Update app.py

app.py

@@ -4,6 +4,13 @@ import requests
 import io
 import tempfile
 import sys
+from PIL import Image
+
+def compress_image(image, quality):
+    img_buffer = io.BytesIO()
+    image.save(img_buffer, format='JPEG', quality=quality)
+    img_buffer.seek(0)
+    return img_buffer
 
 def compress_pdf(input_file, url, strength):
     if input_file is None and (url is None or url.strip() == ""):
@@ -26,24 +33,49 @@ def compress_pdf(input_file, url, strength):
         reader = PyPDF2.PdfReader(pdf_content)
         writer = PyPDF2.PdfWriter()
 
-        for i, page in enumerate(reader.pages):
-            try:
-                if strength == "High":
-                    page.compress_content_streams()  # Maximum compression
-                elif strength == "Medium":
-                    page.compress_content_streams()  # Medium compression (same as high for now)
-                else:  # Low strength
-                    writer.add_page(page)  # No compression
-                
-                if strength != "Low":
-                    writer.add_page(page)
-            except Exception as e:
-                return None, f"Error compressing page {i+1}: {str(e)}"
+        if strength == "Low":
+            image_quality = 65
+            compression_level = 1
+        elif strength == "Medium":
+            image_quality = 40
+            compression_level = 2
+        else:  # High
+            image_quality = 20
+            compression_level = 3
+
+        for page in reader.pages:
+            page.compress_content_streams()  # Apply content stream compression
+            
+            # Compress images on the page
+            for img_index, img in enumerate(page.images):
+                if img.image is not None:
+                    try:
+                        pil_image = Image.open(io.BytesIO(img.image))
+                        compressed_image = compress_image(pil_image, image_quality)
+                        page.replace_image(img_index, compressed_image)
+                    except Exception as e:
+                        print(f"Error compressing image: {e}")
+
+            writer.add_page(page)
 
         with tempfile.NamedTemporaryFile(delete=False, suffix=".pdf") as temp_file:
             writer.write(temp_file)
             temp_file_path = temp_file.name
 
+        # Apply additional compression using PyPDF2's built-in compression
+        with open(temp_file_path, 'rb') as file:
+            reader = PyPDF2.PdfReader(file)
+            writer = PyPDF2.PdfWriter()
+            
+            for page in reader.pages:
+                page.compress_content_streams()
+                writer.add_page(page)
+            
+            writer.add_metadata(reader.metadata)
+            
+            with open(temp_file_path, 'wb') as output_file:
+                writer.write(output_file)
+
         return temp_file_path, "PDF compressed successfully!"
     except Exception as e:
         return None, f"Error compressing PDF: {str(e)}"
@@ -62,7 +94,7 @@ with gr.Blocks() as demo:
     with gr.Row():
         input_file = gr.File(label="Upload PDF")
         url_input = gr.Textbox(label="Or enter PDF URL")
-    strength = gr.Radio(["High", "Medium", "Low"], label="Compression Strength", value="Medium", info="High: Maximum compression, Low: Least compression")
+    strength = gr.Radio(["Low", "Medium", "High"], label="Compression Strength", value="Medium", info="Low: ~25% compression, Medium: ~50% compression, High: ~75% compression")
     compress_btn = gr.Button("Compress")
     output_file = gr.File(label="Download Compressed PDF")
     message = gr.Textbox(label="Message")