Update app.py

app.py

@@ -1,14 +1,32 @@
 import gradio as gr
 import os
 from openai import OpenAI
+import PyPDF2
+from docx import Document
 
 # ==========================================
-# 1. CORE ENGINE (WITH AUTO-FALLBACK & LANGUAGE LOCK)
+# CORE LOGIC (Updated functionality)
 # ==========================================
 def get_client():
     api_key = os.environ.get("GROQ_API_KEY")
     return OpenAI(api_key=api_key, base_url="https://api.groq.com/openai/v1")
 
+def extract_text(file_path):
+    if not file_path: return ""
+    text = ""
+    try:
+        if file_path.endswith('.pdf'):
+            with open(file_path, 'rb') as f:
+                reader = PyPDF2.PdfReader(f)
+                for page in reader.pages:
+                    content = page.extract_text()
+                    if content: text += content + "\n"
+        elif file_path.endswith('.docx'):
+            doc = Document(file_path)
+            for para in doc.paragraphs: text += para.text + "\n"
+    except: return "Could not parse file."
+    return text
+
 def transcribe_audio(audio_path):
     if audio_path is None: return ""
     try:
@@ -16,157 +34,74 @@ def transcribe_audio(audio_path):
         with open(audio_path, "rb") as audio_file:
             transcript = client.audio.transcriptions.create(model="whisper-large-v3", file=audio_file)
         return transcript.text
-    except Exception: return ""
+    except: return ""
 
-def stream_llm(system_role, user_text, audio_path=None, file_path=None):
-    voice_content = transcribe_audio(audio_path) if audio_path else ""
-    file_context = f"\n[Document Provided: {os.path.basename(file_path)}]" if file_path else ""
+def stream_llm(system_role, user_text, audio_path, file_path):
+    voice_content = transcribe_audio(audio_path)
+    file_content = extract_text(file_path)
+    combined_input = f"User Request: {user_text}\nVoice: {voice_content}\nDocument: {file_content}"
     
-    combined_input = f"User Request: {user_text}\nVoice: {voice_content}\n{file_context}".strip()
-    
-    if not combined_input and not file_path:
-        yield "### ⚠️ Attention Required\nPlease upload your CV in the Nexus tab or state your query."
-        return
-
-    # LIST OF MODELS TO PREVENT 429 ERRORS[cite: 1, 3]
-    models_to_try = ["llama-3.3-70b-versatile", "llama-3.1-8b-instant"]
+    # Updated System Rules
+    full_prompt = f"{system_role}\n\nRULES: 1. Respond in the exact language of user input. 2. Include 5-10 clickable links. 3. Use bold headings."
     
-    # UNIVERSAL LANGUAGE MIRRORING GUARD[cite: 1]
-    detail_cmd = """
-    \n\nCRITICAL SYSTEM RULES:
-    1. STRICT LANGUAGE MATCH: Identify the input language (English, Urdu, or Roman-Urdu) and respond ONLY in that exact language. NO SWITCHING.
-    2. HIGH-QUALITY VOICE: Professional, human, and empathetic tone.
-    3. EXTREME DETAIL: Structure with bold headings (###).
-    4. CLICKABLE RESOURCES: Include 5-10 specific clickable links.
-    """
-
     client = get_client()
-    success = False
-    
-    for model_name in models_to_try:
-        try:
-            response = client.chat.completions.create(
-                model=model_name,
-                messages=[
-                    {"role": "system", "content": system_role + detail_cmd},
-                    {"role": "user", "content": combined_input}
-                ],
-                stream=True,
-                temperature=0.75 
-            )
-            
-            partial_text = ""
-            for chunk in response:
-                if chunk.choices[0].delta.content is not None:
-                    partial_text += chunk.choices[0].delta.content
-                    yield partial_text
-            success = True
-            break 
-        except Exception as e:
-            if "429" in str(e): # Switch to next model if rate limit reached[cite: 3]
-                continue
-            else:
-                yield f"### ⚠️ Connection Error\n{str(e)}"
-                return
-
-    if not success:
-        yield "### ⚠️ Servers Overloaded\nPlease try again in a few minutes."
+    try:
+        stream = client.chat.completions.create(
+            model="llama-3.3-70b-versatile",
+            messages=[{"role": "system", "content": full_prompt}, {"role": "user", "content": combined_input}],
+            stream=True
+        )
+        for chunk in stream:
+            if chunk.choices[0].delta.content: yield chunk.choices[0].delta.content
+    except Exception as e: yield f"Error: {str(e)}"
 
 # ==========================================
-# 2. STRATEGIC SYSTEM PROMPTS
+# PROMPTS
 # ==========================================
-
-NEXUS_PROMPT = """You are an Elite Technical CV Auditor. Analyze the CV with extreme depth:
-1. CORE STRENGTHS. 2. CRITICAL WEAKNESSES. 3. TECH SKILL MATRIX. 4. NETWORKING STRATEGY. 5. IMPROVEMENT ROADMAP."""
-
-ZEN_PROMPT = """You are a Master Motivational Strategist. 
-STRICT LIMITATION: Your ONLY role is to encourage and motivate. Do not perform any other task.
-CORE MESSAGE: Remind the user that failure is natural. 
-MANDATORY EXAMPLES: Always include real-life examples of leaders like Steve Jobs, Elon Musk, and Jack Ma. Respond ONLY in user's input language."""
-
-ORBIT_PROMPT = """You are a Lead Hiring Strategist. Create a 6-Month Roadmap (Month 1-2: Skill Gaps, Month 3-4: Projects, Month 5: Branding, Month 6: Applications). Respond ONLY in user's input language."""
-
-# PATHFINDER: STRICT LANGUAGE REINFORCEMENT[cite: 1, 3]
-PATH_PROMPT = """You are a Global Market Scout. 
-STRICT RULE: Detect the user's language and respond EXCLUSIVELY in that language. If the user asks in English, respond in English. If they use Roman-Urdu, respond in Roman-Urdu. 
-TASK: Identify Industry and Companies where the user fits perfectly based on their skills."""
+NEXUS_PROMPT = "You are a CV Auditor. Analyze the CV (up to 6 pages), provide a skill matrix, and actionable feedback with 5-10 clickable links."
+ZEN_PROMPT = "You are a Mentor. Provide 7 distinct success stories of leaders who failed before succeeding. Focus on their journey. Include links."
+ORBIT_PROMPT = "You are a Career Strategist. Provide a detailed 6-month roadmap with weekly milestones and 10 clickable learning resources."
+PATH_PROMPT = "You are a Hiring Expert. Recommend companies based on skills, and provide a clear step-by-step application guide with links."
 
 # ==========================================
-# 3. UI DESIGN (REMAINING STABLE)
+# UI (Matching your Sidebar Layout)
 # ==========================================
-custom_css = """
-footer {display: none !important;}
-body { margin: 0; padding: 0; }
-.gradio-container { 
-    background: linear-gradient(135deg, #0a0a2e 0%, #1a1a4b 25%, #4b0082 50%, #800080 75%, #ff007f 100%) !important; 
-    background-attachment: fixed !important;
-    background-size: cover !important;
-    min-height: 100vh !important;
-}
-.glass-panel { background: rgba(10, 10, 30, 0.88) !important; backdrop-filter: blur(25px); border: 1px solid rgba(0, 242, 254, 0.3); border-radius: 20px; padding: 25px; height: fit-content; }
-.shimmer-title {
-    font-size: 4rem; font-weight: 900; color: #ffffff; text-transform: uppercase;
-    text-shadow: 0 0 10px #ff007f, 0 0 20px #ff007f, 0 0 40px #00f2fe;
-    animation: pulse-glow 2s ease-in-out infinite alternate; letter-spacing: 5px; margin-bottom: 0px;
-}
-@keyframes pulse-glow {
-    from { text-shadow: 0 0 10px #ff007f, 0 0 20px #ff007f, 0 0 40px #00f2fe; transform: scale(1); }
-    to { text-shadow: 0 0 20px #00f2fe, 0 0 40px #ff007f, 0 0 60px #00f2fe; transform: scale(1.02); }
-}
-.new-tagline { font-size: 1.1rem; color: #00f2fe; font-weight: bold; letter-spacing: 8px; text-transform: uppercase; margin-top: -10px; opacity: 0.9; }
-.gr-button-primary { background: linear-gradient(90deg, #00f2fe 0%, #ff007f 100%) !important; border: none !important; font-weight: bold !important; }
-input, textarea { background: rgba(0, 0, 0, 0.7) !important; color: white !important; border: 1px solid #00f2fe !important; }
-"""
-
-with gr.Blocks(css=custom_css) as demo:
-    gr.HTML("""
-        <div style="text-align: center; padding: 30px 10px;">
-            <h1 class="shimmer-title">AURAPATH AI</h1>
-            <p class="new-tagline">◈ ENGINEERING YOUR DESTINY ◈</p>
-        </div>
-    """)
-    
+with gr.Blocks() as demo:
     with gr.Tabs():
-        with gr.Tab("📄 Nexus"):
+        with gr.Tab("Nexus"):
             with gr.Row():
-                with gr.Column(elem_classes="glass-panel", scale=1):
+                with gr.Column(scale=1):
                     f1 = gr.File(label="Upload CV / Portfolio")
-                    t1 = gr.Textbox(label="TARGET GOAL", placeholder="e.g. AI Developer...", lines=4)
-                    a1 = gr.Audio(label="VOICE SCAN", sources=["microphone"], type="filepath")
-                    btn1 = gr.Button("INITIATE DEEP AUDIT", variant="primary")
-                with gr.Column(elem_classes="glass-panel", scale=1.5):
-                    out1 = gr.Markdown("### 🔍 Technical Audit & Skill Matrix")
+                    t1 = gr.Textbox(label="TARGET GOAL", placeholder="e.g. AI Developer...")
+                    a1 = gr.Audio(label="VOICE SCAN", type="filepath")
+                    btn1 = gr.Button("INITIATE DEEP AUDIT")
+                with gr.Column(scale=2):
+                    out1 = gr.Markdown(label="Technical Audit & Skill Matrix")
             btn1.click(stream_llm, [gr.State(NEXUS_PROMPT), t1, a1, f1], out1)
 
-        with gr.Tab("🧠 Zen"):
+        with gr.Tab("Zen"):
             with gr.Row():
-                with gr.Column(elem_classes="glass-panel", scale=1):
-                    gr.Markdown("#### 🔥 WARRIOR MODE\nPure motivation in your own language.")
-                    a2 = gr.Audio(label="VOICE LOG", sources=["microphone"], type="filepath")
-                    btn2 = gr.Button("ACTIVATE ZEN POWER", variant="primary")
-                with gr.Column(elem_classes="glass-panel", scale=1.5):
-                    out2 = gr.Markdown("### 🔥 High-Intensity Vision Protocol")
-            btn2.click(stream_llm, [gr.State(ZEN_PROMPT), t1, a2, f1], out2)
+                with gr.Column(scale=1):
+                    btn2 = gr.Button("GET MOTIVATED")
+                with gr.Column(scale=2):
+                    out2 = gr.Markdown()
+            btn2.click(stream_llm, [gr.State(ZEN_PROMPT), t1, a1, f1], out2)
 
-        with gr.Tab("📅 Orbit"):
+        with gr.Tab("Orbit"):
             with gr.Row():
-                with gr.Column(elem_classes="glass-panel", scale=1):
-                    gr.Markdown("#### 🛰️ MISSION ROADMAP\nRoadmap strictly matching your input language.")
-                    a3 = gr.Audio(label="VOICE COMMAND", sources=["microphone"], type="filepath")
-                    btn3 = gr.Button("LAUNCH MASTER PLAN", variant="primary")
-                with gr.Column(elem_classes="glass-panel", scale=1.5):
-                    out3 = gr.Markdown("### 🛣️ 6-Month Strategic Execution Roadmap")
-            btn3.click(stream_llm, [gr.State(ORBIT_PROMPT), t1, a3, f1], out3)
+                with gr.Column(scale=1):
+                    btn3 = gr.Button("CREATE ROADMAP")
+                with gr.Column(scale=2):
+                    out3 = gr.Markdown()
+            btn3.click(stream_llm, [gr.State(ORBIT_PROMPT), t1, a1, f1], out3)
 
-        with gr.Tab("🎯 Pathfinder"):
+        with gr.Tab("Pathfinder"):
             with gr.Row():
-                with gr.Column(elem_classes="glass-panel", scale=1):
-                    gr.Markdown("#### 🔭 MARKET MATCHING\nIndustry match in your exact language.")
-                    a_p = gr.Audio(label="VOICE COMMAND", sources=["microphone"], type="filepath")
-                    btn_p = gr.Button("FIND MY INDUSTRY MATCH", variant="primary")
-                with gr.Column(elem_classes="glass-panel", scale=1.5):
-                    out_p = gr.Markdown("### 🏢 Skill-Based Industry & Company Match")
-            btn_p.click(stream_llm, [gr.State(PATH_PROMPT), t1, a_p, f1], out_p)
+                with gr.Column(scale=1):
+                    btn4 = gr.Button("FIND MY MATCH")
+                with gr.Column(scale=2):
+                    out4 = gr.Markdown()
+            btn4.click(stream_llm, [gr.State(PATH_PROMPT), t1, a1, f1], out4)
 
 if __name__ == "__main__":
     demo.launch()