pyfunda 2.2.0__tar.gz → 2.3.0__tar.gz

{pyfunda-2.2.0 → pyfunda-2.3.0}/.github/workflows/publish.yml

@@ -1,7 +1,9 @@
 name: Publish to PyPI
 
 on:
-  workflow_dispatch:
+  push:
+    branches: [main]
+    paths: [pyproject.toml]
 
 jobs:
   publish:
@@ -18,13 +20,27 @@ jobs:
         id: version
         run: echo "version=$(grep '^version' pyproject.toml | cut -d'"' -f2)" >> $GITHUB_OUTPUT
 
+      - name: Check if release exists
+        id: check
+        run: |
+          if gh release view "v${{ steps.version.outputs.version }}" &>/dev/null; then
+            echo "exists=true" >> $GITHUB_OUTPUT
+          else
+            echo "exists=false" >> $GITHUB_OUTPUT
+          fi
+        env:
+          GH_TOKEN: ${{ github.token }}
+
       - name: Build
+        if: steps.check.outputs.exists == 'false'
         run: uv build
 
       - name: Publish to PyPI
+        if: steps.check.outputs.exists == 'false'
         run: uv publish --token ${{ secrets.PYPI_TOKEN }}
 
       - name: Create GitHub Release
+        if: steps.check.outputs.exists == 'false'
         run: gh release create "v${{ steps.version.outputs.version }}" --title "v${{ steps.version.outputs.version }}" --generate-notes
         env:
           GH_TOKEN: ${{ github.token }}

pyfunda-2.2.0/README.md → pyfunda-2.3.0/PKG-INFO

@@ -1,3 +1,27 @@
+Metadata-Version: 2.4
+Name: pyfunda
+Version: 2.3.0
+Summary: Python API for Funda.nl real estate listings
+Project-URL: Homepage, https://github.com/0xMH/pyfunda
+Project-URL: Repository, https://github.com/0xMH/pyfunda
+Project-URL: Issues, https://github.com/0xMH/pyfunda/issues
+Author: 0xMH
+License-Expression: AGPL-3.0-or-later
+License-File: LICENSE
+Keywords: api,funda,housing,netherlands,real-estate,scraper
+Classifier: Development Status :: 4 - Beta
+Classifier: Intended Audience :: Developers
+Classifier: License :: OSI Approved :: GNU Affero General Public License v3 or later (AGPLv3+)
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Requires-Python: >=3.10
+Requires-Dist: curl-cffi>=0.14.0
+Requires-Dist: tls-client>=1.0.1
+Requires-Dist: typing-extensions>=4.0.0
+Description-Content-Type: text/markdown
+
 # pyfunda
 
 [![PyPI version](https://img.shields.io/pypi/v/pyfunda)](https://pypi.org/project/pyfunda/)
@@ -10,11 +34,17 @@ The only working real Python API for Funda ([funda.nl](https://www.funda.nl))
 
 [![Star History Chart](https://api.star-history.com/svg?repos=0xMH/pyfunda&type=Date)](https://star-history.com/#0xMH/pyfunda&Date)
 
-## Installation
+## Why I'm open-sourcing this?
 
-```bash
-pip install pyfunda
-```
+After pyfunda, I got messages asking why I'd give this away when aggregators will just take it and sell it. They're right, every week there's a new "revolutionary AI-powered housing finder" charging €40/month or a €250 "success fee.". They all pull from the same one or two sources and wrap it in a fancy UI completely built with AI.
+
+That's exactly why I'm open-sourcing it.
+
+These services are selling air to people who are looking for any kind of hope. The data is public. The APIs aren't hard to figure out. You shouldn't have to pay someone to refresh a webpage for you. Funda could kill this entire market overnight by offering a public API. They don't, so here we are.
+
+Here's the code, do it yourself. Send my library link to any AI service you use and ask it to build whatever tool you think will make your life easier while searching for your next home.
+
+With pyfunda, I've already done all the heavy lifting for you.
 
 ## Why pyfunda?
 
@@ -36,6 +66,12 @@ Funda has no public API. If you want Dutch real estate data programmatically, yo
 - 70+ fields including photos, floorplans, coordinates, and listing dates
 - Stable mobile API that doesn't break when the website changes
 
+## Installation
+
+```bash
+pip install pyfunda
+```
+
 ## Quick Start
 
 ```python
@@ -545,6 +581,18 @@ if len(funda_prices) >= 2:
     print(f"\nPrice change: {change_pct:+.1f}%")
 ```
 
+## Disclaimer
+
+This is an unofficial library and is not affiliated with, authorized, maintained, sponsored, or endorsed by Funda or any of its affiliates. Use at your own risk.
+
+This library only accesses publicly available listing data through Funda's undocumented internal API. Using this library may violate Funda's Terms of Service. The authors are not responsible for any consequences of using this software.
+
+This project is intended for personal use, research, and educational purposes only.
+
+- The API is undocumented and may change or break at any time without notice.
+- Please use this library responsibly and avoid excessive requests that could burden Funda's infrastructure.
+- Scraped data may be subject to copyright and usage restrictions. Ensure your use complies with applicable laws.
+
 ## License
 
 AGPL-3.0

pyfunda-2.2.0/PKG-INFO → pyfunda-2.3.0/README.md

@@ -1,25 +1,3 @@
-Metadata-Version: 2.4
-Name: pyfunda
-Version: 2.2.0
-Summary: Python API for Funda.nl real estate listings
-Project-URL: Homepage, https://github.com/0xMH/pyfunda
-Project-URL: Repository, https://github.com/0xMH/pyfunda
-Project-URL: Issues, https://github.com/0xMH/pyfunda/issues
-Author: 0xMH
-License-Expression: AGPL-3.0-or-later
-License-File: LICENSE
-Keywords: api,funda,housing,netherlands,real-estate,scraper
-Classifier: Development Status :: 4 - Beta
-Classifier: Intended Audience :: Developers
-Classifier: License :: OSI Approved :: GNU Affero General Public License v3 or later (AGPLv3+)
-Classifier: Programming Language :: Python :: 3
-Classifier: Programming Language :: Python :: 3.10
-Classifier: Programming Language :: Python :: 3.11
-Classifier: Programming Language :: Python :: 3.12
-Requires-Python: >=3.10
-Requires-Dist: curl-cffi>=0.14.0
-Description-Content-Type: text/markdown
-
 # pyfunda
 
 [![PyPI version](https://img.shields.io/pypi/v/pyfunda)](https://pypi.org/project/pyfunda/)
@@ -32,11 +10,17 @@ The only working real Python API for Funda ([funda.nl](https://www.funda.nl))
 
 [![Star History Chart](https://api.star-history.com/svg?repos=0xMH/pyfunda&type=Date)](https://star-history.com/#0xMH/pyfunda&Date)
 
-## Installation
+## Why I'm open-sourcing this?
 
-```bash
-pip install pyfunda
-```
+After pyfunda, I got messages asking why I'd give this away when aggregators will just take it and sell it. They're right, every week there's a new "revolutionary AI-powered housing finder" charging €40/month or a €250 "success fee.". They all pull from the same one or two sources and wrap it in a fancy UI completely built with AI.
+
+That's exactly why I'm open-sourcing it.
+
+These services are selling air to people who are looking for any kind of hope. The data is public. The APIs aren't hard to figure out. You shouldn't have to pay someone to refresh a webpage for you. Funda could kill this entire market overnight by offering a public API. They don't, so here we are.
+
+Here's the code, do it yourself. Send my library link to any AI service you use and ask it to build whatever tool you think will make your life easier while searching for your next home.
+
+With pyfunda, I've already done all the heavy lifting for you.
 
 ## Why pyfunda?
 
@@ -58,6 +42,12 @@ Funda has no public API. If you want Dutch real estate data programmatically, yo
 - 70+ fields including photos, floorplans, coordinates, and listing dates
 - Stable mobile API that doesn't break when the website changes
 
+## Installation
+
+```bash
+pip install pyfunda
+```
+
 ## Quick Start
 
 ```python
@@ -567,6 +557,18 @@ if len(funda_prices) >= 2:
     print(f"\nPrice change: {change_pct:+.1f}%")
 ```
 
+## Disclaimer
+
+This is an unofficial library and is not affiliated with, authorized, maintained, sponsored, or endorsed by Funda or any of its affiliates. Use at your own risk.
+
+This library only accesses publicly available listing data through Funda's undocumented internal API. Using this library may violate Funda's Terms of Service. The authors are not responsible for any consequences of using this software.
+
+This project is intended for personal use, research, and educational purposes only.
+
+- The API is undocumented and may change or break at any time without notice.
+- Please use this library responsibly and avoid excessive requests that could burden Funda's infrastructure.
+- Scraped data may be subject to copyright and usage restrictions. Ensure your use complies with applicable laws.
+
 ## License
 
 AGPL-3.0

{pyfunda-2.2.0 → pyfunda-2.3.0}/funda/__init__.py

@@ -13,8 +13,10 @@ Example usage:
     ...     print(r['title'], r['city'])
 """
 
+from importlib.metadata import version
+
 from funda.funda import Funda, FundaAPI
 from funda.listing import Listing
 
-__version__ = "2.2.0"
+__version__ = version("pyfunda")
 __all__ = ["Funda", "FundaAPI", "Listing", "__version__"]

{pyfunda-2.2.0 → pyfunda-2.3.0}/funda/funda.py

@@ -1,10 +1,12 @@
 """Main Funda API class."""
 
+import random
 import re
 import time
 from typing import Any
 
 from curl_cffi import requests
+import tls_client
 
 from funda.listing import Listing
 
@@ -16,17 +18,71 @@ API_LISTING_TINY = f"{API_BASE}/tinyId/{{tiny_id}}"
 API_SEARCH = "https://listing-search-wonen.funda.io/_msearch/template"
 API_WALTER = "https://api.walterliving.com/hunter/lookup"
 
-# Headers for mobile API
-HEADERS = {
-    "x-funda-app-platform": "android",
-    "content-type": "application/json",
-}
+# Funda mobile app JA3 fingerprints (captured from real Dart/Flutter app traffic)
+# JA3 without extension 21 - from favourites.funda.io
+# JA3 hash: 9225d95490794840d9d5f1f94d339285
+FUNDA_JA3 = "771,4867-4865-4866-52393-52392-49195-49199-49196-49200-49161-49171-49162-49172-156-157-47-53,0-23-65281-10-11-35-13-51-45-43,29-23-24,0"
 
-SEARCH_HEADERS = {
-    "content-type": "application/json",
-    "accept": "application/json",
-    "referer": "https://www.funda.nl/",
-}
+# JA3 with extension 21 (padding) - from cdn-settings.segment.com
+# JA3 hash: 4bf8cdd8919b07d35ca824c20efb3537
+FUNDA_JA3_EXT21 = "771,4867-4865-4866-52393-52392-49195-49199-49196-49200-49161-49171-49162-49172-156-157-47-53,0-23-65281-10-11-35-13-51-45-43-21,29-23-24,0"
+
+# Fingerprint pool - tried in order until one works
+# Types: "tls_ja3", "curl_impersonate", "tls_client"
+FINGERPRINT_POOL = [
+    # tls_client with exact Funda app JA3 fingerprints
+    {"type": "tls_ja3", "ja3": FUNDA_JA3},
+    {"type": "tls_ja3", "ja3": FUNDA_JA3_EXT21},
+    # curl_cffi impersonate fallback - Safari works best with Funda headers
+    {"type": "curl_impersonate", "target": "safari15_5"},
+    {"type": "curl_impersonate", "target": "safari15_3"},
+    # tls_client preset profiles as final fallback
+    {"type": "tls_client", "identifier": "okhttp4_android_13"},
+    {"type": "tls_client", "identifier": "chrome_120"},
+]
+
+# Test endpoint to verify fingerprint works
+TEST_URL = f"{API_BASE}/tinyId/43117443"
+
+
+
+def _make_headers(for_search: bool = False) -> list[tuple[str, str]]:
+    """Generate headers matching the Funda Android app exactly.
+
+    Header order and values are captured from real Funda app traffic.
+    """
+    trace_id = str(random.randint(10**18, 10**19))
+    parent_id = hex(random.randint(10**15, 10**16))[2:]
+    tid = hex(int(time.time()))[2:] + "00000000"
+
+    # Base headers in exact order from app traffic
+    headers = [
+        ("user-agent", "Dart/3.9 (dart:io)"),
+        ("x-datadog-sampling-priority", "0"),
+        ("x-datadog-origin", "rum"),
+        ("tracestate", f"dd=s:0;o:rum;p:{parent_id}"),
+        ("accept-encoding", "gzip"),
+        ("x-datadog-parent-id", trace_id),
+    ]
+
+    if for_search:
+        # Search endpoint: content-type, referer, accept, then traceparent
+        headers.extend([
+            ("content-type", "application/json"),
+            ("referer", "https://www.funda.nl/"),
+            ("accept", "application/json"),
+        ])
+    else:
+        # Listing endpoint: x-funda-app-platform, content-type, then traceparent
+        headers.extend([
+            ("x-funda-app-platform", "android"),
+            ("content-type", "application/json"),
+        ])
+
+    # traceparent is always last
+    headers.append(("traceparent", f"00-{tid}{trace_id[:16]}-{parent_id}-00"))
+
+    return headers
 
 
 def _parse_area(value: str | None) -> int | None:
@@ -63,21 +119,145 @@ class Funda:
             timeout: Request timeout in seconds
         """
         self.timeout = timeout
-        self._session: requests.Session | None = None
+        self._curl_session: requests.Session | None = None
+        self._tls_session: tls_client.Session | None = None
+        self._fingerprint: dict | None = None
 
-    @property
-    def session(self) -> requests.Session:
-        """Lazily create HTTP session."""
-        if self._session is None:
-            self._session = requests.Session(impersonate="chrome")
-            self._session.headers.update(HEADERS)
-        return self._session
+    def _make_headers_dict(self, for_search: bool = False) -> dict[str, str]:
+        """Generate headers as dict for tls_client.
+
+        Header order and values match the real Funda Android app exactly.
+        """
+        trace_id = str(random.randint(10**18, 10**19))
+        parent_id = hex(random.randint(10**15, 10**16))[2:]
+        tid = hex(int(time.time()))[2:] + "00000000"
+
+        # Build headers in exact order from app traffic
+        # Python 3.7+ dicts preserve insertion order
+        headers = {
+            "user-agent": "Dart/3.9 (dart:io)",
+            "x-datadog-sampling-priority": "0",
+            "x-datadog-origin": "rum",
+            "tracestate": f"dd=s:0;o:rum;p:{parent_id}",
+            "accept-encoding": "gzip",
+            "x-datadog-parent-id": trace_id,
+        }
+
+        if for_search:
+            # Search endpoint: content-type, referer, accept, then traceparent
+            headers["content-type"] = "application/json"
+            headers["referer"] = "https://www.funda.nl/"
+            headers["accept"] = "application/json"
+        else:
+            # Listing endpoint: x-funda-app-platform, content-type, then traceparent
+            headers["x-funda-app-platform"] = "android"
+            headers["content-type"] = "application/json"
+
+        # traceparent is always last
+        headers["traceparent"] = f"00-{tid}{trace_id[:16]}-{parent_id}-00"
+
+        return headers
+
+    def _test_fingerprint(self, fingerprint: dict) -> bool:
+        """Test if a fingerprint works against Funda API."""
+        try:
+            fp_type = fingerprint["type"]
+            if fp_type == "tls_ja3":
+                # tls_client with custom JA3 - primary method for Funda app fingerprint
+                session = tls_client.Session(ja3_string=fingerprint["ja3"], random_tls_extension_order=False)
+                headers = self._make_headers_dict()
+                response = session.get(TEST_URL, headers=headers, timeout_seconds=5)
+            elif fp_type == "curl_ja3":
+                session = requests.Session()
+                headers = _make_headers()
+                response = session.get(TEST_URL, headers=headers, ja3=fingerprint["ja3"], timeout=5)
+                session.close()
+            elif fp_type == "curl_impersonate":
+                session = requests.Session(impersonate=fingerprint["target"])
+                headers = _make_headers()
+                response = session.get(TEST_URL, headers=headers, timeout=5)
+                session.close()
+            elif fp_type == "tls_client":
+                session = tls_client.Session(client_identifier=fingerprint["identifier"], random_tls_extension_order=False)
+                headers = self._make_headers_dict()
+                response = session.get(TEST_URL, headers=headers, timeout_seconds=5)
+            else:
+                return False
+            return response.status_code == 200
+        except Exception:
+            return False
+
+    def _find_working_fingerprint(self) -> dict:
+        """Find a working fingerprint from the pool."""
+        for fp in FINGERPRINT_POOL:
+            if self._test_fingerprint(fp):
+                return fp
+        raise RuntimeError("No working fingerprint found. Funda may have updated their bot detection.")
+
+    def _ensure_session(self) -> None:
+        """Ensure a working session is created."""
+        if self._fingerprint is None:
+            self._fingerprint = self._find_working_fingerprint()
+
+        fp_type = self._fingerprint["type"]
+        if fp_type == "tls_ja3":
+            # tls_client with custom JA3 - exact Funda app fingerprint
+            if self._tls_session is None:
+                self._tls_session = tls_client.Session(
+                    ja3_string=self._fingerprint["ja3"],
+                    random_tls_extension_order=False
+                )
+        elif fp_type == "curl_ja3":
+            if self._curl_session is None:
+                self._curl_session = requests.Session()
+        elif fp_type == "curl_impersonate":
+            if self._curl_session is None:
+                self._curl_session = requests.Session(impersonate=self._fingerprint["target"])
+        elif fp_type == "tls_client":
+            if self._tls_session is None:
+                self._tls_session = tls_client.Session(
+                    client_identifier=self._fingerprint["identifier"],
+                    random_tls_extension_order=False
+                )
+
+    def _get(self, url: str, headers_list: list[tuple[str, str]]) -> Any:
+        """Make GET request using the active session."""
+        self._ensure_session()
+        fp_type = self._fingerprint["type"]
+
+        if fp_type in ("tls_ja3", "tls_client"):
+            headers = self._make_headers_dict()
+            return self._tls_session.get(url, headers=headers, timeout_seconds=self.timeout)
+        elif fp_type == "curl_ja3":
+            return self._curl_session.get(url, headers=headers_list, ja3=self._fingerprint["ja3"], timeout=self.timeout)
+        else:
+            return self._curl_session.get(url, headers=headers_list, timeout=self.timeout)
+
+    def _post(self, url: str, headers_list: list[tuple[str, str]], data: str = None, json_data: dict = None, for_search: bool = False) -> Any:
+        """Make POST request using the active session."""
+        self._ensure_session()
+        fp_type = self._fingerprint["type"]
+
+        if fp_type in ("tls_ja3", "tls_client"):
+            headers = self._make_headers_dict(for_search=for_search)
+            if json_data:
+                return self._tls_session.post(url, headers=headers, json=json_data, timeout_seconds=self.timeout)
+            return self._tls_session.post(url, headers=headers, data=data, timeout_seconds=self.timeout)
+        elif fp_type == "curl_ja3":
+            if json_data:
+                return self._curl_session.post(url, headers=headers_list, json=json_data, ja3=self._fingerprint["ja3"], timeout=self.timeout)
+            return self._curl_session.post(url, headers=headers_list, data=data, ja3=self._fingerprint["ja3"], timeout=self.timeout)
+        else:
+            if json_data:
+                return self._curl_session.post(url, headers=headers_list, json=json_data, timeout=self.timeout)
+            return self._curl_session.post(url, headers=headers_list, data=data, timeout=self.timeout)
 
     def close(self) -> None:
         """Close the HTTP session."""
-        if self._session:
-            self._session.close()
-            self._session = None
+        if self._curl_session:
+            self._curl_session.close()
+            self._curl_session = None
+        self._tls_session = None
 
     def __enter__(self) -> "Funda":
         return self
@@ -116,12 +296,14 @@ class Funda:
         else:
             url = API_LISTING.format(listing_id=listing_id_str)
 
-        response = self.session.get(url, timeout=self.timeout)
+        headers = _make_headers()
+        response = self._get(url, headers)
 
         # If tinyId fails, try as globalId
         if response.status_code == 404 and len(listing_id_str) >= 8:
             url = API_LISTING.format(listing_id=listing_id_str)
-            response = self.session.get(url, timeout=self.timeout)
+            headers = _make_headers()
+            response = self._get(url, headers)
 
         if response.status_code != 200:
             raise LookupError(f"Listing {listing_id} not found")
@@ -260,12 +442,8 @@ class Funda:
 
         # Retry on intermittent 400 errors from API
         for attempt in range(3):
-            response = self.session.post(
-                API_SEARCH,
-                headers=SEARCH_HEADERS,
-                data=query,
-                timeout=self.timeout,
-            )
+            headers = _make_headers(for_search=True)
+            response = self._post(API_SEARCH, headers, data=query, for_search=True)
             if response.status_code == 200:
                 break
             if response.status_code == 400 and attempt < 2:
@@ -462,7 +640,8 @@ class Funda:
         while consecutive_404s < max_consecutive_404s:
             url = API_LISTING.format(listing_id=current_id)
             try:
-                response = self.session.get(url, timeout=self.timeout)
+                headers = _make_headers()
+                response = self._get(url, headers)
 
                 if response.status_code == 200:
                     consecutive_404s = 0
@@ -480,7 +659,7 @@ class Funda:
                 else:
                     consecutive_404s += 1
 
-            except requests.RequestException:
+            except requests.errors.RequestsError:
                 consecutive_404s += 1
 
             current_id += 1
@@ -531,12 +710,11 @@ class Funda:
             "zipcode": postcode,
         }
 
-        response = self.session.post(
-            API_WALTER,
-            json=payload,
-            headers={"Accept": "application/json", "Content-Type": "application/json"},
-            timeout=self.timeout,
-        )
+        walter_headers = [
+            ("Accept", "application/json"),
+            ("Content-Type", "application/json"),
+        ]
+        response = self._post(API_WALTER, walter_headers, json_data=payload)
 
         if response.status_code != 200:
             raise LookupError(f"Could not fetch price history (status {response.status_code})")

{pyfunda-2.2.0 → pyfunda-2.3.0}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "hatchling.build"
 
 [project]
 name = "pyfunda"
-version = "2.2.0"
+version = "2.3.0"
 description = "Python API for Funda.nl real estate listings"
 readme = "README.md"
 license = "AGPL-3.0-or-later"
@@ -24,6 +24,8 @@ classifiers = [
 ]
 dependencies = [
     "curl-cffi>=0.14.0",
+    "tls-client>=1.0.1",
+    "typing_extensions>=4.0.0",
 ]
 
 [project.urls]

{pyfunda-2.2.0 → pyfunda-2.3.0}/test_all_flows.py

@@ -626,18 +626,18 @@ def test_session_lazy_loading():
     print("  Session lazy loading working correctly")
 
 
-@test("Session headers are set correctly")
+@test("Session is created correctly")
 def test_session_headers():
     from funda import Funda
 
     f = Funda()
     session = f.session
 
-    assert 'user-agent' in session.headers, "Should have user-agent header"
-    assert 'Dart' in session.headers['user-agent'], "User-agent should contain 'Dart'"
+    # Session exists (headers are generated per-request with Dart fingerprint)
+    assert session is not None, "Session should exist"
 
     f.close()
-    print("  Session headers set correctly")
+    print("  Session created correctly")
 
 
 @test("Custom timeout is respected")