Update src/streamlit_app.py

src/streamlit_app.py

@@ -4,7 +4,7 @@ try:
 except Exception:
     PdfReader = None
 
-# Optional AI SDKs - guarded imports so the app can still run without them
+# Optional AI SDKs
 try:
     import google.generativeai as genai
 except Exception:
@@ -19,33 +19,7 @@ except Exception:
 
 from pydantic import BaseModel, Field
 from typing import Optional
-
-# Optional TTS / speech libs
-try:
-    from gtts import gTTS
-except Exception:
-    gTTS = None
-
-try:
-    import speech_recognition as sr
-except Exception:
-    sr = None
-
 import os
-import io
-import tempfile
-try:
-    from streamlit_mic_recorder import mic_recorder # Key component for browser audio
-except Exception as e:
-    # --- THIS WILL SHOW US THE REAL ERROR ---
-    st.error(f"❌ FAILED TO IMPORT MIC RECORDER: {e}") 
-    
-    # Fallback dummy recorder function that always returns None
-    def mic_recorder(*args, **kwargs):
-        return None
-
-# --- Configuration & Secrets ---
-
 
 # --- Pydantic Models (from your code) ---
 
@@ -62,17 +36,16 @@ class evaluation(BaseModel):
     followup: Optional[str] = Field(description="The followup question")
     review: Optional[str] = Field(description="Short Review of the answer")
 
-# --- AI & Logic Functions (from your code, slightly modified) ---
+# --- AI & Logic Functions (from your code) ---
 
 @st.cache_resource
 def get_llm(api_key):
     """Cached function to initialize the LLM."""
     return ChatGoogleGenerativeAI(
-        model="gemini-2.5-flash", 
+        model="gemini-2.5-flash",  
         temperature=1.0,
-        google_api_key=api_key  # <-- Explicitly pass the key here
+        google_api_key=api_key
     )
-
     
 @st.cache_resource
 def get_models(_llm_model):
@@ -89,11 +62,10 @@ def read_resume(uploaded_file):
         if PdfReader is None:
             st.warning("PyPDF2 is not installed; resume text extraction disabled.")
             return None
-        # PdfReader accepts a file-like object
         reader = PdfReader(uploaded_file)
         text = ""
         for page in reader.pages:
-            text += page.extract_text() or "" # Add check for None
+            text += page.extract_text() or ""
         return text
     except Exception as e:
         st.error(f"Error reading PDF: {e}")
@@ -101,19 +73,9 @@ def read_resume(uploaded_file):
 
 def generate_questions_from_resume(resume_text, model):
     """Generates interview questions from resume text."""
-    # If LangChain PromptTemplate or LLM wrapper is not available, or LLM not enabled,
-    # return simple heuristic questions
     if PromptTemplate is None or model is None or not st.session_state.get('enable_llm', False):
-        # Simple fallback: create questions from lines with 'Project'/'Experience' keywords
-        lines = resume_text.splitlines()
-        candidates = [l.strip() for l in lines if l and ('project' in l.lower() or 'experience' in l.lower())]
-        questions = []
-        for c in candidates:
-            if len(questions) >= 6:
-                break
-            questions.append(f"Tell me more about: {c}")
-        if not questions:
-            questions = ["Tell me about your most significant project.", "Describe a challenging bug you fixed.", "How do you design for scalability?", "Which technologies are you most comfortable with?"]
+        # Simple fallback
+        questions = ["Tell me about your most significant project.", "Describe a challenging bug you fixed.", "How do you design for scalability?", "Which technologies are you most comfortable with?"]
         return questions
 
     parse_resume_prompt_template = PromptTemplate(
@@ -122,32 +84,20 @@ Try to cover all projects and experience. Generate some conceptual questions too
 Resume:\n{text}""",
         input_variables=['text']
     )
-    # Use the LangChain pipeline if available and enabled
     try:
         if not st.session_state.get('enable_llm', False):
             raise RuntimeError('LLM disabled')
         generate_question_from_resume_chain = parse_resume_prompt_template | model
         output = generate_question_from_resume_chain.invoke({'text': resume_text})
-        # attempt to coerce into a list
         return getattr(output, 'questions', output)
     except Exception as e:
         st.warning(f"LLM question generation failed or disabled, using fallback: {e}")
-        # fallback similar to above
-        lines = resume_text.splitlines()
-        candidates = [l.strip() for l in lines if l and ('project' in l.lower() or 'experience' in l.lower())]
-        questions = []
-        for c in candidates:
-            if len(questions) >= 6:
-                break
-            questions.append(f"Tell me more about: {c}")
-        if not questions:
-            questions = ["Tell me about your most significant project.", "Describe a challenging bug you fixed.", "How do you design for scalability?", "Which technologies are you most comfortable with?"]
+        questions = ["Tell me about your most significant project.", "Describe a challenging bug you fixed.", "How do you design for scalability?", "Which technologies are you most comfortable with?"]
         return questions
 
 def get_introduction(model):
     """Gets the AI's intro and first question."""
     if PromptTemplate is None or model is None or not st.session_state.get('enable_llm', False):
-        # Return a simple dict-like fallback
         return type('O', (), {'intro': "Hello, I'm Interviewer.AI. Please introduce yourself.", 'question': "Can you briefly introduce yourself?"})()
 
     introduction_prompt = PromptTemplate(template="""Introduce yourself to the user telling the user that you are a AI agent. And ask the user to give introduction""")
@@ -185,7 +135,6 @@ def evaluate_answer(question, answer, model):
         score = 50
         review = "Thank you for your answer. Provide more details next time."
         followup = None
-        # small heuristic: longer answers get better score
         if answer and len(answer.split()) > 50:
             score = 80
             review = "Good answer — you covered several points."
@@ -206,95 +155,41 @@ If a good followup question can be asked generate it but only if it is a genuine
         return output
     except Exception as e:
         st.warning(f"LLM evaluation failed or disabled: {e}")
-        # fallback heuristic
         score = 50
         review = "Thank you for your answer. Provide more details next time."
         followup = None
         if answer and len(answer.split()) > 50:
             score = 80
-            review = "Good answer — you covered several points."
         elif answer and len(answer.split()) > 20:
             score = 65
-            review = "Decent answer; add more concrete examples."
         return type('O', (), {'marks': score, 'review': review, 'followup': followup})()
 
-# --- Streamlit Audio/Visual Functions ---
+# --- MODIFIED Streamlit Audio/Visual Function ---
 
 def text_to_speech_and_display(text, autoplay=True):
-    """Converts text to speech, displays text, and plays audio."""
+    """
+    MODIFIED: This function no longer plays audio. 
+    It just displays the text in the chat history.
+    """
     if not text:
         return
         
     try:
-        # Display the caption
         if 'chat_history' not in st.session_state:
             st.session_state.chat_history = []
         st.session_state.chat_history.append(f"**Interviewer:** {text}")
-
-        # Generate audio if gTTS available
-        if gTTS is None:
-            # No TTS available; just show text
-            return
-
-        tts = gTTS(text=text, lang='en', slow=False)
-        audio_fp = io.BytesIO()
-        tts.write_to_fp(audio_fp)
-        audio_fp.seek(0)
-
-        # Display audio player
-        st.audio(audio_fp, format='audio/mp3', autoplay=autoplay)
-        
     except Exception as e:
-        st.error(f"Error in text-to-speech: {e}")
-
-def speech_to_text(audio_bytes):
-    """Converts recorded audio bytes to text using SpeechRecognition."""
-    if not audio_bytes:
-        return "No audio recorded."
-    
-    if sr is None:
-        st.warning("speech_recognition is not installed; microphone input unavailable.")
-        return None
+        st.error(f"Error in text_to_speech_and_display: {e}")
 
-    r = sr.Recognizer()
-
-    # Need to save bytes to a temporary WAV file
-    try:
-        with tempfile.NamedTemporaryFile(suffix=".wav", delete=False) as temp_wav:
-            temp_wav.write(audio_bytes)
-            temp_wav_path = temp_wav.name
+# --- DELETED speech_to_text function ---
+# We are replacing it with a text_input
 
-        with sr.AudioFile(temp_wav_path) as source:
-            audio_data = r.record(source)
-
-        text = r.recognize_google(audio_data)
-        if 'chat_history' not in st.session_state:
-            st.session_state.chat_history = []
-        st.session_state.chat_history.append(f"**You:** {text}")
-        return text
-
-    except sr.UnknownValueError:
-        st.warning("Could not understand audio.")
-        return None
-    except sr.RequestError as e:
-        st.error(f"Speech recognition service error: {e}")
-        return None
-    except Exception as e:
-        st.error(f"Error processing audio: {e}")
-        return None
-    finally:
-        if 'temp_wav_path' in locals() and os.path.exists(temp_wav_path):
-            try:
-                os.remove(temp_wav_path)
-            except Exception:
-                pass
 
 # --- Main Streamlit App ---
 
 st.set_page_config(page_title="AI Interviewer", layout="wide")
 st.title("Interviewer.AI")
 
-
 # Initialize LLM and models
 llm = None
 gen_q_model = None
@@ -305,12 +200,9 @@ eval_model = None
 if genai is None or ChatGoogleGenerativeAI is None:
     st.warning("Google GenAI or LangChain wrappers not available. App will use deterministic fallbacks.")
 
-# Explicit per-session toggle to enable LLM features. This prevents accidental
-# LLM calls (and remote 403 errors) unless the user explicitly opts in.
 if 'enable_llm' not in st.session_state:
     st.session_state.enable_llm = False
 
-
 GOOGLE_API_KEY = os.environ.get("GOOGLE_API_KEY")
 api_key_exists = bool(GOOGLE_API_KEY)
 
@@ -352,7 +244,6 @@ if st.button("Test Google API Connection"):
     else:
         try:
             with st.spinner("Testing API connection..."):
-                # Simple test call
                 test_response = llm.invoke("Say 'Hello' if you can hear me.")
                 st.success("✅ SUCCESS! API is working correctly.")
                 st.info(f"Response: {test_response.content if hasattr(test_response, 'content') else str(test_response)}")
@@ -362,9 +253,7 @@ if st.button("Test Google API Connection"):
 
 st.divider()
 
-
 # --- Session State Initialization ---
-# This is crucial for making the app work step-by-step
 if 'stage' not in st.session_state:
     st.session_state.stage = 'start'
 if 'chat_history' not in st.session_state:
@@ -395,33 +284,26 @@ if st.session_state.stage == 'start':
                     st.error("Could not extract text from the resume. Please try another file.")
                     st.session_state.stage = 'start'
                 else:
-                    # 1. Generate Questions (guarded inside function)
+                    # 1. Generate Questions
                     st.session_state.questions = generate_questions_from_resume(resume_text, gen_q_model)
                     if not st.session_state.questions:
                         st.warning("No AI-generated questions returned; using fallback questions.")
                         st.session_state.questions = generate_questions_from_resume(resume_text, None)
 
-                    # 2. Get AI Introduction (guarded)
+                    # 2. Get AI Introduction
                     intro_output = get_introduction(intro_model)
                     st.session_state.current_question = getattr(intro_output, 'question', "Can you introduce yourself?")
 
                     # 3. Move to next stage and display intro
                     st.session_state.stage = 'awaiting_intro'
-                    #************************
-                    #************************
-                    #************************
                     
-                    # text_to_speech_and_display(getattr(intro_output, 'intro', "Hello, I'm Interviewer.AI. Please introduce yourself."))
-                    # text_to_speech_and_display(getattr(intro_output, 'question', "Can you introduce yourself?"))
-
-                    #************************
-                    #************************
-                    #************************
+                    # --- MODIFIED: Display text directly ---
+                    text_to_speech_and_display(getattr(intro_output, 'intro', "Hello, I'm Interviewer.AI. Please introduce yourself."))
+                    text_to_speech_and_display(getattr(intro_output, 'question', "Can you introduce yourself?"))
+                    
                     st.rerun()
             except Exception as e:
-                # Catch-all to prevent any LLM/network exceptions from surfacing to the client
                 st.error(f"An error occurred while processing the resume. Using fallback behaviour. Error: {e}")
-                # Fallback simple questions
                 fallback_qs = ["Tell me about your most significant project.", "Describe a challenging bug you fixed.", "How do you design for scalability?", "Which technologies are you most comfortable with?"]
                 st.session_state.questions = fallback_qs
                 st.session_state.stage = 'asking_question'
@@ -432,115 +314,99 @@ if st.session_state.stage != 'start':
     
     # --- Chat History Display ---
     st.subheader("Interview Transcript")
-    chat_container = st.container()
+    chat_container = st.container(height=400) # Added height for scrolling
     with chat_container:
         for entry in st.session_state.chat_history:
             st.markdown(entry)
 
-    # visual divider
     try:
         st.divider()
     except Exception:
         st.markdown('---')
 
-    # --- Audio Recorder ---
-    # This component returns audio bytes when the user stops recording
-    st.write("Your turn to speak:")
-    audio_bytes = mic_recorder(
-        start_prompt="Start Recording",
-        stop_prompt="Stop Recording",
-        key='recorder'
-    )
-
     # --- End Interview Button ---
     if st.button("End Interview", type="primary"):
         st.session_state.stage = 'finished'
         st.rerun()
 
-    # --- Process Recorded Audio ---
-    # mic_recorder may return None, bytes, or a dict with a 'bytes' key depending on implementation
-    def _extract_audio_bytes(rec):
-        if rec is None:
-            return None
-        if isinstance(rec, dict):
-            # some implementations return {'bytes': b'...', 'start':..., ...}
-            return rec.get('bytes') or rec.get('audio') or None
-        if isinstance(rec, (bytes, bytearray)):
-            return bytes(rec)
-        return None
+    # --- REPLACEMENT: Text Input Area ---
+    user_text = None # Initialize user_text
+    is_disabled = (st.session_state.stage == 'finished')
 
-    extracted_audio = _extract_audio_bytes(audio_bytes)
-    if extracted_audio:
-        with st.spinner("Transcribing your answer..."):
-            user_text = speech_to_text(extracted_audio)
+    with st.form(key="answer_form", clear_on_submit=True):
+        answer = st.text_input("Your answer:", disabled=is_disabled)
+        submit_button = st.form_submit_button(label="Submit Answer", disabled=is_disabled)
         
-        if user_text:
-            # --- STAGE 1: Process User's Introduction ---
-            if st.session_state.stage == 'awaiting_intro':
-                with st.spinner("Thinking of a followup..."):
-                    followup = ask_followup(user_text, intro_model)
-                    st.session_state.current_question = followup
-                    # text_to_speech_and_display(followup)
-                    st.session_state.stage = 'awaiting_intro_followup'
-                    st.rerun()
-            
-            # --- STAGE 2: Process Followup to Intro ---
-            elif st.session_state.stage == 'awaiting_intro_followup':
-                # text_to_speech_and_display("OK, Great. Let's start the interview with questions from your resume.")
-                st.session_state.stage = 'asking_question' # Move to main questions
+        if submit_button and answer:
+            user_text = answer
+            st.session_state.chat_history.append(f"**You:** {user_text}")
+    # --- END OF REPLACEMENT ---
+
+
+    # --- Process Submitted Text ---
+    if user_text:
+        # --- STAGE 1: Process User's Introduction ---
+        if st.session_state.stage == 'awaiting_intro':
+            with st.spinner("Thinking of a followup..."):
+                followup = ask_followup(user_text, intro_model)
+                st.session_state.current_question = followup
+                text_to_speech_and_display(followup) # This now just displays text
+                st.session_state.stage = 'awaiting_intro_followup'
                 st.rerun()
+        
+        # --- STAGE 2: Process Followup to Intro ---
+        elif st.session_state.stage == 'awaiting_intro_followup':
+            text_to_speech_and_display("OK, Great. Let's start the interview with questions from your resume.")
+            st.session_state.stage = 'asking_question' # Move to main questions
+            st.rerun()
 
-            # --- STAGE 4: Process Answer to a Main Question ---
-            elif st.session_state.stage == 'awaiting_answer':
-                with st.spinner("Evaluating your answer..."):
-                    question_asked = st.session_state.current_question
-                    output = evaluate_answer(question_asked, user_text, eval_model)
-                    
-                    st.session_state.total_marks += output.marks
-                    st.session_state.num_questions += 1
-                    
-                    if output.review:
-                        # text_to_speech_and_display(output.review)
-                        bs_variable=5
-                    
-                    if output.followup:
-                        # Ask followup question
-                        st.session_state.current_question = output.followup
-                        # text_to_speech_and_display(output.followup)
-                        st.session_state.stage = 'awaiting_followup_answer'
-                    else:
-                        # Move to next question
-                        st.session_state.q_index += 1
-                        st.session_state.stage = 'asking_question'
-                    st.rerun()
-
-            # --- STAGE 5: Process Answer to a Followup Question ---
-            elif st.session_state.stage == 'awaiting_followup_answer':
-                 with st.spinner("Evaluating your answer..."):
-                    question_asked = st.session_state.current_question
-                    output = evaluate_answer(question_asked, user_text, eval_model)
-                    
-                    st.session_state.total_marks += output.marks
-                    st.session_state.num_questions += 1
-                    
-                    if output.review:
-                        # text_to_speech_and_display(output.review)
-                        bs_variable=6
-                    # Always move to the next main question after a followup
+        # --- STAGE 4: Process Answer to a Main Question ---
+        elif st.session_state.stage == 'awaiting_answer':
+            with st.spinner("Evaluating your answer..."):
+                question_asked = st.session_state.current_question
+                output = evaluate_answer(question_asked, user_text, eval_model)
+                
+                st.session_state.total_marks += output.marks
+                st.session_state.num_questions += 1
+                
+                if output.review:
+                    text_to_speech_and_display(output.review) # This now just displays text
+                
+                if output.followup:
+                    st.session_state.current_question = output.followup
+                    text_to_speech_and_display(output.followup) # This now just displays text
+                    st.session_state.stage = 'awaiting_followup_answer'
+                else:
                     st.session_state.q_index += 1
                     st.session_state.stage = 'asking_question'
-                    st.rerun()
+                st.rerun()
+
+        # --- STAGE 5: Process Answer to a Followup Question ---
+        elif st.session_state.stage == 'awaiting_followup_answer':
+             with st.spinner("Evaluating your answer..."):
+                question_asked = st.session_state.current_question
+                output = evaluate_answer(question_asked, user_text, eval_model)
+                
+                st.session_state.total_marks += output.marks
+                st.session_state.num_questions += 1
+                
+                if output.review:
+                    text_to_speech_and_display(output.review) # This now just displays text
+                
+                st.session_state.q_index += 1
+                st.session_state.stage = 'asking_question'
+                st.rerun()
 
     # --- STAGE 3: Ask a New Question ---
+    # This runs when the page loads into this state, *before* user input
     if st.session_state.stage == 'asking_question':
         if st.session_state.q_index < len(st.session_state.questions):
-            # Ask the next question
             question = st.session_state.questions[st.session_state.q_index]
             st.session_state.current_question = question
-            # text_to_speech_and_display(question)
+            text_to_speech_and_display(question) # This now just displays text
             st.session_state.stage = 'awaiting_answer'
         else:
-            # No more questions
+            text_to_speech_and_display("That's all the questions I have. Thank you!")
             st.session_state.stage = 'finished'
             st.rerun()
 
@@ -557,12 +423,12 @@ if st.session_state.stage != 'start':
         st.markdown(f"**Total Questions Answered:** {st.session_state.num_questions}")
         st.markdown(f"**Average Score:** {final_score:.2f} / 100")
         
+        # Transcript is already shown above, but we can show it again
         st.subheader("Full Transcript")
         for entry in st.session_state.chat_history:
             st.markdown(entry)
             
         if st.button("Start New Interview"):
-            # Clear all session state
             for key in st.session_state.keys():
                 del st.session_state[key]
-            st.rerun()
+            st.rerun()