Fix Indian Bank extraction and add shared multi-format dates for all banks

app/modules/bank_statement_analyzer/parsers/hdfc.py

@@ -2,7 +2,7 @@ from __future__ import annotations
 import re, pandas as pd
 from pathlib import Path
 from .base import *
-from .common import find
+from .common import DATE_TOKEN_PATTERN, DATE_TOKEN_RE, date_iso, find
 
 class HDFCParser(BaseParser):
     bank_name='HDFC Bank'; parser_name='HDFCParser'
@@ -14,15 +14,15 @@ class HDFCParser(BaseParser):
         meta.account_number=find(r'Account No\s*:?\s*([0-9X*]+)',text)
         meta.customer_id=find(r'Cust ID\s*:?\s*([0-9X*]+)',text)
         meta.ifsc=find(r'(?:RTGS/\s*NEFT IFSC|IFSC)\s*:?\s*([A-Z0-9]+)',text)
-        m=re.search(r'From\s*:\s*(\d{2}/\d{2}/\d{4})\s+To\s*:\s*(\d{2}/\d{2}/\d{4})',text,re.I)
+        m=re.search(rf'From\s*:\s*({DATE_TOKEN_PATTERN})\s+To\s*:\s*({DATE_TOKEN_PATTERN})', text, re.I)
         if m:
-            meta.period_from=pd.to_datetime(m.group(1),dayfirst=True).strftime('%Y-%m-%d'); meta.period_to=pd.to_datetime(m.group(2),dayfirst=True).strftime('%Y-%m-%d')
+            meta.period_from=date_iso(m.group(1)); meta.period_to=date_iso(m.group(2))
         # first non-empty line before address usually contains customer name; allow manual override in UI
         m=re.search(r'\n\s{2,}([^\n:]{3,60})\s*\n.*?Address\s*:',text,re.S|re.I)
         if m: meta.customer_name=norm(m.group(1).splitlines()[-1])
         lines=text.splitlines(); rows=[]; cur=None; page=1
         wd_pos,dep_pos,bal_pos=130,165,190
-        date_re=re.compile(r'^\s*(\d{2}/\d{2}/\d{2})\s+(.*)$')
+        date_re=re.compile(rf'^\s*({DATE_TOKEN_PATTERN})\s+(.*)$', re.I)
         for line in lines:
             if '\f' in line: page += line.count('\f')
             if 'Withdrawal Amt.' in line and 'Deposit Amt.' in line:
@@ -37,7 +37,7 @@ class HDFCParser(BaseParser):
         for r in rows:
             first=r['raw_lines'][0]; body=' '.join(x.strip() for x in r['raw_lines'])
             # identify value date and ref on first line
-            dts=list(re.finditer(r'\d{2}/\d{2}/\d{2}',first))
+            dts=list(DATE_TOKEN_RE.finditer(first))
             if len(dts)>1: r['value_date']=dts[-1].group()
             else: r['value_date']=r['transaction_date']
             nums=list(re.finditer(r'(?<!\d)(?:\d{1,3}(?:,\d{3})+|\d+)\.\d{2}(?!\d)',first))