tradein-mvp/backend/app/services/estimator.py

@@ -34,7 +34,7 @@ logger = logging.getLogger(__name__)
 
 
 # ── Constants ────────────────────────────────────────────────────────────────
-DEFAULT_RADIUS_M = 800
+DEFAULT_RADIUS_M = 1000   # ПО ВСТРЕЧЕ ПТИЦЫ: «локация не дальше 800-1000 м»
 FALLBACK_RADIUS_M = 2000
 AREA_TOLERANCE = 0.15        # ±15% площади
 LISTINGS_FRESH_DAYS = 14     # объявления не старше 14 дней

tradein-mvp/backend/app/services/scrapers/cian.py

@@ -65,17 +65,19 @@ class CianScraper(BaseScraper):
         await super().__aexit__(*args)
 
     async def fetch_around(
-        self, lat: float, lon: float, radius_m: int = 1000, rooms: tuple[int, ...] | None = None
+        self, lat: float, lon: float, radius_m: int = 1000,
+        rooms: tuple[int, ...] | None = None, page: int = 1,
     ) -> list[ScrapedLot]:
         """Найти объявления Циан вокруг (lat, lon) в radius_m метрах.
 
         rooms — список из 1,2,3,4 (Cian's room codes). Если None — все.
+        page — страница выдачи (Cian ~28 на страницу).
         """
         self._anchor_lat = lat
         self._anchor_lon = lon
         self._anchor_radius_m = radius_m
 
-        url = self._build_url(lat, lon, radius_m, rooms)
+        url = self._build_url(lat, lon, radius_m, rooms, page)
         try:
             assert self._cffi is not None
             response = await self._cffi.get(url)
@@ -88,28 +90,34 @@ class CianScraper(BaseScraper):
 
         lots = self._parse_html(response.text)
         logger.info(
-            "cian: %d lots around (%.4f, %.4f) rooms=%s", len(lots), lat, lon, rooms
+            "cian: %d lots around (%.4f, %.4f) rooms=%s page=%d",
+            len(lots), lat, lon, rooms, page,
         )
         await self.sleep_between_requests()
         return lots
 
     async def fetch_around_multi_room(
-        self, lat: float, lon: float, radius_m: int = 1000
+        self, lat: float, lon: float, radius_m: int = 1000, pages: int = 3,
     ) -> list[ScrapedLot]:
-        """Скрейп Циан отдельно по 1к/2к/3к/4+ — каждый запрос ~25 лотов, всего ~100.
-        Циан умеет несколько rooms[]=N сразу, но для расширения выборки делаем отдельно.
+        """Скрейп Циан по 1к/2к/3к/4+ × N страниц — расширяет выборку.
+        4 комнаты × 3 страницы × ~28 = до ~330 лотов до дедупликации.
         """
         seen: dict[str, ScrapedLot] = {}
         for rooms in ((1,), (2,), (3,), (4,)):
-            try:
-                lots = await self.fetch_around(lat, lon, radius_m, rooms=rooms)
-            except Exception:
-                logger.exception("cian multi-room fetch failed for rooms=%s", rooms)
-                continue
-            for lot in lots:
-                key = lot.source_id or lot.source_url
-                if key and key not in seen:
-                    seen[key] = lot
+            for page in range(1, pages + 1):
+                try:
+                    lots = await self.fetch_around(lat, lon, radius_m, rooms=rooms, page=page)
+                except Exception:
+                    logger.exception(
+                        "cian multi-room fetch failed rooms=%s page=%d", rooms, page
+                    )
+                    continue
+                if not lots:
+                    break  # пустая страница → дальше нет смысла
+                for lot in lots:
+                    key = lot.source_id or lot.source_url
+                    if key and key not in seen:
+                        seen[key] = lot
         logger.info(
             "cian multi-room: %d unique lots around (%.4f, %.4f)", len(seen), lat, lon
         )
@@ -121,6 +129,7 @@ class CianScraper(BaseScraper):
         lon: float,
         radius_m: int,
         rooms: tuple[int, ...] | None = None,
+        page: int = 1,
     ) -> str:
         """URL для Cian каталога с регионом ЕКБ.
 
@@ -138,6 +147,8 @@ class CianScraper(BaseScraper):
             for r in rooms:
                 # Cian rooms[]=1 для 1к, rooms[]=2 для 2к, и т.д.
                 params.append((f"room{r}", "1"))
+        if page > 1:
+            params.append(("p", page))
         return f"{self.base_url}/cat.php?{urlencode(params)}"
 
     def _parse_html(self, html: str) -> list[ScrapedLot]:

tradein-mvp/backend/app/services/scrapers/n1.py

@@ -63,13 +63,14 @@ class N1Scraper(BaseScraper):
         await super().__aexit__(*args)
 
     async def fetch_around(
-        self, lat: float, lon: float, radius_m: int = 1000, rooms: int | None = None,
+        self, lat: float, lon: float, radius_m: int = 1000,
+        rooms: int | None = None, page: int = 1,
     ) -> list[ScrapedLot]:
         self._anchor_lat = lat
         self._anchor_lon = lon
         self._anchor_radius_m = radius_m
 
-        url = self._build_url(rooms)
+        url = self._build_url(rooms, page)
         try:
             assert self._cffi is not None
             response = await self._cffi.get(url)
@@ -82,35 +83,42 @@ class N1Scraper(BaseScraper):
 
         lots = self._parse_html(response.text)
         logger.info(
-            "n1: %d lots around (%.4f, %.4f) rooms=%s", len(lots), lat, lon, rooms
+            "n1: %d lots around (%.4f, %.4f) rooms=%s page=%d",
+            len(lots), lat, lon, rooms, page,
         )
         await self.sleep_between_requests()
         return lots
 
     async def fetch_around_multi_room(
-        self, lat: float, lon: float, radius_m: int = 1000
+        self, lat: float, lon: float, radius_m: int = 1000, pages: int = 3,
     ) -> list[ScrapedLot]:
-        """Скрейп N1 по сегментам комнат 1/2/3/4 — расширяет выборку."""
+        """Скрейп N1 по сегментам комнат 1/2/3/4 × N страниц — расширяет выборку."""
         seen: dict[str, ScrapedLot] = {}
         for rooms in (1, 2, 3, 4):
-            try:
-                lots = await self.fetch_around(lat, lon, radius_m, rooms=rooms)
-            except Exception:
-                logger.exception("n1 multi-room fetch failed for rooms=%s", rooms)
-                continue
-            for lot in lots:
-                key = lot.source_id or lot.source_url
-                if key and key not in seen:
-                    seen[key] = lot
+            for page in range(1, pages + 1):
+                try:
+                    lots = await self.fetch_around(lat, lon, radius_m, rooms=rooms, page=page)
+                except Exception:
+                    logger.exception(
+                        "n1 multi-room fetch failed rooms=%s page=%d", rooms, page
+                    )
+                    continue
+                if not lots:
+                    break  # пустая страница → дальше нет смысла
+                for lot in lots:
+                    key = lot.source_id or lot.source_url
+                    if key and key not in seen:
+                        seen[key] = lot
         logger.info(
             "n1 multi-room: %d unique lots around (%.4f, %.4f)", len(seen), lat, lon
         )
         return list(seen.values())
 
-    def _build_url(self, rooms: int | None = None) -> str:
+    def _build_url(self, rooms: int | None = None, page: int = 1) -> str:
         """URL каталога вторички ЕКБ.
 
         N1 фильтр комнат: /kupit/kvartiry/vtorichka/1-komnatnye/ итд.
+        Пагинация: ?page=N.
         """
         rooms_segment = {
             1: "1-komnatnye/",
@@ -118,7 +126,10 @@ class N1Scraper(BaseScraper):
             3: "3-komnatnye/",
             4: "4-komnatnye/",
         }.get(rooms or 0, "")
-        return f"{self.base_url}/kupit/kvartiry/vtorichka/{rooms_segment}"
+        url = f"{self.base_url}/kupit/kvartiry/vtorichka/{rooms_segment}"
+        if page > 1:
+            url += f"?page={page}"
+        return url
 
     def _parse_html(self, html: str) -> list[ScrapedLot]:
         tree = HTMLParser(html)