..

finder/zoopla.py

@@ -1,4 +1,4 @@
-"""Zoopla (zoopla.co.uk) scraper — sale properties.
+"""Zoopla (zoopla.co.uk) scraper: sale properties.
 
 Zoopla is behind Cloudflare Turnstile (managed interactive challenge), which
 blocks non-browser HTTP clients and even Playwright with stealth patches. Only
@@ -522,7 +522,7 @@ def _gluetun_set_vpn_status(client: httpx.Client, status: str) -> bool:
         return False
     if resp.status_code == 401:
         log.warning(
-            "Gluetun vpn/status %s: 401 Unauthorized — the API key must be "
+            "Gluetun vpn/status %s: 401 Unauthorized. The API key must be "
             "authorised for 'PUT /v1/vpn/status' in Gluetun's auth config.toml",
             status,
         )
@@ -593,7 +593,7 @@ def _rotate_and_retry_challenge(page, max_rotations: int) -> bool:
     """Rotate IP and reload until the challenge clears. Returns True on success."""
     for attempt in range(1, max_rotations + 1):
         log.warning(
-            "Cloudflare Turnstile challenge — rotating Gluetun IP (attempt %d/%d)",
+            "Cloudflare Turnstile challenge, rotating Gluetun IP (attempt %d/%d)",
             attempt, max_rotations,
         )
         if not _rotate_gluetun_ip():
@@ -637,7 +637,7 @@ def _wait_for_turnstile(page, headless_mode: bool | str) -> None:
     if not _is_turnstile_challenge(page):
         return
 
-    # Try Gluetun IP rotation first — works in any mode and is the only option
+    # Try Gluetun IP rotation first: works in any mode and is the only option
     # in headless/unattended runs where no human can click the challenge.
     max_rotations = _gluetun_max_rotations()
     if max_rotations > 0 and _rotate_and_retry_challenge(page, max_rotations):
@@ -654,7 +654,7 @@ def _wait_for_turnstile(page, headless_mode: bool | str) -> None:
 
     timeout = _challenge_timeout_seconds()
     log.warning(
-        "Gluetun rotation insufficient — falling back to interactive solve. "
+        "Gluetun rotation insufficient. Falling back to interactive solve. "
         "Complete the Cloudflare challenge in the Zoopla browser window; "
         "waiting up to %ds. Profile: %s",
         timeout,
@@ -727,7 +727,7 @@ def launch_browser():
         page.goto(f"{ZOOPLA_BASE}/", wait_until="domcontentloaded", timeout=60000)
         _wait_for_turnstile(page, headless_mode)
 
-        log.info("Zoopla browser ready — title: %s", page.title())
+        log.info("Zoopla browser ready, title: %s", page.title())
         time.sleep(2)
 
         # Dismiss cookie consent
@@ -1117,14 +1117,14 @@ def _extract_outcode(text: str) -> str | None:
 #               "outcode":...,"postcode":...,"uprn":...} feeding the map widgets.
 # Nearby points of interest (stations, schools, EV chargers) and comparable
 # listings carry their own "coordinates" too, but never inside the property's
-# own "location" / "address":{"fullAddress" wrapper — so the wrapper, not a
+# own "location" / "address":{"fullAddress" wrapper, so the wrapper, not a
 # loose coordinates object, is what we anchor on (see parse_detail_geo).
 
 # listingId -> parsed detail dict (or None). Failures are cached too, so a
 # broken listing is not re-fetched within a run (the same listing reappears
 # across overlapping outcode searches). Seeded from / dumped to a persistent
 # on-disk cache by the orchestrator (see postcode_cache.py) so a recurring
-# scrape only re-fetches newly-listed properties — the biggest saving for
+# scrape only re-fetches newly-listed properties, the biggest saving for
 # Zoopla, whose detail fetch drives a real browser tab.
 _detail_cache: dict[str, dict | None] = {}
 
@@ -1144,7 +1144,7 @@ _LISTING_ID_RE = re.compile(r"/details/(\d+)/?")
 # The property's own location is carried by a `"location":{...}` wrapper and a
 # twin `"address":{"fullAddress":...}` widget object. We anchor on those
 # wrappers (and capture their full object body, which contains exactly one
-# nested object — `coordinates`) rather than scanning for loose coordinate
+# nested object, `coordinates`) rather than scanning for loose coordinate
 # objects: nearby points of interest (stations/schools/EV chargers) and
 # comparable/"similar" listings also embed coordinates, but never inside the
 # property's own `"location"` / `"address":{"fullAddress"` wrapper, so the
@@ -1225,7 +1225,7 @@ def parse_detail_geo(html: str, search_outcode: str | None = None) -> dict | Non
         # twin, not the `location` wrapper we anchor coordinates on. Pull it from
         # the twin that shares this property's uprn; when there is no uprn to
         # disambiguate, fall back to the first twin (document order = primary
-        # listing), but never guess a twin when a uprn exists and none matches —
+        # listing), but never guess a twin when a uprn exists and none matches:
         # that would risk grabbing a comparable listing's address.
         if result is None or result.get("full_address"):
             return result
@@ -1245,7 +1245,7 @@ def parse_detail_geo(html: str, search_outcode: str | None = None) -> dict | Non
             result["full_address"] = first
         return result
 
-    # Strategy 1 — the property's own `location` wrapper (authoritative). Take
+    # Strategy 1: the property's own `location` wrapper (authoritative). Take
     # the first match (the primary listing precedes any comparables in the
     # flight stream), but prefer one whose outcode matches the searched outcode.
     first_location = None
@@ -1269,7 +1269,7 @@ def parse_detail_geo(html: str, search_outcode: str | None = None) -> dict | Non
     if first_location is not None:
         return attach_full_address(first_location)
 
-    # Strategy 2 — the `address` map-widget twin (same coordinates, backup).
+    # Strategy 2: the `address` map-widget twin (same coordinates, backup).
     for match in _DETAIL_ADDRESS_RE.finditer(buf):
         full_address = match.group(1) or None
         body = match.group(2)
@@ -1284,7 +1284,7 @@ def parse_detail_geo(html: str, search_outcode: str | None = None) -> dict | Non
 
 
 def _detail_cache_key(listing_url: str) -> str:
-    """Cache key for a listing detail page — its numeric id when present."""
+    """Cache key for a listing detail page: its numeric id when present."""
     id_match = _LISTING_ID_RE.search(listing_url)
     return id_match.group(1) if id_match else listing_url
 
@@ -1388,7 +1388,7 @@ def transform_property(
     location comes from the listing's detail page (see ``parse_detail_geo`` /
     ``_fetch_listing_detail``), passed in as ``detail``. When detail-page
     coordinates are available we resolve the nearest postcode via the spatial
-    index — mirroring rightmove/onthemarket — and only fall back to the coarse
+    index (mirroring rightmove/onthemarket) and only fall back to the coarse
     outcode centroid when no detail location could be obtained."""
     price = parse_int_value(raw.get("price")) or 0
 
@@ -1563,7 +1563,7 @@ def search_outcode(
         if detail_budget_seconds is not None:
             detail_deadline = time.monotonic() + detail_budget_seconds
 
-    # Always try extraction even if result count is 0 — the count regex may
+    # Always try extraction even if result count is 0: the count regex may
     # not match Zoopla's current text format, but listings may still be in DOM
     raw_listings = _paginate(
         page,
@@ -1577,7 +1577,7 @@ def search_outcode(
     if not raw_listings:
         if total_results > 0:
             log.debug(
-                "Zoopla %s %s: page claims %d results but extraction found 0 — "
+                "Zoopla %s %s: page claims %d results but extraction found 0; "
                 "DOM selectors may need updating",
                 outcode, "BUY", total_results,
             )