app.py

import google.generativeai as genai
import fitz  # PyMuPDF for PDF text extraction
import streamlit as st
import spacy
from transformers import AutoTokenizer, AutoModelForTokenClassification, pipeline
from docx import Document
import re
import dateparser
from datetime import datetime
import os

# Load SpaCy model for dependency parsing
nlp_spacy = spacy.load('en_core_web_sm')

# Load the NER model
tokenizer = AutoTokenizer.from_pretrained("Babelscape/wikineural-multilingual-ner")
model = AutoModelForTokenClassification.from_pretrained("Babelscape/wikineural-multilingual-ner")
nlp_ner = pipeline('ner', model=model, tokenizer=tokenizer, aggregation_strategy="simple")

# Function to authenticate with Gemini API
def authenticate_gemini():
    api_key = os.environ.get("GOOGLE_GEMINI_API_KEY")
    if not api_key:
        st.error("Google Gemini API key not found. Please set it in the Hugging Face Spaces secrets.")
        return None
    try:
        genai.configure(api_key=api_key)
        model = genai.GenerativeModel(model_name="gemini-pro")
        st.success("Gemini API successfully configured.")
        return model
    except Exception as e:
        st.error(f"Error configuring Gemini API: {e}")
        return None

# Function to filter and refine extracted ORG entities
def refine_org_entities(entities):
    refined_entities = set()
    company_suffixes = ['Inc', 'LLC', 'Corporation', 'Corp', 'Ltd', 'Co', 'GmbH', 'S.A.']

    for entity in entities:
        if any(entity.endswith(suffix) for suffix in company_suffixes):
            refined_entities.add(entity)
        elif re.match(r'([A-Z][a-z]+)\s([A-Z][a-z]+)', entity):
            refined_entities.add(entity)
    return list(refined_entities)

# Function to extract ORG entities using NER
def extract_orgs(text):
    ner_results = nlp_ner(text)
    orgs = set()
    for entity in ner_results:
        if entity['entity_group'] == 'ORG':
            orgs.add(entity['word'])

    return refine_org_entities(orgs)

# Extract text from PDF
def extract_text_from_pdf(pdf_file):
    doc = fitz.open(stream=pdf_file.read(), filetype="pdf")
    text = ""
    for page_num in range(doc.page_count):
        page = doc.load_page(page_num)
        text += page.get_text()
    return text

# Extract text from DOCX
def extract_text_from_doc(doc_file):
    doc = Document(doc_file)
    text = '\n'.join([para.text for para in doc.paragraphs])
    return text

# Summary generation function
def generate_summary(text, model):
    prompt = f"Can you summarize the following document in 100 words?\n\n{text}"
    try:
        response = model.generate_content(prompt)
        return response.text
    except Exception as e:
        return f"Error generating summary: {str(e)}"

# Additional resume parsing functions
def extract_experience(doc):
    experience = 0
    for ent in doc.ents:
        if ent.label_ == "DATE":
            date = dateparser.parse(ent.text)
            if date:
                experience = max(experience, datetime.now().year - date.year)
    return experience

def extract_phone(text):
    phone_patterns = [
        r'\b(?:\+?1[-.\s]?)?(?:\(\d{3}\)|\d{3})[-.\s]?\d{3}[-.\s]?\d{4}\b',
        r'\b\d{3}[-.\s]?\d{3}[-.\s]?\d{4}\b'
    ]
    for pattern in phone_patterns:
        match = re.search(pattern, text)
        if match:
            return match.group()
    return "Not found"

def extract_email(text):
    email_pattern = r'\b[A-Za-z0-9._%+-]+@[A-Za-z0-9.-]+\.[A-Z|a-z]{2,}\b'
    match = re.search(email_pattern, text)
    return match.group() if match else "Not found"

def extract_colleges(doc):
    colleges = set()
    edu_keywords = ["university", "college", "institute", "school"]
    for ent in doc.ents:
        if ent.label_ == "ORG" and any(keyword in ent.text.lower() for keyword in edu_keywords):
            colleges.add(ent.text)
    return list(colleges)

def extract_linkedin(text):
    linkedin_pattern = r'(?:https?:)?\/\/(?:[\w]+\.)?linkedin\.com\/in\/[A-z0-9_-]+\/?'
    match = re.search(linkedin_pattern, text)
    return match.group() if match else "Not found"

# Main function to process the resume and return the analysis
def main():
    st.title("Comprehensive Resume Analyzer")
    st.write("Upload a resume to extract information, generate a summary, and analyze details.")

    # Authenticate with Gemini API
    model = authenticate_gemini()
    if model is None:
        return

    # File uploader for resume input
    uploaded_file = st.file_uploader("Choose a PDF or DOCX file", type=["pdf", "docx", "doc"])

    if uploaded_file is not None:
        try:
            # Extract text from the uploaded resume
            file_ext = uploaded_file.name.split('.')[-1].lower()
            if file_ext == 'pdf':
                resume_text = extract_text_from_pdf(uploaded_file)
            elif file_ext in ['docx', 'doc']:
                resume_text = extract_text_from_doc(uploaded_file)
            else:
                st.error("Unsupported file format.")
                return

            if not resume_text.strip():
                st.error("The resume appears to be empty.")
                return

            # Process the resume
            doc = nlp_spacy(resume_text)

            # Extract information
            companies = extract_orgs(resume_text)
            summary = generate_summary(resume_text, model)
            experience = extract_experience(doc)
            phone = extract_phone(resume_text)
            email = extract_email(resume_text)
            colleges = extract_colleges(doc)
            linkedin = extract_linkedin(resume_text)

            # Display results
            st.subheader("Extracted Information")
            st.write(f"*Years of Experience:* {experience}")
            st.write("*Companies Worked For:*")
            st.write(", ".join(companies))
            st.write(f"*Phone Number:* {phone}")
            st.write(f"*Email ID:* {email}")
            st.write("*Colleges Attended:*")
            st.write(", ".join(colleges))
            st.write(f"*LinkedIn ID:* {linkedin}")

            st.subheader("Generated Summary")
            st.write(summary)

        except Exception as e:
            st.error(f"Error during processing: {e}")

if __name__ == "__main__":
    main()