first version of gmail invoice extractor

2025-09-16 19:36:05 +02:00
parent a71b3adb40
commit 1df43ae405
13 changed files with 436 additions and 513 deletions
@@ -0,0 +1,39 @@
+import yaml
+import os
+from typing import List
+
+class Config:
+    def __init__(self, config_file: str = "config.yaml"):
+        self.config_file = config_file
+        self.load_config()
+
+    def load_config(self):
+        """Load configuration from YAML file."""
+        if os.path.exists(self.config_file):
+            with open(self.config_file, 'r') as f:
+                config = yaml.safe_load(f)
+        else:
+            # Default configuration
+            config = {
+                'vat_patterns': [
+                    'BE0?\\d{9}',
+                    'NL\\d{9}B\\d{2}',
+                    'DE\\d{9}',
+                    'FR[A-Z0-9]{2}\\d{9}',
+                    'GB\\d{9,12}'
+                ],
+                'gmail_query': 'has:attachment filename:pdf',
+                'max_messages': 100,
+                'save_folder': 'invoices',
+                'log_level': 'INFO'
+            }
+
+        self.vat_patterns = config.get('vat_patterns', [])
+        self.gmail_query = config.get('gmail_query', 'has:attachment filename:pdf')
+        self.max_messages = config.get('max_messages', 100)
+        self.save_folder = config.get('save_folder', 'invoices')
+        self.log_level = config.get('log_level', 'INFO')
+
+    def get_vat_patterns(self) -> List[str]:
+        """Get VAT number patterns as compiled regex patterns."""
+        return self.vat_patterns