feat: 为出站请求统一接入 HTTP 代理

This commit is contained in:
2026-08-14 16:08:39 +08:00
parent 3dc2aeb3a2
commit fabca9510d
23 changed files with 213 additions and 14 deletions
+2 -1
View File
@@ -16,6 +16,7 @@ from dataclasses import dataclass, field
from typing import Any
from app.shared.config import Settings
from app.shared.proxy import playwright_launch_proxy
from app.scraping.core import site
logger = logging.getLogger(__name__)
@@ -100,7 +101,7 @@ class BrowserFallback:
self._browser = await self._playwright.chromium.launch(
headless=self._settings.browser_headless_effective,
channel=self._settings.browser_channel or None,
proxy=self._settings.playwright_proxy,
proxy=playwright_launch_proxy(self._settings),
timeout=self._settings.browser_launch_timeout_seconds * 1000,
args=["--no-first-run", "--disable-blink-features=AutomationControlled"],
)
+2 -1
View File
@@ -22,6 +22,7 @@ from opentelemetry import trace
from app.scraping.core import rakuma_site as site
from app.shared.config import Settings
from app.shared.proxy import httpx_client_options
from app.shared.errors import (
ItemNotFoundError,
ResourceBusyError,
@@ -51,7 +52,7 @@ class RakumaSession:
headers=site.default_headers(),
timeout=self._settings.request_timeout_seconds,
follow_redirects=True,
proxy=self._settings.httpx_proxy,
**httpx_client_options(self._settings),
http2=True,
)
logger.info(
+2 -1
View File
@@ -23,6 +23,7 @@ from opentelemetry import trace
from app.scraping.core import site
from app.shared.config import Settings
from app.shared.proxy import httpx_client_options
from app.shared.errors import (
ItemNotFoundError,
ResourceBusyError,
@@ -91,7 +92,7 @@ class SiteSession:
headers=site.default_headers(mobile=mobile),
timeout=self._settings.request_timeout_seconds,
follow_redirects=True,
proxy=self._settings.httpx_proxy,
**httpx_client_options(self._settings),
http2=True,
),
)
+27 -1
View File
@@ -8,9 +8,11 @@
「通用 / 仅抓取 / 仅交易 / 仅网关 / 仅交易 worker」分区标注,各进程只读自己那部分。
"""
import socket
from fnmatch import fnmatchcase
from functools import lru_cache
from pathlib import Path
from typing import Literal
from urllib.parse import quote, urlsplit
from pydantic import Field
from pydantic_settings import BaseSettings, SettingsConfigDict
@@ -88,6 +90,8 @@ class Settings(BaseSettings):
proxy_server: str | None = None
proxy_username: str | None = None
proxy_password: str | None = None
# 逗号分隔的直连主机名或通配符。服务间请求不应绕到公网代理。
proxy_bypass: str = "localhost,127.0.0.1,::1,rakuten-api,rakuten-trading,rakuten-gateway"
# ---- OpenTelemetry traces(通用,可选;默认关闭)----
# 启用后把抓取-解析链路以 span 导出到 OTLP/HTTP endpoint,用于排查"抓到
@@ -192,6 +196,8 @@ class Settings(BaseSettings):
proxy["username"] = self.proxy_username
if self.proxy_password:
proxy["password"] = self.proxy_password
if self.proxy_bypass.strip():
proxy["bypass"] = self.proxy_bypass
return proxy
@property
@@ -205,9 +211,29 @@ class Settings(BaseSettings):
scheme, _, rest = self.proxy_server.partition("://")
if not rest:
return self.proxy_server
credentials = f"{self.proxy_username}:{self.proxy_password or ''}"
credentials = f"{quote(self.proxy_username, safe='')}:{quote(self.proxy_password or '', safe='')}"
return f"{scheme}://{credentials}@{rest}"
def proxy_bypasses(self, url: str) -> bool:
"""目标 URL 的主机是否应绕过外部代理。"""
host = urlsplit(url).hostname
if not host:
return False
normalized_host = host.rstrip(".").lower()
for raw_pattern in self.proxy_bypass.split(","):
pattern = raw_pattern.strip().rstrip(".").lower()
if not pattern:
continue
if pattern.startswith("."):
suffix = pattern[1:]
if normalized_host == suffix or normalized_host.endswith(f".{suffix}"):
return True
continue
if fnmatchcase(normalized_host, pattern):
return True
return False
@property
def auth_state_path(self) -> Path:
"""登录态目录的绝对路径,不存在时创建"""
+23
View File
@@ -0,0 +1,23 @@
"""项目出站代理策略;OTel exporter 不使用这里的配置。"""
from __future__ import annotations
from typing import TYPE_CHECKING
if TYPE_CHECKING:
from app.shared.config import Settings
def httpx_client_options(
settings: "Settings", *, target_url: str | None = None
) -> dict[str, object]:
"""返回 HTTPX 客户端统一的代理与环境变量策略。"""
proxy = None if target_url and settings.proxy_bypasses(target_url) else settings.httpx_proxy
options: dict[str, object] = {"trust_env": False}
if proxy is not None:
options["proxy"] = proxy
return options
def playwright_launch_proxy(settings: "Settings") -> dict[str, str] | None:
"""返回 Playwright 启动参数使用的统一代理设置。"""
return settings.playwright_proxy
+1
View File
@@ -61,6 +61,7 @@ def build_container() -> TradingContainer:
client = GatewayClient(
settings.order_gateway_url,
settings.bearer_token,
settings=settings,
timeout=max(60.0, settings.lease_max_wait_seconds + 10),
)
local_db = LocalDB(settings.trading_db_path_resolved)
+2 -1
View File
@@ -37,6 +37,7 @@ from typing import Any
import httpx
from app.shared.config import Settings
from app.shared.proxy import httpx_client_options
from app.shared.errors import NotLoggedInError, UpstreamRequestError
from app.trading.core import auth_site
@@ -110,7 +111,7 @@ class AuthSession:
headers=profile.headers(),
timeout=self._settings.request_timeout_seconds,
follow_redirects=True,
proxy=self._settings.httpx_proxy,
**httpx_client_options(self._settings),
http2=True,
)
self._sites[name] = _SiteAuth(name=name, client=client)
+2 -1
View File
@@ -25,6 +25,7 @@ from typing import Any
import yaml
from app.shared.config import BASE_DIR, Settings
from app.shared.proxy import playwright_launch_proxy
from app.trading.core import auth_site
logger = logging.getLogger(__name__)
@@ -355,7 +356,7 @@ async def login_one(
user_data_dir=str(account.user_data_dir),
headless=False, # 自动登录仍需有头:撞验证码要人工接管
channel=settings.browser_channel or None,
proxy=settings.playwright_proxy,
proxy=playwright_launch_proxy(settings),
user_agent=profile.user_agent,
locale="ja-JP",
timezone_id="Asia/Tokyo",
+6 -1
View File
@@ -16,7 +16,9 @@ from typing import Any
import httpx
from app.shared.config import Settings
from app.shared.errors import AppError
from app.shared.proxy import httpx_client_options
from app.shared.task_state import OrderState, TaskStatus
from app.trading.worker.models import LeaseTask
@@ -29,13 +31,16 @@ class GatewayClient:
一份 AsyncClient 实例贯穿 worker 整个生命周期,连接池由 httpx 管理。
"""
def __init__(self, base_url: str, bearer_token: str, *, timeout: float = 60.0):
def __init__(
self, base_url: str, bearer_token: str, *, settings: Settings, timeout: float = 60.0
):
# 末尾去斜杠,避免 base + "/api/..." 拼出双斜杠
self._base_url = base_url.rstrip("/")
self._client = httpx.AsyncClient(
base_url=self._base_url,
headers={"Authorization": f"Bearer {bearer_token}"},
timeout=timeout,
**httpx_client_options(settings, target_url=self._base_url),
)
async def aclose(self) -> None:
+2
View File
@@ -97,6 +97,7 @@ from app.shared.purchase_contract import (
basket_domain_of,
inventory_flag_for,
)
from app.shared.proxy import playwright_launch_proxy
from app.shared.task_state import OrderState
from app.trading.core import auth_site
from app.trading.worker.models import LeaseTask
@@ -556,6 +557,7 @@ class SiteInteractor:
# 状态的操作(加购/结算/支付)都必须走非无头浏览器
headless=False,
channel=self._settings.browser_channel or None,
proxy=playwright_launch_proxy(self._settings),
args=["--no-first-run", "--disable-blink-features=AutomationControlled"],
)
self._context = await self._browser.new_context(