I’m working on a script that will add employee names to a table in a PDF by first extracting that table, modifying it, and reentering it into the PDF. My script is as follows:

import camelot
import fitz
import pandas as pd

def extract_tables_from_pdf(pdf_path):
    try:
        tables = camelot.read_pdf(pdf_path, pages='all')
        print(f"Total tables extracted: {len(tables)}")
        
        pdf_document = fitz.open(pdf_path)
        
        header_coordinates = {}

        for i, table in enumerate(tables):
            print(f"--- Table {i + 1} ---")
            
            table.to_csv(f'table_{i + 1}.csv')
            
            page_num = table.page
            page = pdf_document.load_page(page_num - 1)
            
            header_texts = ["Print", "Sign", "Company"]
            
            for text in header_texts:
                text_instances = page.search_for(text)
                for inst in text_instances:
                    print(f"Header '{text}' found at page {page_num}, coordinates: {inst}")
                    header_coordinates[text] = inst
        
        pdf_document.close()
        return tables, header_coordinates
    
    except Exception as e:
        print(f"An error occurred: {str(e)}")
        return None, None

def modify_and_insert_table(pdf_path, tables, header_coordinates, output_pdf_path):
    try:
        pdf_document = fitz.open(pdf_path)

        for i, table in enumerate(tables):
            df = table.df

            df_no_headers = df.iloc[1:].copy()
            
            new_data = pd.Series(["Employee Name", "Employee Name", "Company Name"], index=[0, 1, 2])
            df_no_headers.iloc[0, 0:3] = new_data
            
            table_text = df_no_headers.to_string(index=False, header=False)

            page_num = table.page
            page = pdf_document.load_page(page_num - 1)
            
            for col_idx, header in enumerate(["Print", "Sign", "Company"]):
                if header in header_coordinates:
                    inst = header_coordinates[header]
                    rect = fitz.Rect(inst.x0, inst.y1, inst.x1, inst.y1 + 20)
                    text = new_data[col_idx]
                      
                    page.insert_textbox(rect, text, fontsize=12, fontname="helvetica", color=(0, 0, 0), align=0)

        pdf_document.save(output_pdf_path)
        pdf_document.close()
        print(f"Modified PDF saved to {output_pdf_path}")
    
    except Exception as e:
        print(f"An error occurred: {str(e)}")

if __name__ == "__main__":
    pdf_path = "/path/to/Adobe.pdf"
    output_pdf_path = "/path/to/Signed Adobe.pdf"
    
    tables, header_coordinates = extract_tables_from_pdf(pdf_path)
    if tables and header_coordinates:
        modify_and_insert_table(pdf_path, tables, header_coordinates, output_pdf_path)

I noticed that new_data = pd.Series(["Employee Name", "Employee Name", "Company Name"], index=[0, 1, 2]) didn’t work, so I tried new_data = pd.Series(["Employee", "Employee", "Company"], index=[0, 1, 2]) and that appears to have worked fine.

Right before page.insert_textbox(rect, text, fontsize=12, fontname="helvetica", color=(0, 0, 0), align=0), I put print(text) and am able to see Employee Name and Company Name, but when I inspect the Signed Adobe.pdf, the text is not there. I’m not sure why it would work with a single string containing no spaces, and not with a string containing spaces. Any assistance is greatly appreciated!