
cmd = [
"yt-dlp",
f"ytsearch{count}:{core_topic}",
"--dump-json",
"--no-warnings",
"--no-download",
]
preexec = os.setsid if hasattr(os, 'setsid') else None
try:
proc = subprocess.Popen(
cmd,
stdout=subprocess.PIPE,
stderr=subprocess.PIPE,
text=True,
preexec_fn=preexec,
)
try:
stdout, stderr = proc.communicate(timeout=120)
except subprocess.TimeoutExpired:
try:
os.killpg(os.getpgid(proc.pid), signal.SIGTERM)
except (ProcessLookupError, PermissionError, OSError):
proc.kill()
proc.wait(timeout=5)
_log("YouTube search timed out (120s)")
return {"items": [], "error": "Search timed out"}
except FileNotFoundError:
return {"items": [], "error": "yt-dlp not found"}
if not (stdout or "").strip():
_log("YouTube search returned 0 results")
return {"items": []}
# Parse JSON-per-line output
items = []
for line in stdout.strip().split("\n"):
line = line.strip()
if not line:
continue
try:
video = json.loads(line)
except json.JSONDecodeError:
continue
video_id = video.get("id", "")
view_count = video.get("view_count") or 0
like_count = video.get("like_count") or 0
comment_count = video.get("comment_count") or 0
upload_date = video.get("upload_date", "") # YYYYMMDD
# Convert YYYYMMDD to YYYY-MM-DD
date_str = None
if upload_date and len(upload_date) == 8:
date_str = f"{upload_date[:4]}-{upload_date[4:6]}-{upload_date[6:8]}"
items.append({
"video_id": video_id,
"title": video.get("title", ""),
"url": f"https://www.youtube.com/watch?v={video_id}",
"channel_name": video.get("channel", video.get("uploader", "")),
"date": date_str,
"engagement": {
"views": view_count,
"likes": like_count,
"comments": comment_count,
},
"duration": video.get("duration"),
"relevance": _compute_relevance(core_topic, video.get("title", "")),
"why_relevant": f"YouTube: {video.get('title', core_topic)[:60]}",
})
# Soft date filter: prefer recent items but fall back to all if too few
recent = [i for i in items if i["date"] and i["date"] >= from_date]
if len(recent) >= 3:
items = recent
_log(f"Found {len(items)} videos within date range")
else:
_log(f"Found {len(items)} videos ({len(recent)} within date range, keeping all)")
# Sort by views descending
items.sort(key=lambda x: x["engagement"]["views"], reverse=True)
return {"items": items}
def _clean_vtt(vtt_text: str) -> str:
"""Convert VTT subtitle format to clean plaintext."""
# Strip VTT header
text = re.sub(r'^WEBVTT.*?\n\n', '', vtt_text, flags=re.DOTALL)
# Strip timestamps
text = re.sub(r'\d{2}:\d{2}:\d{2}\.\d{3}\s*-->\s*\d{2}:\d{2}:\d{2}\.\d{3}.*\n', '', text)
# Strip position/alignment tags
text = re.sub(r'<[^>]+>', '', text)
# Strip cue numbers
text = re.sub(r'^\d+\s*$', '', text, flags=re.MULTILINE)
# Deduplicate overlapping lines
lines = text.strip().split('\n')
seen = set()
unique = []
for line in lines:
stripped = line.strip()
if stripped and stripped not in seen:
seen.add(stripped)
unique.append(stripped)
return re.sub(r'\s+', ' ', ' '.join(unique)).strip()
def fetch_transcript(video_id: str, temp_dir: str) -> Optional[str]:
"""Fetch auto-generated transcript for a YouTube video.
Args:
video_id: YouTube video ID
temp_dir: Temporary directory for subtitle files
Returns:
Plaintext transcript string, or None if no captions available.
"""
cmd = [
"yt-dlp",
"--write-auto-subs",
"--sub-lang", "en",
"--sub-format", "vtt",
"--skip-download",
"--no-warnings",
"-o", f"{temp_dir}/%(id)s",
f"https://www.youtube.com/watch?v={video_id}",
]
preexec = os.setsid if hasattr(os, 'setsid') else None
try:
proc = subprocess.Popen(
cmd,
stdout=subprocess.PIPE,
stderr=subprocess.PIPE,
text=True,
preexec_fn=preexec,
)
try:
proc.communicate(timeout=30)
except subprocess.TimeoutExpired:
try:
os.killpg(os.getpgid(proc.pid), signal.SIGTERM)
except (ProcessLookupError, PermissionError, OSError):
proc.kill()
proc.wait(timeout=5)
return None
except FileNotFoundError:
return None
# yt-dlp may save as .en.vtt or .en-orig.vtt
vtt_path = Path(temp_dir) / f"{video_id}.en.vtt"
if not vtt_path.exists():
# Try alternate naming
for p in Path(temp_dir).glob(f"{video_id}*.vtt"):
vtt_path = p
break
else:
return None
try:
raw = vtt_path.read_text(encoding="utf-8", errors="replace")
except OSError:
return None
transcript = _clean_vtt(raw)
# Truncate to max words
words = transcript.split()
if len(words) > TRANSCRIPT_MAX_WORDS:
transcript = ' '.join(words[:TRANSCRIPT_MAX_WORDS]) + '...'
return transcript if transcript else None
def fetch_transcripts_parallel(
video_ids: List[str],
max_workers: int = 5,
) -> Dict[str, Optional[str]]:
"""Fetch transcripts for multiple videos in parallel.
Args:
video_ids: List of YouTube video IDs
max_workers: Max parallel fetches
Returns:
Dict mapping video_id to transcript text (or None).
"""
if not video_ids:
return {}
_log(f"Fetching transcripts for {len(video_ids)} videos")
results = {}
with tempfile.TemporaryDirectory() as temp_dir:
with ThreadPoolExecutor(max_workers=max_workers) as executor:
futures = {
executor.submit(fetch_transcript, vid, temp_dir): vid
for vid in video_ids
}
for future in as_completed(futures):
vid = futures[future]
try:
results[vid] = future.result()
except Exception:
results[vid] = None
got = sum(1 for v in results.values() if v)
_log(f"Got transcripts for {got}/{len(video_ids)} videos")
return results
def search_and_transcribe(
topic: str,
from_date: str,
to_date: str,
depth: str = "default",
) -> Dict[str, Any]:
"""Full YouTube search: find videos, then fetch transcripts for top results.
Args:
topic: Search topic
from_date: Start date (YYYY-MM-DD)
to_date: End date (YYYY-MM-DD)
depth: 'quick', 'default', or 'deep'
Returns:
Dict with 'items' list. Each item has a 'transcript_snippet' field.
"""
# Step 1: Search
search_result = search_youtube(topic, from_date, to_date, depth)
items = search_result.get("items", [])
if not items:
return search_result
# Step 2: Fetch transcripts for top N by views
transcript_limit = TRANSCRIPT_LIMITS.get(depth, TRANSCRIPT_LIMITS["default"])
top_ids = [item["video_id"] for item in items[:transcript_limit]]
transcripts = fetch_transcripts_parallel(top_ids)
# Step 3: Attach transcripts to items
for item in items:
vid = item["video_id"]
transcript = transcripts.get(vid)
item["transcript_snippet"] = transcript or ""
return {"items": items}
def parse_youtube_response(response: Dict[str, Any]) -> List[Dict[str, Any]]:
"""Parse YouTube search response to normalized format.
Returns:
List of item dicts ready for normalization.
"""
return response.get("items", [])