Pourquoi construire un calque mandataire
Lorsque votre projet de grattage dépasse un seul script, vous avez besoin d'une couche d'abstraction entre votre code d'application et le fournisseur de proxy. Une couche d'intergiciel gère la sélection de proxy, l'authentification, réessayer la logique, la défectuosité et l'enregistrement.
Sans intergiciel, chaque racleur de votre base de code duplique la configuration proxy, la gestion des erreurs et la logique de rotation. Ce guide vous montre comment concevoir et construire un intergiciel proxy de qualité de production dans Python et Node.js, en utilisant Proxies résidentielles tournantes de ProxyHat.
Un intergiciel mandataire se situe entre votre application et Internet. Il intercepte les demandes sortantes, les canalise à travers les procurations, gère les échecs et renvoie les réponses propres à votre code d'application.
Aperçu de l'architecture
Un intergiciel proxy bien conçu comporte quatre composantes principales :
| Composante | Responsabilité | Pourquoi ça compte |
|---|---|---|
| Gestionnaire mandataire | Gère les références, les sessions et le géo-ciblage par procuration | Centralise la configuration de proxy |
| Intercepteur de demande | Enveloppe les requêtes sortantes avec les paramètres et en-têtes de proxy | Élimine la plaque de chaudière dans les racleurs |
| Moteur de réessayer | Les requêtes ont échoué avec rétrocession et rotation par procuration | Augmente le taux de réussite de 70% à 95%+ |
| Collecteur de métriques | Suivre la latence, les taux de réussite et les modèles d'erreur | Permet d'optimiser les données |
Mise en œuvre du Middleware Python
Gestionnaire de proxy de base
import uuid
import time
import logging
from dataclasses import dataclass, field
from typing import Optional
logger = logging.getLogger(__name__)
@dataclass
class ProxyConfig:
gateway: str = "gate.proxyhat.com"
http_port: int = 8080
socks5_port: int = 1080
username: str = "USERNAME"
password: str = "PASSWORD"
@dataclass
class RequestMetrics:
total: int = 0
success: int = 0
failed: int = 0
retries: int = 0
total_latency: float = 0.0
@property
def success_rate(self) -> float:
return (self.success / self.total * 100) if self.total > 0 else 0.0
@property
def avg_latency(self) -> float:
return (self.total_latency / self.success) if self.success > 0 else 0.0
class ProxyManager:
"""Manages proxy sessions, geo-targeting, and protocol selection."""
def __init__(self, config: Optional[ProxyConfig] = None):
self.config = config or ProxyConfig()
self.metrics = RequestMetrics()
def get_proxy_url(
self,
protocol: str = "http",
country: Optional[str] = None,
session_id: Optional[str] = None,
sticky: bool = False,
) -> str:
username = self.config.username
# Add session for sticky IPs
if sticky and not session_id:
session_id = uuid.uuid4().hex[:8]
if session_id:
username = f"{username}-session-{session_id}"
# Add country targeting
if country:
username = f"{username}-country-{country}"
if protocol == "socks5":
port = self.config.socks5_port
scheme = "socks5h"
else:
port = self.config.http_port
scheme = "http"
return f"{scheme}://{username}:{self.config.password}@{self.config.gateway}:{port}"
def get_requests_proxies(self, **kwargs) -> dict:
"""Return a proxies dict compatible with the requests library."""
url = self.get_proxy_url(**kwargs)
return {"http": url, "https": url}
Moteur de réessayer avec moteur de recul
import requests
from requests.exceptions import RequestException
RETRYABLE_STATUS_CODES = {403, 429, 500, 502, 503, 504}
class RetryEngine:
"""Retries failed requests with exponential backoff and proxy rotation."""
def __init__(
self,
proxy_manager: ProxyManager,
max_retries: int = 3,
base_delay: float = 1.0,
max_delay: float = 30.0,
rotate_on_failure: bool = True,
):
self.proxy_manager = proxy_manager
self.max_retries = max_retries
self.base_delay = base_delay
self.max_delay = max_delay
self.rotate_on_failure = rotate_on_failure
def _should_retry(self, response: Optional[requests.Response], error: Optional[Exception]) -> bool:
if error:
return True
if response and response.status_code in RETRYABLE_STATUS_CODES:
return True
return False
def _get_delay(self, attempt: int) -> float:
delay = self.base_delay * (2 ** attempt)
return min(delay, self.max_delay)
def execute(
self,
method: str,
url: str,
country: Optional[str] = None,
sticky: bool = False,
**request_kwargs,
) -> requests.Response:
session_id = uuid.uuid4().hex[:8] if sticky else None
last_error = None
for attempt in range(self.max_retries + 1):
# Rotate proxy on retry if enabled
if attempt > 0 and self.rotate_on_failure:
session_id = uuid.uuid4().hex[:8]
self.proxy_manager.metrics.retries += 1
proxies = self.proxy_manager.get_requests_proxies(
country=country, session_id=session_id
)
self.proxy_manager.metrics.total += 1
start_time = time.time()
try:
response = requests.request(
method, url,
proxies=proxies,
timeout=request_kwargs.pop("timeout", 30),
**request_kwargs,
)
latency = time.time() - start_time
self.proxy_manager.metrics.total_latency += latency
if self._should_retry(response, None) and attempt < self.max_retries:
logger.warning(
f"Retryable status {response.status_code} for {url} "
f"(attempt {attempt + 1}/{self.max_retries + 1})"
)
time.sleep(self._get_delay(attempt))
continue
self.proxy_manager.metrics.success += 1
return response
except RequestException as e:
last_error = e
latency = time.time() - start_time
logger.error(f"Request failed for {url}: {e} (attempt {attempt + 1})")
if attempt < self.max_retries:
time.sleep(self._get_delay(attempt))
continue
self.proxy_manager.metrics.failed += 1
raise last_error or Exception(f"All retries exhausted for {url}")
Classe de Middleware complète
class ProxyMiddleware:
"""High-level middleware that wraps all proxy logic."""
def __init__(
self,
config: Optional[ProxyConfig] = None,
max_retries: int = 3,
default_country: Optional[str] = None,
):
self.proxy_manager = ProxyManager(config)
self.retry_engine = RetryEngine(
proxy_manager=self.proxy_manager,
max_retries=max_retries,
)
self.default_country = default_country
self.default_headers = {
"Accept": "text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8",
"Accept-Language": "en-US,en;q=0.9",
"Accept-Encoding": "gzip, deflate, br",
}
def get(self, url: str, **kwargs) -> requests.Response:
kwargs.setdefault("headers", self.default_headers)
kwargs.setdefault("country", self.default_country)
return self.retry_engine.execute("GET", url, **kwargs)
def post(self, url: str, **kwargs) -> requests.Response:
kwargs.setdefault("headers", self.default_headers)
kwargs.setdefault("country", self.default_country)
return self.retry_engine.execute("POST", url, **kwargs)
@property
def stats(self) -> dict:
m = self.proxy_manager.metrics
return {
"total_requests": m.total,
"successful": m.success,
"failed": m.failed,
"retries": m.retries,
"success_rate": f"{m.success_rate:.1f}%",
"avg_latency": f"{m.avg_latency:.3f}s",
}
# Usage
middleware = ProxyMiddleware(max_retries=3, default_country="us")
# Simple GET — all proxy logic is hidden
response = middleware.get("https://httpbin.org/ip")
print(response.json())
# Geo-targeted request
response = middleware.get("https://example.com/pricing", country="de")
print(response.text[:200])
# Check performance
print(middleware.stats)
Node.js Mise en œuvre du Middleware
Gestionnaire de proxy de base
const crypto = require('crypto');
const { HttpsProxyAgent } = require('https-proxy-agent');
const { SocksProxyAgent } = require('socks-proxy-agent');
class ProxyManager {
constructor(config = {}) {
this.gateway = config.gateway || 'gate.proxyhat.com';
this.httpPort = config.httpPort || 8080;
this.socks5Port = config.socks5Port || 1080;
this.username = config.username || 'USERNAME';
this.password = config.password || 'PASSWORD';
this.metrics = {
total: 0,
success: 0,
failed: 0,
retries: 0,
totalLatency: 0,
};
}
getProxyUrl({ protocol = 'http', country, sessionId, sticky = false } = {}) {
let username = this.username;
if (sticky && !sessionId) {
sessionId = crypto.randomBytes(4).toString('hex');
}
if (sessionId) username += `-session-${sessionId}`;
if (country) username += `-country-${country}`;
const port = protocol === 'socks5' ? this.socks5Port : this.httpPort;
const scheme = protocol === 'socks5' ? 'socks5h' : 'http';
return `${scheme}://${username}:${this.password}@${this.gateway}:${port}`;
}
createAgent(options = {}) {
const url = this.getProxyUrl(options);
if (options.protocol === 'socks5') {
return new SocksProxyAgent(url);
}
return new HttpsProxyAgent(url);
}
get stats() {
const m = this.metrics;
return {
total: m.total,
success: m.success,
failed: m.failed,
retries: m.retries,
successRate: m.total > 0 ? ((m.success / m.total) * 100).toFixed(1) + '%' : '0%',
avgLatency: m.success > 0 ? (m.totalLatency / m.success).toFixed(3) + 's' : '0s',
};
}
}
Moteur de réessayer
const RETRYABLE_STATUS_CODES = new Set([403, 429, 500, 502, 503, 504]);
class RetryEngine {
constructor(proxyManager, options = {}) {
this.proxyManager = proxyManager;
this.maxRetries = options.maxRetries || 3;
this.baseDelay = options.baseDelay || 1000;
this.maxDelay = options.maxDelay || 30000;
this.rotateOnFailure = options.rotateOnFailure !== false;
}
_getDelay(attempt) {
return Math.min(this.baseDelay * Math.pow(2, attempt), this.maxDelay);
}
_sleep(ms) {
return new Promise(resolve => setTimeout(resolve, ms));
}
async execute(url, options = {}) {
const { country, sticky, ...fetchOptions } = options;
let sessionId = sticky ? crypto.randomBytes(4).toString('hex') : undefined;
let lastError;
for (let attempt = 0; attempt <= this.maxRetries; attempt++) {
if (attempt > 0 && this.rotateOnFailure) {
sessionId = crypto.randomBytes(4).toString('hex');
this.proxyManager.metrics.retries++;
}
const agent = this.proxyManager.createAgent({ country, sessionId });
this.proxyManager.metrics.total++;
const startTime = Date.now();
try {
const response = await fetch(url, {
...fetchOptions,
agent,
signal: AbortSignal.timeout(fetchOptions.timeout || 30000),
});
const latency = (Date.now() - startTime) / 1000;
this.proxyManager.metrics.totalLatency += latency;
if (RETRYABLE_STATUS_CODES.has(response.status) && attempt < this.maxRetries) {
console.warn(
`Retryable status ${response.status} for ${url} ` +
`(attempt ${attempt + 1}/${this.maxRetries + 1})`
);
await this._sleep(this._getDelay(attempt));
continue;
}
this.proxyManager.metrics.success++;
return response;
} catch (err) {
lastError = err;
console.error(`Request failed for ${url}: ${err.message} (attempt ${attempt + 1})`);
if (attempt < this.maxRetries) {
await this._sleep(this._getDelay(attempt));
continue;
}
this.proxyManager.metrics.failed++;
}
}
throw lastError || new Error(`All retries exhausted for ${url}`);
}
}
Classe de Middleware complète
class ProxyMiddleware {
constructor(options = {}) {
this.proxyManager = new ProxyManager(options.proxy);
this.retryEngine = new RetryEngine(this.proxyManager, {
maxRetries: options.maxRetries || 3,
});
this.defaultCountry = options.defaultCountry || null;
this.defaultHeaders = {
'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
'Accept-Language': 'en-US,en;q=0.9',
'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36',
};
}
async get(url, options = {}) {
return this.retryEngine.execute(url, {
method: 'GET',
headers: { ...this.defaultHeaders, ...options.headers },
country: options.country || this.defaultCountry,
...options,
});
}
async post(url, body, options = {}) {
return this.retryEngine.execute(url, {
method: 'POST',
headers: { ...this.defaultHeaders, ...options.headers },
body: typeof body === 'string' ? body : JSON.stringify(body),
country: options.country || this.defaultCountry,
...options,
});
}
get stats() {
return this.proxyManager.stats;
}
}
// Usage
const middleware = new ProxyMiddleware({
maxRetries: 3,
defaultCountry: 'us',
});
// Simple GET
const response = await middleware.get('https://httpbin.org/ip');
const data = await response.json();
console.log(data);
// Check stats
console.log(middleware.stats);
Ajout de l'échec du fournisseur
Pour le grattage critique de mission, vous pouvez étendre l'intergiciel pour prendre en charge plusieurs fournisseurs de proxy avec un basculement automatique.
class MultiProviderMiddleware:
"""Falls back to secondary proxy providers on failure."""
def __init__(self, providers: list[ProxyConfig], max_retries: int = 3):
self.providers = [
ProxyMiddleware(config=cfg, max_retries=max_retries)
for cfg in providers
]
def get(self, url: str, **kwargs) -> requests.Response:
last_error = None
for i, provider in enumerate(self.providers):
try:
response = provider.get(url, **kwargs)
if response.status_code < 400:
return response
except Exception as e:
last_error = e
logger.warning(f"Provider {i} failed for {url}: {e}")
continue
raise last_error or Exception(f"All providers failed for {url}")
# Primary: ProxyHat | Fallback: backup provider
providers = [
ProxyConfig(
gateway="gate.proxyhat.com",
username="USERNAME",
password="PASSWORD",
),
ProxyConfig(
gateway="backup-proxy.example.com",
username="BACKUP_USER",
password="BACKUP_PASS",
),
]
middleware = MultiProviderMiddleware(providers)
response = middleware.get("https://example.com/data")
Le Middleware Async (Python)
Pour le grattage à haut débit, utilisez la version async avec aiohttp. Pour plus de modèles de concordance, voir notre guide sur à l'échelle des demandes de procuration avec contrôle de la concordance.
import aiohttp
import asyncio
class AsyncProxyMiddleware:
def __init__(self, config: Optional[ProxyConfig] = None, max_retries: int = 3):
self.proxy_manager = ProxyManager(config)
self.max_retries = max_retries
async def get(self, url: str, country: Optional[str] = None) -> str:
for attempt in range(self.max_retries + 1):
session_id = uuid.uuid4().hex[:8]
proxy_url = self.proxy_manager.get_proxy_url(
country=country, session_id=session_id
)
try:
async with aiohttp.ClientSession() as session:
async with session.get(
url, proxy=proxy_url, timeout=aiohttp.ClientTimeout(total=30)
) as response:
if response.status in RETRYABLE_STATUS_CODES and attempt < self.max_retries:
await asyncio.sleep(2 ** attempt)
continue
self.proxy_manager.metrics.success += 1
return await response.text()
except Exception as e:
if attempt == self.max_retries:
self.proxy_manager.metrics.failed += 1
raise
await asyncio.sleep(2 ** attempt)
# Usage
async def main():
middleware = AsyncProxyMiddleware(max_retries=3)
urls = [f"https://example.com/page/{i}" for i in range(100)]
tasks = [middleware.get(url) for url in urls]
results = await asyncio.gather(*tasks, return_exceptions=True)
success = sum(1 for r in results if not isinstance(r, Exception))
print(f"Success: {success}/{len(results)}")
print(middleware.proxy_manager.metrics)
asyncio.run(main())
Intégration aux cadres de scraping
Le middleware s'intègre proprement aux motifs de grattage communs. Voici comment l'utiliser avec Playwright et avec standard Demandes de Python.
# With Playwright (Node.js)
const { chromium } = require('playwright');
class PlaywrightProxyMiddleware {
constructor(proxyManager) {
this.proxyManager = proxyManager;
this.browser = null;
}
async init() {
this.browser = await chromium.launch();
}
async scrape(url, options = {}) {
const proxyUrl = this.proxyManager.getProxyUrl({
sessionId: crypto.randomBytes(4).toString('hex'),
country: options.country,
});
const parsed = new URL(proxyUrl);
const context = await this.browser.newContext({
proxy: {
server: `${parsed.protocol}//${parsed.hostname}:${parsed.port}`,
username: decodeURIComponent(parsed.username),
password: decodeURIComponent(parsed.password),
},
});
const page = await context.newPage();
try {
await page.goto(url, { timeout: 30000 });
return await page.content();
} finally {
await context.close();
}
}
async close() {
if (this.browser) await this.browser.close();
}
}
Pour les modèles de proxy spécifiques à Playwright, consultez notre Guide de rotation proxy de Playwright. Pour surveiller les mesures que votre intergiciel collecte, voir Surveillance de la performance proxy.
Explorer Python SDK, Numéro SDKou Allez au SDK pour les fonctionnalités d'intergiciel intégré, et vérifier Prix ProxyHat pour commencer. Pour une documentation complète sur l'intégration, visitez Docs.proxyhat.com.






