Add robust dependency handling with fallbacks for missing packages

structured_ocr.py

@@ -6,14 +6,29 @@ from enum import Enum
 from pathlib import Path
 import json
 import base64
-import pycountry
 import logging
 from functools import lru_cache
 from typing import Optional, Dict, Any, List, Union, Tuple
+
+# Try to import pycountry, provide fallback if not available
+try:
+    import pycountry
+    PYCOUNTRY_AVAILABLE = True
+except ImportError:
+    PYCOUNTRY_AVAILABLE = False
+    logging.warning("pycountry module not available - using language code fallback")
+
 from pydantic import BaseModel
-from mistralai import Mistral
-from mistralai import DocumentURLChunk, ImageURLChunk, TextChunk
-from mistralai.models import OCRImageObject
+
+# Try to import Mistral AI, provide fallback if not available
+try:
+    from mistralai import Mistral
+    from mistralai import DocumentURLChunk, ImageURLChunk, TextChunk
+    from mistralai.models import OCRImageObject
+    MISTRAL_AVAILABLE = True
+except ImportError:
+    MISTRAL_AVAILABLE = False
+    logging.warning("mistralai module not available - OCR functionality will be limited")
 
 # Configure logging
 logging.basicConfig(level=logging.INFO, 
@@ -41,7 +56,17 @@ except ImportError:
         return "\n\n".join(markdowns)
 
 # Import config directly (now local to historical-ocr)
-from config import MISTRAL_API_KEY, OCR_MODEL, TEXT_MODEL, VISION_MODEL, TEST_MODE
+try:
+    from config import MISTRAL_API_KEY, OCR_MODEL, TEXT_MODEL, VISION_MODEL, TEST_MODE
+except ImportError:
+    # Fallback defaults if config is not available
+    import os
+    MISTRAL_API_KEY = os.environ.get("MISTRAL_API_KEY", "")
+    OCR_MODEL = "mistral-ocr-latest"
+    TEXT_MODEL = "mistral-large-latest"
+    VISION_MODEL = "mistral-large-latest"
+    TEST_MODE = True
+    logging.warning("Config module not found. Using environment variables and defaults.")
 
 # Helper function to make OCR objects JSON serializable
 # Removed caching to fix unhashable type error
@@ -89,7 +114,24 @@ def serialize_ocr_response(obj):
 # Create language enum for structured output - cache language lookup to avoid repeated processing
 @lru_cache(maxsize=1)
 def get_language_dict():
-    return {lang.alpha_2: lang.name for lang in pycountry.languages if hasattr(lang, 'alpha_2')}
+    if PYCOUNTRY_AVAILABLE:
+        return {lang.alpha_2: lang.name for lang in pycountry.languages if hasattr(lang, 'alpha_2')}
+    else:
+        # Fallback with basic languages when pycountry is not available
+        return {
+            "en": "English",
+            "es": "Spanish",
+            "fr": "French",
+            "de": "German",
+            "it": "Italian",
+            "pt": "Portuguese",
+            "ru": "Russian",
+            "zh": "Chinese",
+            "ja": "Japanese",
+            "ar": "Arabic",
+            "hi": "Hindi",
+            "la": "Latin"
+        }
 
 class LanguageMeta(Enum.__class__):
     def __new__(metacls, cls, bases, classdict):
@@ -99,7 +141,12 @@ class LanguageMeta(Enum.__class__):
         return super().__new__(metacls, cls, bases, classdict)
 
 class Language(Enum, metaclass=LanguageMeta):
-    pass
+    # Fallback languages in case enum creation fails
+    ENGLISH = "English"
+    SPANISH = "Spanish"
+    FRENCH = "French"
+    GERMAN = "German"
+    LATIN = "Latin"
 
 class StructuredOCRModel(BaseModel):
     file_name: str
@@ -110,8 +157,15 @@ class StructuredOCRModel(BaseModel):
 class StructuredOCR:
     def __init__(self, api_key=None):
         """Initialize the OCR processor with API key"""
-        # Check if we're running in test mode
-        self.test_mode = TEST_MODE
+        # Check if we're running in test mode or if Mistral is not available
+        self.test_mode = TEST_MODE or not MISTRAL_AVAILABLE
+        
+        if not MISTRAL_AVAILABLE:
+            logger = logging.getLogger("api_validator")
+            logger.warning("Mistral AI package not available - running in test mode")
+            self.api_key = "placeholder_key"
+            self.client = None
+            return
         
         # Initialize API key - use provided key, or environment var
         if self.test_mode and not api_key:
@@ -140,7 +194,10 @@ class StructuredOCR:
             if "unauthorized" in error_msg or "401" in error_msg:
                 raise ValueError(f"API key authentication failed. Please check your Mistral API key: {str(e)}")
             else:
-                raise
+                logger = logging.getLogger("api_validator")
+                logger.warning(f"Failed to initialize Mistral client: {str(e)}")
+                self.test_mode = True
+                self.client = None
     
     def process_file(self, file_path, file_type=None, use_vision=True, max_pages=None, file_size_mb=None, custom_pages=None, custom_prompt=None):
         """Process a file and return structured OCR results