Update app.py

app.py

@@ -19,15 +19,22 @@ from datetime import datetime
 
 PDF_PATH = "2VBMAPP.pdf"  # Your pre-loaded PDF file
 
-class ABA Tutor:
+class ABATutor:  # FIXED: No space in class name
     def __init__(self):
         # Load CLIP for computer vision
         self.clip_model = CLIPModel.from_pretrained("openai/clip-vit-base-patch32")
         self.clip_processor = CLIPProcessor.from_pretrained("openai/clip-vit-base-patch32")
         
         # Load PDF
-        self.doc = fitz.open(PDF_PATH)
-        self.total_pages = len(self.doc)
+        try:
+            self.doc = fitz.open(PDF_PATH)
+            self.total_pages = len(self.doc)
+            print(f"✅ Loaded PDF with {self.total_pages} pages")
+        except Exception as e:
+            print(f"⚠️ Could not load PDF: {e}")
+            self.doc = None
+            self.total_pages = 1
+        
         self.current_page = 6  # Start at first Tact page (page 7 in PDF, index 6)
         
         # Session data
@@ -37,7 +44,6 @@ class ABA Tutor:
         # Expected labels for each page (built from book content)
         self.page_labels = self.build_label_map()
         
-        print(f"✅ Loaded PDF with {self.total_pages} pages")
         print(f"✅ CLIP model ready")
     
     def build_label_map(self):
@@ -54,6 +60,10 @@ class ABA Tutor:
     
     def get_current_image(self):
         """Extract current page as PIL Image for CV"""
+        if self.doc is None:
+            # Return a blank image if PDF not loaded
+            return Image.new('RGB', (800, 600), color='white')
+        
         page = self.doc.load_page(self.current_page)
         pix = page.get_pixmap(dpi=100)
         img_data = pix.tobytes("png")
@@ -62,6 +72,9 @@ class ABA Tutor:
     
     def get_current_image_html(self):
         """Get HTML image for display"""
+        if self.doc is None:
+            return '<p style="text-align:center;padding:50px;">⚠️ PDF not loaded. Please ensure 2VBMAPP.pdf is in the Space.</p>'
+        
         page = self.doc.load_page(self.current_page)
         pix = page.get_pixmap(dpi=100)
         img_data = pix.tobytes("png")
@@ -83,17 +96,22 @@ class ABA Tutor:
         candidates = expected + [spoken_word]
         candidates = list(set(candidates))  # Remove duplicates
         
-        # Run CLIP
-        inputs = self.clip_processor(text=candidates, images=image, return_tensors="pt", padding=True)
-        with torch.no_grad():
-            outputs = self.clip_model(**inputs)
-            logits_per_image = outputs.logits_per_image
-            probs = logits_per_image.softmax(dim=1).numpy()
-        
-        # Get best match
-        best_idx = probs[0].argmax()
-        best_label = candidates[best_idx]
-        confidence = float(probs[0][best_idx])
+        try:
+            # Run CLIP
+            inputs = self.clip_processor(text=candidates, images=image, return_tensors="pt", padding=True)
+            with torch.no_grad():
+                outputs = self.clip_model(**inputs)
+                logits_per_image = outputs.logits_per_image
+                probs = logits_per_image.softmax(dim=1).numpy()
+            
+            # Get best match
+            best_idx = probs[0].argmax()
+            best_label = candidates[best_idx]
+            confidence = float(probs[0][best_idx])
+        except Exception as e:
+            print(f"CLIP error: {e}")
+            best_label = spoken_word
+            confidence = 0.5
         
         # Check if spoken word matches any expected label
         spoken_lower = spoken_word.lower().strip()
@@ -112,10 +130,14 @@ class ABA Tutor:
     
     def speak(self, text):
         """Convert text to speech and return audio file"""
-        tts = gTTS(text=text, lang="en", slow=False)
-        audio_path = tempfile.NamedTemporaryFile(delete=False, suffix=".mp3").name
-        tts.save(audio_path)
-        return audio_path
+        try:
+            tts = gTTS(text=text, lang="en", slow=False)
+            audio_path = tempfile.NamedTemporaryFile(delete=False, suffix=".mp3").name
+            tts.save(audio_path)
+            return audio_path
+        except Exception as e:
+            print(f"TTS error: {e}")
+            return None
     
     def get_sd_prompt(self):
         """Get the SD prompt for current page based on book content"""
@@ -134,18 +156,18 @@ class ABA Tutor:
     def listen(self):
         """Capture and transcribe child's spoken response"""
         recognizer = sr.Recognizer()
-        with sr.Microphone() as source:
-            try:
+        try:
+            with sr.Microphone() as source:
                 recognizer.adjust_for_ambient_noise(source, duration=0.5)
                 audio = recognizer.listen(source, timeout=5, phrase_time_limit=3)
                 text = recognizer.recognize_google(audio)
                 return text.lower().strip()
-            except sr.WaitTimeoutError:
-                return "[no response]"
-            except sr.UnknownValueError:
-                return "[could not understand]"
-            except Exception as e:
-                return f"[error: {str(e)}]"
+        except sr.WaitTimeoutError:
+            return "[no response]"
+        except sr.UnknownValueError:
+            return "[could not understand]"
+        except Exception as e:
+            return f"[error: {str(e)}]"
     
     def run_trial(self):
         """Run one complete assessment trial with CV validation"""
@@ -245,7 +267,7 @@ class ABA Tutor:
 # 2. INITIALIZE TUTOR
 # ============================================
 
-tutor = ABA Tutor()
+tutor = ABATutor()  # FIXED: Class name without space
 
 # ============================================
 # 3. GRADIO INTERFACE