import streamlit as st
import PyPDF2
import docx
import spacy

# Load NLP model
@st.cache_resource
def load_model():
    return spacy.load("en_core_web_sm")

nlp = load_model()

# Function to extract text
def extract_text(file):
    if file.type == "application/pdf":
        reader = PyPDF2.PdfFileReader(file)
        return "".join([reader.getPage(i).extract_text() for i in range(reader.numPages)])
    elif file.type == "application/vnd.openxmlformats-officedocument.wordprocessingml.document":
        return "\n".join([para.text for para in docx.Document(file).paragraphs])

# Function to extract and display information
def analyze_resume(text, job_desc):
    doc = nlp(text)
    extracted_info = {"skills": [], "education": [], "experience": []}
    
    for ent in doc.ents:
        if ent.label_ in ["ORG", "GPE"]:
            extracted_info["education"].append(ent.text)
        elif ent.label_ == "DATE":
            extracted_info["experience"].append(ent.text)

    match_score = sum(1 for token in nlp(job_desc) if token.text in text) / len(job_desc.split()) * 100
    return extracted_info, match_score

# Streamlit interface
st.title("Resume Scanner with NLP")
uploaded_file = st.file_uploader("Upload a resume (PDF or DOCX)", type=["pdf", "docx"])
job_description = st.text_area("Paste the job description here")

if uploaded_file and job_description:
    resume_text = extract_text(uploaded_file)
    extracted_info, match_score = analyze_resume(resume_text, job_description)
    
    st.subheader("Extracted Information:")
    st.write(extracted_info)
    st.subheader(f"Match Score: {match_score:.2f}%")