ashcodes/pdf-table-extractor-tabula

1import streamlit as st2import numpy as np 3import pandas as pd 4import subprocess 5from subprocess import STDOUT, check_call 6import os 7import base64  8import camelot 9 10# to run this only once and it's cached11@st.cache12def ghostscript():13    """install ghostscript on the linux machine"""14    proc = subprocess.Popen('apt-get install -y ghostscript', shell=True, stdin=None, stdout=open(os.devnull,"wb"), stderr=STDOUT, executable="/bin/bash")15    proc.wait()16 17ghostscript()18 19#heading20html_temp = """21    <div style="background-color:tomato;padding:10px">22    <h2 style="color:white;text-align:center;">PDF Table Extractor WebApp </h2>23    </div>24    """25st.markdown(html_temp,unsafe_allow_html=True)26 27 28# file uploader on streamlit 29#st.sidebar.markdown('Upload PDF files')30input_pdf = st.sidebar.file_uploader(label = "Upload PDF files here", type = 'pdf')31 32# run this only when a PDF is uploaded33if input_pdf is not None:34    # byte object into a PDF file 35    with open("input.pdf", "wb") as f:36        base64_pdf = base64.b64encode(input_pdf.read()).decode('utf-8')37        f.write(base64.b64decode(base64_pdf))38    f.close()39 40#To print uploaded pdf    41def show_pdf(file_path):42    with open(file_path,"rb") as f:43        base64_pdf = base64.b64encode(f.read()).decode('utf-8')44        pdf_display = f'<iframe src="data:application/pdf;base64,{base64_pdf}" width="800" height="800" type="application/pdf"></iframe>'45    st.markdown('## Uploaded PDF')46    st.markdown(pdf_display, unsafe_allow_html=True)47 48#st.sidebar.markdown('Display Uploaded PDF')    49#if st.sidebar.button('Show'):50    #show_pdf("input.pdf")51 52# read the pdf and parse it using stream53if input_pdf is not None:54    table = camelot.read_pdf('input.pdf', flavor='stream',layout_kwargs={'detect_vertical':True},backend='poppler')55    csv_table = table[0].df   56 57st.sidebar.markdown('Extract tables from PDF')58if st.sidebar.button('Extract Table'):59    st.markdown('## Extracted table from PDF')60    st.dataframe(csv_table)61    62if input_pdf is not None:63    st.sidebar.markdown('Download Extracted Table as CSV file')64    st.sidebar.download_button("Download",csv_table.to_csv(),file_name = 'extracted_table.csv', mime = 'text/csv')