Flatten repo structure: move crawler/ to root, remove vqa/ and immoweb/

The crawler subdirectory was the only active project. Moving it to the
repo root simplifies paths and removes the unnecessary nesting. The
vqa/ and immoweb/ directories were legacy/unused and have been removed.

Updated .drone.yml, .gitignore, .claude/ docs, and skills to reflect
the new flat structure.
This commit is contained in:
Viktor Barzin 2026-02-07 23:01:20 +00:00
parent e2247be700
commit eafbc1ac52
No known key found for this signature in database
GPG key ID: 0EB088298288D958
221 changed files with 70 additions and 146140 deletions

507
rec/query.py Normal file
View file

@ -0,0 +1,507 @@
import enum
import logging
import time
from typing import Any
from contextlib import asynccontextmanager
from collections.abc import AsyncIterator
import aiohttp
from models.listing import FurnishType, ListingType
from rec import districts
from rec.exceptions import (
CircuitBreakerOpenError,
ThrottlingError,
)
from rec.throttle_detector import get_throttle_metrics, validate_response
from rec.circuit_breaker import CircuitBreaker
from tenacity import (
retry,
retry_if_exception_type,
stop_after_attempt,
wait_exponential,
wait_random,
)
from config.scraper_config import ScraperConfig
logger = logging.getLogger("uvicorn.error")
# Global circuit breaker instance
_circuit_breaker: CircuitBreaker | None = None
# API constants
ANDROID_APP_VERSION = "3.70.0"
ANDROID_APP_VERSION_LISTING = "4.28.0"
RIGHTMOVE_API_BASE = "https://api.rightmove.co.uk/api"
PROPERTY_LISTING_ENDPOINT = f"{RIGHTMOVE_API_BASE}/property-listing"
DEFAULT_HEADERS = {
"Host": "api.rightmove.co.uk",
"User-Agent": "okhttp/4.12.0",
"Connection": "keep-alive",
}
LISTING_HEADERS = {
**DEFAULT_HEADERS,
"Accept-Encoding": "gzip, deflate, br",
}
class PropertyType(enum.StrEnum):
BUNGALOW = "bungalow"
DETACHED = "detached"
FLAT = "flat"
LAND = "land"
PARK_HOME = "park-home"
SEMI_DETACHED = "semi-detached"
TERRACED = "terraced"
@asynccontextmanager
async def create_session(
config: ScraperConfig | None = None,
) -> AsyncIterator[aiohttp.ClientSession]:
"""Create an aiohttp session with optional proxy support.
Args:
config: Scraper configuration. Loads from environment if not provided.
Yields:
Configured aiohttp ClientSession.
"""
if config is None:
config = ScraperConfig.from_env()
connector = None
if config.proxy_url:
try:
from aiohttp_socks import ProxyConnector
connector = ProxyConnector.from_url(config.proxy_url)
except ImportError:
raise ImportError(
"aiohttp-socks is required for proxy support. "
"Install with: pip install aiohttp-socks"
)
session = aiohttp.ClientSession(
trust_env=True,
connector=connector,
headers=DEFAULT_HEADERS,
)
try:
yield session
finally:
await session.close()
def get_circuit_breaker(config: ScraperConfig | None = None) -> CircuitBreaker | None:
"""Get the global circuit breaker instance.
Args:
config: Configuration for initializing the circuit breaker.
Returns:
CircuitBreaker instance if enabled, None otherwise.
"""
global _circuit_breaker
if config is None:
config = ScraperConfig.from_env()
if not config.enable_circuit_breaker:
return None
if _circuit_breaker is None:
_circuit_breaker = CircuitBreaker(
failure_threshold=config.circuit_breaker_failure_threshold,
recovery_timeout=config.circuit_breaker_recovery_timeout,
)
return _circuit_breaker
def reset_circuit_breaker() -> None:
"""Reset the global circuit breaker."""
global _circuit_breaker
if _circuit_breaker is not None:
_circuit_breaker.reset()
def check_circuit_breaker(config: ScraperConfig | None = None) -> None:
"""Check if the circuit breaker allows requests.
Args:
config: Configuration for the circuit breaker.
Raises:
CircuitBreakerOpenError: If the circuit is open.
"""
cb = get_circuit_breaker(config)
if cb is not None:
cb.call()
def _build_base_params(
*,
channel: ListingType,
page: int,
page_size: int,
radius: float,
min_price: int,
max_price: int,
min_bedrooms: int,
max_bedrooms: int,
district: str,
) -> dict[str, str]:
return {
"locationIdentifier": districts.get_districts()[district],
"channel": str(channel).upper(),
"page": str(page),
"numberOfPropertiesPerPage": str(page_size),
"radius": str(radius),
"sortBy": "distance",
"includeUnavailableProperties": "false",
"minPrice": str(min_price),
"maxPrice": str(max_price),
"minBedrooms": str(min_bedrooms),
"maxBedrooms": str(max_bedrooms),
"apiApplication": "ANDROID",
"appVersion": ANDROID_APP_VERSION_LISTING,
}
def _build_listing_params(
*,
page: int,
channel: ListingType,
min_bedrooms: int,
max_bedrooms: int,
radius: float,
min_price: int,
max_price: int,
district: str,
mustNewHome: bool,
max_days_since_added: int,
property_type: list[PropertyType],
page_size: int,
furnish_types: list[FurnishType],
) -> dict[str, str]:
params = _build_base_params(
channel=channel,
page=page,
page_size=page_size,
radius=radius,
min_price=min_price,
max_price=max_price,
min_bedrooms=min_bedrooms,
max_bedrooms=max_bedrooms,
district=district,
)
if channel is ListingType.BUY:
params["dontShow"] = "sharedOwnership,retirement"
if len(property_type) > 0:
params["propertyTypes"] = ",".join(property_type)
if max_days_since_added is not None and max_days_since_added not in [
1,
3,
7,
14,
]:
raise Exception(
f"Invalid max days - {max_days_since_added} Can only be got",
[1, 3, 7, 14],
)
params["maxDaysSinceAdded"] = str(max_days_since_added)
if mustNewHome:
params["mustHave"] = "newHome"
if channel is ListingType.RENT:
if furnish_types:
params["furnishTypes"] = ",".join(furnish_types)
return params
def _build_probe_params(
*,
channel: ListingType,
min_bedrooms: int,
max_bedrooms: int,
radius: float,
min_price: int,
max_price: int,
district: str,
max_days_since_added: int,
furnish_types: list[FurnishType],
) -> dict[str, str]:
params = _build_base_params(
channel=channel,
page=1,
page_size=1, # Minimal page size for probing
radius=radius,
min_price=min_price,
max_price=max_price,
min_bedrooms=min_bedrooms,
max_bedrooms=max_bedrooms,
district=district,
)
if channel is ListingType.BUY:
params["dontShow"] = "sharedOwnership,retirement"
if max_days_since_added is not None and max_days_since_added in [
1,
3,
7,
14,
]:
params["maxDaysSinceAdded"] = str(max_days_since_added)
if channel is ListingType.RENT:
if furnish_types:
params["furnishTypes"] = ",".join(furnish_types)
return params
async def _execute_api_request(
*,
url: str,
params: dict[str, str],
headers: dict[str, str],
session: aiohttp.ClientSession | None,
config: ScraperConfig,
expect_data: bool = True,
error_context: str = "",
) -> dict[str, Any]:
check_circuit_breaker(config)
cb = get_circuit_breaker(config)
async def do_request(s: aiohttp.ClientSession) -> dict[str, Any]:
start_time = time.time()
try:
async with s.get(url, params=params, headers=headers) as response:
response_time = time.time() - start_time
body = await response.json() if response.status == 200 else None
validate_response(
response,
response_time,
body,
config.slow_response_threshold,
expect_data=expect_data,
)
if response.status != 200:
raise Exception(
f"{error_context}Failed due to: {await response.text()}"
)
if cb is not None:
cb.record_success()
return body # type: ignore
except ThrottlingError:
if cb is not None:
cb.record_failure()
raise
except Exception as e:
if cb is not None:
cb.record_failure()
raise e
if session:
return await do_request(session)
else:
async with aiohttp.ClientSession(trust_env=True) as new_session:
return await do_request(new_session)
@retry(
retry=retry_if_exception_type(ThrottlingError),
wait=wait_exponential(multiplier=2, min=2, max=120),
stop=stop_after_attempt(5),
)
async def detail_query(
detail_id: int,
session: aiohttp.ClientSession | None = None,
config: ScraperConfig | None = None,
) -> dict[str, Any]:
"""Fetch detailed property information.
Args:
detail_id: The property identifier.
session: Optional aiohttp session. Creates new one if not provided.
config: Scraper configuration. Loads from environment if not provided.
Returns:
Property details as a dictionary.
Raises:
CircuitBreakerOpenError: If the circuit breaker is open.
ThrottlingError: If the request is throttled.
"""
if config is None:
config = ScraperConfig.from_env()
params = {
"apiApplication": "ANDROID",
"appVersion": ANDROID_APP_VERSION,
}
url = f"{RIGHTMOVE_API_BASE}/property/{detail_id}"
return await _execute_api_request(
url=url,
params=params,
headers=DEFAULT_HEADERS,
session=session,
config=config,
expect_data=True,
error_context=f"id: {detail_id}. Status Code: ",
)
@retry(
retry=retry_if_exception_type(ThrottlingError),
wait=wait_exponential(multiplier=2, min=2, max=120),
stop=stop_after_attempt(5),
)
async def listing_query(
*,
page: int,
channel: ListingType,
min_bedrooms: int,
max_bedrooms: int,
radius: float,
min_price: int,
max_price: int,
district: str, # = "STATION^5168", # kings cross station
mustNewHome: bool = False,
max_days_since_added: int = 30,
property_type: list[PropertyType] | None = None,
page_size: int = 25,
furnish_types: list[FurnishType] | None = None,
session: aiohttp.ClientSession | None = None,
config: ScraperConfig | None = None,
) -> dict[str, Any]:
"""Execute a listing search query.
Args:
page: Page number to fetch (1-indexed).
channel: Listing type (BUY or RENT).
min_bedrooms: Minimum number of bedrooms.
max_bedrooms: Maximum number of bedrooms.
radius: Search radius.
min_price: Minimum price.
max_price: Maximum price.
district: District identifier string.
mustNewHome: Filter for new homes only (BUY only).
max_days_since_added: Maximum days since listing was added (BUY only).
property_type: List of property types to filter (BUY only).
page_size: Number of results per page (default 25).
furnish_types: List of furnish types to filter (RENT only).
session: Optional aiohttp session. Creates new one if not provided.
config: Scraper configuration. Loads from environment if not provided.
Returns:
API response as a dictionary.
Raises:
CircuitBreakerOpenError: If the circuit breaker is open.
ThrottlingError: If the request is throttled.
"""
if config is None:
config = ScraperConfig.from_env()
if property_type is None:
property_type = []
if furnish_types is None:
furnish_types = []
params = _build_listing_params(
page=page,
channel=channel,
min_bedrooms=min_bedrooms,
max_bedrooms=max_bedrooms,
radius=radius,
min_price=min_price,
max_price=max_price,
district=district,
mustNewHome=mustNewHome,
max_days_since_added=max_days_since_added,
property_type=property_type,
page_size=page_size,
furnish_types=furnish_types,
)
return await _execute_api_request(
url=PROPERTY_LISTING_ENDPOINT,
params=params,
headers=LISTING_HEADERS,
session=session,
config=config,
expect_data=(page == 1),
)
@retry(
retry=retry_if_exception_type(ThrottlingError),
wait=wait_exponential(multiplier=2, min=2, max=60),
stop=stop_after_attempt(5),
)
async def probe_query(
*,
session: aiohttp.ClientSession,
channel: ListingType,
min_bedrooms: int,
max_bedrooms: int,
radius: float,
min_price: int,
max_price: int,
district: str,
max_days_since_added: int = 30,
furnish_types: list[FurnishType] | None = None,
config: ScraperConfig | None = None,
) -> dict[str, Any]:
"""Probe the API to get result count without fetching full results.
Makes a minimal request (page_size=1) to efficiently get totalAvailableResults.
Args:
session: aiohttp session for making requests.
channel: Listing type (BUY or RENT).
min_bedrooms: Minimum number of bedrooms.
max_bedrooms: Maximum number of bedrooms.
radius: Search radius.
min_price: Minimum price.
max_price: Maximum price.
district: District identifier string.
max_days_since_added: Maximum days since listing was added (BUY only).
furnish_types: List of furnish types to filter (RENT only).
config: Scraper configuration. Loads from environment if not provided.
Returns:
API response containing totalAvailableResults.
Raises:
CircuitBreakerOpenError: If the circuit breaker is open.
ThrottlingError: If the request is throttled.
"""
if config is None:
config = ScraperConfig.from_env()
if furnish_types is None:
furnish_types = []
params = _build_probe_params(
channel=channel,
min_bedrooms=min_bedrooms,
max_bedrooms=max_bedrooms,
radius=radius,
min_price=min_price,
max_price=max_price,
district=district,
max_days_since_added=max_days_since_added,
furnish_types=furnish_types,
)
return await _execute_api_request(
url=PROPERTY_LISTING_ENDPOINT,
params=params,
headers=LISTING_HEADERS,
session=session,
config=config,
expect_data=False,
error_context="Probe failed: ",
)