app.py

import gradio as gr
import os
import shutil
from recommender import rank_resumes, summarize_resume_flan, extract_applicant_name
from docx import Document
import pdfplumber

UPLOAD_FOLDER = "uploads"
os.makedirs(UPLOAD_FOLDER, exist_ok=True)


def process_resumes(job_description, uploaded_files):
    if not job_description.strip():
        return "Please provide a job description.", None

    resume_texts = []

    for uploaded_file in uploaded_files:
        filename = os.path.basename(uploaded_file.name)
        ext = filename.lower().split(".")[-1]

        # Copying the file from Gradio temp folder to uploads folder
        file_path = os.path.join(UPLOAD_FOLDER, filename)
        shutil.copy(uploaded_file.name, file_path)

        if ext == "txt":
            with open(file_path, "r", encoding="utf-8") as f:
                text = f.read()

        elif ext == "pdf":
            try:
                with pdfplumber.open(file_path) as pdf:
                    pages = [page.extract_text() for page in pdf.pages if page.extract_text() is not None]
                    if not pages:
                        return f"No extractable text found in PDF: {filename}. Is it scanned or image-only?", None
                    text = "\n".join(pages)
            except Exception as e:
                return f"Failed to process PDF {filename}: {str(e)}", None

        elif ext == "docx":
            try:
                doc = Document(file_path)
                text = "\n".join([p.text for p in doc.paragraphs])
            except Exception as e:
                return f"Failed to process DOCX {filename}: {str(e)}", None

        else:
            return f"Unsupported file format: {filename}", None

        resume_texts.append((filename, text))

    # Rank resumes and generate summaries
    results = rank_resumes(job_description, resume_texts)

    # Attach filename to each candidate for display
    for i, candidate in enumerate(results):
        candidate["filename"] = resume_texts[i][0]

    for candidate in results:
        candidate["summary"] = summarize_resume_flan(candidate["text"], job_description)

    table_data = [
        [
            candidate.get("applicant_name", extract_applicant_name(candidate["text"], candidate.get("filename", "Unknown"))),
            candidate.get("filename", "Unknown"),
            f"{candidate['score']:.4f}",
            candidate["summary"]
        ] for candidate in results
    ]

    return "", table_data


with gr.Blocks() as demo:
    gr.Markdown("## Candidate Recommendation Engine")
    with gr.Row():
        job_desc = gr.Textbox(label="Job Description", lines=10, placeholder="Paste job description here...")

    resumes = gr.Files(label="Upload Resumes (.txt, .pdf, .docx)", file_types=[".txt", ".pdf", ".docx"])

    btn = gr.Button("Rank Candidates")

    msg = gr.Markdown()
    output_table = gr.Dataframe(headers=["Candidate", "File Name", "Similarity Score", "Why a Good Fit"], wrap=True)

    btn.click(process_resumes, inputs=[job_desc, resumes], outputs=[msg, output_table])


if __name__ == "__main__":
    demo.launch()