Update app.py

app.py

@@ -163,7 +163,11 @@ def initialize_data_and_model():
         
         def extract_skills_llm(text: str) -> list[str]:
             if not isinstance(text, str) or len(text.strip()) < 20 or not LLM_PIPELINE: return []
-            prompt = f"""Instruct: You are an expert technical recruiter... [omitted for brevity]"""
+            prompt = f"""
+Instruct: You are an expert technical recruiter. Extract the key skills from the job description text. List technical and soft skills as a comma-separated string.
+[Example 1] Text: "Requires 3+ years of experience in cloud infrastructure. Must be proficient in AWS, particularly EC2 and S3. Experience with Terraform for IaC is a plus." Extracted Skills: cloud infrastructure, aws, ec2, s3, terraform, infrastructure as code
+[Example 2] Text: "Seeking a team lead with strong project management abilities. Must communicate effectively with stakeholders and manage timelines using Agile methodologies like Scrum." Extracted Skills: project management, leadership, stakeholder communication, agile, scrum
+[Actual Task] Text: "{text}" Extracted Skills:"""
             try:
                 response = LLM_PIPELINE(prompt, max_new_tokens=150, do_sample=False, temperature=0.1, return_full_text=False)
                 skills_text = response[0]['generated_text'].strip()
@@ -198,7 +202,7 @@ def initialize_data_and_model():
         def expand_skills_with_llm(job_title: str, existing_skills: list) -> list:
             if not LLM_PIPELINE or not job_title or not existing_skills: return []
             skills_to_add = 6 - len(existing_skills)
-            prompt = f"""Instruct: A job has the title "{job_title}"... [omitted for brevity]"""
+            prompt = f"""Instruct: A job has the title "{job_title}" and requires the skills: {', '.join(existing_skills)}. Based on this, what are {skills_to_add} additional, closely related skills typically required for such a role? List only the new skills, separated by commas. Do not repeat skills from the original list. Additional Skills:"""
             try:
                 response = LLM_PIPELINE(prompt, max_new_tokens=50, do_sample=True, temperature=0.5, return_full_text=False)
                 new_skills_text = response[0]['generated_text'].strip()
@@ -270,11 +274,13 @@ def get_job_matches(dream_job: str, top_n: int):
     dropdown_options = [(f"{i+1}. {row['job_title']} - {row['company']}", row.name) for i, row in display_df.iterrows()]
     dropdown_value = dropdown_options[0][1] if dropdown_options else None
     
-    return status, emb_matches, table_to_show, gr.update(choices=dropdown_options, value=dropdown_value, visible=True), gr.update(visible=True)
+    # When initially finding matches, only df_output is used. skill_jobs_output is cleared/hidden.
+    return status, emb_matches, table_to_show, gr.update(choices=dropdown_options, value=dropdown_value, visible=True), gr.update(visible=True), pd.DataFrame(), gr.update(visible=False)
 
 def analyze_skills(dream_job, initial_matches_df, skills_text, top_n):
     user_skills = [_norm_skill_token(s) for s in skills_text.split(',') if _norm_skill_token(s)]
     if not user_skills:
+        # If skills are cleared, just show the original table without skill scores and hide the second table
         table1_df = pd.DataFrame(initial_matches_df).head(top_n)
         table1_to_show = table1_df[['job_title', 'company', 'Similarity Score']]
         table1_to_show['Similarity Score'] = table1_to_show['Similarity Score'].map('{:.2%}'.format)
@@ -282,27 +288,28 @@ def analyze_skills(dream_job, initial_matches_df, skills_text, top_n):
 
     status = "Analyzing skills and finding new job matches..."
     
-    # --- LOGIC FOR TABLE 1 (Reranked Initial Jobs) ---
-    reranked_initial_jobs = pd.DataFrame(initial_matches_df)
+    # --- LOGIC FOR TOP TABLE (Reranked Initial Jobs) ---
+    reranked_initial_jobs = pd.DataFrame(initial_matches_df) # Ensure it's a DataFrame
     reranked_initial_jobs['Skill Match Score'] = reranked_initial_jobs['Skills'].apply(lambda js: calculate_skill_match_score(user_skills, js))
     reranked_initial_jobs = reranked_initial_jobs.sort_values(by='Skill Match Score', ascending=False)
     
-    table1_df = reranked_initial_jobs.head(top_n)[['job_title', 'company', 'Similarity Score', 'Skill Match Score']]
-    table1_df['Similarity Score'] = table1_df['Similarity Score'].map('{:.2%}'.format)
-    table1_df['Skill Match Score'] = table1_df['Skill Match Score'].map('{:.2%}'.format)
+    top_table_df = reranked_initial_jobs.head(top_n)[['job_title', 'company', 'Similarity Score', 'Skill Match Score']]
+    top_table_df['Similarity Score'] = top_table_df['Similarity Score'].map('{:.2%}'.format)
+    top_table_df['Skill Match Score'] = top_table_df['Skill Match Score'].map('{:.2%}'.format)
 
-    # --- LOGIC FOR TABLE 2 (New Skill-Based Jobs) ---
+    # --- LOGIC FOR BOTTOM TABLE (New Skill-Based Jobs) ---
     combined_query = dream_job + ". My current skills are: " + skills_text
     newly_found_jobs = find_job_matches(combined_query, top_k=top_n)
     newly_found_jobs['Skill Match Score'] = newly_found_jobs['Skills'].apply(lambda js: calculate_skill_match_score(user_skills, js))
     
-    table2_df = newly_found_jobs[['job_title', 'company', 'Similarity Score', 'Skill Match Score']]
-    table2_df['Similarity Score'] = table2_df['Similarity Score'].map('{:.2%}'.format)
-    table2_df['Skill Match Score'] = table2_df['Skill Match Score'].map('{:.2%}'.format)
+    bottom_table_df = newly_found_jobs[['job_title', 'company', 'Similarity Score', 'Skill Match Score']]
+    bottom_table_df['Similarity Score'] = bottom_table_df['Similarity Score'].map('{:.2%}'.format)
+    bottom_table_df['Skill Match Score'] = bottom_table_df['Skill Match Score'].map('{:.2%}'.format)
 
     status = f"Re-ranked initial jobs and found new jobs for your skills."
-    # Return in the correct order for the UI components
-    return status, table1_df, table2_df, gr.update(visible=True)
+    
+    # Corrected return order: top_table_df for df_output, bottom_table_df for skill_jobs_output
+    return status, top_table_df, bottom_table_df, gr.update(visible=True)
 
 def find_matches_and_rank_with_check(dream_job: str, top_n: int):
     if not dream_job:
@@ -314,11 +321,13 @@ def find_matches_and_rank_with_check(dream_job: str, top_n: int):
         alert_message = f"<b><span style='color: #F87171;'>⚠️ Possible Spelling Error:</span></b> Unrecognized: {word_list_html}."
         return "Status: Awaiting confirmation.", None, pd.DataFrame(), gr.update(visible=False), gr.update(visible=False), gr.update(value=alert_message, visible=True), gr.update(visible=True), pd.DataFrame(), gr.update(visible=False)
     
-    status, emb_matches, table_to_show, dropdown_update, accordion_update = get_job_matches(dream_job, top_n)
+    status, emb_matches, table_to_show, dropdown_update, accordion_update, _, _ = get_job_matches(dream_job, top_n) # Adjusted to unpack the 2 extra outputs
+    # Hide the second table on a new search
     return status, emb_matches, table_to_show, dropdown_update, accordion_update, gr.update(visible=False), gr.update(visible=False), pd.DataFrame(), gr.update(visible=False)
 
+
 def find_matches_and_rank_anyway(dream_job: str, top_n: int):
-    status, emb_matches, table_to_show, dropdown_update, accordion_update = get_job_matches(dream_job, top_n)
+    status, emb_matches, table_to_show, dropdown_update, accordion_update, _, _ = get_job_matches(dream_job, top_n) # Adjusted to unpack the 2 extra outputs
     return status, emb_matches, table_to_show, dropdown_update, accordion_update, gr.update(visible=False), gr.update(visible=False), pd.DataFrame(), gr.update(visible=False)
 
 def on_select_job(job_id, skills_text):
@@ -416,12 +425,28 @@ with gr.Blocks(theme=gr.themes.Soft()) as ui:
         learning_plan_output = gr.HTML(label="Learning Plan")
         load_more_btn = gr.Button("Load More Skills", visible=False)
     
-    search_btn.click(fn=find_matches_and_rank_with_check, inputs=[dream_text, topk_slider], outputs=[status_text, initial_matches_state, df_output, job_selector, details_accordion, spelling_alert, spelling_row, skill_jobs_output, skill_jobs_col])
-    search_anyway_btn.click(fn=find_matches_and_rank_anyway, inputs=[dream_text, topk_slider], outputs=[status_text, initial_matches_state, df_output, job_selector, details_accordion, spelling_alert, spelling_row, skill_jobs_output, skill_jobs_col])
-    analyze_btn.click(fn=analyze_skills, inputs=[dream_text, initial_matches_state, skills_text, topk_slider], outputs=[status_text, df_output, skill_jobs_output, skill_jobs_col])
-    retype_btn.click(lambda: ("Status: Ready for you to retype.", None, pd.DataFrame(), gr.update(visible=False), gr.update(visible=False), gr.update(visible=False), gr.update(visible=False), pd.DataFrame(), gr.update(visible=False)), outputs=[status_text, initial_matches_state, df_output, job_selector, details_accordion, spelling_alert, spelling_row, skill_jobs_output, skill_jobs_col])
-    reset_btn.click(fn=on_reset, outputs=[dream_text, topk_slider, skills_text, df_output, initial_matches_state, job_selector, details_accordion, status_text, job_details_markdown, duties_markdown, qualifications_markdown, description_markdown, spelling_alert, spelling_row, missing_skills_state, skills_offset_state, load_more_btn, skill_jobs_output, skill_jobs_col], queue=False)
-    job_selector.change(fn=on_select_job, inputs=[job_selector, skills_text], outputs=[job_details_markdown, duties_markdown, qualifications_markdown, description_markdown, learning_plan_output, details_accordion, missing_skills_state, skills_offset_state, load_more_btn])
-    load_more_btn.click(fn=load_more_skills, inputs=[missing_skills_state, skills_offset_state], outputs=[learning_plan_output, skills_offset_state, load_more_btn])
+    # Updated 'search_btn' and 'search_anyway_btn' to correctly unpack outputs from get_job_matches
+    search_btn.click(fn=find_matches_and_rank_with_check, inputs=[dream_text, topk_slider], 
+                     outputs=[status_text, initial_matches_state, df_output, job_selector, details_accordion, 
+                              spelling_alert, spelling_row, skill_jobs_output, skill_jobs_col])
+    search_anyway_btn.click(fn=find_matches_and_rank_anyway, inputs=[dream_text, topk_slider], 
+                            outputs=[status_text, initial_matches_state, df_output, job_selector, details_accordion, 
+                                     spelling_alert, spelling_row, skill_jobs_output, skill_jobs_col])
+    
+    # The analyze_btn output order is now correct: top_table_df to df_output, bottom_table_df to skill_jobs_output
+    analyze_btn.click(fn=analyze_skills, inputs=[dream_text, initial_matches_state, skills_text, topk_slider], 
+                      outputs=[status_text, df_output, skill_jobs_output, skill_jobs_col])
+    
+    retype_btn.click(lambda: ("Status: Ready for you to retype.", None, pd.DataFrame(), gr.update(visible=False), gr.update(visible=False), gr.update(visible=False), gr.update(visible=False), pd.DataFrame(), gr.update(visible=False)), 
+                     outputs=[status_text, initial_matches_state, df_output, job_selector, details_accordion, spelling_alert, spelling_row, skill_jobs_output, skill_jobs_col])
+    reset_btn.click(fn=on_reset, 
+                    outputs=[dream_text, topk_slider, skills_text, df_output, initial_matches_state, job_selector, details_accordion, status_text, 
+                             job_details_markdown, duties_markdown, qualifications_markdown, description_markdown, spelling_alert, spelling_row, 
+                             missing_skills_state, skills_offset_state, load_more_btn, skill_jobs_output, skill_jobs_col], queue=False)
+    job_selector.change(fn=on_select_job, inputs=[job_selector, skills_text], 
+                        outputs=[job_details_markdown, duties_markdown, qualifications_markdown, description_markdown, learning_plan_output, 
+                                 details_accordion, missing_skills_state, skills_offset_state, load_more_btn])
+    load_more_btn.click(fn=load_more_skills, inputs=[missing_skills_state, skills_offset_state], 
+                        outputs=[learning_plan_output, skills_offset_state, load_more_btn])
 
 ui.launch()