rapidata 2.21.3__tar.gz → 2.21.5__tar.gz

{rapidata-2.21.3 → rapidata-2.21.5}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: rapidata
-Version: 2.21.3
+Version: 2.21.5
 Summary: Rapidata package containing the Rapidata Python Client to interact with the Rapidata Web API in an easy way.
 License: Apache-2.0
 Author: Rapidata AG

{rapidata-2.21.3 → rapidata-2.21.5}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "rapidata"
-version = "2.21.3"
+version = "2.21.5"
 description = "Rapidata package containing the Rapidata Python Client to interact with the Rapidata Web API in an easy way."
 authors = ["Rapidata AG <info@rapidata.ai>"]
 license = "Apache-2.0"

{rapidata-2.21.3 → rapidata-2.21.5}/src/rapidata/api_client/rest.py

@@ -18,7 +18,7 @@ from typing import Dict, Optional
 
 import httpx
 from authlib.integrations.httpx_client import OAuth2Client
-from httpx import Timeout
+from httpx import Timeout, ConnectError
 
 from rapidata.api_client.exceptions import ApiException, ApiValueError
 
@@ -67,7 +67,13 @@ class RESTClientObject:
             **client_args,
         )
 
-        self.session.fetch_token()
+        try:
+            self.session.fetch_token()
+        except ConnectError as e:
+            if self._is_certificate_validation_error(e):
+                exit(self._get_ssl_verify_error_message())
+            else:
+                raise
 
     def setup_oauth_with_token(
         self,
@@ -218,14 +224,18 @@ class RESTClientObject:
                     raise ApiException(status=0, reason=msg)
 
             else:
-                r = session.request(
-                    method, url, timeout=timeout, headers=headers
-                )
+                r = session.request(method, url, timeout=timeout, headers=headers)
 
         except httpx.HTTPError as e:
             msg = "\n".join([type(e).__name__, str(e)])
             raise ApiException(status=0, reason=msg)
 
+        except ConnectError as e:
+            if self._is_certificate_validation_error(e):
+                exit(self._get_ssl_verify_error_message())
+            else:
+                raise
+
         return RESTResponse(r)
 
     def _get_session_defaults(self):
@@ -251,3 +261,28 @@ class RESTClientObject:
             client_kwargs["transport"] = transport
 
         return client_kwargs
+
+    @staticmethod
+    def _is_certificate_validation_error(error: ConnectError) -> bool:
+        """
+        Check if the error is related to certificate validation.
+        """
+        return error.args[0].startswith("[SSL: CERTIFICATE_VERIFY_FAILED]")
+
+    @staticmethod
+    def _get_ssl_verify_error_message() -> str:
+        return """
+        We encountered an issue while trying to verify the SSL certificate.
+        This often happens on macOS when using the default Python installation provided by Apple,
+        which lacks the required certificates to perform secure HTTPS requests.
+        To resolve this, please perform the following steps:
+        1. Make sure you are using the latest version of the Rapidata Package: `pip install --upgrade rapidata`
+        2. If you are using the default Python installation from Apple, consider switching to a different Python distribution, such as Homebrew or pyenv.
+        3. If you prefer to continue using the default Python, you can run the following command in your terminal:
+           `/Applications/Python\\ 3.1*/Install\\ Certificates.command`
+        
+        For more details on this issue, please refer to the following link:
+        https://stackoverflow.com/questions/42098126/mac-osx-python-ssl-sslerror-ssl-certificate-verify-failed-certificate-verify
+        
+        If the issue persists, please reach out to us at 'info@rapidata.ai', we're happy to help you.
+        """

{rapidata-2.21.3 → rapidata-2.21.5}/src/rapidata/rapidata_client/assets/__init__.py

@@ -7,3 +7,4 @@ from ._media_asset import MediaAsset
 from ._text_asset import TextAsset
 from ._multi_asset import MultiAsset
 from .data_type_enum import RapidataDataTypes
+from ._sessions import SessionManager

{rapidata-2.21.3 → rapidata-2.21.5}/src/rapidata/rapidata_client/assets/_media_asset.py

@@ -4,7 +4,7 @@ Defines the MediaAsset class for handling media file paths within assets.
 Implements lazy loading for URL-based media to prevent unnecessary downloads.
 """
 
-from typing import Optional, cast, Sequence
+from typing import Optional, cast
 import os
 from io import BytesIO
 from rapidata.rapidata_client.assets._base_asset import BaseAsset
@@ -16,8 +16,7 @@ import tempfile
 from pydantic import StrictStr, StrictBytes
 import logging
 from functools import cached_property
-from requests.adapters import HTTPAdapter
-from urllib3.util.retry import Retry
+from rapidata.rapidata_client.assets._sessions import SessionManager
 
 class MediaAsset(BaseAsset):
     """MediaAsset Class with Lazy Loading
@@ -78,7 +77,7 @@ class MediaAsset(BaseAsset):
         
         self._url = None
         self._content = None
-        self.session: None | requests.Session  = None
+        self.session: requests.Session  = SessionManager.get_session()
         
         if re.match(r'^https?://', path):
             self._url = path
@@ -130,11 +129,11 @@ class MediaAsset(BaseAsset):
                 tmp.flush()
                 tmp_path = tmp.name
                 
-                try:
-                    tag = TinyTag.get(tmp_path)
-                finally:
-                    # Clean up the temporary file
-                    os.unlink(tmp_path)
+            try:
+                tag = TinyTag.get(tmp_path)
+            finally:
+                # Clean up the temporary file
+                os.unlink(tmp_path)
             
             if tag.duration is None:
                 raise ValueError("Could not read duration from file")

rapidata-2.21.5/src/rapidata/rapidata_client/assets/_sessions.py

@@ -0,0 +1,35 @@
+import requests
+from requests.adapters import HTTPAdapter
+from urllib3.util.retry import Retry
+
+class SessionManager:
+    _session = None
+    
+    @classmethod
+    def get_session(cls, ) -> requests.Session:
+        """Get a singleton requests session with retry logic.
+
+        Returns:
+            requests.Session: A singleton requests session with retry logic.
+        """
+        if cls._session is None:
+            max_retries: int = 5 
+            max_workers: int = 10
+            cls._session = requests.Session()
+            retries = Retry(
+                total=max_retries,
+                backoff_factor=1,
+                status_forcelist=[500, 502, 503, 504],
+                allowed_methods=["GET"],
+                respect_retry_after_header=True
+            )
+
+            adapter = HTTPAdapter(
+                pool_connections=max_workers * 2,
+                pool_maxsize=max_workers * 4,
+                max_retries=retries
+            )
+            cls._session.mount('http://', adapter)
+            cls._session.mount('https://', adapter)
+
+        return cls._session

{rapidata-2.21.3 → rapidata-2.21.5}/src/rapidata/rapidata_client/demographic/demographic_manager.py

@@ -25,38 +25,6 @@ class DemographicManager:
                 title=instruction
             )
         )
-        session = self._get_session()
-        media.session = session
         
         self._openapi_service.rapid_api.rapid_demographic_post(model=model, file=[media.to_file()])
-
-    def _get_session(self, max_retries: int = 5, max_workers: int = 10) -> requests.Session:   
-        """Get a requests session with retry logic.
-
-
-        Args:
-            max_retries (int): The maximum number of retries.
-            max_workers (int): The maximum number of workers.
-
-        Returns:
-            requests.Session: A requests session with retry logic.
-        """
-        session = requests.Session()
-        retries = Retry(
-            total=max_retries,
-            backoff_factor=1,
-            status_forcelist=[500, 502, 503, 504],
-            allowed_methods=["GET"],
-            respect_retry_after_header=True
-        )
-
-        adapter = HTTPAdapter(
-            pool_connections=max_workers * 2,
-            pool_maxsize=max_workers * 4,
-            max_retries=retries
-        )
-        session.mount('http://', adapter)
-        session.mount('https://', adapter)
-
-        return session
         

{rapidata-2.21.3 → rapidata-2.21.5}/src/rapidata/rapidata_client/order/_rapidata_dataset.py

@@ -89,7 +89,6 @@ class RapidataDataset:
             media_asset: MediaAsset or MultiAsset to upload
             meta: Optional metadata for the asset
             index: Sort index for the upload
-            session: Requests session for HTTP requests
             
         Returns:
             tuple[list[str], list[str]]: Lists of successful and failed identifiers
@@ -272,7 +271,6 @@ class RapidataDataset:
         Args:
             media_paths: List of assets to upload
             metadata: Optional sequence of metadata
-            session: Requests session for HTTP requests
             max_workers: Maximum number of concurrent workers
             chunk_size: Number of items to process in each batch
             stop_progress_tracking: Event to signal progress tracking to stop

{rapidata-2.21.3 → rapidata-2.21.5}/src/rapidata/rapidata_client/order/rapidata_order.py

@@ -50,7 +50,7 @@ class RapidataOrder:
 
     def run(self, print_link: bool = True) -> "RapidataOrder":
         """Runs the order to start collecting responses."""
-        self.__openapi_service.order_api.order_submit_post(self.order_id)
+        self.__openapi_service.order_api.order_order_id_submit_post(self.order_id)
         if print_link:
             print(f"Order '{self.name}' is now viewable under: {self.order_details_page}")
         return self

{rapidata-2.21.3 → rapidata-2.21.5}/src/rapidata/rapidata_client/validation/rapidata_validation_set.py

@@ -3,6 +3,7 @@ from rapidata.service.openapi_service import OpenAPIService
 from requests.adapters import HTTPAdapter, Retry
 import requests
 from rapidata.api_client.models.update_dimensions_model import UpdateDimensionsModel
+from rapidata.rapidata_client.assets._sessions import SessionManager
 
 class RapidataValidationSet:
     """A class for interacting with a Rapidata validation set.
@@ -20,7 +21,6 @@ class RapidataValidationSet:
         self.id = validation_set_id
         self.name = name
         self.__openapi_service = openapi_service
-        self.__session = self._get_session()
     
     def add_rapid(self, rapid: Rapid):
         """Add a Rapid to the validation set.
@@ -28,7 +28,7 @@ class RapidataValidationSet:
         Args:
             rapid (Rapid): The Rapid to add to the validation set.
         """
-        rapid._add_to_validation_set(self.id, self.__openapi_service, self.__session)
+        rapid._add_to_validation_set(self.id, self.__openapi_service)
         return self
 
     def update_dimensions(self, dimensions: list[str]):
@@ -40,36 +40,6 @@ class RapidataValidationSet:
         self.__openapi_service.validation_api.validation_validation_set_id_dimensions_patch(self.id, UpdateDimensionsModel(dimensions=dimensions))
         return self
 
-    def _get_session(self, max_retries: int = 5, max_workers: int = 10) -> requests.Session:   
-        """Get a requests session with retry logic.
-
-
-        Args:
-            max_retries (int): The maximum number of retries.
-            max_workers (int): The maximum number of workers.
-
-        Returns:
-            requests.Session: A requests session with retry logic.
-        """
-        session = requests.Session()
-        retries = Retry(
-            total=max_retries,
-            backoff_factor=1,
-            status_forcelist=[500, 502, 503, 504],
-            allowed_methods=["GET"],
-            respect_retry_after_header=True
-        )
-
-        adapter = HTTPAdapter(
-            pool_connections=max_workers * 2,
-            pool_maxsize=max_workers * 4,
-            max_retries=retries
-        )
-        session.mount('http://', adapter)
-        session.mount('https://', adapter)
-
-        return session
-
     def __str__(self):
         return f"name: '{self.name}' id: {self.id}"
 

{rapidata-2.21.3 → rapidata-2.21.5}/src/rapidata/rapidata_client/validation/rapids/rapids.py

@@ -30,14 +30,14 @@ class Rapid():
         self.randomCorrectProbability = randomCorrectProbability
         self.explanation = explanation 
 
-    def _add_to_validation_set(self, validationSetId: str, openapi_service: OpenAPIService, session: requests.Session) -> None:
+    def _add_to_validation_set(self, validationSetId: str, openapi_service: OpenAPIService) -> None:
         if isinstance(self.asset, TextAsset) or (isinstance(self.asset, MultiAsset) and isinstance(self.asset.assets[0], TextAsset)):
             openapi_service.validation_api.validation_add_validation_text_rapid_post(
                 add_validation_text_rapid_model=self.__to_text_model(validationSetId)
             )
 
         elif isinstance(self.asset, MediaAsset) or (isinstance(self.asset, MultiAsset) and isinstance(self.asset.assets[0], MediaAsset)):
-            model = self.__to_media_model(validationSetId, session=session)
+            model = self.__to_media_model(validationSetId)
             openapi_service.validation_api.validation_add_validation_rapid_post(
                 model=model[0], files=model[1]
             )
@@ -45,7 +45,7 @@ class Rapid():
         else:
             raise TypeError("The asset must be a MediaAsset, TextAsset, or MultiAsset")
 
-    def __to_media_model(self, validationSetId: str, session: requests.Session) -> tuple[AddValidationRapidModel, list[StrictStr | tuple[StrictStr, StrictBytes] | StrictBytes]]:
+    def __to_media_model(self, validationSetId: str) -> tuple[AddValidationRapidModel, list[StrictStr | tuple[StrictStr, StrictBytes] | StrictBytes]]:
         assets: list[MediaAsset] = [] 
         if isinstance(self.asset, MultiAsset):
             for asset in self.asset.assets:
@@ -60,9 +60,6 @@ class Rapid():
         if isinstance(self.asset, MediaAsset):
             assets = [self.asset]
 
-        for asset in assets:
-            asset.session = session
-
         return (AddValidationRapidModel(
             validationSetId=validationSetId,
             payload=AddValidationRapidModelPayload(self.payload),