From 364798ca5281e7ed26bd415ba3a207bad2fc0ffb Mon Sep 17 00:00:00 2001
From: Zuzanna Rachuba <zuziarachuba@gmail.com>
Date: Sun, 3 Nov 2024 12:15:47 +0100
Subject: [PATCH] Added regex for postal-codes

---
 projekt.py | 20 ++++++++------------
 1 file changed, 8 insertions(+), 12 deletions(-)

diff --git a/projekt.py b/projekt.py
index 2d9ce13..795d1e3 100644
--- a/projekt.py
+++ b/projekt.py
@@ -2,9 +2,7 @@ import fitz
 import re 
 
 my_path = r"C:\Users\DELL\Downloads\A-24VU-00511.PDF.pdf"
-
-postal_code_pattern = r'\b(?!0)([1-9]\d?-\d{3}|[1-9]\d{4}|[1-9]\d? \d{3})\b'
-
+postal_code_pattern = r"(?<!\S)(?:(?:[A-Za-z]{3,}\s+)(\d{5}|\d{2}-\d{3})|(\d{5}|\d{2}-\d{3})(?:\s+[A-Za-z]{3,}))(?!\S)"
 with fitz.open(my_path) as doc:
     order_numbers = []
     order_dates = []
@@ -33,15 +31,13 @@ with fitz.open(my_path) as doc:
                 line_2 = lines[i-2] if i>0 else ''
                 line_above = lines[i - 1] if i > 0 else ''
                 line_below = lines[i + 1] if i + 1 < len(lines) else ''
-                
-         
-                if not re.search(r'[^0-9a-zA-Z \-]', line):
-                    buyers.append({
-                        'line_2': line_2,
-                        'line_above': line_above,
-                        'postal_code_line': line,
-                        'line_below': line_below
-                    })
+            
+                buyers.append({
+                    'line_2': line_2,
+                    'line_above': line_above,
+                    'postal_code_line': line,
+                    'line_below': line_below
+                })
 
 print("Reference numbers:", order_numbers)
 print("Document dates:", order_dates)