app.py

import io
import os
import pandas as pd
import streamlit as st

from concurrent.futures import ThreadPoolExecutor
from datetime import datetime
from langchain_community.document_loaders.pdf import PyPDFLoader
from langchain_core.documents.base import Document
from langchain_text_splitters import TokenTextSplitter
from process import Process
from tempfile import NamedTemporaryFile
from stqdm import stqdm
from validate import Validation

buffer = io.BytesIO()

st.cache_data()
st.set_page_config(page_title="NutriGenMe Paper Extractor")
st.title("NutriGenMe - Paper Extractor")
st.markdown("<div style='text-align: justify;text-justify: inter-word;'>NutriGenMe Paper Extractor is a tool designed to extract relevant information from genomic papers related to the NutriGenMe project. It utilizes natural language processing techniques to parse through documents and extract key data points, enabling researchers and practitioners to efficiently gather insights from a large corpus of literature.</div>", unsafe_allow_html=True)
st.divider()

st.markdown("<h4>Extraction</h4>", unsafe_allow_html=True)
col1, col2 = st.columns(2)
st.markdown("<h4>Validation</h4>", unsafe_allow_html=True)
col3, col4 = st.columns(2)

with col1:
    models = (
        'gpt-4-turbo',
        'gemini-1.5-pro-latest',
        # 'llama-3-sonar-large-32k-chat',
        # 'mixtral-8x7b-instruct',
    )
    model = st.selectbox('Model selection:', models, key='model')

with col2:
    tokens = (
        8000,
        16000,
        24000
    )
    chunk_option = st.selectbox('Token amounts per process:', tokens, key='token')
    chunk_overlap = 0

with col3:
    models_val = (
        'gpt-4-turbo',
        'gemini-1.5-pro-latest',
        'mixtral-8x7b-instruct',
        # 'llama-3-sonar-large-32k-chat',
    )
    model_val = st.selectbox('Model validator selection:', models_val, key='model_val')

with col4:
    api = st.toggle('Validate with API')

if api:
    st.warning("""This validation process leverage external application programming interfaces (APIs) from NCBI and EBI to verify information. 
               These APIs may have limitations on their usage, so please exercise responsible use of this functionality. 
               If you opt to employ API validation and the process takes a long time (more than 1 hour), consider refreshing the page and proceeding without API validation.""", icon="⚠️")

st.divider()
st.markdown("<h4>Process</h4>", unsafe_allow_html=True)
uploaded_files = st.file_uploader("Upload Paper(s) here :", type="pdf", accept_multiple_files=True)

if uploaded_files:
    submit = st.button("Get Result", key='submit')

if uploaded_files and submit:

    with st.status("Extraction in progress ...", expanded=True) as status:
        for uploaded_file in stqdm(uploaded_files):
            start_time = datetime.now()
            with NamedTemporaryFile(dir='.', suffix=".pdf", delete=eval(os.getenv('DELETE_TEMP_PDF', 'True'))) as pdf:

                pdf.write(uploaded_file.getbuffer())
                st.markdown(f"Start Extraction process at <code>{datetime.now().strftime('%H:%M')}</code>", unsafe_allow_html=True)

                # Load Documents
                loader = PyPDFLoader(pdf.name)
                pages = loader.load()

                chunk_size = 120000
                chunk_overlap = 0
                docs = pages

                # Split Documents
                if chunk_option:
                    docs = [Document('\n'.join([page.page_content for page in pages]))]
                    docs[0].metadata = {'source': pages[0].metadata['source']}

                    chunk_size = chunk_option
                    chunk_overlap = int(0.25 * chunk_size)

                text_splitter = TokenTextSplitter.from_tiktoken_encoder(
                    chunk_size=chunk_size, chunk_overlap=chunk_overlap
                )
                chunks = text_splitter.split_documents(docs)

                # Start extraction process in parallel
                process = Process(model)
                with ThreadPoolExecutor() as executor:
                    result_gsd = executor.submit(process.get_entity, (chunks, 'gsd'))
                    result_summ = executor.submit(process.get_entity, (chunks, 'summ'))
                    result = executor.submit(process.get_entity, (chunks, 'all'))
                    result_one = executor.submit(process.get_entity_one, [c.page_content for c in chunks[:1]])
                    result_table = executor.submit(process.get_table, pdf.name)

                    result_gsd = result_gsd.result()
                    result_summ = result_summ.result()
                    result = result.result()
                    result_one = result_one.result()
                    res_gene, res_snp, res_dis = result_table.result()

                # Combine Result
                result['Genes'] = res_gene + result_gsd['Genes']
                result['SNPs'] = res_snp + result_gsd['SNPs']
                result['Diseases'] = res_dis + result_gsd['Diseases']
                result['Conclusion'] = result_summ
                for k in result_one.keys():
                    result[k] = result_one[k]

                if len(result['Genes']) == 0:
                    result['Genes'] = ['']
                
                # Adjust Genes, SNPs, Diseases
                num_rows = max(max(len(result['Genes']), len(result['SNPs'])), len(result['Diseases']))
                for k in ['Genes', 'SNPs', 'Diseases']:
                    while len(result[k]) < num_rows:
                        result[k].append('')

                    # Temporary handling
                    result[k] = result[k][:num_rows]

                # Arrange Column 
                result = {key: value if isinstance(value, list) else [value] * num_rows for key, value in result.items()}
                dataframe = pd.DataFrame(result)
                dataframe = dataframe[['Genes', 'SNPs', 'Diseases', 'Title', 'Authors', 'Publisher Name', 'Publication Year', 'Population', 'Sample Size', 'Study Methodology', 'Study Level', 'Conclusion']]
                dataframe = dataframe[dataframe['Genes'].astype(bool)].reset_index(drop=True)
                dataframe.reset_index(drop=True, inplace=True)
                
                # Validate Result
                st.markdown(f"Start Validation process at <code>{datetime.now().strftime('%H:%M')}</code>", unsafe_allow_html=True)
                validation = Validation(model_val)
                df, df_no_llm, df_clean = validation.validate(dataframe, api)
                df.drop_duplicates(['Genes', 'SNPs'], inplace=True)
                st.write("Success in ", round((datetime.now().timestamp() - start_time.timestamp()) / 60, 2), "minutes")

                st.dataframe(df)
                with pd.ExcelWriter(buffer, engine='xlsxwriter') as writer:
                    if api:
                        df.to_excel(writer, sheet_name='Result Cleaned API LLM')
                        df_no_llm.to_excel(writer, sheet_name='Result Cleaned API')
                    else:
                        df.to_excel(writer, sheet_name='Result Cleaned LLM')
                    df_clean.to_excel(writer, sheet_name='Result Cleaned')
                    dataframe.to_excel(writer, sheet_name='Original')
                    writer.close()

                st.download_button(
                    label="Save Result",
                    data=buffer,
                    file_name=f"{uploaded_file.name.replace('.pdf', '')}_{chunk_option}_{model.split('-')[0]}_{model_val.split('-')[0]}.xlsx",
                    mime='application/vnd.ms-excel'
                )