app.py

from fastapi import FastAPI, HTTPException, Query
from fastapi.middleware.cors import CORSMiddleware
import requests
from bs4 import BeautifulSoup
import uvicorn
import urllib.parse
import json
from duckduckgo_search import DDGS

app = FastAPI(title="Glimpse Scraping API", description="API de Scraping directo e Imágenes multi-motor")

app.add_middleware(
    CORSMiddleware,
    allow_origins=["*"],
    allow_credentials=True,
    allow_methods=["*"],
    allow_headers=["*"],
)

HEADERS = {
    "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36",
    "Accept-Language": "es-ES,es;q=0.9,en;q=0.8"
}

# --- FUNCIONES DE SCRAPING WEB (TEXTO) ---

def scrape_duckduckgo(query: str):
    url = f"https://html.duckduckgo.com/html/?q={urllib.parse.quote(query)}"
    resp = requests.get(url, headers=HEADERS, timeout=10)
    resp.raise_for_status()
    soup = BeautifulSoup(resp.text, 'lxml')
    results = []
    for result in soup.select('.result__body'):
        title_tag = result.select_one('.result__title a')
        snippet_tag = result.select_one('.result__snippet')
        if title_tag:
            raw_url = title_tag.get('href', '')
            clean_url = urllib.parse.unquote(raw_url.replace('//duckduckgo.com/l/?uddg=', '').split('&')[0])
            results.append({
                "title": title_tag.text.strip(),
                "url": clean_url if clean_url.startswith('http') else raw_url,
                "content": snippet_tag.text.strip() if snippet_tag else ""
            })
    return results

def scrape_mojeek(query: str):
    url = f"https://www.mojeek.com/search?q={urllib.parse.quote(query)}&fmt=html"
    resp = requests.get(url, headers=HEADERS, timeout=10)
    resp.raise_for_status()
    soup = BeautifulSoup(resp.text, 'lxml')
    results = []
    for li in soup.select('ul.results-standard > li'):
        a_tag = li.select_one('a.ob')
        p_tag = li.select_one('p.s')
        if a_tag:
            results.append({
                "title": a_tag.text.strip(),
                "url": a_tag.get('href', ''),
                "content": p_tag.text.strip() if p_tag else ""
            })
    return results

def scrape_qwant(query: str):
    url = f"https://lite.qwant.com/?q={urllib.parse.quote(query)}"
    resp = requests.get(url, headers=HEADERS, timeout=10)
    resp.raise_for_status()
    soup = BeautifulSoup(resp.text, 'lxml')
    results = []
    for article in soup.select('article.result'):
        title_tag = article.select_one('h2 a')
        snippet_tag = article.select_one('.result-snippet')
        if title_tag:
            results.append({
                "title": title_tag.text.strip(),
                "url": title_tag.get('href', ''),
                "content": snippet_tag.text.strip() if snippet_tag else ""
            })
    return results

def scrape_brave(query: str):
    url = f"https://search.brave.com/search?q={urllib.parse.quote(query)}"
    resp = requests.get(url, headers=HEADERS, timeout=10)
    resp.raise_for_status()
    soup = BeautifulSoup(resp.text, 'lxml')
    results = []
    for snippet in soup.select('.snippet'):
        title_tag = snippet.select_one('.heading')
        link_tag = snippet.select_one('a')
        desc_tag = snippet.select_one('.snippet-content, .snippet-description')
        if title_tag and link_tag:
            results.append({
                "title": title_tag.text.strip(),
                "url": link_tag.get('href', ''),
                "content": desc_tag.text.strip() if desc_tag else ""
            })
    return results

# --- ENDPOINTS ---

@app.get("/")
def read_root():
    return {"status": "Glimpse API is running!", "endpoints": ["/search", "/images"]}

@app.get("/search")
def search(
    q: str = Query(..., description="Término de búsqueda"), 
    engine: str = Query("duckduckgo", description="Motor de búsqueda")
):
    engine = engine.lower()
    results = []
    try:
        if engine == "mojeek":
            results = scrape_mojeek(q)
        elif engine in ["qwant", "tapnav"]:
            results = scrape_qwant(q)
        elif engine == "brave":
            results = scrape_brave(q)
        else:
            results = scrape_duckduckgo(q)
    except Exception as e:
        raise HTTPException(status_code=500, detail=f"Error en {engine}: {str(e)}")
    return {"results": results}

@app.get("/images")
def search_images(
    q: str = Query(..., description="Término de búsqueda para imágenes"),
    max_results: int = Query(40, description="Cantidad máxima de imágenes")
):
    results = []

    # INTENTO 1: Qwant API (Muy estable para imágenes)
    try:
        url = "https://api.qwant.com/v3/search/images"
        params = {"q": q, "t": "images", "locale": "es_ES", "count": max_results}
        resp = requests.get(url, params=params, headers=HEADERS, timeout=5)
        if resp.status_code == 200:
            items = resp.json().get("data", {}).get("result", {}).get("items", [])
            for item in items:
                if item.get("media"):
                    results.append({
                        "title": item.get("title", "Imagen"),
                        "image_url": item.get("media", ""),
                        "thumbnail_url": item.get("thumbnail", ""),
                        "source_url": item.get("url", ""),
                        "source_name": item.get("domain", "")
                    })
            if results:
                return {"results": results}
    except Exception as e:
        print(f"Intento 1 (Qwant) falló: {e}")

    # INTENTO 2: Yahoo Images (Scraping a la estructura JSON oculta)
    try:
        url = f"https://images.search.yahoo.com/search/images?p={urllib.parse.quote(q)}"
        resp = requests.get(url, headers=HEADERS, timeout=5)
        if resp.status_code == 200:
            soup = BeautifulSoup(resp.text, 'lxml')
            for li in soup.select('li.ld'):
                data_attr = li.get('data')
                if data_attr:
                    try:
                        item = json.loads(data_attr)
                        if item.get('iurl'):
                            results.append({
                                "title": item.get('title', 'Imagen'),
                                "image_url": item.get('iurl', ''),
                                "thumbnail_url": item.get('ith', '') or item.get('iurl', ''),
                                "source_url": item.get('rurl', ''),
                                "source_name": item.get('surl', '')
                            })
                    except:
                        continue
            if results:
                return {"results": results[:max_results]}
    except Exception as e:
        print(f"Intento 2 (Yahoo) falló: {e}")

    # INTENTO 3: DuckDuckGo (El original, por si los demás fallan)
    try:
        with DDGS() as ddgs:
            ddg_images = list(ddgs.images(keywords=q, max_results=max_results))
            for img in ddg_images:
                img_url = img.get("image", "") or img.get("url", "")
                if img_url:
                    results.append({
                        "title": img.get("title", "Imagen"),
                        "image_url": img_url,
                        "thumbnail_url": img.get("thumbnail", "") or img_url,
                        "source_url": img.get("url", ""),
                        "source_name": img.get("source", "")
                    })
            if results:
                return {"results": results}
    except Exception as e:
        print(f"Intento 3 (DDG) falló: {e}")

    # Si todo falla, enviamos el error real al Frontend
    raise HTTPException(status_code=503, detail="Los servidores proxy de imágenes están temporalmente bloqueados. Intenta en unos minutos.")

if __name__ == "__main__":
    uvicorn.run(app, host="0.0.0.0", port=7860)