Update app.py

app.py

@@ -1,8 +1,15 @@
 import gradio as gr
 from gtts import gTTS
 import pdfplumber
+from sumy.parsers.plaintext import PlaintextParser
+from sumy.nlp.tokenizers import Tokenizer
+from sumy.summarizers.lsa import LsaSummarizer
+import nltk
 import os
 
+# Download NLTK data for sumy
+nltk.download('punkt')
+
 def extract_text_from_pdf(pdf_file):
     """
     Extract text from a PDF file using pdfplumber.
@@ -22,45 +29,84 @@ def extract_text_from_pdf(pdf_file):
     except Exception as e:
         return f"Error extracting text: {str(e)}"
 
-def text_to_speech(pdf_file, lang="en"):
+def summarize_text(text, sentences_count=12):
     """
-    Convert text from a PDF to speech using gTTS and return the audio file path.
+    Summarize text to approximately four paragraphs using sumy LSA summarizer.
+    
+    Args:
+        text (str): Text to summarize.
+        sentences_count (int): Number of sentences in summary (approx. 3 sentences per paragraph).
+    
+    Returns:
+        str: Summarized text.
+    """
+    try:
+        # Initialize parser and tokenizer
+        parser = PlaintextParser.from_string(text, Tokenizer("english"))
+        summarizer = LsaSummarizer()
+        
+        # Summarize to specified number of sentences
+        summary = summarizer(parser.document, sentences_count)
+        
+        # Combine sentences and format into paragraphs (approx. 3 sentences per paragraph)
+        summary_text = ""
+        for i, sentence in enumerate(summary):
+            summary_text += str(sentence) + " "
+            if (i + 1) % 3 == 0:  # Add paragraph break every 3 sentences
+                summary_text += "\n\n"
+        
+        return summary_text.strip() if summary_text else "No summary generated."
+    except Exception as e:
+        return f"Error summarizing text: {str(e)}"
+
+def pdf_to_speech(pdf_file, lang="en"):
+    """
+    Convert text from a PDF to summarized speech using gTTS.
     
     Args:
         pdf_file: Uploaded PDF file.
         lang (str): Language code (default is 'en' for English).
     
     Returns:
-        str: Path to the generated audio file or error message.
+        tuple: (Path to audio file, summarized text) or (error message, error message).
     """
     try:
         # Extract text from PDF
         text = extract_text_from_pdf(pdf_file)
         if "Error" in text:
-            return text
+            return text, text
+        
+        # Summarize text (approx. 12 sentences for 4 paragraphs)
+        summarized_text = summarize_text(text, sentences_count=12)
+        if "Error" in summarized_text:
+            return summarized_text, summarized_text
         
         # Create gTTS object
-        tts = gTTS(text=text, lang=lang, slow=False)
+        tts = gTTS(text=summarized_text, lang=lang, slow=False)
         
         # Save the audio file
         output_file = "output.mp3"
         tts.save(output_file)
         
-        return output_file
+        return output_file, summarized_text
     
     except Exception as e:
-        return f"An error occurred: {str(e)}"
+        error_msg = f"An error occurred: {str(e)}"
+        return error_msg, error_msg
 
 # Define Gradio interface
 demo = gr.Interface(
-    fn=text_to_speech,
+    fn=pdf_to_speech,
     inputs=[
         gr.File(label="Upload a PDF file", file_types=[".pdf"]),
         gr.Dropdown(choices=["en", "es", "fr"], label="Select Language", value="en")
     ],
-    outputs=gr.Audio(label="Generated Speech"),
-    title="PDF to Speech with gTTS",
-    description="Upload a PDF file, select a language, and generate speech from the extracted text."
+    outputs=[
+        gr.Audio(label="Generated Speech"),
+        gr.Textbox(label="Summarized Text")
+    ],
+    title="PDF Summary to Speech",
+    description="Upload an English PDF file, select a language, and generate speech from a summarized version (approx. 4 paragraphs). The summarized text is also displayed."
 )
 
 # Launch the app