feat: support multiple repos

app.py

@@ -7,29 +7,28 @@ import shutil
 from openai import OpenAI
 import dotenv
 
-def generate_prompt(repo_url, role, seniority, assignment_details, repo_data, focus_areas=""):
-    """Generates the AI analysis prompt with repo details."""
-    return f"""
-    You are an AI expert in evaluating software engineering candidates based on their GitHub repositories. Your goal is to assess the quality, organization, and best practices of the submitted code repository. Below is the candidate's information:
+def generate_prompt(repos_data, role, seniority, assignment_details, focus_areas=""):
+    """Generates the AI analysis prompt with multiple repo details."""
+    prompt = f"""
+    You are an AI expert in evaluating software engineering candidates based on their GitHub repositories. Your goal is to assess the quality, organization, and best practices of the submitted code repositories. Below is the candidate's information:
 
     ### Candidate Information:
     - **Role Applied For**: {role}
     - **Seniority Level**: {seniority}
-    - **GitHub Repository URL**: {repo_url}
 
     ### Guidelines:
     - Focus on the candidate's ability to write clean, efficient, and maintainable code.
-    - Take into consideration the seniority level when evaluating the code.
-    - Focus on the assignment details provided. What was requested of him and how he responded to it.
+    - Consider the seniority level when evaluating the code.
+    - Consider the assignment details and how the candidate responded.
 
     ### Assignment Details:
     {assignment_details}
 
     ### Repository Analysis:
-    You are provided with the repository's cloned structure and its contents. Analyze the following aspects:
-    {repo_data}
+    Below are the repositories submitted by the candidate:
+    {repos_data}
 
-    ### Additional Evaluation Criteria:
+    ### Evaluation Criteria:
     1. **Code Organization & Architecture**
     2. **Code Quality & Best Practices**
     3. **Language Proficiency & Best Practices**
@@ -39,15 +38,15 @@ def generate_prompt(repo_url, role, seniority, assignment_details, repo_data, fo
     {focus_areas}
 
     ### **Final Score Calculation**
-    - Score the repository **out of 100** based on the criteria above.
+    - Score each repository **out of 100** and provide an overall weighted score.
     - Justify the **score** by explaining the candidate’s strengths and weaknesses.
 
     ### **Expected Output:**
-    - **Strengths**: What is done well?
-    - **Weaknesses**: What needs improvement?
-    - **Final Score (0-100)**: Provide a numeric score with an explanation.
-    - **Summary**: Briefly summarize the candidate’s proficiency based on this analysis.
+    - **Per Repository Analysis**: A detailed breakdown of strengths and weaknesses.
+    - **Overall Candidate Summary**: A final evaluation of their coding skills across all repositories.
+    - **Final Score (0-100)**: A numeric score with justification.
     """
+    return prompt
 
 # Load environment variables
 dotenv.load_dotenv()
@@ -61,49 +60,57 @@ def authenticate(password):
         return "❌ Incorrect password! Access denied.", None
     return None, "✅ Access granted! You may proceed."
 
-def analyze_repo(repo_url, role, seniority, assignment_details, focus_areas, password):
+def analyze_repos(repo_urls, role, seniority, assignment_details, focus_areas, password):
     auth_error, auth_success = authenticate(password)
     if auth_error:
         return auth_error, gr.update(visible=False)  # If incorrect password, return error
         
-    """Clone and analyze a GitHub repository with a loading state."""
-    if not repo_url.startswith("https://github.com/"):
-        return "❌ Invalid GitHub URL!", gr.update(visible=False)
+    """Clone and analyze multiple GitHub repositories."""
+    repo_urls = [url.strip() for url in repo_urls.split(",") if url.strip()]
+    
+    if not all(url.startswith("https://github.com/") for url in repo_urls):
+        return "❌ One or more URLs are invalid!", gr.update(visible=False)
 
-    # Extract repo details
-    repo_name = repo_url.split("/")[-1]
     temp_dir = tempfile.mkdtemp()
+    repos_data = ""
+    total_files = 0
 
     try:
-        progress = gr.update(value="🔄 Cloning repository...", visible=True)
-        
-        # Clone the repo
-        repo_path = os.path.join(temp_dir, repo_name)
-        git.Repo.clone_from(repo_url, repo_path)
+        progress = gr.update(value="🔄 Cloning repositories...", visible=True)
         
-        progress = gr.update(value="📂 Analyzing repository structure...", visible=True)
-        
-        # Gather repository file structure and contents
-        repo_data = ""
-        file_count = 0  # Initialize file counter
-
-        for root, _, filenames in os.walk(repo_path):
-            for file in filenames:
-                file_count += 1  # Increment file counter
-                file_path = os.path.join(root, file)
-
-                try:
-                    with open(file_path, "r", encoding="utf-8") as f:
-                        file_content = f.read()
-                    repo_data += f"\n**File {file_count}:** {file_path.replace(repo_path, '')}\n```\n{file_content[:1000]}\n```\n"
-                except Exception:
-                    repo_data += f"\n**File {file_count}:** {file_path.replace(repo_path, '')} (⚠️ Cannot read binary file)\n"
+        for repo_url in repo_urls:
+            repo_name = repo_url.split("/")[-1]
+            repo_path = os.path.join(temp_dir, repo_name)
+            
+            try:
+                git.Repo.clone_from(repo_url, repo_path)
+            except Exception as e:
+                repos_data += f"\n❌ Failed to clone `{repo_name}`: {str(e)}\n"
+                continue
+
+            repo_data = f"\n📂 **Repository: {repo_name}**\n"
+            file_count = 0  
+
+            for root, _, filenames in os.walk(repo_path):
+                for file in filenames:
+                    file_count += 1  
+                    file_path = os.path.join(root, file)
+
+                    try:
+                        with open(file_path, "r", encoding="utf-8") as f:
+                            file_content = f.read()
+                        repo_data += f"\n**File {file_count}:** {file_path.replace(repo_path, '')}\n```\n{file_content[:1000]}\n```\n"
+                    except Exception:
+                        repo_data += f"\n**File {file_count}:** {file_path.replace(repo_path, '')} (⚠️ Cannot read binary file)\n"
+
+            repos_data += repo_data
+            total_files += file_count
 
         progress = gr.update(value="🤖 Sending data to AI for evaluation...", visible=True)
 
         # AI-based evaluation
         evaluation = f"✅ **Evaluation for Role: {role}**\n\n"
-        evaluation += f"📂 Repository `{repo_name}` has `{file_count}` files.\n"
+        evaluation += f"📂 `{len(repo_urls)}` repositories analyzed, containing `{total_files}` files.\n"
         evaluation += f"💡 Key focus areas: {focus_areas}\n\n"
         evaluation += "**🔍 Code Quality Analysis:**\n"
         
@@ -112,7 +119,7 @@ def analyze_repo(repo_url, role, seniority, assignment_details, focus_areas, pas
             messages=[
                 {
                     "role": "user",
-                    "content": generate_prompt(repo_url, role, seniority, assignment_details, repo_data, focus_areas)
+                    "content": generate_prompt(repos_data, role, seniority, assignment_details, focus_areas)
                 }
             ]
         )
@@ -123,7 +130,7 @@ def analyze_repo(repo_url, role, seniority, assignment_details, focus_areas, pas
         return evaluation, progress
 
     except Exception as e:
-        return f"❌ Error analyzing repository: {str(e)}", gr.update(visible=False)
+        return f"❌ Error analyzing repositories: {str(e)}", gr.update(visible=False)
     
     finally:
         shutil.rmtree(temp_dir)  # Cleanup
@@ -138,10 +145,9 @@ with gr.Blocks() as app:
         ["Junior", "Mid", "Senior"],
         label="Seniority Level",
         value="Mid"
-
     )
     assignment_details = gr.Textbox(label="Assignment Details", lines=8)
-    repo_url = gr.Textbox(label="GitHub Repository URL")
+    repo_urls = gr.Textbox(label="GitHub Repository URLs (comma-separated)")
     focus_areas = gr.Textbox(label="Optional Focus Areas (e.g., Clean Code, Performance)")
 
     output = gr.Markdown()
@@ -150,8 +156,8 @@ with gr.Blocks() as app:
     submit_btn = gr.Button("🔍 Evaluate")
 
     submit_btn.click(
-        fn=analyze_repo,
-        inputs=[repo_url, role, seniority, assignment_details, focus_areas, password],
+        fn=analyze_repos,
+        inputs=[repo_urls, role, seniority, assignment_details, focus_areas, password],
         outputs=[output, progress]
     )