update app.py

.gitignore

@@ -0,0 +1,2 @@
+hf-env
+.gradio

app.py

@@ -1,19 +1,78 @@
-import streamlit as st
+import gradio as gr
 from transformers import pipeline
+import PyPDF2
 
 # Load the token classification pipeline
 model_name = "jjzha/jobbert_knowledge_extraction"
 pipe = pipeline("token-classification", model=model_name)
 
-# Streamlit UI
-st.title("Token Classification with Hugging Face")
-text_input = st.text_area("Enter text for token classification:")
+# Function to extract and highlight key skills/words from the job posting
+def extract_keywords_with_highlights(job_posting_text):
+    results = pipe(job_posting_text)
 
-if text_input:
-    # Perform token classification
-    results = pipe(text_input)
-
-    # Display the results
-    st.write("Token Classification Results:")
+    # Fix the `##` issue by reconstructing full words
+    reconstructed_text = ""
+    highlighted_words = set()
+    previous_end = 0
     for result in results:
-        st.write(f"Entity: {result['entity']}, Word: {result['word']}, Score: {result['score']:.2f}")
+        start, end, word = result['start'], result['end'], result['word']
+        # Remove `##` for subwords
+        clean_word = word.replace("##", "")
+        highlighted_words.add(clean_word.lower())
+        # Add text before the current word
+        reconstructed_text += job_posting_text[previous_end:start]
+        # Highlight the cleaned word
+        reconstructed_text += (
+            f'<span style="background-color:yellow; font-weight:bold;" '
+            f'title="Entity: {result["entity"]} (Score: {result["score"]:.2f})">'
+            f"{clean_word}</span>"
+        )
+        previous_end = end
+    # Add the remaining text
+    reconstructed_text += job_posting_text[previous_end:]
+    # Replace newline characters with <br> to preserve line breaks
+    reconstructed_text = reconstructed_text.replace("\n", "<br>")
+
+    return (
+        f'<div style="font-family:Arial, sans-serif; line-height:1.5;">{reconstructed_text}</div>',
+        highlighted_words,
+    )
+
+# Function to check if highlighted words are in the resume
+def check_keywords_in_resume(resume_file_path, job_posting_text):
+    # Extract text from the uploaded PDF resume
+    with open(resume_file_path, "rb") as file:
+        pdf_reader = PyPDF2.PdfReader(file)
+        resume_text = " ".join(page.extract_text() for page in pdf_reader.pages)
+
+    # Extract highlighted keywords from the job posting
+    highlighted_html, highlighted_words = extract_keywords_with_highlights(job_posting_text)
+
+    # Check if each highlighted word is in the resume
+    resume_words = set(resume_text.lower().split())
+    matched_words = highlighted_words.intersection(resume_words)
+    missing_words = highlighted_words - matched_words
+
+    # Prepare a summary
+    matched_summary = f"Matched Keywords: {', '.join(matched_words)}"
+    missing_summary = f"Missing Keywords: {', '.join(missing_words)}"
+    return highlighted_html, matched_summary, missing_summary
+
+# Set up Gradio interface
+interface = gr.Interface(
+    fn=check_keywords_in_resume,
+    inputs=[
+        gr.File(label="Upload Resume PDF", type="filepath"),
+        gr.Textbox(label="Enter Job Posting Text", lines=30, placeholder="Paste job posting text here..."),
+    ],
+    outputs=[
+        gr.HTML(label="Highlighted Key Skills/Words in Job Posting"),
+        gr.Textbox(label="Matched Keywords"),
+        gr.Textbox(label="Missing Keywords"),
+    ],
+    title="Resume vs Job Posting Skill Match with Highlights",
+    description="Upload your resume and enter a job posting. The app will highlight key skills from the job posting and check if they are present in your resume.",
+)
+
+# Launch the Gradio app
+interface.launch()