surfdataverse 4.1.2__tar.gz → 4.2.1__tar.gz

{surfdataverse-4.1.2 → surfdataverse-4.2.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: surfdataverse
-Version: 4.1.2
+Version: 4.2.1
 Summary: A Python package for ionysis Microsoft Dataverse integration
 Keywords: dataverse,microsoft,crm,api
 Author: ionysis
@@ -14,6 +14,7 @@ Classifier: Programming Language :: Python :: 3.8
 Classifier: Programming Language :: Python :: 3.9
 Classifier: Programming Language :: Python :: 3.10
 Classifier: Programming Language :: Python :: 3.11
+Requires-Dist: azure-identity>=1.25.3
 Requires-Dist: dependency-injector>=4.42.0
 Requires-Dist: msal>=1.33.0
 Requires-Dist: numpy>=2.3.3

{surfdataverse-4.1.2 → surfdataverse-4.2.1}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "uv_build"
 
 [project]
 name = "surfdataverse"
-version = "4.1.2"
+version = "4.2.1"
 description = "A Python package for ionysis Microsoft Dataverse integration"
 readme = "README.md"
 authors = [
@@ -24,6 +24,7 @@ classifiers = [
 keywords = ["dataverse", "microsoft", "crm", "api"]
 requires-python = ">=3.11"
 dependencies = [
+    "azure-identity>=1.25.3",
     "dependency-injector>=4.42.0",
     "msal>=1.33.0",
     "numpy>=2.3.3",
@@ -50,6 +51,7 @@ typing-test = [
 ]
 dev = [
     "ruff>=0.0.241",
+    "ty>=0.0.35",
     "uv-build>=0.8.18",
 ]
 

{surfdataverse-4.1.2 → surfdataverse-4.2.1}/src/surfdataverse/core.py

@@ -1,16 +1,16 @@
-from collections import defaultdict
 import io
 import json
 import logging
 import sys
+import time
 import uuid
 from decimal import Decimal
 from pathlib import Path
-from typing import Optional
 
-import msal  # type: ignore
+import msal
 import pandas as pd
 import requests
+from azure.core.credentials import AccessToken
 
 from .exceptions import (
     AuthenticationError,
@@ -24,10 +24,10 @@ from .exceptions import (
 class DecimalEncoder(json.JSONEncoder):
     """Custom JSON encoder to handle Decimal objects"""
 
-    def default(self, obj):
-        if isinstance(obj, Decimal):
-            return float(obj)
-        return super().default(obj)
+    def default(self, o):
+        if isinstance(o, Decimal):
+            return float(o)
+        return super().default(o)
 
 
 logger = logging.getLogger(__name__)
@@ -61,24 +61,69 @@ def is_valid_guid(value: str) -> bool:
         return False
 
 
+def get_token_cache():
+    """Returns a singleton token cache to prevent cache conflicts between client instances."""
+    global _global_token_cache, _global_cache_file
+
+    # Return existing singleton if already initialized
+    if _global_token_cache is not None:
+        logger.debug(f"Returning existing token cache: {_global_cache_file}")
+        return _global_token_cache, _global_cache_file
+
+    # Initialize singleton cache
+    logger.info("Initializing singleton token cache...")
+
+    # Create app-specific config directory
+    app_config_dir = user_config_dir("SurfDataverse", "ionysis")
+    app_config_dir.mkdir(parents=True, exist_ok=True)
+
+    # Token cache file path
+    _global_cache_file = app_config_dir / "token_cache.bin"
+    _global_token_cache = msal.SerializableTokenCache()
+
+    if _global_cache_file.exists():
+        logger.info(f"Found existing cache file: {_global_cache_file}")
+        try:
+            with open(_global_cache_file, "r") as f:
+                cache_content = f.read()
+                logger.info(f"Cache file size: {len(cache_content)} bytes")
+                if not cache_content.strip():
+                    logger.warning("Cache file is empty, removing it")
+                    _global_cache_file.unlink()
+                else:
+                    _global_token_cache.deserialize(cache_content)
+        except (json.JSONDecodeError, ValueError) as e:
+            logger.warning(f"Corrupted cache file detected: {e}. Removing and starting fresh.")
+            _global_cache_file.unlink()
+        except Exception as e:
+            logger.warning(f"Error reading cache file: {e}. Removing and starting fresh.")
+            _global_cache_file.unlink()
+
+    logger.info(f"Using singleton token cache: {_global_cache_file}")
+    return _global_token_cache, _global_cache_file
+
+
 class DataverseClient:
-    def __init__(self, config_path=None):
+    def __init__(self, config_path: str | None = None):
         self.connected = False
         self.session = requests.Session()
         self.environment_uri: str = ""
         self.app = None  # Store MSAL application instance
         self.token_cache = None  # Store the token cache
         self.account = None  # Store the logged-in account
-        self.config_path = config_path
+        self.config_path: str | None = config_path
 
     # === CONNECTION, AUTHENTICATION, TEST
-    def get_authenticated_session(self, config_json: Optional[str | Path] = None):
+    def get_authenticated_session(self, config_json: str | Path | None = None):
         if config_json is None and self.config_path is None:
             raise ConfigurationError(
                 "No configuration file provided. Please specify config_json or set config_path during initialization."
             )
 
         config_file = config_json or self.config_path
+        if config_file is None:
+            raise ConfigurationError("Configuration file path is not set.")
+
         logger.info(f"Loading config from: {config_file}")
         config = json.load(open(config_file))
         logger.info("Config loaded successfully")
@@ -88,7 +133,7 @@ class DataverseClient:
         scope = [self.environment_uri + config["scopeSuffix"]]
 
         logger.info("Getting token cache...")
-        self.token_cache, cache_file = self.get_token_cache()
+        self.token_cache, cache_file = get_token_cache()
         logger.info(f"Token cache initialized, cache file: {cache_file}")
         self.app = msal.PublicClientApplication(
             config["clientID"], authority=authority, token_cache=self.token_cache
@@ -138,48 +183,6 @@ class DataverseClient:
             logger.error(f"Description: {result.get('error_description', 'No description')}")
             raise AuthenticationError(error_msg)
 
-    @staticmethod
-    def get_token_cache():
-        """Returns a singleton token cache to prevent cache conflicts between client instances."""
-        global _global_token_cache, _global_cache_file
-
-        # Return existing singleton if already initialized
-        if _global_token_cache is not None:
-            logger.debug(f"Returning existing token cache: {_global_cache_file}")
-            return _global_token_cache, _global_cache_file
-
-        # Initialize singleton cache
-        logger.info("Initializing singleton token cache...")
-
-        # Create app-specific config directory
-        app_config_dir = user_config_dir("SurfDataverse", "ionysis")
-        app_config_dir.mkdir(parents=True, exist_ok=True)
-
-        # Token cache file path
-        _global_cache_file = app_config_dir / "token_cache.bin"
-        _global_token_cache = msal.SerializableTokenCache()
-
-        if _global_cache_file.exists():
-            logger.info(f"Found existing cache file: {_global_cache_file}")
-            try:
-                with open(_global_cache_file, "r") as f:
-                    cache_content = f.read()
-                    logger.info(f"Cache file size: {len(cache_content)} bytes")
-                    if not cache_content.strip():
-                        logger.warning("Cache file is empty, removing it")
-                        _global_cache_file.unlink()
-                    else:
-                        _global_token_cache.deserialize(cache_content)
-            except (json.JSONDecodeError, ValueError) as e:
-                logger.warning(f"Corrupted cache file detected: {e}. Removing and starting fresh.")
-                _global_cache_file.unlink()
-            except Exception as e:
-                logger.warning(f"Error reading cache file: {e}. Removing and starting fresh.")
-                _global_cache_file.unlink()
-
-        logger.info(f"Using singleton token cache: {_global_cache_file}")
-        return _global_token_cache, _global_cache_file
-
     def test_connection(self):
         request_uri = f"{self.environment_uri}api/data/v9.2/"
         try:
@@ -196,6 +199,35 @@ class DataverseClient:
             raise ConnectionError(f"Network error during connection test: {str(e)}")
 
 
+class MsalStorageCredential:
+    """Azure TokenCredential backed by MSAL.
+
+    If the user already authenticated for Dataverse in this session, acquiring a
+    storage token is silent (no browser popup). Otherwise the browser opens once.
+    """
+
+    def __init__(self, client_id: str, authority: str):
+        token_cache, _ = get_token_cache()
+        self._app = msal.PublicClientApplication(
+            client_id, authority=authority, token_cache=token_cache
+        )
+
+    def get_token(self, *scopes, **kwargs) -> AccessToken:
+        scope = list(scopes) if scopes else [kwargs.get("scope")]
+        result = None
+        accounts = self._app.get_accounts()
+        if accounts:
+            result = self._app.acquire_token_silent(scopes=scope, account=accounts[0])
+        if not result:
+            result = self._app.acquire_token_interactive(scope)
+        if "access_token" not in result:
+            raise RuntimeError(
+                f"Blob storage authentication failed: {result.get('error_description', result.get('error'))}"
+            )
+        expires_on = int(time.time()) + result.get("expires_in", 3600)
+        return AccessToken(result["access_token"], expires_on)
+
+
 class DataverseBase:
     """Base class for Dataverse operations with shared functionality"""
 
@@ -495,7 +527,9 @@ class DataverseTable(DataverseBase):
             logger.error(error_msg)
             raise DataverseAPIError(error_msg, response.status_code, response.text)
 
-    def get_column_data(self, columns, entity_set_name=None, polars=False, infer_schema_length=None):
+    def get_column_data(
+        self, columns, entity_set_name=None, polars=False, infer_schema_length=None
+    ):
         """Fetches specific column data from this table
         Args:
             columns: Column name (string) or list of column names to retrieve