refactor extraction date from string

helperfunctions.py

@@ -1,57 +1,74 @@
 from flask import session
-from datetime import datetime
 import re
 import os
 import sqlite3
 from datetime import datetime, timedelta
+from typing import Optional
 
 log_db = sqlite3.connect("access_log.db", check_same_thread=False)
 
-def extract_date_from_string(string_with_date):
-    # grab X.Y.Z where X,Y,Z are 1–4 digits
-    m = re.search(r'(\d{1,4}\.\d{1,2}\.\d{1,4})', string_with_date)
-    if not m:
-        return None
 
-    date_str = m.group(1)
-    parts = date_str.split('.')
-    
-    # 1) Unambiguous “last group = YYYY”
-    if len(parts) == 3 and len(parts[2]) == 4:
-        fmt = '%d.%m.%Y'
 
-    # 2) Unambiguous “first group = YYYY”
-    elif len(parts) == 3 and len(parts[0]) == 4:
-        fmt = '%Y.%m.%d'
+# Precompiled regex to find date-like patterns: either dotted X.Y.Z or ISO dashed YYYY-MM-DD
+_DATE_REGEX = re.compile(
+    r"("  # start group
+    r"\d{4}-\d{1,2}-\d{1,2}"          # ISO: YYYY-M-D or YYYY-MM-DD
+    r"|"                              # or
+    r"\d{1,4}\.\d{1,2}\.\d{1,4}"      # dotted: X.Y.Z, where each is 1–4 digits (year may be 1–4)
+    r")"
+)
 
-    # 3) Ambiguous “XX.XX.XX” → prefer DD.MM.YY, fallback to YY.MM.DD
-    elif len(parts) == 3 and all(len(p) == 2 for p in parts):
-        # try last-group-as-year first
-        try:
-            dt = datetime.strptime(date_str, '%d.%m.%y')
-            return dt.strftime('%Y-%m-%d')
-        except ValueError:
-            # fallback to first-group-as-year
-            fmt = '%y.%m.%d'
 
-    else:
-        # optional: handle ISO with dashes
-        if '-' in date_str:
-            try:
-                dt = datetime.strptime(date_str, '%Y-%m-%d')
-                return dt.strftime('%Y-%m-%d')
-            except ValueError:
-                return None
-        return None
-
-    # parse with whichever fmt we settled on
+def _try_parse(date_str: str, fmt: str) -> Optional[datetime]:
+    """Try to parse date_str with fmt, return datetime or None."""
     try:
-        dt = datetime.strptime(date_str, fmt)
-        return dt.strftime('%Y-%m-%d')
+        return datetime.strptime(date_str, fmt)
     except ValueError:
         return None
 
 
+def extract_date_from_string(text: str) -> Optional[str]:
+    """
+    Extract the first date-like substring from text and return it in ISO format (YYYY-MM-DD).
+    Supports:
+    - ISO-style dates with dashes (YYYY-M-D or YYYY-MM-DD)
+    - Dotted dates (DD.MM.YYYY, YYYY.MM.DD, DD.MM.YY, YY.MM.DD)
+    """
+    match = _DATE_REGEX.search(text)
+    if not match:
+        return None
+
+    date_str = match.group(1)
+
+    # 1) ISO dashed format takes priority
+    if '-' in date_str:
+        dt = _try_parse(date_str, '%Y-%m-%d')
+        return dt.strftime('%Y-%m-%d') if dt else None
+
+    # 2) Dotted formats
+    parts = date_str.split('.')
+    candidates = []
+
+    # Unambiguous: last part 4 digits → DD.MM.YYYY
+    if len(parts) == 3 and len(parts[2]) == 4:
+        candidates.append('%d.%m.%Y')
+    # Unambiguous: first part 4 digits → YYYY.MM.DD
+    if len(parts) == 3 and len(parts[0]) == 4:
+        candidates.append('%Y.%m.%d')
+    # Ambiguous two-digit groups: try DD.MM.YY, then YY.MM.DD
+    if len(parts) == 3 and all(len(p) == 2 for p in parts):
+        candidates.extend(['%d.%m.%y', '%y.%m.%d'])
+
+    # Try each candidate
+    for fmt in candidates:
+        dt = _try_parse(date_str, fmt)
+        if dt:
+            return dt.strftime('%Y-%m-%d')
+
+    # no valid parse
+    return None
+
+
 def extract_structure_from_string(input_string):
     # extract category and titel from filename
     filepathname_ext = os.path.splitext(input_string)[0] # remove file extension