scrapely-python 0.1.24__tar.gz → 0.1.26__tar.gz

{scrapely_python-0.1.24/src/scrapely_python.egg-info → scrapely_python-0.1.26}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: scrapely-python
-Version: 0.1.24
+Version: 0.1.26
 Summary: Scrapely Python SDK - A toolkit for building Actors on the Scrapely platform
 Author-email: Scrapely <support@scrape.ly>
 License-Expression: MIT AND Apache-2.0

{scrapely_python-0.1.24 → scrapely_python-0.1.26}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "scrapely-python"
-version = "0.1.24"
+version = "0.1.26"
 description = "Scrapely Python SDK - A toolkit for building Actors on the Scrapely platform"
 readme = "README.md"
 license = "MIT AND Apache-2.0"

{scrapely_python-0.1.24 → scrapely_python-0.1.26}/src/scrapely/__init__.py

@@ -227,7 +227,7 @@ if TYPE_CHECKING:  # Static analysers and IDEs see the lazy names as plain impor
         set_api_storage,
     )
 
-__version__ = '0.1.24'
+__version__ = '0.1.26'
 __all__ = [
     # Core
     'Actor',

{scrapely_python-0.1.24 → scrapely_python-0.1.26}/src/scrapely/actor.py

@@ -360,6 +360,7 @@ class Actor(metaclass=_ActorMeta):
 
     @classmethod
     async def init(cls, options: Optional[InitOptions] = None) -> None:
+        cls._main_loop = asyncio.get_running_loop()  # for debug_capture_sync from worker threads
         if cls._initialized:
             return
 
@@ -934,6 +935,29 @@ class Actor(metaclass=_ActorMeta):
         import hashlib
         return 'DEBUG_CAPTURE_' + hashlib.sha1(str(key).strip().encode('utf-8')).hexdigest()[:24]
 
+    _main_loop: Optional[asyncio.AbstractEventLoop] = None
+
+    @classmethod
+    def debug_capture_sync(cls, key: str, payload: Any) -> None:
+        """debug_capture for synchronous code (also from worker threads, e.g. asyncio.to_thread).
+        Does nothing unless the platform enabled capture for this run."""
+        if os.environ.get('SCRAPELY_DEBUG_CAPTURE') != '1' or not key:
+            return
+        loop = cls._main_loop
+        if loop is None or loop.is_closed():
+            return
+        try:
+            try:
+                running = asyncio.get_running_loop()
+            except RuntimeError:
+                running = None
+            if running is loop:
+                loop.create_task(cls.debug_capture(key, payload))
+            else:
+                asyncio.run_coroutine_threadsafe(cls.debug_capture(key, payload), loop).result(timeout=15)
+        except Exception as err:  # never break the run
+            logging.getLogger('scrapely').debug('debug_capture_sync failed: %s', err)
+
     @classmethod
     async def debug_capture(cls, key: str, payload: Any) -> None:
         """Save the raw source payload (API JSON or HTML) an item was extracted from, keyed by

{scrapely_python-0.1.24 → scrapely_python-0.1.26}/src/scrapely/solver.py

@@ -1,6 +1,6 @@
 """Anti-bot solver helpers for Scrapely Actors.
 
-The platform runs the challenge solver (xhr.dev today) behind
+The platform runs the challenge solver behind
 ``/v2/solver``. Actors never hold a solver key: the platform injects
 ``SCRAPELY_SOLVER_URL`` and ``SCRAPELY_SOLVER_API_KEY`` (the run's own token)
 into runs of Actors that may use it.
@@ -88,7 +88,7 @@ DEGRADED_STATUS_MESSAGE = "Site protection temporarily unavailable, results may
 # Codes the platform answers with when a vendor cannot be solved right now.
 UNAVAILABLE_CODES = ("vendor_degraded", "vendor_down", "provider_unavailable", "vendor_unsupported", "solver_not_enabled")
 
-_COMPAT_SUFFIX = "/compat/xhrdev"
+_COMPAT_SUFFIXES = ("/compat/v1", "/compat/xhrdev")
 _STATUS_TTL_SECS = 15.0
 _MAX_BODY = 4 * 1024 * 1024
 
@@ -118,9 +118,10 @@ def _gateway() -> tuple[str, str] | None:
     if env is None:
         return None
     url, key = env
-    if not url.endswith(_COMPAT_SUFFIX):
-        return None
-    return url[: -len(_COMPAT_SUFFIX)], key
+    for suffix in _COMPAT_SUFFIXES:
+        if url.endswith(suffix):
+            return url[: -len(suffix)], key
+    return None
 
 
 # ─── errors ──────────────────────────────────────────────────────────────────

{scrapely_python-0.1.24 → scrapely_python-0.1.26/src/scrapely_python.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: scrapely-python
-Version: 0.1.24
+Version: 0.1.26
 Summary: Scrapely Python SDK - A toolkit for building Actors on the Scrapely platform
 Author-email: Scrapely <support@scrape.ly>
 License-Expression: MIT AND Apache-2.0

{scrapely_python-0.1.24 → scrapely_python-0.1.26}/tests/test_debug_capture.py

@@ -29,3 +29,26 @@ def test_debug_capture_writes_keyed_capped_record(monkeypatch):
     assert key.startswith('DEBUG_CAPTURE_') and len(key) == len('DEBUG_CAPTURE_') + 24
     assert saved[key]['key'] == 'https://x/item/1'
     assert len(saved[key]['payload']) == Actor._DEBUG_CAPTURE_MAX_BYTES
+
+
+def test_debug_capture_sync_from_thread(monkeypatch):
+    monkeypatch.setenv('SCRAPELY_DEBUG_CAPTURE', '1')
+    saved = {}
+
+    async def fake_set_value(key, value, content_type=None):
+        saved[key] = value
+
+    monkeypatch.setattr(Actor, 'set_value', fake_set_value)
+    monkeypatch.setattr(Actor, '_debug_capture_count', 0)
+
+    async def main():
+        Actor._main_loop = asyncio.get_running_loop()
+        await asyncio.to_thread(Actor.debug_capture_sync, 'https://x/item/2', {'a': 1})
+
+    asyncio.run(main())
+    assert saved[Actor.debug_capture_record_key('https://x/item/2')]['key'] == 'https://x/item/2'
+
+
+def test_debug_capture_sync_noop_without_flag(monkeypatch):
+    monkeypatch.delenv('SCRAPELY_DEBUG_CAPTURE', raising=False)
+    Actor.debug_capture_sync('k', {'a': 1})  # no loop, no flag: nothing happens

{scrapely_python-0.1.24 → scrapely_python-0.1.26}/tests/test_solver.py

@@ -177,7 +177,7 @@ class FakeSession:
 @pytest.fixture
 def gateway(monkeypatch):
     gw = Gateway()
-    monkeypatch.setenv("SCRAPELY_SOLVER_URL", gw.base + "/v2/solver/compat/xhrdev")
+    monkeypatch.setenv("SCRAPELY_SOLVER_URL", gw.base + "/v2/solver/compat/v1")
     monkeypatch.setenv("SCRAPELY_SOLVER_API_KEY", "st_test_token")
     S._status_cache.clear()
     yield gw