fix: remove some requirements

.gitignore

@@ -1 +1,3 @@
-.env
+.env
+.DS_Store
+output/*

extract_single.py

@@ -263,7 +263,7 @@ TEXTE DE LA PAGE {page_num}:
             logger.warning(f"⚠️  Page {page_num} is empty")
             return []
 
-        max_chars = 8000 # todo
+        max_chars = 8000
         user_prompt = user_prompt_template.format(
             page_num=page_num,
             text=page_text[:max_chars]
@@ -278,8 +278,8 @@ TEXTE DE LA PAGE {page_num}:
                     {"role": "system", "content": system_prompt},
                     {"role": "user", "content": user_prompt}
                 ],
-                temperature=0.1,
+                temperature=1.0,
-                max_tokens=4000, # todo
+                max_tokens=4000,
                 response_format={"type": "json_object"}
             )
 

requirements.txt

@@ -1,36 +1,16 @@
-# Azure services
+# Azure services (required)
 azure-ai-formrecognizer==3.3.0
 azure-core>=1.29.0
 openai>=1.0.0
 
-# PDF processing
+# Environment variables
-pdf2image==1.17.0
+python-dotenv==1.0.0
-Pillow>=10.0.0
+
-pytesseract==0.3.13  # Fallback OCR
+# PDF processing (required for page counting)
-pdfplumber==0.10.3  # Enhanced table extraction
 PyPDF2>=3.0.1
 
-# CLI and utilities
+# CLI interface
 click==8.1.7
-pathlib2==2.3.7.post1
 
-# Data handling
+# Python built-in modules extensions (if needed for older Python versions)
-numpy>=1.24.0
+typing-extensions>=4.0.0  # Only if Python < 3.8
-python-dateutil>=2.8.2
-
-# For better logging
-colorlog>=6.7.0
-
-# Development tools (optional)
-pytest>=7.4.0
-black>=23.0.0
-flake8>=6.0.0
-
-# Type hints
-typing-extensions>=4.0.0
-dataclasses>=0.6
-
-# For text processing
-nltk>=3.8.1
-textstat>=0.7.3
-python-dotenv==1.0.0