Update repo

scripts/epub_page_renderer.py

@@ -0,0 +1,434 @@
+#!/usr/bin/env python3
+"""
+Simple EPUB page renderer tool.
+
+This tool uses the pyWebLayout epub_reader and layout modules to:
+1. Load an EPUB file
+2. Render the first X pages according to command line arguments
+3. Save the pages as PNG images
+
+Usage:
+    python epub_page_renderer.py book.epub --pages 5 --output-dir rendered_pages
+"""
+
+import os
+import sys
+import argparse
+from pathlib import Path
+from typing import Optional, List
+
+# Add the parent directory to sys.path to import pyWebLayout
+sys.path.insert(0, str(Path(__file__).parent.parent))
+
+try:
+    from pyWebLayout.io.readers.epub_reader import read_epub
+    from pyWebLayout.layout.ereader_layout import BidirectionalLayouter, RenderingPosition
+    from pyWebLayout.layout.document_layouter import paragraph_layouter
+    from pyWebLayout.concrete.page import Page
+    from pyWebLayout.style.page_style import PageStyle
+    from pyWebLayout.style.fonts import Font
+    from pyWebLayout.abstract.block import Block
+    from PIL import Image, ImageDraw
+except ImportError as e:
+    print(f"Error importing required modules: {e}")
+    print("Make sure pyWebLayout is properly installed and PIL is available")
+    sys.exit(1)
+
+
+def render_page_to_image(page: Page) -> Image.Image:
+    """
+    Render a Page object to a PIL Image using pyWebLayout's built-in rendering.
+    
+    Args:
+        page: The Page object to render
+        
+    Returns:
+        PIL Image object
+    """
+    try:
+        # Use the Page's built-in render method
+        rendered_image = page.render()
+        if isinstance(rendered_image, Image.Image):
+            return rendered_image
+        else:
+            # If render() doesn't return a PIL Image, create error image
+            error_image = Image.new('RGB', page.size, 'white')
+            draw = ImageDraw.Draw(error_image)
+            draw.text((20, 20), "Error: Page.render() did not return PIL Image", fill='red')
+            return error_image
+            
+    except Exception as e:
+        # Create error image if rendering fails
+        error_image = Image.new('RGB', page.size, 'white')
+        draw = ImageDraw.Draw(error_image)
+        draw.text((20, 20), f"Rendering error: {str(e)}", fill='red')
+        print(f"Warning: Error rendering page: {e}")
+        return error_image
+
+
+def extract_text_from_page(page: Page) -> str:
+    """
+    Extract text content from a Page object for verification purposes.
+    
+    Args:
+        page: The Page object to extract text from
+        
+    Returns:
+        String containing the page's text content
+    """
+    text_lines = []
+    text_lines.append(f"=== PAGE CONTENT ===")
+    text_lines.append("")
+    
+    try:
+        # Recursively extract text from page children
+        def extract_from_element(element, indent_level=0):
+            indent = "  " * indent_level
+            
+            # Import abstract block types
+            from pyWebLayout.abstract.block import Paragraph, Heading, HList, Table, Image as AbstractImage
+            from pyWebLayout.concrete.text import Line
+            
+            # Handle Line objects (concrete)
+            if isinstance(element, Line):
+                line_text = []
+                if hasattr(element, '_text_objects') and element._text_objects:
+                    for text_obj in element._text_objects:
+                        if hasattr(text_obj, 'text'):
+                            line_text.append(str(text_obj.text))
+                if line_text:
+                    text_lines.append(f"{indent}{' '.join(line_text)}")
+            
+            # Handle abstract block objects
+            elif isinstance(element, (Paragraph, Heading)):
+                # Extract text from paragraph/heading
+                paragraph_text = extract_text_from_paragraph(element)
+                if paragraph_text:
+                    block_type = "HEADING" if isinstance(element, Heading) else "PARAGRAPH"
+                    text_lines.append(f"{indent}{block_type}: {paragraph_text}")
+            
+            elif isinstance(element, HList):
+                text_lines.append(f"{indent}LIST:")
+                # Extract text from list items
+                try:
+                    for item in element.items():
+                        item_text = extract_text_from_paragraph(item)
+                        if item_text:
+                            text_lines.append(f"{indent}  - {item_text}")
+                except:
+                    text_lines.append(f"{indent}  (List content extraction failed)")
+            
+            elif isinstance(element, Table):
+                text_lines.append(f"{indent}[TABLE]")
+            
+            elif isinstance(element, AbstractImage):
+                alt_text = getattr(element, 'alt_text', '')
+                src = getattr(element, 'source', 'Unknown')
+                text_lines.append(f"{indent}[IMAGE: {alt_text or src}]")
+            
+            # Handle containers with children
+            elif hasattr(element, '_children') and element._children:
+                for child in element._children:
+                    extract_from_element(child, indent_level + 1)
+            
+            # Handle text elements
+            elif hasattr(element, 'text'):
+                text = str(element.text).strip()
+                if text:
+                    text_lines.append(f"{indent}{text}")
+            
+            # Handle other object types by showing their class name
+            else:
+                class_name = element.__class__.__name__
+                text_lines.append(f"{indent}[{class_name}]")
+        
+        # Helper function to extract text from paragraph-like objects
+        def extract_text_from_paragraph(para_obj):
+            words = []
+            try:
+                # Try to get words from the paragraph
+                if hasattr(para_obj, 'words_iter') and callable(para_obj.words_iter):
+                    for _, word in para_obj.words_iter():
+                        if hasattr(word, 'text'):
+                            words.append(word.text)
+                        else:
+                            words.append(str(word))
+                elif hasattr(para_obj, '_words'):
+                    # Direct access to words list
+                    for word in para_obj._words:
+                        if hasattr(word, 'text'):
+                            words.append(word.text)
+                        else:
+                            words.append(str(word))
+            except Exception as e:
+                return f"(Text extraction error: {str(e)})"
+            
+            return ' '.join(words) if words else "(No text)"
+        
+        # Extract text from page children
+        if hasattr(page, '_children'):
+            for child in page._children:
+                extract_from_element(child)
+        
+        # If no text was extracted, add a note
+        if len(text_lines) <= 2:  # Only header and empty line
+            text_lines.append("(No text content found)")
+        
+    except Exception as e:
+        text_lines.append(f"Error extracting text: {str(e)}")
+        import traceback
+        text_lines.append(traceback.format_exc())
+    
+    return "\n".join(text_lines)
+
+
+def get_all_blocks_from_book(book) -> List[Block]:
+    """
+    Extract all blocks from all chapters in the book.
+    
+    Args:
+        book: The Book object from epub_reader
+        
+    Returns:
+        List of all Block objects
+    """
+    all_blocks = []
+    
+    # Iterate through all chapters
+    for chapter in book.chapters:
+        # Get blocks from the chapter
+        if hasattr(chapter, '_blocks'):
+            all_blocks.extend(chapter._blocks)
+    
+    return all_blocks
+
+
+def main():
+    """Main function to handle command line arguments and process the EPUB."""
+    parser = argparse.ArgumentParser(
+        description='Render EPUB pages to images using pyWebLayout',
+        formatter_class=argparse.RawDescriptionHelpFormatter,
+        epilog="""
+Examples:
+  python epub_page_renderer.py book.epub --pages 5
+  python epub_page_renderer.py book.epub --pages 10 --output-dir my_output --width 600 --height 800
+        """
+    )
+    
+    parser.add_argument(
+        'epub_file',
+        help='Path to the EPUB file to render'
+    )
+    
+    parser.add_argument(
+        '--pages', '-p',
+        type=int,
+        default=5,
+        help='Number of pages to render (default: 5)'
+    )
+    
+    parser.add_argument(
+        '--output-dir', '-o',
+        default='rendered_pages',
+        help='Output directory for rendered images (default: rendered_pages)'
+    )
+    
+    parser.add_argument(
+        '--width', '-w',
+        type=int,
+        default=800,
+        help='Page width in pixels (default: 800)'
+    )
+    
+    parser.add_argument(
+        '--height', '-t',
+        type=int,
+        default=1000,
+        help='Page height in pixels (default: 1000)'
+    )
+    
+    parser.add_argument(
+        '--margin', '-m',
+        type=int,
+        default=40,
+        help='Page margin in pixels (default: 40)'
+    )
+    
+    parser.add_argument(
+        '--align', '-a',
+        choices=['left', 'justify'],
+        default='left',
+        help='Text alignment: left or justify (default: left)'
+    )
+    
+    args = parser.parse_args()
+    
+    # Validate arguments
+    if not os.path.exists(args.epub_file):
+        print(f"Error: EPUB file '{args.epub_file}' not found")
+        return 1
+    
+    if args.pages <= 0:
+        print("Error: Number of pages must be positive")
+        return 1
+    
+    # Create output directory
+    try:
+        os.makedirs(args.output_dir, exist_ok=True)
+    except OSError as e:
+        print(f"Error creating output directory: {e}")
+        return 1
+    
+    print(f"Loading EPUB file: {args.epub_file}")
+    
+    # Load the EPUB file
+    try:
+        book = read_epub(args.epub_file)
+        print(f"Successfully loaded EPUB: {book.get_title() or 'Unknown Title'}")
+        
+        # Print book information
+        author = book.get_metadata('AUTHOR')
+        if author:
+            print(f"Author: {author}")
+        
+        print(f"Chapters: {len(book.chapters) if hasattr(book, 'chapters') else 'Unknown'}")
+        
+    except Exception as e:
+        print(f"Error loading EPUB file: {e}")
+        import traceback
+        traceback.print_exc()
+        return 1
+    
+    # Extract all blocks from the book
+    print("Extracting content blocks...")
+    try:
+        all_blocks = get_all_blocks_from_book(book)
+        print(f"Extracted {len(all_blocks)} content blocks")
+        
+        if not all_blocks:
+            print("No content blocks found in EPUB. The book might be empty.")
+            return 1
+        
+        # Apply alignment setting to all paragraphs and headings
+        from pyWebLayout.abstract.block import Paragraph, Heading
+        from pyWebLayout.style.alignment import Alignment
+        
+        alignment = Alignment.JUSTIFY if args.align == 'justify' else Alignment.LEFT
+        print(f"Applying {args.align} alignment to all text blocks...")
+        
+        # Note: We'll pass alignment to the layouter which will handle it during rendering
+        # The alignment is applied at the Line level in paragraph_layouter
+            
+    except Exception as e:
+        print(f"Error extracting blocks: {e}")
+        import traceback
+        traceback.print_exc()
+        return 1
+    
+    # Set up page style and layouter
+    page_size = (args.width, args.height)
+    page_style = PageStyle(
+        background_color=(255, 255, 255),
+        border_width=args.margin,
+        border_color=(200, 200, 200),
+        padding=(10, 10, 10, 10),  # top, right, bottom, left
+        line_spacing=5,
+        inter_block_spacing=15
+    )
+    
+    print(f"Setting up layouter with page size {page_size} and {args.align} alignment")
+    
+    try:
+        layouter = BidirectionalLayouter(
+            blocks=all_blocks,
+            page_style=page_style,
+            page_size=page_size,
+            alignment_override=alignment
+        )
+    except Exception as e:
+        print(f"Error setting up layouter: {e}")
+        import traceback
+        traceback.print_exc()
+        return 1
+    
+    # Render pages
+    print(f"Rendering up to {args.pages} pages...")
+    
+    try:
+        pages = []
+        current_position = RenderingPosition()  # Start from beginning
+        
+        for page_num in range(args.pages):
+            print(f"Rendering page {page_num + 1}/{args.pages}...")
+            
+            try:
+                # Render the page
+                page, next_position = layouter.render_page_forward(current_position)
+                pages.append(page)
+                
+                # Check if we've reached the end of the document
+                if next_position.block_index >= len(all_blocks):
+                    print(f"Reached end of document after {page_num + 1} pages")
+                    break
+                
+                # Update position for next page
+                current_position = next_position
+                
+            except Exception as e:
+                print(f"Error rendering page {page_num + 1}: {e}")
+                import traceback
+                traceback.print_exc()
+                break
+        
+        if not pages:
+            print("No pages were generated.")
+            return 1
+        
+        print(f"Generated {len(pages)} pages")
+        
+        # Save each page to an image and extract text
+        for i, page in enumerate(pages):
+            print(f"Saving page {i + 1}/{len(pages)}...")
+            
+            try:
+                # Create image from page using pyWebLayout's built-in rendering
+                image = render_page_to_image(page)
+                
+                # Save the image
+                output_filename = f"page_{i + 1:03d}.png"
+                output_path = os.path.join(args.output_dir, output_filename)
+                image.save(output_path, 'PNG')
+                
+                # Extract and save text content for verification
+                page_text = extract_text_from_page(page)
+                text_filename = f"page_{i + 1:03d}.txt"
+                text_path = os.path.join(args.output_dir, text_filename)
+                with open(text_path, 'w', encoding='utf-8') as f:
+                    f.write(page_text)
+                
+                print(f"Saved: {output_path} and {text_path}")
+                
+            except Exception as e:
+                print(f"Error saving page {i + 1}: {e}")
+                import traceback
+                traceback.print_exc()
+                continue
+        
+        print(f"\nCompleted! Rendered {len(pages)} pages to {args.output_dir}")
+        
+        # Calculate progress through the book
+        if len(all_blocks) > 0:
+            progress = (current_position.block_index / len(all_blocks)) * 100
+            print(f"Progress through book: {progress:.1f}%")
+        
+    except Exception as e:
+        print(f"Error during pagination/rendering: {e}")
+        import traceback
+        traceback.print_exc()
+        return 1
+    
+    return 0
+
+
+if __name__ == "__main__":
+    sys.exit(main())