app.py

import streamlit as st
import os
import pandas as pd
from PyPDF2 import PdfReader
import openai
from collections import defaultdict
from io import StringIO
from pdfminer.high_level import extract_text  
import json
from openai import OpenAI
import re

# 1. Initialization
api_key = "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"  
openai.api_key = api_key
client = OpenAI(api_key=api_key)
pdf_folder = "pdf" 

st.title("Mahkamah Agung: NER & Summarization of Legal Documents")



#---------------------PDF OVERVIEW----------------------
st.subheader("PDF Folder Overview")
def get_pdf_details(folder_path):
    pdf_details = []
    for filename in os.listdir(folder_path):
        if filename.lower().endswith('.pdf'):
            pdf_path = os.path.join(folder_path, filename)
            try:
                with open(pdf_path, "rb") as file:
                    pdf_reader = PdfReader(file)
                    page_count = len(pdf_reader.pages)
                    pdf_details.append({"Filename": filename, "Page Count": page_count})
            except Exception as e:
                st.warning(f"Could not read {filename}: {str(e)}")
    return pdf_details
pdf_list = get_pdf_details(pdf_folder)
pdf_df = pd.DataFrame(pdf_list)
if not pdf_df.empty:
    with st.expander('PDF Overview'):
       st.dataframe(pdf_df)
else:
    st.warning("No PDFs found in the specified folder.")



#---------------------PDF SEARCH AND EXTRACT----------------------
st.subheader("PDF to Text Conversion")

# Function to read and extract text from a PDF using PdfReader
def extract_text_from_pdf_pypdf2(pdf_path):
    text = ""
    with open(pdf_path, "rb") as file:
        pdf_reader = PdfReader(file)
        for page in pdf_reader.pages:
            page_text = page.extract_text()
            if page_text:
                text += page_text
    return text




# Function to read and extract text from a PDF using pdfminer
def extract_text_from_pdf_pdfminer(pdf_path):
    # Extract text using pdfminer.six
    try:
        text = extract_text(pdf_path)
    except Exception as e:
        st.error(f"Error extracting text from {pdf_path}: {str(e)}")
        text = ""
    return text

pdf_files = [f for f in os.listdir(pdf_folder) if f.lower().endswith('.pdf')]
search_query = st.text_input("Search for a PDF")
filtered_pdfs = [pdf for pdf in pdf_files if search_query.lower() in pdf.lower()]

if filtered_pdfs:
    selected_pdf = st.selectbox("Select a PDF to convert to text", filtered_pdfs)
else:
    st.warning("No PDFs found matching your search.")

if st.button("analyze The Document"):
    pdf_path = os.path.join(pdf_folder, selected_pdf)
    extracted_text = extract_text_from_pdf_pdfminer(pdf_path)


    # Display the extracted text
    if extracted_text:
        with st.expander('Extracted Text'):
            st.text_area("Extracted Text", value=extracted_text, height=300)
    else:
        st.warning("No text extracted. The PDF might contain images or other non-text content.")





    # template = """
    #
    # #   Anda adalah seorang hakim agung di Mahkamah Agung di Indonesia. Dari hasil putusan dibawah ini berikan aku kesimpulannya:
    # {}
    # variabel yang harus ada adalah sebagai berikut: presiding judge, member judge, clerk, ruling, other rulings, note of ruling, date of deliberation, date read out, type of judicial institution, date of register, judicial institution, case_number, court, defendants.name, defendants.place_of_birth, defendants.date_of_birth, defendants.age, defendants.gender, defendants.nationality, defendants.religion, defendants.occupation, charges.article, charges.offense, verdict.sentence, verdict.assets_confiscated.description, verdict.assets_confiscated.weight, fine dan conclusion
    # # """

    template = """
        
        # Anda Adalah Seorang Hakim Agung Di Mahkamah Agung Di Indonesia. Berdasarkan Putusan Di Bawah Ini, Berikan Kesimpulannya:
        {}
        Variabel Yang Harus Ada Adalah Sebagai Berikut: Hakim Ketua, Hakim Anggota, Panitera, Putusan, Putusan Lainnya, Catatan Putusan, Tanggal Musyawarah, Tanggal Pembacaan, Jenis Institusi Yudisial, Tanggal Pendaftaran, Institusi Yudisial, Nomor Kasus, Pengadilan, Terdakwa.Nama, Terdakwa.Tempat_Lahir, Terdakwa.Tanggal_Lahir, Terdakwa.Usia, Terdakwa.Jenis_Kelamin, Terdakwa.Kebangsaan, Terdakwa.Agama, Terdakwa.Pekerjaan, Pasal_Dakwaan, Pelanggaran_Dakwaan, Vonis.Hukuman, Vonis.Atribut_Disita.Deskripsi, Vonis.Atribut_Disita.Berat, Denda, Dan Kesimpulan.
    # """



    #---------------------NER & SUMMARIZATION----------------------
    response = client.chat.completions.create(
    model="gpt-3.5-turbo-0125",
    response_format={ "type": "json_object" },
    messages=[
        {"role": "system", "content": "You are a helpful assistant designed to output JSON."},
        {"role": "user", "content": template.format(extracted_text)}
    ]
    )



    data= json.loads(response.choices[0].message.content)
    df = pd.json_normalize(data)
    df=df.T
    df.columns = ["Kesimpulan Putusan"]
    st.dataframe(df)