Update app.py
Browse files
app.py
CHANGED
@@ -1,144 +1,109 @@
|
|
1 |
-
import re
|
2 |
-
import pandas as pd
|
3 |
import pdfplumber
|
|
|
|
|
4 |
import gradio as gr
|
5 |
|
6 |
-
|
7 |
def extract_text_from_pdf(pdf_file):
|
8 |
-
"""
|
9 |
-
Extracts text from an uploaded PDF file.
|
10 |
-
Args:
|
11 |
-
pdf_file: The uploaded PDF file.
|
12 |
-
Returns:
|
13 |
-
str: The extracted text from the PDF.
|
14 |
-
"""
|
15 |
with pdfplumber.open(pdf_file.name) as pdf:
|
16 |
text = ""
|
17 |
for page in pdf.pages:
|
18 |
-
text += page.extract_text()
|
19 |
-
print("\nExtracted Text:\n", text) # Debugging: Print
|
20 |
return text
|
21 |
|
22 |
-
|
23 |
-
def
|
24 |
-
"""
|
25 |
-
|
26 |
-
|
27 |
-
|
28 |
-
|
29 |
-
|
30 |
-
"""
|
31 |
-
|
32 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
33 |
|
34 |
for line in lines:
|
35 |
-
|
36 |
-
|
37 |
-
|
38 |
-
|
39 |
-
|
40 |
-
|
41 |
-
|
42 |
-
if current_row:
|
43 |
-
combined_rows.append(current_row.strip())
|
44 |
-
|
45 |
-
return combined_rows
|
46 |
-
|
47 |
-
|
48 |
-
def parse_po_items(rows):
|
49 |
-
"""
|
50 |
-
Parses purchase order items from reconstructed rows.
|
51 |
-
Args:
|
52 |
-
rows (list): List of reconstructed rows.
|
53 |
-
Returns:
|
54 |
-
tuple: DataFrame with extracted data and a status message.
|
55 |
-
"""
|
56 |
-
data = []
|
57 |
-
for row in rows:
|
58 |
-
try:
|
59 |
-
# Match ITEM, DESCRIPTION, QTY, UNIT, UNIT PRICE, TOTAL PRICE
|
60 |
-
match = re.match(
|
61 |
-
r"^(?P<Item>\d+)\s+(?P<Description>.+?)\s+(?P<Qty>\d+)\s+(?P<Unit>\S+)\s+(?P<UnitPrice>[\d.]+)\s+(?P<TotalPrice>[\d.]+)$",
|
62 |
-
row,
|
63 |
-
)
|
64 |
-
if match:
|
65 |
-
data.append(
|
66 |
-
{
|
67 |
-
"ITEM": match.group("Item"),
|
68 |
-
"DESCRIPTION": match.group("Description"),
|
69 |
-
"QTY": match.group("Qty"),
|
70 |
-
"UNIT": match.group("Unit"),
|
71 |
-
"UNIT PRICE": match.group("UnitPrice"),
|
72 |
-
"TOTAL PRICE": match.group("TotalPrice"),
|
73 |
-
}
|
74 |
)
|
75 |
-
|
76 |
-
|
77 |
-
|
78 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
79 |
|
80 |
if not data:
|
81 |
-
|
|
|
82 |
return pd.DataFrame(data), "Data extracted successfully."
|
83 |
|
84 |
-
|
85 |
def save_to_excel(df, output_path="extracted_po_data.xlsx"):
|
86 |
-
"""
|
87 |
-
Saves the extracted data to an Excel file.
|
88 |
-
Args:
|
89 |
-
df (pd.DataFrame): DataFrame containing the structured data.
|
90 |
-
output_path (str): Path to save the Excel file.
|
91 |
-
Returns:
|
92 |
-
str: Path to the saved file.
|
93 |
-
"""
|
94 |
df.to_excel(output_path, index=False)
|
95 |
return output_path
|
96 |
|
97 |
-
|
98 |
def process_pdf(file):
|
99 |
-
"""
|
100 |
-
Processes the uploaded PDF file, extracts data, and saves it to an Excel file.
|
101 |
-
Args:
|
102 |
-
file: The uploaded PDF file.
|
103 |
-
Returns:
|
104 |
-
tuple: Path to the saved Excel file and a status message.
|
105 |
-
"""
|
106 |
try:
|
107 |
-
# Extract text from the uploaded PDF
|
108 |
text = extract_text_from_pdf(file)
|
109 |
-
|
110 |
-
lines = text.splitlines()
|
111 |
-
# Preprocess lines to reconstruct rows
|
112 |
-
rows = preprocess_lines(lines)
|
113 |
-
# Parse reconstructed rows
|
114 |
-
df, status = parse_po_items(rows)
|
115 |
if df is not None:
|
116 |
output_path = save_to_excel(df)
|
117 |
return output_path, status
|
118 |
return None, status
|
119 |
except Exception as e:
|
120 |
-
return None, f"Error: {str(e)}"
|
121 |
|
122 |
-
|
123 |
-
|
124 |
-
|
125 |
-
"""
|
126 |
-
Creates a Gradio interface for processing PO data from PDF files.
|
127 |
-
"""
|
128 |
-
interface = gr.Interface(
|
129 |
fn=process_pdf,
|
130 |
inputs=gr.File(label="Upload PDF", file_types=[".pdf"]),
|
131 |
outputs=[
|
132 |
-
gr.File(label="Download Extracted
|
133 |
gr.Textbox(label="Status"),
|
134 |
],
|
135 |
title="PO Data Extraction",
|
136 |
-
description="Upload a
|
137 |
)
|
138 |
-
return interface
|
139 |
-
|
140 |
|
141 |
if __name__ == "__main__":
|
142 |
-
|
143 |
-
|
144 |
-
app.launch()
|
|
|
|
|
|
|
1 |
import pdfplumber
|
2 |
+
import pandas as pd
|
3 |
+
import re
|
4 |
import gradio as gr
|
5 |
|
6 |
+
# Function: Extract Text from PDF
|
7 |
def extract_text_from_pdf(pdf_file):
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
8 |
with pdfplumber.open(pdf_file.name) as pdf:
|
9 |
text = ""
|
10 |
for page in pdf.pages:
|
11 |
+
text += page.extract_text()
|
12 |
+
print("\nExtracted Text:\n", text) # Debugging: Print extracted text
|
13 |
return text
|
14 |
|
15 |
+
# Function: Clean Description
|
16 |
+
def clean_description(description, item_number=None):
|
17 |
+
description = re.sub(r"\d+\s+(Nos\.|Set)\s+[\d.]+\s+[\d.]+", "", description) # Remove Qty + Unit + Price
|
18 |
+
description = re.sub(r"Page \d+ of \d+.*", "", description) # Remove page references
|
19 |
+
description = re.sub(r"\(Q\. No:.*?\)", "", description) # Remove Q.No-related data
|
20 |
+
description = re.sub(r"TOTAL EX-WORK.*", "", description) # Remove EX-WORK-related text
|
21 |
+
description = re.sub(r"NOTES:.*", "", description) # Remove notes section
|
22 |
+
description = re.sub(r"HS CODE.*", "", description) # Remove HS CODE-related data
|
23 |
+
description = re.sub(r"DELIVERY:.*", "", description) # Remove delivery instructions
|
24 |
+
return description.strip()
|
25 |
+
|
26 |
+
# Function: Parse PO Items with Filters
|
27 |
+
def parse_po_items_with_filters(text):
|
28 |
+
lines = text.splitlines()
|
29 |
+
data = []
|
30 |
+
current_item = {}
|
31 |
+
description_accumulator = []
|
32 |
|
33 |
for line in lines:
|
34 |
+
print(f"Processing Line: {line}") # Debugging
|
35 |
+
item_match = re.match(r"^\s*(?P<Item>\d+)\s+(?P<Description>.+)", line)
|
36 |
+
if item_match:
|
37 |
+
if current_item:
|
38 |
+
current_item["Description"] = clean_description(
|
39 |
+
" ".join(description_accumulator).strip(), item_number=int(current_item["Item"])
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
40 |
)
|
41 |
+
data.append(current_item)
|
42 |
+
description_accumulator = []
|
43 |
+
|
44 |
+
current_item = {
|
45 |
+
"Item": item_match.group("Item"),
|
46 |
+
"Description": "",
|
47 |
+
"Qty": "",
|
48 |
+
"Unit": "",
|
49 |
+
"Unit Price": "",
|
50 |
+
"Total Price": "",
|
51 |
+
}
|
52 |
+
description_accumulator.append(item_match.group("Description"))
|
53 |
+
elif current_item:
|
54 |
+
description_accumulator.append(line.strip())
|
55 |
+
|
56 |
+
qty_match = re.search(r"(?P<Qty>\d+)\s+(Nos\.|Set)", line)
|
57 |
+
if qty_match:
|
58 |
+
current_item["Qty"] = qty_match.group("Qty")
|
59 |
+
current_item["Unit"] = qty_match.group(2)
|
60 |
+
|
61 |
+
price_match = re.search(r"(?P<UnitPrice>[\d.]+)\s+(?P<TotalPrice>[\d.]+)$", line)
|
62 |
+
if price_match:
|
63 |
+
current_item["Unit Price"] = price_match.group("UnitPrice")
|
64 |
+
current_item["Total Price"] = price_match.group("TotalPrice")
|
65 |
+
|
66 |
+
if current_item:
|
67 |
+
current_item["Description"] = clean_description(
|
68 |
+
" ".join(description_accumulator).strip(), item_number=int(current_item["Item"])
|
69 |
+
)
|
70 |
+
data.append(current_item)
|
71 |
|
72 |
if not data:
|
73 |
+
print("No items found. Check PDF format.") # Debugging
|
74 |
+
return None, "No items found. Please check the PDF file format."
|
75 |
return pd.DataFrame(data), "Data extracted successfully."
|
76 |
|
77 |
+
# Function: Save to Excel
|
78 |
def save_to_excel(df, output_path="extracted_po_data.xlsx"):
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
79 |
df.to_excel(output_path, index=False)
|
80 |
return output_path
|
81 |
|
82 |
+
# Gradio Interface Function
|
83 |
def process_pdf(file):
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
84 |
try:
|
|
|
85 |
text = extract_text_from_pdf(file)
|
86 |
+
df, status = parse_po_items_with_filters(text)
|
|
|
|
|
|
|
|
|
|
|
87 |
if df is not None:
|
88 |
output_path = save_to_excel(df)
|
89 |
return output_path, status
|
90 |
return None, status
|
91 |
except Exception as e:
|
92 |
+
return None, f"Error during processing: {str(e)}"
|
93 |
|
94 |
+
# Gradio Interface Setup
|
95 |
+
def create_gradio_interface():
|
96 |
+
return gr.Interface(
|
|
|
|
|
|
|
|
|
97 |
fn=process_pdf,
|
98 |
inputs=gr.File(label="Upload PDF", file_types=[".pdf"]),
|
99 |
outputs=[
|
100 |
+
gr.File(label="Download Extracted Data"),
|
101 |
gr.Textbox(label="Status"),
|
102 |
],
|
103 |
title="PO Data Extraction",
|
104 |
+
description="Upload a Purchase Order PDF to extract items into an Excel file.",
|
105 |
)
|
|
|
|
|
106 |
|
107 |
if __name__ == "__main__":
|
108 |
+
interface = create_gradio_interface()
|
109 |
+
interface.launch()
|
|