YouTube Comment Monitoring Pipeline

import requests
import json
from pathlib import Path
from datetime import datetime

API_KEY = "your_scavio_api_key"

POSITIVE_WORDS = ["love", "great", "amazing", "helpful", "best", "awesome", "excellent"]
NEGATIVE_WORDS = ["bad", "terrible", "broken", "hate", "worst", "scam", "awful", "disappointed"]

def search_youtube(query: str) -> list[dict]:
    res = requests.post(
        "https://api.scavio.dev/api/v1/search",
        headers={"x-api-key": API_KEY},
        json={"platform": "youtube", "query": query},
        timeout=15,
    )
    res.raise_for_status()
    return res.json().get("organic", [])

def score_sentiment(text: str) -> str:
    lower = text.lower()
    pos = sum(1 for w in POSITIVE_WORDS if w in lower)
    neg = sum(1 for w in NEGATIVE_WORDS if w in lower)
    if neg > pos:
        return "negative"
    if pos > neg:
        return "positive"
    return "neutral"

def extract_comments(videos: list[dict]) -> list[dict]:
    comments = []
    for video in videos:
        for comment in video.get("comments", []):
            sentiment = score_sentiment(comment.get("text", ""))
            comments.append({
                "video_title": video.get("title", ""),
                "author": comment.get("author", ""),
                "text": comment.get("text", ""),
                "likes": comment.get("likes", 0),
                "sentiment": sentiment,
                "video_link": video.get("link", ""),
            })
    return comments

def run():
    config = json.loads(Path("monitor_config.json").read_text())
    keywords = config.get("keywords", ["your brand name"])
    all_comments = []

    for kw in keywords:
        videos = search_youtube(kw)
        comments = extract_comments(videos)
        all_comments.extend(comments)

    # Flag notable comments
    notable = [c for c in all_comments if c["sentiment"] == "negative" or c["likes"] >= 10]
    notable.sort(key=lambda x: x["likes"], reverse=True)

    date = datetime.utcnow().strftime("%Y-%m-%d")
    report = {
        "date": date,
        "total_comments": len(all_comments),
        "positive": sum(1 for c in all_comments if c["sentiment"] == "positive"),
        "negative": sum(1 for c in all_comments if c["sentiment"] == "negative"),
        "neutral": sum(1 for c in all_comments if c["sentiment"] == "neutral"),
        "notable": notable[:20],
    }

    Path(f"yt_comments_{date}.json").write_text(json.dumps(report, indent=2))
    print(f"Collected {len(all_comments)} comments, {len(notable)} notable")
    for c in notable[:5]:
        print(f"  [{c['sentiment']}] {c['text'][:80]}")

if __name__ == "__main__":
    run()

const API_KEY = "your_scavio_api_key";

const POSITIVE_WORDS = ["love", "great", "amazing", "helpful", "best", "awesome", "excellent"];
const NEGATIVE_WORDS = ["bad", "terrible", "broken", "hate", "worst", "scam", "awful", "disappointed"];

async function searchYouTube(query) {
  const res = await fetch("https://api.scavio.dev/api/v1/search", {
    method: "POST",
    headers: { "x-api-key": API_KEY, "content-type": "application/json" },
    body: JSON.stringify({ platform: "youtube", query }),
  });
  if (!res.ok) throw new Error(`scavio ${res.status}`);
  return (await res.json()).organic ?? [];
}

function scoreSentiment(text) {
  const lower = text.toLowerCase();
  const pos = POSITIVE_WORDS.filter((w) => lower.includes(w)).length;
  const neg = NEGATIVE_WORDS.filter((w) => lower.includes(w)).length;
  if (neg > pos) return "negative";
  if (pos > neg) return "positive";
  return "neutral";
}

function extractComments(videos) {
  const comments = [];
  for (const video of videos) {
    for (const comment of video.comments ?? []) {
      comments.push({
        videoTitle: video.title ?? "",
        author: comment.author ?? "",
        text: comment.text ?? "",
        likes: comment.likes ?? 0,
        sentiment: scoreSentiment(comment.text ?? ""),
        videoLink: video.link ?? "",
      });
    }
  }
  return comments;
}

async function run() {
  const fs = await import("fs/promises");
  const config = JSON.parse(await fs.readFile("monitor_config.json", "utf8"));
  const keywords = config.keywords ?? ["your brand name"];
  const allComments = [];

  for (const kw of keywords) {
    const videos = await searchYouTube(kw);
    allComments.push(...extractComments(videos));
  }

  const notable = allComments
    .filter((c) => c.sentiment === "negative" || c.likes >= 10)
    .sort((a, b) => b.likes - a.likes);

  const date = new Date().toISOString().slice(0, 10);
  const report = {
    date,
    totalComments: allComments.length,
    positive: allComments.filter((c) => c.sentiment === "positive").length,
    negative: allComments.filter((c) => c.sentiment === "negative").length,
    neutral: allComments.filter((c) => c.sentiment === "neutral").length,
    notable: notable.slice(0, 20),
  };

  await fs.writeFile(`yt_comments_${date}.json`, JSON.stringify(report, null, 2));
  console.log(`Collected ${allComments.length} comments, ${notable.length} notable`);
  for (const c of notable.slice(0, 5)) console.log(`  [${c.sentiment}] ${c.text.slice(0, 80)}`);
}

run();