Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
@@ -63,7 +63,121 @@ def extract_expiry_dates(text):
|
|
63 |
r'(\d{1,2}\s*[A-Za-z]{3,}\s*\d{4})', # 20 MAY 2024
|
64 |
r'([A-Za-z]{3,}\s*\d{1,2}[,\s]*\d{4})', # July 20, 2024
|
65 |
r'(\d{4}[\/\-]\d{1,2}[\/\-]\d{1,2})', # 2024/07/20 or 2024-07-20
|
66 |
-
r'([A-Za-z]{3}[\-]\d{1,2}[\-]\d{4})'
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
67 |
]
|
68 |
dates = []
|
69 |
for pattern in patterns:
|
|
|
63 |
r'(\d{1,2}\s*[A-Za-z]{3,}\s*\d{4})', # 20 MAY 2024
|
64 |
r'([A-Za-z]{3,}\s*\d{1,2}[,\s]*\d{4})', # July 20, 2024
|
65 |
r'(\d{4}[\/\-]\d{1,2}[\/\-]\d{1,2})', # 2024/07/20 or 2024-07-20
|
66 |
+
r'([A-Za-z]{3}[\-]\d{1,2}[\-]\d{4})',
|
67 |
+
r'(?:exp(?:iry)?\.?\s*date\s*[:\-]?\s*.*?(\d{2}[\/\-]\d{2}[\/\-][0O]\d{2}))', # Expiry Date: 20/07/2O24
|
68 |
+
r'(?:exp(?:iry)?\.?\s*date\s*[:\-]?\s*.*?(\d{2}[\/\-]\d{2}[\/\-]\d{4}))', # Expiry Date: 20/07/2024
|
69 |
+
r'(?:exp(?:iry)?\.?\s*date\s*[:\-]?\s*.*?(\d{2}[\/\-]\d{2}[\/\-][0O]\d{2}))', # Expiry Date: 20/07/2O24
|
70 |
+
r'(?:exp(?:iry)?\.?\s*date\s*[:\-]?\s*.?(\d{2}\s[A-Za-z]{3,}\s*[0O]\d{2}))', # Expiry Date: 20 MAY 2O24
|
71 |
+
r'(?:exp(?:iry)?\.?\s*date\s*[:\-]?\s*.?(\d{2}\s[A-Za-z]{3,}\s*\d{4}))', # Expiry Date: 20 MAY 2024
|
72 |
+
r'(?:exp(?:iry)?\.?\s*date\s*[:\-]?\s*.?(\d{2}\s[A-Za-z]{3,}\s*[0O]\d{2}))', # Expiry Date: 20 MAY 2O24
|
73 |
+
r'(?:exp(?:iry)?\.?\s*date\s*[:\-]?\s*.*?(\d{4}[\/\-]\d{2}[\/\-][0O]\d{2}))', # Expiry Date: 2024/07/2O24
|
74 |
+
r'(?:exp(?:iry)?\.?\s*date\s*[:\-]?\s*.*?(\d{4}[\/\-]\d{2}[\/\-]\d{2}))', # Expiry Date: 2024/07/20
|
75 |
+
r'(?:best\s*before\s*[:\-]?\s*.*?(\d{4}))', # Best Before: 2025
|
76 |
+
r'(?:best\s*before\s*[:\-]?\s*.*?(\d{2}[\/\-]\d{2}[\/\-][0O]\d{2}))', # Best Before: 20/07/2O24
|
77 |
+
r'(?:best\s*before\s*[:\-]?\s*.*?(\d{2}[\/\-]\d{2}[\/\-]\d{4}))', # Best Before: 20/07/2024
|
78 |
+
r'(?:best\s*before\s*[:\-]?\s*.*?(\d{2}[\/\-]\d{2}[\/\-][0O]\d{2}))', # Best Before: 20/07/2O24
|
79 |
+
r'(?:best\s*before\s*[:\-]?\s*.?(\d{2}\s[A-Za-z]{3,}\s*[0O]\d{2}))', # Best Before: 20 MAY 2O24
|
80 |
+
r'(?:best\s*before\s*[:\-]?\s*.?(\d{2}\s[A-Za-z]{3,}\s*\d{4}))', # Best Before: 20 MAY 2024
|
81 |
+
r'(?:best\s*before\s*[:\-]?\s*.?(\d{2}\s[A-Za-z]{3,}\s*[0O]\d{2}))', # Best Before: 20 MAY 2O24
|
82 |
+
r'(?:best\s*before\s*[:\-]?\s*.*?(\d{4}[\/\-]\d{2}[\/\-][0O]\d{2}))', # Best Before: 2024/07/2O24
|
83 |
+
r'(?:best\s*before\s*[:\-]?\s*.*?(\d{4}[\/\-]\d{2}[\/\-]\d{2}))', # Best Before: 2024/07/20
|
84 |
+
r'(?:best\s*before\s*[:\-]?\s*.*?(\d{1,2}\d{2}\d{2}))',
|
85 |
+
r'(?:best\s*before\s*[:\-]?\s*(\d{6}))',
|
86 |
+
r'(?:consume\s*before\s*[:\-]?\s*.*?(\d{1,2}[A-Za-z]{3,}[0O]\d{2}))', # Consume Before: 3ODEC2O24
|
87 |
+
r'(?:consume\s*before\s*[:\-]?\s*.*?(\d{1,2}[A-Za-z]{3,}\d{2}))', # Consume Before: 30DEC23
|
88 |
+
r'(?:consume\s*before\s*[:\-]?\s*.*?(\d{2}[\/\-]\d{2}[\/\-][0O]\d{2}))', # Consume Before: 20/07/2O24
|
89 |
+
r'(?:consume\s*before\s*[:\-]?\s*.*?(\d{2}[\/\-]\d{2}[\/\-]\d{4}))', # Consume Before: 20/07/2024
|
90 |
+
r'(?:consume\s*before\s*[:\-]?\s*.*?(\d{2}[\/\-]\d{2}[\/\-][0O]\d{2}))', # Consume Before: 20/07/2O24
|
91 |
+
r'(?:consume\s*before\s*[:\-]?\s*.?(\d{2}\s[A-Za-z]{3,}\s*[0O]\d{2}))', # Consume Before: 20 MAY 2O24
|
92 |
+
r'(?:consume\s*before\s*[:\-]?\s*.?(\d{2}\s[A-Za-z]{3,}\s*\d{4}))', # Consume Before: 20 MAY 2024
|
93 |
+
r'(?:consume\s*before\s*[:\-]?\s*.?(\d{2}\s[A-Za-z]{3,}\s*[0O]\d{2}))', # Consume Before: 20 MAY 2O24
|
94 |
+
r'(?:consume\s*before\s*[:\-]?\s*.*?(\d{4}[\/\-]\d{2}[\/\-][0O]\d{2}))', # Consume Before: 2024/07/2O24
|
95 |
+
r'(?:consume\s*before\s*[:\-]?\s*.*?(\d{4}[\/\-]\d{2}[\/\-]\d{2}))', # Consume Before: 2024/07/20
|
96 |
+
r'(?:exp\s*[:\-]?\s*.*?(\d{2}[\/\-]\d{2}[\/\-][0O]\d{2}))', # Exp: 20/07/2O24
|
97 |
+
r'(?:exp\s*[:\-]?\s*.*?(\d{2}[\/\-]\d{2}[\/\-]\d{4}))', # Exp: 20/07/2024
|
98 |
+
r'(?:exp\s*[:\-]?\s*.*?(\d{2}[\/\-]\d{2}[\/\-][0O]\d{2}))', # Exp: 20/07/2O24
|
99 |
+
r'(?:exp\s*[:\-]?\s*.?(\d{2}\s[A-Za-z]{3,}\s*[0O]\d{2}))', # Exp: 20 MAY 2O24
|
100 |
+
r'(?:exp\s*[:\-]?\s*.?(\d{2}\s[A-Za-z]{3,}\s*\d{4}))', # Exp: 20 MAY 2024
|
101 |
+
r'(?:exp\s*[:\-]?\s*.?(\d{2}\s[A-Za-z]{3,}\s*[0O]\d{2}))', # Exp: 20 MAY 2O24
|
102 |
+
r'(?:exp\s*[:\-]?\s*.*?(\d{4}[\/\-]\d{2}[\/\-][0O]\d{2}))', # Exp: 2024/07/2O24
|
103 |
+
r'(?:exp\s*[:\-]?\s*.*?(\d{4}[\/\-]\d{2}[\/\-]\d{2}))', # Exp: 2024/07/20
|
104 |
+
r'Exp\.Date\s+(\d{2}[A-Z]{3}\d{4})',
|
105 |
+
r'(?:exp\s*\.?\s*date\s*[:\-]?\s*.?(\d{2}\s[A-Za-z]{3,}\s*[0O]\d{2}))', # Exp. Date: 16 MAR 2O30 (with typo)
|
106 |
+
r'(?:exp\s*\.?\s*date\s*[:\-]?\s*.*?(\d{2}[\/\-]\d{2}[\/\-][0O]\d{2}))', # Exp. Date: 15/12/2O30 (with typo)
|
107 |
+
r'(?:exp\s*\.?\s*date\s*[:\-]?\s*.?(\d{2}\s[A-Za-z]{3,}\s*[0O]\d{2}))', # Exp. Date: 15 MAR 2O30 (with typo)
|
108 |
+
r'(?:exp\s*\.?\s*date\s*[:\-]?\s*.?(\d{2}\s[A-Za-z]{3,}\s*[0O]\d{2}))', # Exp. Date cdsyubfuyef 15 MAR 2O30 (with typo)
|
109 |
+
r'(\d{2}[\/\-]\d{2}[\/\-]\d{4})', # 20/07/2024
|
110 |
+
r'(\d{2}[\/\-]\d{2}[\/\-]\d{2})', # 20/07/24
|
111 |
+
r'(\d{2}\s*[A-Za-z]{3,}\s*\d{4})', # 20 MAY 2024
|
112 |
+
r'(\d{2}\s*[A-Za-z]{3,}\s*\d{2})', # 20 MAY 24
|
113 |
+
r'(\d{4}[\/\-]\d{2}[\/\-]\d{2})', # 2024/07/20
|
114 |
+
r'(\d{4}[\/\-]\d{2}[\/\-]\d{2})', # 2024-07-20
|
115 |
+
r'(\d{4}[A-Za-z]{3,}\d{2})', # 2024MAY20
|
116 |
+
r'(\d{2}[A-Za-z]{3,}\d{4})', # 20MAY2024
|
117 |
+
r'(?:exp\.?\s*date\s*[:\-]?\s*(\d{2}\s*[A-Za-z]{3,}\s*(\d{4}|\d{2})))',
|
118 |
+
r'(?:exp\.?\s*date\s*[:\-]?\s*(\d{2}\s*\d{2}\s*\d{4}))', # Exp. Date: 20 05 2025
|
119 |
+
r'(\d{4}[A-Za-z]{3}\d{2})', # 2025MAY11
|
120 |
+
r'(?:best\s*before\s*[:\-]?\s*(\d+)\s*(days?|months?|years?))', # Best Before: 6 months
|
121 |
+
r'(?:best\s*before\s*[:\-]?\s*(three)\s*(months?))',
|
122 |
+
r'(\b(?:Jan|Feb|Mar|Apr|May|Jun|Jul|Aug|Sep|Oct|Nov|Dec)\b\s*\d{4})',
|
123 |
+
r'\bUSE BY\s+(\d{1,2}[A-Za-z]{3}\d{4})\b',
|
124 |
+
r'Exp\.Date\s*(\d{2}[A-Z]{3}\d{4})',
|
125 |
+
r'EXP:\d{4}/\d{2}/\d{4}/\d{1}/[A-Z]', # JAN-15-2024
|
126 |
+
r'USE BY[:\-]?\s*(\d{1,2}[\/\-]\d{1,2}[\/\-]\d{4})', # Use by date
|
127 |
+
r'BEST BEFORE[:\-]?\s*(\d{1,2}[\/\-]\d{1,2}[\/\-]\d{4})', # Best before date
|
128 |
+
r'EXPIRY DATE[:\-]?\s*(\d{1,2}[\/\-]\d{1,2}[\/\-]\d{4})', # Expiry Date
|
129 |
+
r'EXPIRY[:\-]?\s*(\d{1,2}[\/\-]\d{1,2}[\/\-]\d{4})', # Expiry
|
130 |
+
r'EXP[:\-]?\s*(\d{1,2}[\/\-]\d{1,2}[\/\-]\d{4})', # Exp
|
131 |
+
r'VALID UNTIL[:\-]?\s*(\d{1,2}[\/\-]\d{1,2}[\/\-]\d{4})', # Valid Until
|
132 |
+
r'CONSUME BY[:\-]?\s*(\d{1,2}[\/\-]\d{1,2}[\/\-]\d{4})', # Consume By
|
133 |
+
r'EXPIRES ON[:\-]?\s*(\d{1,2}[\/\-]\d{1,2}[\/\-]\d{4})', # Expires On
|
134 |
+
# DDMMMYYYY format
|
135 |
+
r'(\d{1,2}[A-Za-z]{3}\d{4})', # DDMMMYYYY format
|
136 |
+
# Short year formats
|
137 |
+
r'(\d{1,2}[\/\-]\d{1,2}[\/\-]\d{2})', # Short year date format (DD/MM/YY)
|
138 |
+
r'(\d{1,2}[\/\-]\d{1,2}[\/\-]\d{4})', # General date format (DD/MM/YYYY)
|
139 |
+
# Year-month-day formats
|
140 |
+
r'(\d{4}[\/\-]\d{1,2}[\/\-]\d{1,2})', # Year-month-day format (YYYY/MM/DD)
|
141 |
+
# Month/Year formats
|
142 |
+
r'(\d{1,2}[\/\-]\d{1,2})', # MM/DD format
|
143 |
+
r'(\d{1,2}[\/\-]\d{2})', # MM/YY format
|
144 |
+
# Month name formats
|
145 |
+
r'(\d{1,2}\s*[A-Za-z]{3,}\s*\d{4})', # Month name with day and year
|
146 |
+
r'(\d{1,2}\s*[A-Za-z]{3,}\s*\d{2})', # Month name with day and short year
|
147 |
+
# Year with month name
|
148 |
+
r'(\d{4}[A-Za-z]{3,}\d{1,2})', # Year with month name
|
149 |
+
r'(\d{1,2}[A-Za-z]{3,}\d{4})', # Day with month name and full year
|
150 |
+
# Additional formats
|
151 |
+
r'(\d{1,2}[\/\-]\d{1,2}[\/\-]\d{2})', # MM/DD/YY format
|
152 |
+
r'(\d{1,2}[\/\-]\d{1,2}[\/\-]\d{4})', # MM/DD/YYYY format
|
153 |
+
r'(\d{1,2}[\/\-]\d{1,2})', # MM/DD format
|
154 |
+
r'(\d{1,2}[\/\-]\d{2})', # MM/YY format
|
155 |
+
# Best before phrases
|
156 |
+
r'Best before (\d+) months', # Best before in months
|
157 |
+
r'Expiration Date[:\-]?\s*(\d{1,2}[\/\-]\d{1,2}[\/\-]\d{4})', # Expiration Date
|
158 |
+
r'Expires[:\-]?\s*(\d{1,2}[\/\-]\d{1,2}[\/\-]\d{4})', # Expires
|
159 |
+
# Additional variations
|
160 |
+
r'(\d{1,2}\s*[A-Za-z]{3,}\s*\d{4})', # Month name with day and year
|
161 |
+
r'(\d{1,2}\s*[A-Za-z]{3,}\s*\d{2})', # Month name with day and short year
|
162 |
+
# More variations
|
163 |
+
r'(\d{1,2}[\/\-]\d{1,2}[\/\-]\d{4})', # MM/DD/YYYY format
|
164 |
+
r'(\d{1,2}[\/\-]\d{1,2})', # MM/DD format
|
165 |
+
r'(\d{1,2}[\/\-]\d{2})', # MM/YY format
|
166 |
+
# Additional expiry phrases
|
167 |
+
r'(\d{1,2}[\/\-]\d{1,2}[\/\-]\d{4})', # Expiry in various formats
|
168 |
+
r'(\d{1,2}[\/\-]\d{1,2}[\/\-]\d{2})', # Expiry in short year formats
|
169 |
+
r'(\d{1,2}[\/\-]\d{1,2})', # Expiry in MM/DD format
|
170 |
+
r'(\d{1,2}[\/\-]\d{2})', # Expiry in MM/YY format
|
171 |
+
# Additional phrases
|
172 |
+
r'(\d{1,2}\s*[A-Za-z]{3,}\s*\d{4})', # Month name with day and year
|
173 |
+
r'(\d{1,2}\s*[A-Za-z]{3,}\s*\d{2})', # Month name with day and short year
|
174 |
+
r'(\d{4}[A-Za-z]{3,}\d{1,2})', # Year with month name
|
175 |
+
r'(\d{1,2}[A-Za-z]{3,}\d{4})', # Day with month name and full year
|
176 |
+
# Additional expiry phrases
|
177 |
+
r'(\d{1,2}[\/\-]\d{1,2}[\/\-]\d{4})', # Expiry in various formats
|
178 |
+
r'(\d{1,2}[\/\-]\d{1,2}[\/\-]\d{2})', # Expiry in short year formats
|
179 |
+
r'(\d{1,2}[\/\-]\d{1,2})', # Expiry in MM/DD format
|
180 |
+
r'(\d{1,2}[\/\-]\d{2})', # Expiry in MM/YY format
|
181 |
]
|
182 |
dates = []
|
183 |
for pattern in patterns:
|