app.py

import streamlit as st

from file_reader import ResumeReader
from text_extractor import ResumeExtractor

st.set_page_config(page_title="Read Your Resume", page_icon = 'img/app_icon.png')

#@st.cache(allow_output_mutation=True, suppress_st_warning=True)
def load_model():
    model = ResumeExtractor(ner_model_path='./resume_ner_model/model-best')
    return model

header_container = st.empty()

if 'loaded' not in st.session_state:
    with header_container.container():
        header_container.info('Starting to download model from cloud, please wait...')
        extractor = load_model()
        st.session_state['loaded'] = extractor
else:
    extractor = st.session_state['loaded']


#Ready to go
st.success("Model is ready to go!")
with header_container.container():
    header_container.header("Let's Read Your Resume")

resume_file = st.file_uploader("Choose your resume", accept_multiple_files=False, type=['pdf', 'docx'])
reader = ResumeReader()

if resume_file:
    with st.spinner("Reading resumes..."):
        resume_content = reader.read_text_from_file(resume_file)
        resume_summary = extractor.get_summary_from_text(resume_content)

    st.markdown("### Resume summary")
    st.write(resume_summary)

    #Render html entities on resume
    st.markdown("### Coloring resume")
    html = extractor.render_html_entities(resume_content)
    st.markdown(html, unsafe_allow_html=True)