Update app.py (#1)

- Update app.py (6d9fb153198c7384359509bf59c73c1dc50b2893)


Co-authored-by: K Neeraj Kalyan <neerajkalyank@users.noreply.huggingface.co>

app.py

@@ -2,6 +2,7 @@ import pdfplumber
 import pandas as pd
 import gradio as gr
 import re
+import tempfile
 
 # Define function to extract data
 def extract_data(pdf_file):
@@ -12,71 +13,84 @@ def extract_data(pdf_file):
     purchase_order_no = "Not Found"
     purchase_order_date = "Not Found"
 
-    with pdfplumber.open(pdf_file) as pdf:
-        for page in pdf.pages:
-            text = page.extract_text().splitlines()
-
-            # Attempt to dynamically extract Purchase Order No and Date from the first page
-            for line in text:
-                # Search for Purchase Order No
-                po_match = re.search(r'Purchase Order No[:\s]+(\d+)', line, re.IGNORECASE)
-                if po_match:
-                    purchase_order_no = po_match.group(1)
-
-                # Search for Date
-                date_match = re.search(r'Date[:\s]+(\d{2}\.\d{2}\.\d{4})', line, re.IGNORECASE)
-                if date_match:
-                    purchase_order_date = date_match.group(1)
-
-                # Stop if both values are found
-                if purchase_order_no != "Not Found" and purchase_order_date != "Not Found":
-                    break
-
-            # Process lines to extract row data, looking for rows that start with SI No
-            for line in text:
-                try:
-                    # Match lines that start with an SI number (e.g., "10", "20")
-                    si_no_match = re.match(r'^(\d+)\s', line)
-                    if si_no_match:
-                        parts = line.split()
-
-                        # Extract SI No
-                        si_no = parts[0]
-
-                        # Extract Material Number and format the Material Description
-                        material_number = parts[2] if len(parts) > 2 else "Unknown"
-                        material_desc = f"BPS 017507\nMaterial Number: {material_number}\nHSN Code: 8310\nIGST: 18%"
-
-                        # Extract Unit, Quantity, Dely Qty, Dely Date, Unit Rate, and Value
-                        unit = parts[3] if len(parts) > 3 else "NO"  # Default to "NO" if not found
-                        quantity = int(parts[4]) if len(parts) > 4 else 0
-                        dely_qty = int(parts[5]) if len(parts) > 5 else 0
-                        dely_date = parts[6] if len(parts) > 6 else "Unknown"
-                        unit_rate = float(parts[7]) if len(parts) > 7 else 0.0
-                        value = float(parts[8]) if len(parts) > 8 else 0.0
-
-                        # Append extracted data in the specified order
-                        data.append([
-                            purchase_order_no,
-                            purchase_order_date,
-                            si_no,
-                            material_desc,
-                            unit,
-                            quantity,
-                            dely_qty,
-                            dely_date,
-                            unit_rate,
-                            value
-                        ])
-                except (ValueError, IndexError) as e:
-                    print(f"Error processing line: {line} - {e}")
-                    continue  # Skip lines that do not match the expected format
-
-    # Convert data to DataFrame and save as Excel
-    df = pd.DataFrame(data, columns=columns)
-    excel_path = "/tmp/Extracted_Purchase_Order_Data.xlsx"
-    df.to_excel(excel_path, index=False)
-    
+    try:
+        with pdfplumber.open(pdf_file) as pdf:
+            for page in pdf.pages:
+                text = page.extract_text()
+                
+                if not text:
+                    continue  # Skip pages without text
+                
+                lines = text.splitlines()
+
+                # Attempt to dynamically extract Purchase Order No and Date from the first page
+                for line in lines:
+                    # Search for Purchase Order No
+                    po_match = re.search(r'Purchase Order No[:\s]+(\d+)', line, re.IGNORECASE)
+                    if po_match:
+                        purchase_order_no = po_match.group(1)
+
+                    # Search for Date
+                    date_match = re.search(r'Date[:\s]+(\d{2}\.\d{2}\.\d{4})', line, re.IGNORECASE)
+                    if date_match:
+                        purchase_order_date = date_match.group(1)
+
+                    # Stop if both values are found
+                    if purchase_order_no != "Not Found" and purchase_order_date != "Not Found":
+                        break
+
+                # Process lines to extract row data, looking for rows that start with SI No
+                for line in lines:
+                    try:
+                        # Match lines that start with an SI number (e.g., "10", "20")
+                        si_no_match = re.match(r'^(\d+)\s', line)
+                        if si_no_match:
+                            parts = line.split()
+
+                            # Extract SI No
+                            si_no = parts[0]
+
+                            # Extract Material Number and format the Material Description
+                            material_number = parts[2] if len(parts) > 2 else "Unknown"
+                            material_desc = f"BPS 017507\nMaterial Number: {material_number}\nHSN Code: 8310\nIGST: 18%"
+
+                            # Extract Unit, Quantity, Dely Qty, Dely Date, Unit Rate, and Value
+                            unit = parts[3] if len(parts) > 3 else "NO"  # Default to "NO" if not found
+                            quantity = int(parts[4]) if len(parts) > 4 else 0
+                            dely_qty = int(parts[5]) if len(parts) > 5 else 0
+                            dely_date = parts[6] if len(parts) > 6 else "Unknown"
+                            unit_rate = float(parts[7]) if len(parts) > 7 else 0.0
+                            value = float(parts[8]) if len(parts) > 8 else 0.0
+
+                            # Append extracted data in the specified order
+                            data.append([
+                                purchase_order_no,
+                                purchase_order_date,
+                                si_no,
+                                material_desc,
+                                unit,
+                                quantity,
+                                dely_qty,
+                                dely_date,
+                                unit_rate,
+                                value
+                            ])
+                    except (ValueError, IndexError) as e:
+                        print(f"Error processing line: {line} - {e}")
+                        continue  # Skip lines that do not match the expected format
+
+        # Convert data to DataFrame and save as Excel
+        df = pd.DataFrame(data, columns=columns)
+        
+        # Generate a temporary file path for the Excel file
+        with tempfile.NamedTemporaryFile(suffix=".xlsx", delete=False) as tmp_file:
+            excel_path = tmp_file.name
+            df.to_excel(excel_path, index=False)
+
+    except Exception as e:
+        print(f"An error occurred while processing the PDF: {e}")
+        return None
+
     # Log warning if data was not found for Purchase Order No or Date
     if purchase_order_no == "Not Found" or purchase_order_date == "Not Found":
         print("Warning: 'Purchase Order No' or 'Date' was not found in the PDF.")