Cómo crear un resumen diario de noticias con la API de búsqueda

import os, requests, json
from datetime import datetime
from collections import Counter

API_KEY = os.environ['SCAVIO_API_KEY']
SH = {'x-api-key': API_KEY, 'Content-Type': 'application/json'}

TOPICS = [
    'AI agent frameworks',
    'search API market',
    'MCP protocol updates',
    'LLM cost optimization',
    'developer tools funding',
]

def search_topic(topic):
    data = requests.post('https://api.scavio.dev/api/v1/search',
        headers=SH, json={'query': f'{topic} news 2026', 'country_code': 'us'}).json()
    return [{'title': r.get('title', ''), 'link': r.get('link', ''),
             'snippet': r.get('snippet', ''), 'source': r.get('displayed_link', ''),
             'topic': topic} for r in data.get('organic_results', [])[:5]]

all_articles = []
for topic in TOPICS:
    articles = search_topic(topic)
    all_articles.extend(articles)
    print(f'  {topic}: {len(articles)} articles')
print(f'\nTotal: {len(all_articles)} articles. Cost: ${len(TOPICS) * 0.005:.3f}')

def deduplicate(articles):
    seen_links = set()
    unique = []
    for a in articles:
        link = a.get('link', '')
        if link and link not in seen_links:
            seen_links.add(link)
            unique.append(a)
    print(f'Deduplication: {len(articles)} -> {len(unique)} unique articles')
    return unique

def score_article(article):
    score = 50  # base
    title = article.get('title', '').lower()
    snippet = article.get('snippet', '').lower()
    # Recency signals
    if '2026' in title or '2026' in snippet: score += 15
    if any(w in title for w in ['today', 'just', 'breaking', 'new']): score += 10
    # Quality signals
    source = article.get('source', '').lower()
    quality_sources = ['techcrunch', 'reuters', 'bloomberg', 'theverge', 'arstechnica', 'github']
    if any(s in source for s in quality_sources): score += 20
    # Relevance signals
    if any(w in title for w in ['launch', 'release', 'raise', 'acquire']): score += 10
    return min(score, 100)

unique = deduplicate(all_articles)
for a in unique:
    a['score'] = score_article(a)
unique.sort(key=lambda x: x['score'], reverse=True)
print(f'\nTop scored articles:')
for a in unique[:5]:
    print(f'  [{a["score"]:3}] {a["title"][:55]}')

def format_digest(articles, max_per_topic=3):
    print(f'\n{"=" * 60}')
    print(f'  DAILY NEWS DIGEST - {datetime.now().strftime("%Y-%m-%d")}')
    print(f'  {len(articles)} articles across {len(set(a["topic"] for a in articles))} topics')
    print(f'{"=" * 60}')
    # Group by topic
    by_topic = {}
    for a in articles:
        t = a['topic']
        if t not in by_topic:
            by_topic[t] = []
        by_topic[t].append(a)
    for topic, items in by_topic.items():
        items.sort(key=lambda x: x.get('score', 0), reverse=True)
        print(f'\n  ## {topic}')
        for item in items[:max_per_topic]:
            print(f'  - {item["title"][:55]}')
            print(f'    {item["snippet"][:70]}')
            print(f'    Source: {item["source"]} | Score: {item.get("score", 0)}')
    # Slack format
    print(f'\n--- Slack Format ---')
    for a in articles[:5]:
        print(f'*{a["title"][:50]}*')
        print(f'>{a["snippet"][:70]}')
        print(f'<{a["link"]}|Read more>')
    print(f'\n  Digest cost: ${len(TOPICS) * 0.005:.3f}/day')
    print(f'  Monthly: ${len(TOPICS) * 0.005 * 30:.2f}/month')

format_digest(unique)