From 9f3e466b23c20886faedd599e31e4d8bf1c0eb12 Mon Sep 17 00:00:00 2001
From: Viktor Barzin <vbarzin@gmail.com>
Date: Sun, 18 May 2025 17:22:48 +0000
Subject: [PATCH] add filter for furnished/unfurnished type for rented listings

---
 crawler/1_dump_listings.py |  9 +++++++--
 crawler/main.py            | 33 +++++++++++++++++++++------------
 crawler/rec/query.py       | 15 +++++++++++++--
 3 files changed, 41 insertions(+), 16 deletions(-)

diff --git a/crawler/1_dump_listings.py b/crawler/1_dump_listings.py
index 6f76e04..da36915 100644
--- a/crawler/1_dump_listings.py
+++ b/crawler/1_dump_listings.py
@@ -1,7 +1,7 @@
 import asyncio
 from dataclasses import dataclass
 import pathlib
-from rec.query import ListingType, listing_query
+from rec.query import ListingType, listing_query, FurnishType
 from rec.districts import get_districts
 from data_access import Listing
 
@@ -17,7 +17,11 @@ class QueryParameters:
     radius: float = 0
     page_size: int = 500  # items per page
     max_days_since_added: int = 30
-    # available from; furnished/unfurnished; council tax
+    furnish_types: list[FurnishType] | None = None
+
+    # The values below are not supported by rightmove
+    # hence we apply them after fetching
+    # available from; council tax
 
 
 async def dump_listings(
@@ -44,6 +48,7 @@ async def dump_listings(
             location_id=locid,
             page_size=parameters.page_size,
             max_days_since_added=parameters.max_days_since_added,
+            furnish_types=parameters.furnish_types or [],
         ) for locid in districts.values() for i in [1, 2]
     ])
     listings = []
diff --git a/crawler/main.py b/crawler/main.py
index 570ef31..f6136b6 100644
--- a/crawler/main.py
+++ b/crawler/main.py
@@ -6,7 +6,7 @@ import importlib
 from rec.districts import get_districts
 from data_access import Listing
 import csv_exporter
-from rec.query import ListingType
+from rec.query import ListingType, FurnishType
 
 dump_listings_module = importlib.import_module('1_dump_listings')
 dump_detail_module = importlib.import_module('2_dump_detail')
@@ -14,14 +14,6 @@ dump_images_module = importlib.import_module('3_dump_images')
 detect_floorplan_module = importlib.import_module('4_detect_floorplan')
 routing_module = importlib.import_module('5_routing')
 
-steps_to_handlers = {
-    'dump_listings': dump_listings_module.dump_listings,
-    'dump_detail': dump_detail_module.dump_detail,
-    'dump_images': dump_images_module.dump_images,
-    'detect_floorplan': detect_floorplan_module.detect_floorplan,
-    'routing': routing_module.calculate_route,
-}
-
 
 @click.group()
 @click.option(
@@ -84,6 +76,19 @@ def cli(ctx, data_dir: str):
     type=click.Choice(get_districts().keys(), case_sensitive=False),
     multiple=True,
 )
+@click.option(
+    '--furnish-types',
+    '-f',
+    help='Furnish types for rented listings',
+    type=click.Choice(
+        [
+            furnish_type.name
+            for furnish_type in FurnishType.__members__.values()
+        ],
+        case_sensitive=False,
+    ),
+    multiple=True,
+)
 @click.pass_context
 def dump_listings(
     ctx: click.core.Context,
@@ -93,6 +98,7 @@ def dump_listings(
     min_price: int,
     max_price: int,
     type: str,
+    furnish_types: list[str],
 ):
     data_dir: str = ctx.obj['data_dir']
     query_parameters = dump_listings_module.QueryParameters(
@@ -102,13 +108,16 @@ def dump_listings(
         max_bedrooms=max_bedrooms,
         min_price=min_price,
         max_price=max_price,
+        furnish_types=[
+            FurnishType[furnish_type] for furnish_type in furnish_types
+        ],
     )
     click.echo(
         f'Running dump_listings for districts {district}, data dir {data_dir} and parameters: '
-        f'{query_parameters}'
-    )
+        f'{query_parameters}')
     data_dir_path = pathlib.Path(data_dir)
-    asyncio.run(dump_listings_module.dump_listings(query_parameters, data_dir_path))
+    asyncio.run(
+        dump_listings_module.dump_listings(query_parameters, data_dir_path))
 
 
 @cli.command()
diff --git a/crawler/rec/query.py b/crawler/rec/query.py
index 4e7b2f4..ce0a2bc 100644
--- a/crawler/rec/query.py
+++ b/crawler/rec/query.py
@@ -1,6 +1,6 @@
 # from diskcache import Cache
 import enum
-from typing import Any, List
+from typing import Any
 import aiohttp
 import requests
 import urllib3
@@ -13,6 +13,12 @@ class ListingType(enum.StrEnum):
     RENT = "RENT"
 
 
+class FurnishType(enum.StrEnum):
+    FURNISHED = "furnished"
+    UNFURNISHED = "unfurnished"
+    PART_FURNISHED = "partFurnished"
+
+
 headers = {
     "Host": "api.rightmove.co.uk",
     # 'Accept-Encoding': 'gzip, deflate, br',
@@ -48,6 +54,7 @@ async def detail_query(detail_id: int):
 
 
 async def listing_query(
+    *,
     page: int,
     channel: ListingType,
     min_bedrooms: int,
@@ -58,8 +65,9 @@ async def listing_query(
     location_id: str = "STATION^5168",  # kings cross station
     mustNewHome: bool = False,
     max_days_since_added: int = 30,
-    property_type: List["PropertyType"] = [],
+    property_type: list[PropertyType] = [],
     page_size: int = 25,
+    furnish_types: list[FurnishType] = [],
 ) -> dict[str, Any]:
     params: dict[str, str] = {
         "locationIdentifier": location_id,
@@ -88,6 +96,9 @@ async def listing_query(
 
         if mustNewHome:
             params["mustHave"] = "newHome"
+    if channel is ListingType.RENT:
+        if furnish_types:
+            params["furnishTypes"] = ",".join(furnish_types)
 
     headers = {
         "Host": "api.rightmove.co.uk",