reader4.py

import fitz  # PyMuPDF

def extract_transactions_table_only(input_pdf_path, output_pdf_path):
    doc = fitz.open(input_pdf_path)
    output_doc = fitz.open()

    for page_num in range(len(doc)):
        page = doc.load_page(page_num)
        blocks = page.get_text("blocks")  # Get text blocks with positions

        # Define header region (top 15% of page height)
        header_threshold = page.rect.height * 0.15
        found_header = False

        for block in blocks:
            x0, y0, x1, y1, text, *_ = block
            if text.strip() == "TRANSACTIONS" and y1 < header_threshold:
                found_header = True
                break

        if found_header:
            # Define table region (below header line)
            table_start_y = header_threshold + 10  # Start just below header
            table_blocks = [
                block for block in blocks
                if block[1] > table_start_y and block[4].strip() != ""
            ]

            if table_blocks:
                # Sort blocks top to bottom
                table_blocks.sort(key=lambda b: b[1])

                # Combine text lines
                table_text = "\n".join(block[4].strip() for block in table_blocks)

                # Create new page and write table
                new_page = output_doc.new_page(width=595, height=842)  # A4
                new_page.insert_text((50, 50), f"Page {page_num + 1} - TRANSACTIONS Table", fontsize=12)
                new_page.insert_textbox(fitz.Rect(50, 80, 545, 800), table_text, fontsize=10, fontname="courier")

    if output_doc.page_count > 0:
        output_doc.save(output_pdf_path)
        print(f"✅ Extracted tables saved to: {output_pdf_path}")
    else:
        print("⚠️ No pages with exact 'TRANSACTIONS' header found.")

    doc.close()
    output_doc.close()

# 🔧 Replace with your actual paths
input_pdf = r"C:\Users\Suren\Downloads\input.pdf"
output_pdf = r"C:\Users\Suren\Downloads\transactions_output.pdf"

extract_transactions_table_only(input_pdf, output_pdf)