Extract shared proxy module; make it available to direct scraper; remove dead code

Proxy config, DataImpulse sticky-session username syntax, and Playwright/
httpx proxy builders now live in one place (services/proxy.py) instead of
being duplicated across the Booking.com backend and the /config/proxy test
endpoint. Both scrapers consume it.

- services/proxy.py: load_config/normalize (DB-authoritative, env fallback),
  new_session_id, username, playwright_proxy, httpx_proxy_url
- PlaywrightLocalBackend delegates proxy building to the module
- get_scraper_backend factory uses proxy.load_config (one resolution path)
- test_proxy_config endpoint uses the shared URL builder; httpx proxies= ->
  proxy= (forward-compatible, 0.28-safe)
- Direct booking-engine scraper (httpx) can now route through the same proxy,
  gated by the direct_scraper_use_proxy flag (default off, plumbing ready)

Dead code removed: set_scraper_paused (never called — rotate-on-block
replaced pause-on-block), get_competitor_matrix / get_hotels_list /
update_hotel_tier (endpoints have their own SQL), unused PROXY_KEYS tuple.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

backend/services/scraper_backends/playwright_local.py

@@ -7,7 +7,6 @@ No proxy - direct connection. Suitable for low-volume scraping.
 
 import asyncio
 import logging
-import os
 import random
 import re
 from datetime import date
@@ -15,6 +14,8 @@ from decimal import Decimal, InvalidOperation
 from typing import List, Optional
 from urllib.parse import urlencode, urlparse, parse_qs
 
+from services import proxy as proxy_util
+
 
 def location_params_from_url(url: str) -> dict:
     """Pull the stable destination params (ss, dest_id, dest_type) out of a
@@ -109,8 +110,9 @@ class PlaywrightLocalBackend(ScraperBackend):
         Args:
             proxy_config: Optional proxy configuration. Falls back to the
                 BOOKING_PROXY_* environment variables when not supplied.
+                See services/proxy.py for the shared config/build helpers.
         """
-        self.proxy_config = proxy_config if proxy_config is not None else self._proxy_from_env()
+        self.proxy_config = proxy_config if proxy_config is not None else proxy_util.config_from_env()
         self._playwright = None
         self._browser: Optional[Browser] = None
         self._context: Optional[BrowserContext] = None
@@ -122,40 +124,12 @@ class PlaywrightLocalBackend(ScraperBackend):
                 f"(country={self.proxy_config.get('country', 'gb')}, session={self._session_id})"
             )
 
-    @staticmethod
-    def _proxy_from_env() -> dict:
-        """Read proxy settings from BOOKING_PROXY_* env vars (empty = disabled)."""
-        host = os.getenv("BOOKING_PROXY_HOST", "").strip()
-        if not host:
-            return {}
-        return {
-            "host": host,
-            "port": os.getenv("BOOKING_PROXY_PORT", "823").strip(),
-            "username": os.getenv("BOOKING_PROXY_USERNAME", "").strip(),
-            "password": os.getenv("BOOKING_PROXY_PASSWORD", "").strip(),
-            "country": os.getenv("BOOKING_PROXY_COUNTRY", "gb").strip(),
-        }
-
     def _proxy_enabled(self) -> bool:
-        return bool(self.proxy_config.get("host") and self.proxy_config.get("username"))
+        return proxy_util.is_enabled(self.proxy_config)
 
     def _new_session_id(self):
         """Pick a fresh sticky-session id — DataImpulse pins one IP per id."""
-        self._session_id = f"hnf{random.randint(100000, 999999)}"
-
-    def _proxy_launch_arg(self) -> Optional[dict]:
-        """Build Playwright's proxy dict, encoding country + sticky session
-        into the username per DataImpulse's syntax: LOGIN__cr.gb;sessid.ID."""
-        if not self._proxy_enabled():
-            return None
-        username = f"{self.proxy_config['username']}__cr.{self.proxy_config.get('country', 'gb')}"
-        if self._session_id:
-            username += f";sessid.{self._session_id}"
-        return {
-            "server": f"http://{self.proxy_config['host']}:{self.proxy_config['port']}",
-            "username": username,
-            "password": self.proxy_config["password"],
-        }
+        self._session_id = proxy_util.new_session_id()
 
     async def rotate_session(self):
         """Burn the current proxy IP and warm cache; the next scrape gets a
@@ -198,7 +172,7 @@ class PlaywrightLocalBackend(ScraperBackend):
             # Proxy is set at launch (Chromium binds the sticky session, which
             # lives in the username, at the network layer). Rotating the IP
             # therefore relaunches the browser — see rotate_session().
-            proxy = self._proxy_launch_arg()
+            proxy = proxy_util.playwright_proxy(self.proxy_config, self._session_id)
             if proxy:
                 launch_kwargs['proxy'] = proxy
             self._browser = await self._playwright.chromium.launch(**launch_kwargs)