Merge pull request #13 from harmonydata/fix-env-google

Fix loading env GOOGLE_APPLICATION_CREDENTIALS
harmonydata · Dec 15, 2024 · 5407710 · 5407710
2 parents 6fd3e88 + c1c2d9a
commit 5407710
Show file tree

Hide file tree

Showing 3 changed files with 48 additions and 38 deletions.
diff --git a/harmony_api/core/settings.py b/harmony_api/core/settings.py
@@ -24,60 +24,57 @@
 SOFTWARE.
 """
 
-import json
 import os
 from typing import Union
 
+from pydantic import Field
 from pydantic_settings import BaseSettings
 
-GOOGLE_APPLICATION_CREDENTIALS = os.getenv("GOOGLE_APPLICATION_CREDENTIALS", '{}')
-if GOOGLE_APPLICATION_CREDENTIALS:
-    if GOOGLE_APPLICATION_CREDENTIALS.startswith(
-            "{"):  # only load JSON if it's JSON format
-        GOOGLE_APPLICATION_CREDENTIALS = json.loads(GOOGLE_APPLICATION_CREDENTIALS)
-
 
 class Settings(BaseSettings):
     # General harmony_api config
-    VERSION: str = "2.0"
-    APP_TITLE: str = "Harmony API"
-    TIKA_ENDPOINT: str = os.getenv("TIKA_ENDPOINT", "http://tika:9998")
-    OPENAI_API_KEY: str | None = os.getenv("OPENAI_API_KEY")
-    AZURE_OPENAI_API_KEY: str | None = os.getenv("AZURE_OPENAI_API_KEY")
-    AZURE_OPENAI_ENDPOINT: str | None = os.getenv("AZURE_OPENAI_ENDPOINT")
-    AZURE_STORAGE_URL: str | None = os.getenv("AZURE_STORAGE_URL")
-    GOOGLE_APPLICATION_CREDENTIALS: dict = GOOGLE_APPLICATION_CREDENTIALS
+    VERSION: str = Field(description="Application version.", default="2.0")
+    APP_TITLE: str = Field(description="Application title.", default="Harmony API")
+    TIKA_ENDPOINT: str = Field(description="Tika endpoint.", default="http://tika:9998")
+    OPENAI_API_KEY: str | None = Field(description="OpenAI API key.", default=None)
+    AZURE_OPENAI_API_KEY: str | None = Field(description="Azure OpenAI API key.", default=None)
+    AZURE_OPENAI_ENDPOINT: str | None = Field(description="Azure OpenAI endpoint.", default=None)
+    AZURE_STORAGE_URL: str | None = Field(description="Azure Storage URL.", default=None)
+    GOOGLE_APPLICATION_CREDENTIALS: str | None = Field(
+        description="A JSON string is expected here, this is the content of credentials.json.",
+        default=None
+    )
 
 
 class DevSettings(Settings):
-    SERVER_HOST: str = "0.0.0.0"
-    DEBUG: bool = True
-    PORT: int = 8000
-    RELOAD: bool = True
-    CORS: dict = {
+    SERVER_HOST: str = Field(description="Host.", default="0.0.0.0")
+    DEBUG: bool = Field(description="Debug.", default=True)
+    PORT: int = Field(description="Port.", default=8000)
+    RELOAD: bool = Field(description="Reload.", default=True)
+    CORS: dict = Field(description="CORS.", default={
         "origins": [
             "*",
         ],
         "allow_credentials": True,
         "allow_methods": ["*"],
         "allow_headers": ["*"],
-    }
+    })
 
 
 class ProdSettings(Settings):
     # TODO change
-    SERVER_HOST: str = "0.0.0.0"
-    DEBUG: bool = False
-    PORT: int = 8000
-    RELOAD: bool = False
-    CORS: dict = {
+    SERVER_HOST: str = Field(description="Host.", default="0.0.0.0")
+    DEBUG: bool = Field(description="Debug.", default=False)
+    PORT: int = Field(description="Port.", default=8000)
+    RELOAD: bool = Field(description="Reload.", default=False)
+    CORS: dict = Field(description="CORS.", default={
         "origins": [
             "*",
         ],
         "allow_credentials": True,
         "allow_methods": ["*"],
         "allow_headers": ["*"],
-    }
+    })
 
 
 def get_settings() -> Union[DevSettings | ProdSettings]:

diff --git a/harmony_api/helpers.py b/harmony_api/helpers.py
@@ -419,7 +419,7 @@ def check_model_availability(model: dict) -> bool:
 
     # Google
     elif model["framework"] == "google":
-        if not settings.GOOGLE_APPLICATION_CREDENTIALS:
+        if not google_embeddings.GOOGLE_APPLICATION_CREDENTIALS_DICT:
             return False
 
         # Check model

diff --git a/harmony_api/services/google_embeddings.py b/harmony_api/services/google_embeddings.py
@@ -1,3 +1,4 @@
+import json
 import numpy as np
 import vertexai
 from google.api_core.exceptions import NotFound
@@ -14,20 +15,32 @@
 
 settings = get_settings()
 
-# Authenticate Vertex AI with Google service account
+# Load GOOGLE_APPLICATION_CREDENTIALS as a dictionary
+GOOGLE_APPLICATION_CREDENTIALS_DICT: dict | None = None
 if settings.GOOGLE_APPLICATION_CREDENTIALS:
-    credentials = Credentials.from_service_account_info(
-        settings.GOOGLE_APPLICATION_CREDENTIALS,
-        scopes=["https://www.googleapis.com/auth/cloud-platform"],
-    )
-    vertexai.init(
-        project=settings.GOOGLE_APPLICATION_CREDENTIALS["project_id"],
-        credentials=credentials,
-    )
+    try:
+        GOOGLE_APPLICATION_CREDENTIALS_DICT: dict = json.loads(settings.GOOGLE_APPLICATION_CREDENTIALS)
+    except (Exception,) as e:
+        print(f"The env GOOGLE_APPLICATION_CREDENTIALS is not a valid JSON, Google models will be unavailable. Error: {str(e)}")
+
+# Authenticate Vertex AI with Google service account
+if GOOGLE_APPLICATION_CREDENTIALS_DICT:
+    try:
+        credentials = Credentials.from_service_account_info(
+            GOOGLE_APPLICATION_CREDENTIALS_DICT,
+            scopes=["https://www.googleapis.com/auth/cloud-platform"],
+        )
+        vertexai.init(
+            project=GOOGLE_APPLICATION_CREDENTIALS_DICT["project_id"],
+            credentials=credentials,
+        )
+    except (Exception,) as e:
+        GOOGLE_APPLICATION_CREDENTIALS_DICT = None
+        print(f"Error loading Google credentials: {str(e)}")
 
 # Check available models
 HARMONY_API_AVAILABLE_GOOGLE_MODELS_LIST: List[str] = []
-if settings.GOOGLE_APPLICATION_CREDENTIALS:
+if GOOGLE_APPLICATION_CREDENTIALS_DICT:
     print("INFO:\t  Checking Google models...")
     for harmony_api_google_model in HARMONY_API_GOOGLE_MODELS_LIST:
         try: