rapidata 2.21.5__py3-none-any.whl → 2.23.0__py3-none-any.whl

rapidata/rapidata_client/validation/validation_set_manager.py

@@ -15,7 +15,7 @@ from urllib3._collections import HTTPHeaderDict # type: ignore[import]
 
 from rapidata.rapidata_client.validation.rapids.box import Box
 
-from rapidata.api_client.models.query_validation_set_model import QueryValidationSetModel
+from rapidata.rapidata_client.logging import logger, managed_print
 from tqdm import tqdm
 
 
@@ -29,6 +29,7 @@ class ValidationSetManager:
     def __init__(self, openapi_service: OpenAPIService) -> None:
         self.__openapi_service = openapi_service
         self.rapid = RapidsManager()
+        logger.debug("ValidationSetManager initialized")
 
     def create_classification_set(self,
         name: str,
@@ -40,7 +41,6 @@ class ValidationSetManager:
         contexts: list[str] | None = None,
         explanations: list[str | None] | None = None,
         dimensions: list[str] = [],
-        print_confirmation: bool = True,
     ) -> RapidataValidationSet:
         """Create a classification validation set.
         
@@ -60,7 +60,6 @@ class ValidationSetManager:
                 Will be match up with the datapoints using the list index.
             explanations (list[str | None], optional): The explanations for each datapoint. Will be given to the annotators in case the answer is wrong. Defaults to None.
             dimensions (list[str], optional): The dimensions to add to the validation set accross which users will be tracked. Defaults to [] which is the default dimension.
-            print_confirmation (bool, optional): Whether to print a confirmation message that validation set has been created. Defaults to True.
 
         Example:
             ```python
@@ -83,6 +82,7 @@ class ValidationSetManager:
         if(explanations and len(explanations) != len(datapoints)):
             raise ValueError("The numeber of reasons and datapoints must be equal, the index must align, but can be padded with None")
        
+        logger.debug("Creating classification rapids")
         rapids: list[Rapid] = []
         for i in range(len(datapoints)):
             rapids.append(
@@ -97,7 +97,8 @@ class ValidationSetManager:
                 )
             )
 
-        return self._submit(name=name, rapids=rapids, print_confirmation=print_confirmation, dimensions=dimensions)
+        logger.debug("Submitting classification rapids")
+        return self._submit(name=name, rapids=rapids, dimensions=dimensions)
 
     def create_compare_set(self,
         name: str,
@@ -108,7 +109,6 @@ class ValidationSetManager:
         contexts: list[str] | None = None,
         explanation: list[str | None] | None = None,
         dimensions: list[str] = [],
-        print_confirmation: bool = True,
     ) -> RapidataValidationSet:
         """Create a comparison validation set.
 
@@ -128,7 +128,6 @@ class ValidationSetManager:
                 Will be match up with the datapoints using the list index.
             explanation (list[str | None], optional): The explanations for each datapoint. Will be given to the annotators in case the answer is wrong. Defaults to None.
             dimensions (list[str], optional): The dimensions to add to the validation set accross which users will be tracked. Defaults to [] which is the default dimension.
-            print_confirmation (bool, optional): Whether to print a confirmation message that validation set has been created. Defaults to True.
 
         Example:
             ```python
@@ -150,7 +149,8 @@ class ValidationSetManager:
  
         if(explanation and len(explanation) != len(datapoints)):
             raise ValueError("The numeber of reasons and datapoints must be equal, the index must align, but can be padded with None")
-              
+        
+        logger.debug("Creating comparison rapids")
         rapids: list[Rapid] = []
         for i in range(len(datapoints)):
             rapids.append(
@@ -163,8 +163,9 @@ class ValidationSetManager:
                     explanation=explanation[i] if explanation != None else None
                 )
             )
- 
-        return self._submit(name=name, rapids=rapids, dimensions=dimensions, print_confirmation=print_confirmation)
+        
+        logger.debug("Submitting comparison rapids")
+        return self._submit(name=name, rapids=rapids, dimensions=dimensions)
   
     def create_select_words_set(self,
         name: str,
@@ -176,7 +177,6 @@ class ValidationSetManager:
         required_completeness: float = 1.0,
         explanation: list[str | None] | None = None,
         dimensions: list[str] = [],
-        print_confirmation: bool = True,
     ) -> RapidataValidationSet:
         """Create a select words validation set.
 
@@ -195,7 +195,6 @@ class ValidationSetManager:
             required_completeness (float, optional): The required completeness for the labeler to get the rapid correct (miminum ratio of total correct words selected). Defaults to 1.0 (all correct words need to be selected).
             explanation (list[str | None], optional): The explanations for each datapoint. Will be given to the annotators in case the answer is wrong. Defaults to None.
             dimensions (list[str], optional): The dimensions to add to the validation set accross which users will be tracked. Defaults to [] which is the default dimension.
-            print_confirmation (bool, optional): Whether to print a confirmation message that validation set has been created. Defaults to True.
 
         Example:
             ```python
@@ -214,7 +213,8 @@ class ValidationSetManager:
  
         if(explanation and len(explanation) != len(datapoints)):
             raise ValueError("The numeber of reasons and datapoints must be equal, the index must align, but can be padded with None")
-              
+
+        logger.debug("Creating select words rapids")
         rapids: list[Rapid] = []
         for i in range(len(datapoints)):
             rapids.append(
@@ -229,7 +229,8 @@ class ValidationSetManager:
                 )
             )
 
-        return self._submit(name=name, rapids=rapids, dimensions=dimensions, print_confirmation=print_confirmation)
+        logger.debug("Submitting select words rapids")
+        return self._submit(name=name, rapids=rapids, dimensions=dimensions)
 
     def create_locate_set(self,
         name: str,
@@ -239,7 +240,6 @@ class ValidationSetManager:
         contexts: list[str] | None = None,
         explanation: list[str | None] | None = None,
         dimensions: list[str] = [],
-        print_confirmation: bool = True,
     ) -> RapidataValidationSet:
         """Create a locate validation set.
 
@@ -254,7 +254,6 @@ class ValidationSetManager:
             contexts (list[str], optional): The contexts for each datapoint. Defaults to None.
             explanation (list[str | None], optional): The explanations for each datapoint. Will be given to the annotators in case the answer is wrong. Defaults to None.
             dimensions (list[str], optional): The dimensions to add to the validation set accross which users will be tracked. Defaults to [] which is the default dimension.
-            print_confirmation (bool, optional): Whether to print a confirmation message that validation set has been created. Defaults to True.
 
         Example:
             ```python
@@ -275,7 +274,8 @@ class ValidationSetManager:
  
         if(explanation and len(explanation) != len(datapoints)):
             raise ValueError("The numeber of reasons and datapoints must be equal, the index must align, but can be padded with None")
-              
+        
+        logger.debug("Creating locate rapids")
         rapids = []
         rapids: list[Rapid] = []
         for i in range(len(datapoints)):
@@ -290,7 +290,8 @@ class ValidationSetManager:
                 )
             )
         
-        return self._submit(name=name, rapids=rapids, dimensions=dimensions, print_confirmation=print_confirmation)
+        logger.debug("Submitting locate rapids")
+        return self._submit(name=name, rapids=rapids, dimensions=dimensions)
     
     def create_draw_set(self,
         name: str,
@@ -300,7 +301,6 @@ class ValidationSetManager:
         contexts: list[str] | None = None,
         explanation: list[str | None] | None = None,
         dimensions: list[str] = [],
-        print_confirmation: bool = True,
     ) -> RapidataValidationSet:
         """Create a draw validation set.
 
@@ -315,7 +315,6 @@ class ValidationSetManager:
             contexts (list[str], optional): The contexts for each datapoint. Defaults to None.
             explanation (list[str | None], optional): The explanations for each datapoint. Will be given to the annotators in case the answer is wrong. Defaults to None.
             dimensions (list[str], optional): The dimensions to add to the validation set accross which users will be tracked. Defaults to [] which is the default dimension.
-            print_confirmation (bool, optional): Whether to print a confirmation message that validation set has been created. Defaults to True.
 
         Example:
             ```python
@@ -336,7 +335,8 @@ class ValidationSetManager:
  
         if(explanation and len(explanation) != len(datapoints)):
             raise ValueError("The numeber of reasons and datapoints must be equal, the index must align, but can be padded with None")
-              
+
+        logger.debug("Creating draw rapids")
         rapids: list[Rapid] = []
         for i in range(len(datapoints)):
             rapids.append(
@@ -350,7 +350,8 @@ class ValidationSetManager:
                 )
             )
 
-        return self._submit(name=name, rapids=rapids, dimensions=dimensions, print_confirmation=print_confirmation)
+        logger.debug("Submitting draw rapids")
+        return self._submit(name=name, rapids=rapids, dimensions=dimensions)
 
     def create_timestamp_set(self,
         name: str,
@@ -360,7 +361,6 @@ class ValidationSetManager:
         contexts: list[str] | None = None,
         explanation: list[str | None] | None = None,
         dimensions: list[str] = [],
-        print_confirmation: bool = True,
     ) -> RapidataValidationSet:
         """Create a timestamp validation set.
 
@@ -376,7 +376,6 @@ class ValidationSetManager:
             contexts (list[str], optional): The contexts for each datapoint. Defaults to None.
             explanation (list[str | None], optional): The explanations for each datapoint. Will be given to the annotators in case the answer is wrong. Defaults to None.
             dimensions (list[str], optional): The dimensions to add to the validation set accross which users will be tracked. Defaults to [] which is the default dimension.
-            print_confirmation (bool, optional): Whether to print a confirmation message that validation set has been created. Defaults to True.
 
         Example:
             ```python
@@ -398,7 +397,7 @@ class ValidationSetManager:
         if(explanation and len(explanation) != len(datapoints)):
             raise ValueError("The numeber of reasons and datapoints must be equal, the index must align, but can be padded with None")
               
-
+        logger.debug("Creating timestamp rapids")
         rapids: list[Rapid] = []
         for i in range(len(datapoints)):
             rapids.append(
@@ -411,13 +410,13 @@ class ValidationSetManager:
                 )
             )
 
-        return self._submit(name=name, rapids=rapids, dimensions=dimensions, print_confirmation=print_confirmation)
+        logger.debug("Submitting timestamp rapids")
+        return self._submit(name=name, rapids=rapids, dimensions=dimensions)
     
     def create_mixed_set(self,
         name: str,
         rapids: list[Rapid],
         dimensions: list[str] = [],
-        print_confirmation: bool = True
     ) -> RapidataValidationSet:
         """Create a validation set with a list of rapids.
 
@@ -425,10 +424,9 @@ class ValidationSetManager:
             name (str): The name of the validation set. (will not be shown to the labeler)
             rapids (list[Rapid]): The list of rapids to add to the validation set.
             dimensions (list[str], optional): The dimensions to add to the validation set accross which users will be tracked. Defaults to [] which is the default dimension.
-            print_confirmation (bool, optional): Whether to print a confirmation message that validation set has been created. Defaults to True.
         """
 
-        return self._submit(name=name, rapids=rapids, dimensions=dimensions, print_confirmation=print_confirmation)
+        return self._submit(name=name, rapids=rapids, dimensions=dimensions)
     
     def get_validation_set_by_id(self, validation_set_id: str) -> RapidataValidationSet:
         """Get a validation set by ID.
@@ -444,32 +442,37 @@ class ValidationSetManager:
         except Exception:
             raise ValueError(f"ValidationSet with ID {validation_set_id} not found.")
         
-        return RapidataValidationSet(validation_set_id, validation_set.name, self.__openapi_service)
+        return RapidataValidationSet(validation_set_id, str(validation_set.name), self.__openapi_service)
 
-    def _submit(self, name: str, rapids: list[Rapid], dimensions: list[str] | None, print_confirmation: bool) -> RapidataValidationSet:
+    def _submit(self, name: str, rapids: list[Rapid], dimensions: list[str] | None) -> RapidataValidationSet:
+        logger.debug("Creating validation set")
         validation_set_id = (
             self.__openapi_service.validation_api.validation_create_validation_set_post(
                 name=name
             )
         ).validation_set_id
 
+        logger.debug(f"Validation set created with ID: {validation_set_id}")
+
         if validation_set_id is None:
             raise ValueError("Failed to create validation set")
 
+        logger.debug("Creating validation set instance")
+
         validation_set = RapidataValidationSet(
             name=name,
             validation_set_id=validation_set_id,
             openapi_service=self.__openapi_service
         )
 
+        logger.debug("Adding rapids to validation set")
         for rapid in tqdm(rapids, desc="Uploading validation tasks"):
             validation_set.add_rapid(rapid)
         
-        if print_confirmation:
-            print()
-            print(f"Validation set '{name}' created with ID {validation_set_id}\n",
-                  f"Now viewable under: https://app.{self.__openapi_service.environment}/validation-set/detail/{validation_set_id}",
-                  sep="")
+        managed_print()
+        managed_print(f"Validation set '{name}' created with ID {validation_set_id}\n",
+                f"Now viewable under: https://app.{self.__openapi_service.environment}/validation-set/detail/{validation_set_id}",
+                sep="")
         
         if dimensions:
             validation_set.update_dimensions(dimensions)
@@ -500,6 +503,6 @@ class ValidationSetManager:
         except Exception as e:
             raise ValueError(f"Unknown error occured: {e}")
 
-        validation_sets = [self.get_validation_set_by_id(validation_set.id) for validation_set in validation_page_result.items]
+        validation_sets = [self.get_validation_set_by_id(str(validation_set.id)) for validation_set in validation_page_result.items]
         return validation_sets
 

rapidata/service/credential_manager.py

@@ -11,6 +11,7 @@ from typing import Dict, List, Optional, Tuple
 import requests
 from colorama import Fore
 from pydantic import BaseModel
+from rapidata.rapidata_client.logging import logger, managed_print
 
 
 class ClientCredential(BaseModel):
@@ -21,27 +22,6 @@ class ClientCredential(BaseModel):
     created_at: datetime
     last_used: datetime
 
-    @classmethod
-    def from_dict(cls, data: Dict):
-        return cls(
-            display_name=data["display_name"],
-            client_id=data["client_id"],
-            client_secret=data["client_secret"],
-            endpoint=data["endpoint"],
-            created_at=datetime.fromisoformat(data["created_at"]),
-            last_used=datetime.fromisoformat(data["last_used"]),
-        )
-
-    def to_dict(self) -> Dict:
-        return {
-            "display_name": self.display_name,
-            "client_id": self.client_id,
-            "client_secret": self.client_secret,
-            "endpoint": self.endpoint,
-            "created_at": self.created_at.isoformat(),
-            "last_used": self.last_used.isoformat(),
-        }
-
     def get_display_string(self):
         return f"{self.display_name} - Client ID: {self.client_id} (Created: {self.created_at.strftime('%Y-%m-%d %H:%M:%S')})"
 
@@ -72,6 +52,7 @@ class CredentialManager:
 
     def _read_credentials(self) -> Dict[str, List[ClientCredential]]:
         """Read all stored credentials from the config file."""
+        logger.debug(f"Reading credentials from {self.config_path}")
         if not self.config_path.exists():
             return {}
 
@@ -79,7 +60,7 @@ class CredentialManager:
             with open(self.config_path, "r") as f:
                 data = json.load(f)
                 return {
-                    env: [ClientCredential.from_dict(cred) for cred in creds]
+                    env: [ClientCredential.model_validate(cred) for cred in creds]
                     for env, creds in data.items()
                 }
         except json.JSONDecodeError:
@@ -89,15 +70,18 @@ class CredentialManager:
         self, credentials: Dict[str, List[ClientCredential]]
     ) -> None:
         data = {
-            env: [cred.to_dict() for cred in creds]
+            env: [cred.model_dump(mode="json") for cred in creds]
             for env, creds in credentials.items()
         }
 
         with open(self.config_path, "w") as f:
             json.dump(data, f, indent=2)
 
+        logger.debug(f"Credentials written to {self.config_path} with data: {data}")
+
         # Ensure file is only readable by the user
         os.chmod(self.config_path, 0o600)
+        logger.debug(f"Set permissions for {self.config_path} to read/write for user only.")
 
     def _store_credential(self, credential: ClientCredential) -> None:
         credentials = self._read_credentials()
@@ -123,15 +107,19 @@ class CredentialManager:
     def get_client_credentials(self) -> Optional[ClientCredential]:
         """Gets stored client credentials or create new ones via browser auth."""
         credentials = self._read_credentials()
+        logger.debug(f"Stored credentials: {credentials}")
         env_credentials = credentials.get(self.endpoint, [])
 
         if env_credentials:
+            logger.debug(f"Found credentials for {self.endpoint}: {env_credentials}")
             credential = self._select_credential(env_credentials)
+            logger.debug(f"Selected credential: {credential}")
             if credential:
                 credential.last_used = datetime.now(timezone.utc)
                 self._write_credentials(credentials)
                 return credential
 
+        logger.debug(f"No credentials found for {self.endpoint}. Creating new ones.")
         return self._create_new_credentials()
 
     def reset_credentials(self) -> None:
@@ -140,22 +128,26 @@ class CredentialManager:
         if self.endpoint in credentials:
             del credentials[self.endpoint]
             self._write_credentials(credentials)
+            logger.info(
+                f"Credentials for {self.endpoint} have been reset."
+            )
 
     def _get_bridge_tokens(self) -> Optional[BridgeToken]:
         """Get bridge tokens from the identity endpoint."""
+        logger.debug("Getting bridge tokens")
         try:
             bridge_endpoint = (
                 f"{self.endpoint}/Identity/CreateBridgeToken?clientId=rapidata-cli"
             )
             response = requests.post(bridge_endpoint, verify=self.cert_path)
             if not response.ok:
-                print(f"Failed to get bridge tokens: {response.status_code}")
+                logger.error(f"Failed to get bridge tokens: {response.status_code}")
                 return None
 
             data = response.json()
             return BridgeToken(read_key=data["readKey"], write_key=data["writeKey"])
         except requests.RequestException as e:
-            print(f"Failed to get bridge tokens: {e}")
+            logger.error(f"Failed to get bridge tokens: {e}")
             return None
 
     def _poll_read_key(self, read_key: str) -> Optional[str]:
@@ -177,14 +169,14 @@ class CredentialManager:
                     continue
                 else:
                     # Error occurred
-                    print(f"Error polling read key: {response.status_code}")
+                    logger.error(f"Error polling read key: {response.status_code}")
                     return None
 
             except requests.RequestException as e:
-                print(f"Error polling read key: {e}")
+                logger.error(f"Error polling read key: {e}")
                 return None
 
-        print("Polling timed out")
+        logger.error("Polling timed out")
         return None
 
     def _create_client(self, access_token: str) -> Optional[Tuple[str, str, str]]:
@@ -206,7 +198,7 @@ class CredentialManager:
             data = response.json()
             return data.get("clientId"), data.get("clientSecret"), display_name
         except requests.RequestException as e:
-            print(f"Failed to create client: {e}")
+            logger.error(f"Failed to create client: {e}")
             return None
 
     def _create_new_credentials(self) -> Optional[ClientCredential]:
@@ -219,7 +211,7 @@ class CredentialManager:
 
         if not could_open_browser:
             encoded_url = urllib.parse.quote(auth_url, safe="%/:=&?~#+!$,;'@()*[]")
-            print(
+            managed_print(
                 Fore.RED
                 + f'Please open the following URL in your browser to log in: "{encoded_url}"'
                 + Fore.RESET

rapidata/service/openapi_service.py

@@ -11,6 +11,7 @@ from rapidata.api_client.api.workflow_api import WorkflowApi
 from rapidata.api_client.configuration import Configuration
 from rapidata.service.credential_manager import CredentialManager
 from rapidata.rapidata_client.api.rapidata_exception import RapidataApiClient
+from rapidata.rapidata_client.logging import logger
 
 
 class OpenAPIService:
@@ -31,18 +32,25 @@ class OpenAPIService:
         if environment == "rapidata.dev" and not cert_path:
             cert_path = _get_local_certificate()
 
+        logger.debug(f"Using cert_path: {cert_path} environment: {environment} token: {token} client_id: {client_id} client_secret: {client_secret}")
+        logger.debug("Initializing OpenAPIService")
         self.credential_manager = CredentialManager(
             endpoint=auth_endpoint, cert_path=cert_path
         )
+        logger.debug("CredentialManager initialized")
 
+        logger.debug("Initializing RapidataApiClient")
         client_configuration = Configuration(host=endpoint, ssl_ca_cert=cert_path)
+        logger.debug(f"Client configuration: {client_configuration}")
         self.api_client = RapidataApiClient(
             configuration=client_configuration,
             header_name="X-Client",
             header_value=f"RapidataPythonSDK/{self._get_rapidata_package_version()}",
         )
+        logger.debug("RapidataApiClient initialized")
 
         if token:
+            logger.debug("Using token for authentication")
             self.api_client.rest_client.setup_oauth_with_token(
                 token=token,
                 token_endpoint=f"{auth_endpoint}/connect/token",
@@ -50,9 +58,11 @@ class OpenAPIService:
                 client_secret=client_secret,
                 leeway=leeway,
             )
+            logger.debug("Token authentication setup complete")
             return
 
         if not client_id or not client_secret:
+            logger.debug("Client ID and secret not provided, fetching from credential manager")
             credentials = self.credential_manager.get_client_credentials()
             if not credentials:
                 raise ValueError("Failed to fetch client credentials")
@@ -65,6 +75,7 @@ class OpenAPIService:
             token_endpoint=f"{auth_endpoint}/connect/token",
             scope=oauth_scope,
         )
+        logger.debug("Client credentials authentication setup complete")
 
     def reset_credentials(self):
         self.credential_manager.reset_credentials()

{rapidata-2.21.5.dist-info → rapidata-2.23.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: rapidata
-Version: 2.21.5
+Version: 2.23.0
 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
@@ -27,6 +27,7 @@ Requires-Dist: tqdm (>=4.66.5,<5.0.0)
 Description-Content-Type: text/markdown
 
 # Rapidata Python Client
+
 Python client to interface with the Rapidata API.
 
 Docs: https://docs.rapidata.ai/