Spaces:
Running
Running
File size: 18,401 Bytes
5ec879f f83678a 5ec879f 49c5f61 5ec879f f83678a 5ec879f f83678a 60f50f3 49c5f61 5ec879f 60f50f3 5ec879f 49c5f61 5ec879f 49c5f61 5ec879f f83678a 49c5f61 5ec879f 49c5f61 5ec879f 49c5f61 5ec879f 49c5f61 5ec879f f83678a 49c5f61 f83678a 49c5f61 f83678a 49c5f61 f83678a ca15e7a 60f50f3 ca15e7a 60f50f3 ca15e7a 60f50f3 ca15e7a 60f50f3 ca15e7a 60f50f3 49c5f61 60f50f3 ca15e7a 49c5f61 ca15e7a 49c5f61 ca15e7a 49c5f61 ca15e7a 60f50f3 ca15e7a 60f50f3 49c5f61 5ec879f 49c5f61 5ec879f 49c5f61 05d2c32 49c5f61 5ec879f 49c5f61 05d2c32 49c5f61 5ec879f 49c5f61 5ec879f 49c5f61 5ec879f 49c5f61 5ec879f 49c5f61 5ec879f 49c5f61 5ec879f 49c5f61 5ec879f 49c5f61 5ec879f 49c5f61 5ec879f 49c5f61 5ec879f ca15e7a 49c5f61 ca15e7a 49c5f61 ca15e7a 5ec879f 49c5f61 05d2c32 49c5f61 05d2c32 49c5f61 05d2c32 49c5f61 05d2c32 49c5f61 05d2c32 49c5f61 05d2c32 49c5f61 05d2c32 49c5f61 05d2c32 49c5f61 f83678a 49c5f61 f83678a 49c5f61 f83678a ca15e7a 49c5f61 f83678a ca15e7a |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 |
import os
import re
import random
import asyncio
import httpx
import aiohttp
import trafilatura
import json
import uuid
import time
from pathlib import Path
from urllib.parse import urlparse
from typing import List, Dict, Any, Optional, Set, Tuple
from fastapi import APIRouter, HTTPException, Body
from fastapi.responses import FileResponse
from newspaper import Article
from threading import Timer
from google import genai
from google.genai import types
from asyncio import Queue, create_task, gather
from concurrent.futures import ThreadPoolExecutor
import aiofiles
import ujson # JSON mais rápido
router = APIRouter()
BRAVE_API_KEY = os.getenv("BRAVE_API_KEY")
if not BRAVE_API_KEY:
raise ValueError("BRAVE_API_KEY não está definido!")
GEMINI_API_KEY = os.getenv("GEMINI_API_KEY")
if not GEMINI_API_KEY:
raise ValueError("GEMINI_API_KEY não está definido!")
BRAVE_SEARCH_URL = "https://api.search.brave.com/res/v1/web/search"
BRAVE_HEADERS = {
"Accept": "application/json",
"Accept-Encoding": "gzip",
"x-subscription-token": BRAVE_API_KEY
}
USER_AGENTS = [
"Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.0.0 Safari/537.36",
"Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/17.0 Safari/605.1.15",
"Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.0.0 Safari/537.36",
]
BLOCKED_DOMAINS = frozenset({ # frozenset é mais rápido para lookup
"reddit.com", "www.reddit.com", "old.reddit.com",
"quora.com", "www.quora.com"
})
MAX_TEXT_LENGTH = 4000
MAX_CONCURRENT_SEARCHES = 30 # Aumentado
MAX_CONCURRENT_EXTRACTIONS = 80 # Aumentado significativamente
EXTRACTION_TIMEOUT = 8 # Reduzido
HTTP_TIMEOUT = 10 # Reduzido
# Diretório para arquivos temporários
TEMP_DIR = Path("/tmp")
TEMP_DIR.mkdir(exist_ok=True)
# Dicionário para controlar arquivos temporários
temp_files = {}
# Pool de threads para operações CPU-intensive
thread_pool = ThreadPoolExecutor(max_workers=20)
# Cache de domínios bloqueados para evitar verificações repetidas
domain_cache = {}
def is_blocked_domain(url: str) -> bool:
try:
host = urlparse(url).netloc.lower()
# Cache lookup
if host in domain_cache:
return domain_cache[host]
is_blocked = any(host == b or host.endswith("." + b) for b in BLOCKED_DOMAINS)
domain_cache[host] = is_blocked
return is_blocked
except Exception:
return False
def clamp_text(text: str) -> str:
if not text or len(text) <= MAX_TEXT_LENGTH:
return text
return text[:MAX_TEXT_LENGTH]
def get_realistic_headers() -> Dict[str, str]:
return {
"User-Agent": random.choice(USER_AGENTS),
"Accept": "text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,*/*;q=0.8",
"Accept-Language": "en-US,en;q=0.7,pt-BR;q=0.6",
"Connection": "keep-alive",
"Accept-Encoding": "gzip, deflate, br",
}
def delete_temp_file(file_id: str, file_path: Path):
"""Remove arquivo temporário após expiração"""
try:
if file_path.exists():
file_path.unlink()
temp_files.pop(file_id, None)
print(f"Arquivo temporário removido: {file_path}")
except Exception as e:
print(f"Erro ao remover arquivo temporário: {e}")
async def create_temp_file(data: Dict[str, Any]) -> Dict[str, str]:
"""Cria arquivo temporário assíncrono e agenda sua remoção"""
file_id = str(uuid.uuid4())
file_path = TEMP_DIR / f"fontes_{file_id}.txt"
# Salva o JSON no arquivo de forma assíncrona
async with aiofiles.open(file_path, 'w', encoding='utf-8') as f:
await f.write(ujson.dumps(data, ensure_ascii=False, indent=2))
# Agenda remoção em 24 horas
timer = Timer(86400, delete_temp_file, args=[file_id, file_path])
timer.start()
# Registra o arquivo temporário
temp_files[file_id] = {
"path": file_path,
"created_at": time.time(),
"timer": timer
}
return {
"file_id": file_id,
"download_url": f"/download-temp/{file_id}",
"expires_in_hours": 24
}
async def generate_search_terms(context: str) -> List[str]:
"""Gera termos de pesquisa usando o modelo Gemini"""
try:
client = genai.Client(api_key=GEMINI_API_KEY)
model = "gemini-2.5-flash-lite"
system_prompt = """Com base num contexto inicial, gere termos de pesquisa (até 20 termos, no máximo), em um JSON. Esses textos devem ser interpretados como termos que podem ser usados por outras inteligências artificiais pra pesquisar no google e retornar resultados mais refinados e completos pra busca atual. Analise muito bem o contexto, por exemplo, se está falando de uma série coreana, gere os termos em coreano por que obviamente na mídia coreana terá mais cobertura que a americana, etc.
Deve seguir esse formato: "terms": []
Retorne apenas o JSON, sem mais nenhum texto."""
contents = [
types.Content(
role="user",
parts=[
types.Part.from_text(text="Contexto: Taylor Sheridan's 'Landman' Announces Season 2 Premiere Date"),
],
),
types.Content(
role="model",
parts=[
types.Part.from_text(text='{"terms": [ "imdb landman episodes season 2", "imdb landman series", "landman season 2 release date", "taylor sheridan landman series", "landman season 2 cast sam elliott", "billy bob thornton returns landman", "demi moore landman new season", "andy garcia ali larter landman season 2", "landman texas oil drama", "taylor sheridan tv series schedule", "landman 10 month turnaround new episodes", "landman season 2 november 16 premiere", "sam elliott joins taylor sheridan show", "landman streaming platform premiere", "landman season 2 filming details", "landman new cast and returning actors", "taylor sheridan quick tv show production" ]}'),
],
),
types.Content(
role="user",
parts=[
types.Part.from_text(text="Contexto: Pixar's latest animated feature will arrive on digital (via platforms like Apple TV, Amazon Prime Video, and Fandango at Home) on Aug. 19 and on physical media (4K UHD, Blu-ray and DVD) on Sept. 9. The film has not yet set a Disney+ streaming release date, but that will likely come after the Blu-ray release."),
],
),
types.Content(
role="model",
parts=[
types.Part.from_text(text='{ "terms": [ "pixar elio 2024 movie details", "disney pixar new release elio", "elio animated film august 19 digital", "pixar sci-fi comedy elio home release", "elio movie blu-ray dvd release september", "where to watch elio online", "elio streaming disney plus release date", "elio digital release apple tv amazon prime", "elio physical media 4k uhd blu-ray dvd", "elio movie bonus features", "elio cast voice actors", "elio behind the scenes making of", "elio deleted scenes blu-ray", "elio soundtrack and score", "elio merchandise release date", "upcoming disney pixar movies 2024", "pixar elio critical reviews", "elio movie box office results" ] }'),
],
),
types.Content(
role="user",
parts=[
types.Part.from_text(text=f"Contexto: {context}"),
],
),
]
generate_content_config = types.GenerateContentConfig(
thinking_config=types.ThinkingConfig(thinking_budget=0),
)
# Coletamos toda a resposta em stream
full_response = ""
for chunk in client.models.generate_content_stream(
model=model,
contents=contents,
config=generate_content_config,
):
if chunk.text:
full_response += chunk.text
# Tenta extrair o JSON da resposta
try:
clean_response = full_response.strip()
if clean_response.startswith("```json"):
clean_response = clean_response[7:]
if clean_response.endswith("```"):
clean_response = clean_response[:-3]
clean_response = clean_response.strip()
response_data = ujson.loads(clean_response)
terms = response_data.get("terms", [])
if not isinstance(terms, list):
raise ValueError("Terms deve ser uma lista")
return terms[:20]
except (ujson.JSONDecodeError, ValueError) as e:
print(f"Erro ao parsear resposta do Gemini: {e}")
return []
except Exception as e:
print(f"Erro ao gerar termos de pesquisa: {str(e)}")
return []
async def search_brave_batch(client: httpx.AsyncClient, terms: List[str]) -> List[Tuple[str, List[Dict[str, str]]]]:
"""Busca múltiplos termos em paralelo com otimizações"""
semaphore = asyncio.Semaphore(MAX_CONCURRENT_SEARCHES)
async def search_single_term(term: str) -> Tuple[str, List[Dict[str, str]]]:
async with semaphore:
params = {"q": term, "count": 10, "safesearch": "off", "summary": "false"}
try:
resp = await client.get(BRAVE_SEARCH_URL, headers=BRAVE_HEADERS, params=params)
if resp.status_code != 200:
return (term, [])
data = resp.json()
results = []
if "web" in data and "results" in data["web"]:
for item in data["web"]["results"]:
url = item.get("url")
age = item.get("age", "Unknown")
if url and not is_blocked_domain(url):
results.append({"url": url, "age": age})
return (term, results)
except Exception as e:
print(f"Erro na busca do termo '{term}': {e}")
return (term, [])
# Executa todas as buscas em paralelo
tasks = [search_single_term(term) for term in terms]
return await gather(*tasks, return_exceptions=False)
def extract_with_trafilatura(html: str) -> str:
"""Extração CPU-intensive executada em thread pool"""
try:
extracted = trafilatura.extract(html)
return extracted.strip() if extracted else ""
except Exception:
return ""
def extract_with_newspaper(url: str) -> str:
"""Extração com newspaper executada em thread pool"""
try:
art = Article(url)
art.config.browser_user_agent = random.choice(USER_AGENTS)
art.config.request_timeout = 6
art.config.number_threads = 1
art.download()
art.parse()
return (art.text or "").strip()
except Exception:
return ""
async def extract_article_text_optimized(url: str, session: aiohttp.ClientSession) -> str:
"""Extração de artigo otimizada com paralelização de métodos"""
# Tentativa 1: Newspaper em thread pool (paralelo com download HTTP)
newspaper_task = asyncio.create_task(
asyncio.get_event_loop().run_in_executor(thread_pool, extract_with_newspaper, url)
)
# Tentativa 2: Download HTTP e trafilatura
try:
headers = get_realistic_headers()
async with session.get(url, headers=headers, timeout=EXTRACTION_TIMEOUT) as resp:
if resp.status != 200:
# Aguarda newspaper se HTTP falhou
newspaper_result = await newspaper_task
return clamp_text(newspaper_result) if newspaper_result and len(newspaper_result) > 100 else ""
html = await resp.text()
# Verifica paywall rapidamente
if re.search(r"(paywall|subscribe|metered|registration|captcha|access denied)",
html[:2000], re.I): # Verifica apenas o início
newspaper_result = await newspaper_task
return clamp_text(newspaper_result) if newspaper_result and len(newspaper_result) > 100 else ""
# Extração com trafilatura em thread pool
trafilatura_task = asyncio.create_task(
asyncio.get_event_loop().run_in_executor(thread_pool, extract_with_trafilatura, html)
)
# Aguarda ambos os métodos e pega o melhor resultado
newspaper_result, trafilatura_result = await gather(newspaper_task, trafilatura_task)
# Escolhe o melhor resultado
best_result = ""
if trafilatura_result and len(trafilatura_result) > 100:
best_result = trafilatura_result
elif newspaper_result and len(newspaper_result) > 100:
best_result = newspaper_result
return clamp_text(best_result) if best_result else ""
except Exception:
# Se tudo falhar, tenta pelo menos o newspaper
try:
newspaper_result = await newspaper_task
return clamp_text(newspaper_result) if newspaper_result and len(newspaper_result) > 100 else ""
except Exception:
return ""
async def process_urls_batch(session: aiohttp.ClientSession, urls_data: List[Tuple[str, str, str]]) -> List[Dict[str, Any]]:
"""Processa URLs em lotes otimizados"""
semaphore = asyncio.Semaphore(MAX_CONCURRENT_EXTRACTIONS)
results = []
used_urls: Set[str] = set()
async def process_single_url(term: str, url: str, age: str) -> Optional[Dict[str, Any]]:
async with semaphore:
if url in used_urls:
return None
text = await extract_article_text_optimized(url, session)
if text:
used_urls.add(url)
return {
"term": term,
"age": age,
"url": url,
"text": text
}
return None
# Cria todas as tasks de uma vez
tasks = []
for term, url, age in urls_data:
tasks.append(process_single_url(term, url, age))
# Processa tudo em paralelo
processed_results = await gather(*tasks, return_exceptions=True)
# Filtra resultados válidos
return [r for r in processed_results if r is not None and not isinstance(r, Exception)]
@router.post("/search-terms")
async def search_terms(payload: Dict[str, str] = Body(...)) -> Dict[str, Any]:
start_time = time.time()
context = payload.get("context")
if not context or not isinstance(context, str):
raise HTTPException(status_code=400, detail="Campo 'context' é obrigatório e deve ser uma string.")
if len(context.strip()) == 0:
raise HTTPException(status_code=400, detail="Campo 'context' não pode estar vazio.")
print(f"Iniciando geração de termos...")
# Gera os termos de pesquisa usando o Gemini
terms = await generate_search_terms(context)
if not terms:
raise HTTPException(status_code=500, detail="Não foi possível gerar termos de pesquisa válidos.")
print(f"Termos gerados em {time.time() - start_time:.2f}s. Iniciando buscas...")
# Configurações otimizadas para conexões
connector = aiohttp.TCPConnector(
limit=200, # Aumentado
limit_per_host=30, # Aumentado
ttl_dns_cache=300,
use_dns_cache=True,
enable_cleanup_closed=True
)
timeout = aiohttp.ClientTimeout(total=HTTP_TIMEOUT, connect=5)
# Cliente HTTP otimizado
http_client = httpx.AsyncClient(
timeout=HTTP_TIMEOUT,
limits=httpx.Limits(
max_connections=200, # Aumentado
max_keepalive_connections=50 # Aumentado
),
http2=True # Ativa HTTP/2
)
try:
async with aiohttp.ClientSession(connector=connector, timeout=timeout) as session:
# Fase 1: Busca todos os termos em paralelo
search_results = await search_brave_batch(http_client, terms)
print(f"Buscas concluídas em {time.time() - start_time:.2f}s. Iniciando extrações...")
# Fase 2: Prepara dados para extração em lote
urls_data = []
for term, results in search_results:
for result in results:
urls_data.append((term, result["url"], result["age"]))
print(f"Processando {len(urls_data)} URLs...")
# Fase 3: Processa todas as URLs em paralelo
final_results = await process_urls_batch(session, urls_data)
print(f"Extração concluída em {time.time() - start_time:.2f}s. Salvando arquivo...")
finally:
await http_client.aclose()
# Fase 4: Cria arquivo temporário assíncrono
result_data = {"results": final_results}
temp_file_info = await create_temp_file(result_data)
total_time = time.time() - start_time
print(f"Processo completo em {total_time:.2f}s")
return {
"message": "Dados salvos em arquivo temporário",
"total_results": len(final_results),
"context": context,
"generated_terms": terms,
"file_info": temp_file_info,
"processing_time": f"{total_time:.2f}s"
}
@router.get("/download-temp/{file_id}")
async def download_temp_file(file_id: str):
"""Endpoint para download do arquivo temporário"""
if file_id not in temp_files:
raise HTTPException(status_code=404, detail="Arquivo não encontrado ou expirado")
file_info = temp_files[file_id]
file_path = file_info["path"]
if not file_path.exists():
temp_files.pop(file_id, None)
raise HTTPException(status_code=404, detail="Arquivo não encontrado")
return FileResponse(
path=str(file_path),
filename="fontes.txt",
media_type="text/plain",
headers={"Content-Disposition": "attachment; filename=fontes.txt"}
) |