Diario News Collection for ML

import requests, os, json, hashlib
from pathlib import Path
from datetime import date
from urllib.parse import urlparse

API_KEY = os.environ["SCAVIO_API_KEY"]
SH = {"x-api-key": API_KEY, "Content-Type": "application/json"}
TOPICS_FILE = Path("ml_topics.json")
CORPUS_FILE = Path("ml_corpus.jsonl")
SEEN_FILE = Path("ml_seen_urls.json")

def search_news(query: str) -> list:
    resp = requests.post(
        "https://api.scavio.dev/api/v1/search",
        headers=SH,
        json={"query": query, "platform": "google"},
        timeout=15,
    )
    resp.raise_for_status()
    return resp.json().get("organic", [])

def extract_record(result: dict, topic: str, category: str) -> dict:
    url = result.get("url", "")
    return {
        "title": result.get("title", ""),
        "snippet": result.get("snippet", ""),
        "url": url,
        "source_domain": urlparse(url).netloc if url else "",
        "topic": topic,
        "category": category,
        "collected_date": str(date.today()),
        "url_hash": hashlib.md5(url.encode()).hexdigest(),
    }

def run():
    topics = json.loads(TOPICS_FILE.read_text())
    seen = set()
    if SEEN_FILE.exists():
        seen = set(json.loads(SEEN_FILE.read_text()))

    new_records = []
    for topic_config in topics:
        topic = topic_config["topic"]
        category = topic_config.get("category", "general")
        query = topic_config.get("query", f"{topic} news {date.today().year}")
        results = search_news(query)
        for r in results:
            record = extract_record(r, topic, category)
            if record["url_hash"] not in seen:
                new_records.append(record)
                seen.add(record["url_hash"])

    with open(CORPUS_FILE, "a") as f:
        for record in new_records:
            f.write(json.dumps(record) + "\n")
    SEEN_FILE.write_text(json.dumps(list(seen)))

    print(f"Collected {len(new_records)} new articles on {date.today()}")
    for r in new_records[:5]:
        print(f"  [{r['category']}] {r['title'][:60]}")

run()

const SH = {'x-api-key': process.env.SCAVIO_API_KEY, 'Content-Type': 'application/json'};
const fs = await import('fs');
const crypto = await import('crypto');

const topics = JSON.parse(fs.readFileSync('ml_topics.json', 'utf8'));
let seen = new Set();
try { seen = new Set(JSON.parse(fs.readFileSync('ml_seen_urls.json', 'utf8'))); } catch {}

async function searchNews(query) {
  const r = await fetch('https://api.scavio.dev/api/v1/search', {method:'POST', headers:SH, body:JSON.stringify({query, platform:'google'})});
  return (await r.json()).organic || [];
}

const newRecords = [];
const today = new Date().toISOString().split('T')[0];

for (const tc of topics) {
  const query = tc.query || tc.topic+' news 2026';
  const results = await searchNews(query);
  for (const r of results) {
    const urlHash = crypto.createHash('md5').update(r.url||'').digest('hex');
    if (!seen.has(urlHash)) {
      const domain = r.url ? new URL(r.url).hostname : '';
      newRecords.push({title:r.title||'', snippet:r.snippet||'', url:r.url||'', sourceDomain:domain, topic:tc.topic, category:tc.category||'general', collectedDate:today, urlHash});
      seen.add(urlHash);
    }
  }
}

fs.appendFileSync('ml_corpus.jsonl', newRecords.map(r=>JSON.stringify(r)).join('\n')+(newRecords.length?'\n':''));
fs.writeFileSync('ml_seen_urls.json', JSON.stringify([...seen]));
console.log('Collected '+newRecords.length+' new articles');
newRecords.slice(0,5).forEach(r => console.log('  ['+r.category+'] '+r.title.slice(0,60)));