Refactor scraper to use direct requests and bump to 0.8.18

- Replace session-based approach with direct requests calls
- Move headers to module-level DEFAULT_HEADERS constant
- Temporarily disable extra_property_data feature

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

homeharvest/core/scrapers/__init__.py

@@ -2,8 +2,6 @@ from __future__ import annotations
 from typing import Union
 
 import requests
-from requests.adapters import HTTPAdapter
-from urllib3.util.retry import Retry
 import uuid
 from ...exceptions import AuthenticationError
 from .models import Property, ListingType, SiteName, SearchPropertyType, ReturnType
@@ -11,6 +9,27 @@ import json
 from pydantic import BaseModel
 
 
+DEFAULT_HEADERS = {
+    'Content-Type': 'application/json',
+    'Accept': '*/*',
+    'Accept-Language': 'en-US,en;q=0.9',
+    'Cache-Control': 'no-cache',
+    'Origin': 'https://www.realtor.com',
+    'Pragma': 'no-cache',
+    'Referer': 'https://www.realtor.com/',
+    'rdc-client-name': 'RDC_WEB_SRP_FS_PAGE',
+    'rdc-client-version': '3.0.2515',
+    'sec-ch-ua': '"Google Chrome";v="135", "Not-A.Brand";v="8", "Chromium";v="135"',
+    'sec-ch-ua-mobile': '?0',
+    'sec-ch-ua-platform': '"macOS"',
+    'sec-fetch-dest': 'empty',
+    'sec-fetch-mode': 'cors',
+    'sec-fetch-site': 'same-site',
+    'User-Agent': 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/135.0.0.0 Safari/537.36',
+    'x-is-bot': 'false',
+}
+
+
 class ScraperInput(BaseModel):
     location: str
     listing_type: ListingType | list[ListingType] | None
@@ -60,8 +79,6 @@ class ScraperInput(BaseModel):
 
 
 class Scraper:
-    session = None
-
     def __init__(
         self,
         scraper_input: ScraperInput,
@@ -69,36 +86,8 @@ class Scraper:
         self.location = scraper_input.location
         self.listing_type = scraper_input.listing_type
         self.property_type = scraper_input.property_type
-
+        self.proxy = scraper_input.proxy
-        if not self.session:
+        self.proxies = {"http": self.proxy, "https": self.proxy} if self.proxy else None
-            Scraper.session = requests.Session()
-            retries = Retry(
-                total=3, backoff_factor=4, status_forcelist=[429, 403], allowed_methods=frozenset(["GET", "POST"])
-            )
-
-            adapter = HTTPAdapter(max_retries=retries, pool_connections=10, pool_maxsize=20)
-            Scraper.session.mount("http://", adapter)
-            Scraper.session.mount("https://", adapter)
-            Scraper.session.headers.update(
-                {
-                    'Host': 'api.frontdoor.realtor.com',
-                    'rdc-ab-test-client': 'ios_for_sale',
-                    'Content-Type': 'application/json',
-                    'apollographql-client-version': '26.9.25-26.9.25.0774600',
-                    'Accept': '*/*',
-                    'Accept-Language': 'en-US,en;q=0.9',
-                    'rdc-client-version': '26.9.25',
-                    'X-APOLLO-OPERATION-TYPE': 'query',
-                    'rdc-client-name': 'RDC_NATIVE_MOBILE-iPhone-com.move.Realtor',
-                    'apollographql-client-name': 'com.move.Realtor-apollo-ios',
-                    'User-Agent': 'Realtor.com/26.9.25.0774600 CFNetwork/3860.200.71 Darwin/25.1.0',
-                }
-            )
-
-        if scraper_input.proxy:
-            proxy_url = scraper_input.proxy
-            proxies = {"http": proxy_url, "https": proxy_url}
-            self.session.proxies.update(proxies)
 
         self.listing_type = scraper_input.listing_type
         self.radius = scraper_input.radius
@@ -109,7 +98,7 @@ class Scraper:
         self.date_from_precision = scraper_input.date_from_precision
         self.date_to_precision = scraper_input.date_to_precision
         self.foreclosure = scraper_input.foreclosure
-        self.extra_property_data = scraper_input.extra_property_data
+        self.extra_property_data = False  # TODO: temporarily disabled
         self.exclude_pending = scraper_input.exclude_pending
         self.limit = scraper_input.limit
         self.offset = scraper_input.offset

homeharvest/core/scrapers/realtor/__init__.py

@@ -8,25 +8,27 @@ This module implements the scraper for realtor.com
 from __future__ import annotations
 
 import json
+import requests
 from concurrent.futures import ThreadPoolExecutor, as_completed
-from datetime import datetime
 from json import JSONDecodeError
 from typing import Dict, Union
 
 from tenacity import (
     retry,
     retry_if_exception_type,
+    retry_if_not_exception_type,
     wait_exponential,
     stop_after_attempt,
 )
 
-from .. import Scraper
+from .. import Scraper, DEFAULT_HEADERS
+from ....exceptions import AuthenticationError
 from ..models import (
     Property,
     ListingType,
     ReturnType
 )
-from .queries import GENERAL_RESULTS_QUERY, SEARCH_HOMES_DATA, HOMES_DATA, HOME_FRAGMENT
+from .queries import GENERAL_RESULTS_QUERY, HOMES_DATA, SEARCH_SUGGESTIONS_QUERY
 from .processors import (
     process_property,
     process_extra_property_details,
@@ -35,35 +37,53 @@ from .processors import (
 
 
 class RealtorScraper(Scraper):
-    SEARCH_GQL_URL = "https://api.frontdoor.realtor.com/graphql"
+    SEARCH_GQL_URL = "https://www.realtor.com/frontdoor/graphql"
     NUM_PROPERTY_WORKERS = 20
     DEFAULT_PAGE_SIZE = 200
 
     def __init__(self, scraper_input):
         super().__init__(scraper_input)
 
+    @staticmethod
+    def _minify_query(query: str) -> str:
+        """Minify GraphQL query by collapsing whitespace to single spaces."""
+        # Split on whitespace, filter empty strings, join with single space
+        return ' '.join(query.split())
+
     def _graphql_post(self, query: str, variables: dict, operation_name: str) -> dict:
         """
-        Execute a GraphQL query with operation-specific headers.
+        Execute a GraphQL query.
 
         Args:
             query: GraphQL query string (must include operationName matching operation_name param)
             variables: Query variables dictionary
-            operation_name: Name of the GraphQL operation for Apollo headers
+            operation_name: Name of the GraphQL operation
 
         Returns:
             Response JSON dictionary
         """
-        # Set operation-specific header (must match query's operationName)
-        self.session.headers['X-APOLLO-OPERATION-NAME'] = operation_name
-
         payload = {
-            "operationName": operation_name,  # Include in payload
+            "operationName": operation_name,
-            "query": query,
+            "query": self._minify_query(query),
             "variables": variables,
         }
 
-        response = self.session.post(self.SEARCH_GQL_URL, json=payload)
+        response = requests.post(
+            self.SEARCH_GQL_URL,
+            headers=DEFAULT_HEADERS,
+            data=json.dumps(payload, separators=(',', ':')),
+            proxies=self.proxies
+        )
+
+        if response.status_code == 403:
+            if not self.proxy:
+                raise AuthenticationError(
+                    "Received 403 Forbidden from Realtor.com API.",
+                    response=response
+                )
+            else:
+                raise Exception("Received 403 Forbidden, retrying...")
+
         return response.json()
 
     @retry(
@@ -72,33 +92,13 @@ class RealtorScraper(Scraper):
         stop=stop_after_attempt(3),
     )
     def handle_location(self):
-        query = """query SearchSuggestions($searchInput: SearchSuggestionsInput!) {
-            search_suggestions(search_input: $searchInput) {
-                geo_results {
-                    type
-                    text
-                    geo {
-                        _id
-                        area_type
-                        city
-                        state_code
-                        postal_code
-                        county
-                        centroid { lat lon }
-                        slug_id
-                        geo_id
-                    }
-                }
-            }
-        }"""
-
         variables = {
             "searchInput": {
                 "search_term": self.location
             }
         }
 
-        response_json = self._graphql_post(query, variables, "SearchSuggestions")
+        response_json = self._graphql_post(SEARCH_SUGGESTIONS_QUERY, variables, "Search_suggestions")
 
         if (
             response_json is None
@@ -130,18 +130,27 @@ class RealtorScraper(Scraper):
         }
 
         if geo.get("area_type") == "address":
-            geo_id = geo.get("_id", "")
+            # Try to get mpr_id directly from API response first
-            if geo_id.startswith("addr:"):
+            if geo.get("mpr_id"):
-                result["mpr_id"] = geo_id.replace("addr:", "")
+                result["mpr_id"] = geo.get("mpr_id")
+            else:
+                # Fallback: extract from _id field if it has addr: prefix
+                geo_id = geo.get("_id", "")
+                if geo_id.startswith("addr:"):
+                    result["mpr_id"] = geo_id.replace("addr:", "")
 
         return result
 
     def get_latest_listing_id(self, property_id: str) -> str | None:
-        query = """query GetPropertyListingId($property_id: ID!) {
+        query = """
+                fragment ListingFragment on Listing {
+                    listing_id
+                    primary
+                }
+                query GetPropertyListingId($property_id: ID!) {
                     property(id: $property_id) {
                         listings {
-                            listing_id
+                            ...ListingFragment
-                            primary
                         }
                     }
                 }
@@ -399,7 +408,7 @@ class RealtorScraper(Scraper):
                     $radius: String!
                     $offset: Int!,
                     ) {
-                        home_search(
+                        homeSearch: home_search(
                             query: {
                                 %s
                                 nearby: {
@@ -429,9 +438,9 @@ class RealtorScraper(Scraper):
         elif search_type == "area":  #: general search, came from a general location
             query = """query GetHomeSearch(
                                 $search_location: SearchLocation,
-                                $offset: Int,
+                                $offset: Int
                             ) {
-                                home_search(
+                                homeSearch: home_search(
                                     query: {
                                         %s
                                         search_location: $search_location
@@ -463,7 +472,7 @@ class RealtorScraper(Scraper):
                         $property_id: [ID]!
                         $offset: Int!,
                     ) {
-                        home_search(
+                        homeSearch: home_search(
                             query: {
                                 property_id: $property_id
                             }
@@ -475,7 +484,7 @@ class RealtorScraper(Scraper):
             )
 
         response_json = self._graphql_post(query, variables, "GetHomeSearch")
-        search_key = "home_search" if "home_search" in query else "property_search"
+        search_key = "homeSearch"
 
         properties: list[Union[Property, dict]] = []
 
@@ -1095,7 +1104,7 @@ class RealtorScraper(Scraper):
 
 
     @retry(
-        retry=retry_if_exception_type((JSONDecodeError, Exception)),
+        retry=retry_if_exception_type((JSONDecodeError, Exception)) & retry_if_not_exception_type(AuthenticationError),
         wait=wait_exponential(multiplier=1, min=1, max=10),
         stop=stop_after_attempt(3),
     )
@@ -1109,20 +1118,17 @@ class RealtorScraper(Scraper):
 
         property_ids = list(set(property_ids))
 
-        # Construct the bulk query
         fragments = "\n".join(
-            f'home_{property_id}: home(property_id: {property_id}) {{ ...HomeData }}'
+            f'home_{property_id}: home(property_id: {property_id}) {HOMES_DATA}'
             for property_id in property_ids
         )
-        query = f"""{HOME_FRAGMENT}
+        query = f"""query GetHome {{
-
-query GetBulkPropertyDetails {{
     {fragments}
 }}"""
 
-        data = self._graphql_post(query, {}, "GetBulkPropertyDetails")
+        data = self._graphql_post(query, {}, "GetHome")
 
-        if "data" not in data:
+        if "data" not in data or data["data"] is None:
             # If we got a 400 error with "Required parameter is missing", raise to trigger retry
             if data and "errors" in data:
                 error_msgs = [e.get("message", "") for e in data.get("errors", [])]
@@ -1131,6 +1137,6 @@ query GetBulkPropertyDetails {{
             return {}
 
         properties = data["data"]
-        return {data.replace('home_', ''): properties[data] for data in properties if properties[data]}
+        return {key.replace('home_', ''): properties[key] for key in properties if properties[key]}
 
 

homeharvest/core/scrapers/realtor/queries.py

@@ -1,3 +1,193 @@
+SEARCH_RESULTS_FRAGMENT = """
+fragment PropertyResult on SearchHome {
+    __typename
+    pending_date
+    listing_id
+    property_id
+    href
+    permalink
+    list_date
+    status
+    mls_status
+    last_sold_price
+    last_sold_date
+    last_status_change_date
+    last_update_date
+    list_price
+    list_price_max
+    list_price_min
+    price_per_sqft
+    tags
+    open_houses {
+        start_date
+        end_date
+        description
+        time_zone
+        dst
+        href
+        methods
+    }
+    details {
+        category
+        text
+        parent_category
+    }
+    pet_policy {
+        cats
+        dogs
+        dogs_small
+        dogs_large
+        __typename
+    }
+    units {
+        availability {
+          date
+          __typename
+        }
+        description {
+          baths_consolidated
+          baths
+          beds
+          sqft
+          __typename
+        }
+        photos(https: true) {
+            title
+            href
+            tags {
+                label
+            }
+        }
+        list_price
+        __typename
+    }
+    flags {
+        is_contingent
+        is_pending
+        is_new_construction
+    }
+    description {
+        type
+        sqft
+        beds
+        baths_full
+        baths_half
+        lot_sqft
+        year_built
+        garage
+        type
+        name
+        stories
+        text
+    }
+    source {
+        id
+        listing_id
+    }
+    hoa {
+        fee
+    }
+    location {
+        address {
+            street_direction
+            street_number
+            street_name
+            street_suffix
+            line
+            unit
+            city
+            state_code
+            postal_code
+            coordinate {
+                lon
+                lat
+            }
+        }
+        county {
+            name
+            fips_code
+        }
+        neighborhoods {
+            name
+        }
+    }
+    tax_record {
+        cl_id
+        public_record_id
+        last_update_date
+        apn
+        tax_parcel_id
+    }
+    primary_photo(https: true) {
+        href
+    }
+    advertisers {
+        email
+        broker {
+            name
+            fulfillment_id
+        }
+        type
+        name
+        fulfillment_id
+        builder {
+            name
+            fulfillment_id
+        }
+        phones {
+            ext
+            primary
+            type
+            number
+        }
+        office {
+            name
+            email
+            fulfillment_id
+            href
+            phones {
+                number
+                type
+                primary
+                ext
+            }
+            mls_set
+        }
+        corporation {
+            specialties
+            name
+            bio
+            href
+            fulfillment_id
+        }
+        mls_set
+        nrds_id
+        state_license
+        rental_corporation {
+            fulfillment_id
+        }
+        rental_management {
+            name
+            href
+            fulfillment_id
+        }
+    }
+    current_estimates {
+        __typename
+        source {
+            __typename
+            type
+            name
+        }
+        estimate
+        estimateHigh: estimate_high
+        estimateLow: estimate_low
+        date
+        isBestHomeValue: isbest_homevalue
+    }
+}
+"""
+
 _SEARCH_HOMES_DATA_BASE = """{
     pending_date
     listing_id
@@ -181,8 +371,189 @@ _SEARCH_HOMES_DATA_BASE = """{
 
 
 HOME_FRAGMENT = """
-fragment HomeData on Home {
+fragment PropertyResult on Home {
+    __typename
+    pending_date
+    listing_id
     property_id
+    href
+    permalink
+    list_date
+    status
+    mls_status
+    last_sold_price
+    last_sold_date
+    last_status_change_date
+    last_update_date
+    list_price
+    list_price_max
+    list_price_min
+    price_per_sqft
+    tags
+    open_houses {
+        start_date
+        end_date
+        description
+        time_zone
+        dst
+        href
+        methods
+    }
+    details {
+        category
+        text
+        parent_category
+    }
+    pet_policy {
+        cats
+        dogs
+        dogs_small
+        dogs_large
+        __typename
+    }
+    units {
+        availability {
+          date
+          __typename
+        }
+        description {
+          baths_consolidated
+          baths
+          beds
+          sqft
+          __typename
+        }
+        photos(https: true) {
+            title
+            href
+            tags {
+                label
+            }
+        }
+        list_price
+        __typename
+    }
+    flags {
+        is_contingent
+        is_pending
+        is_new_construction
+    }
+    description {
+        type
+        sqft
+        beds
+        baths_full
+        baths_half
+        lot_sqft
+        year_built
+        garage
+        type
+        name
+        stories
+        text
+    }
+    source {
+        id
+        listing_id
+    }
+    hoa {
+        fee
+    }
+    location {
+        address {
+            street_direction
+            street_number
+            street_name
+            street_suffix
+            line
+            unit
+            city
+            state_code
+            postal_code
+            coordinate {
+                lon
+                lat
+            }
+        }
+        county {
+            name
+            fips_code
+        }
+        neighborhoods {
+            name
+        }
+        parcel {
+            parcel_id
+        }
+    }
+    tax_record {
+        cl_id
+        public_record_id
+        last_update_date
+        apn
+        tax_parcel_id
+    }
+    primary_photo(https: true) {
+        href
+    }
+    photos(https: true) {
+        title
+        href
+        tags {
+            label
+        }
+    }
+    advertisers {
+        email
+        broker {
+            name
+            fulfillment_id
+        }
+        type
+        name
+        fulfillment_id
+        builder {
+            name
+            fulfillment_id
+        }
+        phones {
+            ext
+            primary
+            type
+            number
+        }
+        office {
+            name
+            email
+            fulfillment_id
+            href
+            phones {
+                number
+                type
+                primary
+                ext
+            }
+            mls_set
+        }
+        corporation {
+            specialties
+            name
+            bio
+            href
+            fulfillment_id
+        }
+        mls_set
+        nrds_id
+        state_license
+        rental_corporation {
+            fulfillment_id
+        }
+        rental_management {
+            name
+            href
+            fulfillment_id
+        }
+    }
     nearbySchools: nearby_schools(radius: 5.0, limit_per_level: 3) {
         __typename schools { district { __typename id name } }
     }
@@ -198,11 +569,6 @@ fragment HomeData on Home {
             last_n_days
         }
     }
-    location {
-        parcel {
-            parcel_id
-        }
-    }
     taxHistory: tax_history { __typename tax year assessment { __typename building land total } }
     property_history {
         date
@@ -227,6 +593,18 @@ fragment HomeData on Home {
         text
         category
     }
+    estimates {
+        __typename
+        currentValues: current_values {
+            __typename
+            source { __typename type name }
+            estimate
+            estimateHigh: estimate_high
+            estimateLow: estimate_low
+            date
+            isBestHomeValue: isbest_homevalue
+        }
+    }
 }
 """
 
@@ -300,8 +678,128 @@ current_estimates {
 }
 }""" % _SEARCH_HOMES_DATA_BASE
 
-GENERAL_RESULTS_QUERY = """{
+# Query body using inline fields (kept for backward compatibility)
+GENERAL_RESULTS_QUERY_BODY = """{
                             count
                             total
                             results %s
                         }""" % SEARCH_HOMES_DATA
+
+GENERAL_RESULTS_QUERY = """{
+                            __typename
+                            count
+                            total
+                            results %s
+                        }""" % SEARCH_HOMES_DATA
+
+LISTING_PHOTOS_FRAGMENT = """
+fragment ListingPhotosFragment on SearchHome {
+    __typename
+    photos(https: true) {
+        __typename
+        title
+        href
+        tags {
+            __typename
+            label
+            probability
+        }
+    }
+}
+"""
+
+SEARCH_SUGGESTIONS_QUERY = """query Search_suggestions($searchInput: SearchSuggestionsInput!) {
+  search_suggestions(search_input: $searchInput) {
+    raw_input_parser_result
+    typeahead_results {
+      display_string
+      display_geo
+      geo {
+        _id
+        _score
+        mpr_id
+        area_type
+        city
+        state_code
+        state
+        postal_code
+        country
+        lat
+        lon
+        county
+        counties {
+          name
+          fips
+          state_code
+        }
+        slug_id
+        geo_id
+        score
+        name
+        city_slug_id
+        centroid {
+          lat
+          lon
+        }
+        county_needed_for_uniq
+        street
+        line
+        school
+        school_id
+        school_district
+        has_catchment
+        university
+        university_id
+        neighborhood
+        park
+      }
+      url
+    }
+    geo_results {
+      type
+      text
+      geo {
+        _id
+        _score
+        mpr_id
+        area_type
+        city
+        state_code
+        state
+        postal_code
+        country
+        lat
+        lon
+        county
+        counties {
+          name
+          fips
+          state_code
+        }
+        slug_id
+        geo_id
+        score
+        name
+        city_slug_id
+        centroid {
+          lat
+          lon
+        }
+        county_needed_for_uniq
+        street
+        line
+        school
+        school_id
+        school_district
+        has_catchment
+        university
+        university_id
+        neighborhood
+        park
+      }
+    }
+    no_matches
+    has_results
+    original_string
+  }
+}"""

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "homeharvest"
-version = "0.8.7"
+version = "0.8.18"
 description = "Real estate scraping library"
 authors = ["Zachary Hampton <zachary@bunsly.com>", "Cullen Watson <cullen@bunsly.com>"]
 homepage = "https://github.com/ZacharyHampton/HomeHarvest"