pdf_to_excel / app.py
lodhrangpt's picture
Update app.py
a37c21d verified
import gradio as gr
import fitz # PyMuPDF
import pandas as pd
from transformers import pipeline
import base64
# Function to convert PDF to DataFrame
def pdf_to_dataframe(uploaded_file):
# Open the PDF document
# doc = fitz.open(pdf_path)
# # Initialize an empty list to store text blocks
# text_blocks = []
# # Iterate through each page in the PDF
# for page_num in range(len(doc)):
# page = doc.load_page(page_num)
# text = page.get_text("text")
# print(text)
# text_blocks.append(text)
# # Join all text blocks into a single string
# full_text = "\n".join(text_blocks)
# # Split the text into lines
# lines = full_text.split('\n')
# # Create a DataFrame from the lines
if uploaded_file is not None:
ocr_pipeline = pipeline("text2text-generation", model="google/t5-v1_1-large")
extracted_text = ocr_pipeline(uploaded_file.read(), max_length=1024, do_sample=False)[0]["generated_text"]
lines = extracted_text.split("\n")
data = []
for line in lines:
data.append([line])
df = pd.DataFrame(data, columns=["Text"])
# df = pd.DataFrame(lines, columns=['Text'])
return df
# Function to save DataFrame to Excel
def dataframe_to_excel(df, excel_path):
# Save the DataFrame to an Excel file
df.to_excel(excel_path, index=False)
# Main function
def main():
def pdf_to_excel_function(pdf_file):
# Save the uploaded PDF to a temporary file
pdf_path = "temp.pdf"
# with open(pdf_path, "wb") as f:
# f.write(pdf_file.read())
# Convert PDF to DataFrame
df = pdf_to_dataframe(pdf_file)
# Save DataFrame to Excel
excel_path = "output.xlsx"
dataframe_to_excel(df, excel_path)
return excel_path
# Create the Gradio interface
iface = gr.Interface(
fn=pdf_to_excel_function,
inputs=gr.File(label="Upload PDF File"),
outputs=gr.File(label="Download Excel File"),
title="PDF to Excel Converter",
description="Convert a PDF file to an Excel file."
)
# Launch the interface
iface.launch()
if __name__ == "__main__":
main()