view port based browser and test

demo_viewport_system.py

@@ -0,0 +1,224 @@
+#!/usr/bin/env python3
+"""
+Demo script showing the viewport system in action.
+
+This demonstrates how the viewport provides a movable window into large content,
+enabling efficient scrolling without rendering the entire content at once.
+"""
+
+import os
+from PIL import Image
+from pyWebLayout.concrete import (
+    Viewport, ScrollablePageContent, Text, Box, RenderableImage
+)
+from pyWebLayout.style.fonts import Font, FontWeight
+from pyWebLayout.style.layout import Alignment
+
+
+def create_large_document_content():
+    """Create a large document to demonstrate viewport scrolling"""
+    
+    # Create scrollable content container
+    content = ScrollablePageContent(content_width=800, initial_height=100)
+    
+    # Add a title
+    title_font = Font(font_size=24, weight=FontWeight.BOLD)
+    title = Text("Large Document Demo", title_font)
+    content.add_child(title)
+    
+    # Add spacing
+    content.add_child(Box((0, 0), (1, 20)))
+    
+    # Add many paragraphs to create a long document
+    paragraph_font = Font(font_size=14)
+    
+    for i in range(50):
+        # Section header
+        section_font = Font(font_size=18, weight=FontWeight.BOLD)
+        header = Text(f"Section {i+1}", section_font)
+        content.add_child(header)
+        
+        # Add some spacing
+        content.add_child(Box((0, 0), (1, 10)))
+        
+        # Add paragraph content
+        paragraphs = [
+            f"This is paragraph {i+1}, line 1. Lorem ipsum dolor sit amet, consectetur adipiscing elit.",
+            f"This is paragraph {i+1}, line 2. Sed do eiusmod tempor incididunt ut labore et dolore magna aliqua.",
+            f"This is paragraph {i+1}, line 3. Ut enim ad minim veniam, quis nostrud exercitation ullamco.",
+            f"This is paragraph {i+1}, line 4. Duis aute irure dolor in reprehenderit in voluptate velit esse.",
+            f"This is paragraph {i+1}, line 5. Excepteur sint occaecat cupidatat non proident, sunt in culpa."
+        ]
+        
+        for para_text in paragraphs:
+            para = Text(para_text, paragraph_font)
+            content.add_child(para)
+            content.add_child(Box((0, 0), (1, 5)))  # Line spacing
+        
+        # Add section spacing
+        content.add_child(Box((0, 0), (1, 15)))
+    
+    return content
+
+
+def demo_viewport_rendering():
+    """Demonstrate viewport rendering at different scroll positions"""
+    
+    print("Creating large document content...")
+    content = create_large_document_content()
+    
+    print(f"Content size: {content.get_content_height()} pixels tall")
+    
+    # Create viewport
+    viewport = Viewport(viewport_size=(800, 600), background_color=(255, 255, 255))
+    
+    # Add content to viewport
+    viewport.add_content(content)
+    
+    print(f"Viewport content size: {viewport.content_size}")
+    print(f"Max scroll Y: {viewport.max_scroll_y}")
+    
+    # Create output directory
+    os.makedirs("output/viewport_demo", exist_ok=True)
+    
+    # Render viewport at different scroll positions
+    scroll_positions = [
+        (0, "top"),
+        (viewport.max_scroll_y // 4, "quarter"),
+        (viewport.max_scroll_y // 2, "middle"),
+        (viewport.max_scroll_y * 3 // 4, "three_quarters"),
+        (viewport.max_scroll_y, "bottom")
+    ]
+    
+    for scroll_y, label in scroll_positions:
+        print(f"Rendering viewport at scroll position {scroll_y} ({label})...")
+        
+        # Scroll to position
+        viewport.scroll_to(0, scroll_y)
+        
+        # Get scroll info
+        scroll_info = viewport.get_scroll_info()
+        print(f"  Scroll progress: {scroll_info['scroll_progress_y']:.2%}")
+        print(f"  Visible elements: {len(viewport.get_visible_elements())}")
+        
+        # Render viewport
+        viewport_img = viewport.render()
+        
+        # Save image
+        output_path = f"output/viewport_demo/viewport_{label}.png"
+        viewport_img.save(output_path)
+        print(f"  Saved: {output_path}")
+    
+    print("\nViewport demo complete!")
+    return viewport
+
+
+def demo_hit_testing():
+    """Demonstrate hit testing in the viewport"""
+    
+    print("\nTesting hit detection...")
+    
+    # Create simple content for hit testing
+    content = ScrollablePageContent(content_width=800, initial_height=100)
+    
+    # Add clickable elements
+    for i in range(10):
+        text = Text(f"Clickable text item {i}", Font(font_size=16))
+        content.add_child(text)
+        content.add_child(Box((0, 0), (1, 20)))  # Spacing
+    
+    # Create viewport
+    viewport = Viewport(viewport_size=(800, 300))
+    viewport.add_content(content)
+    
+    # Test hit detection at different scroll positions
+    test_points = [(100, 50), (200, 100), (300, 150)]
+    
+    for scroll_y in [0, 100, 200]:
+        viewport.scroll_to(0, scroll_y)
+        print(f"\nAt scroll position {scroll_y}:")
+        
+        for point in test_points:
+            hit_element = viewport.hit_test(point)
+            if hit_element:
+                element_text = getattr(hit_element, '_text', 'Unknown element')
+                print(f"  Point {point}: Hit '{element_text}'")
+            else:
+                print(f"  Point {point}: No element")
+
+
+def demo_scroll_methods():
+    """Demonstrate different scrolling methods"""
+    
+    print("\nTesting scroll methods...")
+    
+    # Create content
+    content = ScrollablePageContent(content_width=800, initial_height=100)
+    
+    for i in range(20):
+        text = Text(f"Line {i+1}: This is some sample text for scrolling demo", Font(font_size=14))
+        content.add_child(text)
+        content.add_child(Box((0, 0), (1, 5)))
+    
+    # Create viewport
+    viewport = Viewport(viewport_size=(800, 200))
+    viewport.add_content(content)
+    
+    print(f"Content height: {viewport.content_size[1]}")
+    print(f"Viewport height: {viewport.viewport_size[1]}")
+    print(f"Max scroll Y: {viewport.max_scroll_y}")
+    
+    # Test different scroll methods
+    print("\nTesting scroll methods:")
+    
+    # Scroll by lines
+    print("Scrolling down 5 lines...")
+    for i in range(5):
+        viewport.scroll_line_down(20)
+        print(f"  After line {i+1}: offset = {viewport.viewport_offset}")
+    
+    # Scroll by pages
+    print("Scrolling down 1 page...")
+    viewport.scroll_page_down()
+    print(f"  After page down: offset = {viewport.viewport_offset}")
+    
+    # Scroll to bottom
+    print("Scrolling to bottom...")
+    viewport.scroll_to_bottom()
+    print(f"  At bottom: offset = {viewport.viewport_offset}")
+    
+    # Scroll to top
+    print("Scrolling to top...")
+    viewport.scroll_to_top()
+    print(f"  At top: offset = {viewport.viewport_offset}")
+
+
+def main():
+    """Run all viewport demos"""
+    print("=== Viewport System Demo ===")
+    
+    # Demo 1: Basic viewport rendering
+    viewport = demo_viewport_rendering()
+    
+    # Demo 2: Hit testing
+    demo_hit_testing()
+    
+    # Demo 3: Scroll methods
+    demo_scroll_methods()
+    
+    print("\n=== Demo Complete ===")
+    print("Check the output/viewport_demo/ directory for rendered images.")
+    
+    # Show final scroll info
+    scroll_info = viewport.get_scroll_info()
+    print(f"\nFinal viewport state:")
+    print(f"  Content size: {scroll_info['content_size']}")
+    print(f"  Viewport size: {scroll_info['viewport_size']}")
+    print(f"  Current offset: {scroll_info['offset']}")
+    print(f"  Scroll progress: {scroll_info['scroll_progress_y']:.1%}")
+    print(f"  Can scroll up: {scroll_info['can_scroll_up']}")
+    print(f"  Can scroll down: {scroll_info['can_scroll_down']}")
+
+
+if __name__ == "__main__":
+    main()

html_browser_with_viewport.py

@@ -0,0 +1,771 @@
+#!/usr/bin/env python3
+"""
+Enhanced HTML Browser using pyWebLayout with Viewport System
+
+This browser uses the new viewport system to enable efficient scrolling
+within HTML pages, only rendering the visible portion of large documents.
+"""
+
+import re
+import tkinter as tk
+from tkinter import ttk, messagebox, filedialog, simpledialog
+from PIL import Image, ImageTk, ImageDraw
+from typing import Dict, List, Optional, Tuple, Any
+import webbrowser
+import os
+from urllib.parse import urljoin, urlparse
+import requests
+from io import BytesIO
+import pyperclip
+
+# Import pyWebLayout components including the new viewport system
+from pyWebLayout.concrete import (
+    Page, Container, Box, Text, RenderableImage, 
+    RenderableLink, RenderableButton, RenderableForm, RenderableFormField,
+    Viewport, ScrollablePageContent
+)
+from pyWebLayout.abstract.functional import (
+    Link, Button, Form, FormField, LinkType, FormFieldType
+)
+from pyWebLayout.abstract.block import Paragraph
+from pyWebLayout.abstract.inline import Word
+from pyWebLayout.style.fonts import Font, FontWeight, FontStyle, TextDecoration
+from pyWebLayout.style.layout import Alignment
+from pyWebLayout.typesetting.paragraph_layout import ParagraphLayout, ParagraphLayoutResult
+from pyWebLayout.io.readers.html_extraction import parse_html_string
+
+
+class HTMLViewportAdapter:
+    """Adapter to convert HTML to viewport using the proper HTML extraction system"""
+    
+    def __init__(self):
+        pass
+        
+    def parse_html_string(self, html_content: str, base_url: str = "", viewport_size: Tuple[int, int] = (800, 600)) -> Viewport:
+        """Parse HTML string and return a Viewport object with scrollable content using the proper parser"""
+        # Use the proper HTML extraction system
+        base_font = Font(font_size=14)
+        blocks = parse_html_string(html_content, base_font)
+        
+        # Extract title
+        title_match = re.search(r'<title>(.*?)</title>', html_content, re.IGNORECASE)
+        title = title_match.group(1) if title_match else "Untitled"
+        
+        # Create scrollable content container
+        content = ScrollablePageContent(content_width=viewport_size[0] - 20, initial_height=100)
+        
+        # Convert abstract blocks to renderable objects using Page's conversion system
+        page = Page(size=(viewport_size[0], 10000))  # Large temporary page
+        
+        # Add blocks to page and let it handle the conversion
+        for i, block in enumerate(blocks):
+            renderable = page._convert_block_to_renderable(block)
+            if renderable:
+                content.add_child(renderable)
+                # Add spacing between blocks (but not after the last block)
+                if i < len(blocks) - 1:
+                    content.add_child(Box((0, 0), (1, 8)))
+        
+        # Create viewport and add the content
+        viewport = Viewport(viewport_size=viewport_size, background_color=(255, 255, 255))
+        viewport.add_content(content)
+        viewport.title = title
+        
+        return viewport
+    
+    def parse_html_file(self, file_path: str, viewport_size: Tuple[int, int] = (800, 600)) -> Viewport:
+        """Parse HTML file and return a Viewport object"""
+        try:
+            with open(file_path, 'r', encoding='utf-8') as f:
+                html_content = f.read()
+            base_url = os.path.dirname(os.path.abspath(file_path))
+            return self.parse_html_string(html_content, base_url, viewport_size)
+        except Exception as e:
+            # Create error viewport
+            content = ScrollablePageContent(content_width=viewport_size[0] - 20, initial_height=100)
+            error_text = Text(f"Error loading file: {str(e)}", Font(font_size=16, colour=(255, 0, 0)))
+            content.add_child(error_text)
+            
+            viewport = Viewport(viewport_size=viewport_size, background_color=(255, 255, 255))
+            viewport.add_content(content)
+            viewport.title = "Error"
+            return viewport
+
+
+class ViewportBrowserWindow:
+    """Enhanced browser window using Tkinter with Viewport support"""
+    
+    def __init__(self):
+        self.root = tk.Tk()
+        self.root.title("pyWebLayout HTML Browser with Viewport")
+        self.root.geometry("1000x800")
+        
+        self.current_viewport = None
+        self.history = []
+        self.history_index = -1
+        
+        # Scrolling parameters
+        self.scroll_speed = 20  # pixels per scroll
+        self.page_scroll_ratio = 0.9  # fraction of viewport height for page scroll
+        
+        # Text selection variables
+        self.selection_start = None
+        self.selection_end = None
+        self.is_selecting = False
+        self.selected_text = ""
+        self.text_elements = []
+        self.selection_overlay = None
+        
+        self.setup_ui()
+        
+    def setup_ui(self):
+        """Setup the user interface"""
+        # Create main frame
+        main_frame = ttk.Frame(self.root)
+        main_frame.pack(fill=tk.BOTH, expand=True, padx=5, pady=5)
+        
+        # Navigation frame
+        nav_frame = ttk.Frame(main_frame)
+        nav_frame.pack(fill=tk.X, pady=(0, 5))
+        
+        # Navigation buttons
+        self.back_btn = ttk.Button(nav_frame, text="←", command=self.go_back, state=tk.DISABLED)
+        self.back_btn.pack(side=tk.LEFT, padx=(0, 5))
+        
+        self.forward_btn = ttk.Button(nav_frame, text="→", command=self.go_forward, state=tk.DISABLED)
+        self.forward_btn.pack(side=tk.LEFT, padx=(0, 5))
+        
+        self.refresh_btn = ttk.Button(nav_frame, text="⟳", command=self.refresh)
+        self.refresh_btn.pack(side=tk.LEFT, padx=(0, 10))
+        
+        # Address bar
+        ttk.Label(nav_frame, text="URL:").pack(side=tk.LEFT)
+        self.url_var = tk.StringVar()
+        self.url_entry = ttk.Entry(nav_frame, textvariable=self.url_var, width=50)
+        self.url_entry.pack(side=tk.LEFT, fill=tk.X, expand=True, padx=(5, 5))
+        self.url_entry.bind('<Return>', self.navigate_to_url)
+        
+        self.go_btn = ttk.Button(nav_frame, text="Go", command=self.navigate_to_url)
+        self.go_btn.pack(side=tk.LEFT, padx=(0, 10))
+        
+        # File operations
+        self.open_btn = ttk.Button(nav_frame, text="Open File", command=self.open_file)
+        self.open_btn.pack(side=tk.LEFT)
+        
+        # Content frame with scrollbars and viewport controls
+        content_frame = ttk.Frame(main_frame)
+        content_frame.pack(fill=tk.BOTH, expand=True)
+        
+        # Create scrollbar frame
+        scroll_frame = ttk.Frame(content_frame)
+        scroll_frame.pack(side=tk.RIGHT, fill=tk.Y)
+        
+        # Vertical scrollbar
+        self.v_scrollbar = ttk.Scrollbar(scroll_frame, orient=tk.VERTICAL, command=self.on_scrollbar)
+        self.v_scrollbar.pack(fill=tk.Y)
+        
+        # Canvas for displaying viewport content
+        self.canvas = tk.Canvas(content_frame, bg='white')
+        self.canvas.pack(side=tk.LEFT, fill=tk.BOTH, expand=True)
+        
+        # Scroll info frame
+        info_frame = ttk.Frame(main_frame)
+        info_frame.pack(fill=tk.X, pady=(5, 0))
+        
+        self.scroll_info_var = tk.StringVar(value="")
+        ttk.Label(info_frame, textvariable=self.scroll_info_var).pack(side=tk.LEFT)
+        
+        # Status bar
+        self.status_var = tk.StringVar(value="Ready")
+        status_bar = ttk.Label(main_frame, textvariable=self.status_var, relief=tk.SUNKEN)
+        status_bar.pack(fill=tk.X, pady=(5, 0))
+        
+        # Bind events
+        self.canvas.bind('<Button-1>', self.on_click)
+        self.canvas.bind('<B1-Motion>', self.on_drag)
+        self.canvas.bind('<ButtonRelease-1>', self.on_release)
+        self.canvas.bind('<Motion>', self.on_mouse_move)
+        self.canvas.bind('<MouseWheel>', self.on_mouse_wheel)  # Windows/Mac
+        self.canvas.bind('<Button-4>', self.on_mouse_wheel)     # Linux scroll up
+        self.canvas.bind('<Button-5>', self.on_mouse_wheel)     # Linux scroll down
+        
+        # Keyboard shortcuts
+        self.root.bind('<Control-c>', self.copy_selection)
+        self.root.bind('<Control-a>', self.select_all)
+        self.root.bind('<Prior>', self.page_up)         # Page Up
+        self.root.bind('<Next>', self.page_down)        # Page Down
+        self.root.bind('<Home>', self.scroll_to_top)    # Home
+        self.root.bind('<End>', self.scroll_to_bottom)  # End
+        self.root.bind('<Up>', lambda e: self.scroll_by_lines(-1))
+        self.root.bind('<Down>', lambda e: self.scroll_by_lines(1))
+        
+        # Context menu
+        self.setup_context_menu()
+        
+        # Make canvas focusable
+        self.canvas.config(highlightthickness=1)
+        self.canvas.focus_set()
+        
+        # Load default page
+        self.load_default_page()
+    
+    def setup_context_menu(self):
+        """Setup the right-click context menu"""
+        self.context_menu = tk.Menu(self.root, tearoff=0)
+        self.context_menu.add_command(label="Copy", command=self.copy_selection)
+        self.context_menu.add_command(label="Select All", command=self.select_all)
+        self.context_menu.add_separator()
+        self.context_menu.add_command(label="Scroll to Top", command=self.scroll_to_top)
+        self.context_menu.add_command(label="Scroll to Bottom", command=self.scroll_to_bottom)
+        
+        # Bind right-click to show context menu
+        self.canvas.bind('<Button-3>', self.show_context_menu)
+    
+    def show_context_menu(self, event):
+        """Show context menu at mouse position"""
+        try:
+            self.context_menu.tk_popup(event.x_root, event.y_root)
+        finally:
+            self.context_menu.grab_release()
+    
+    def on_scrollbar(self, *args):
+        """Handle scrollbar movement"""
+        if not self.current_viewport:
+            return
+        
+        action = args[0]
+        
+        if action == "moveto":
+            # Absolute position (0.0 to 1.0)
+            fraction = float(args[1])
+            max_scroll = self.current_viewport.max_scroll_y
+            new_y = int(fraction * max_scroll)
+            self.current_viewport.scroll_to(0, new_y)
+            self.update_viewport_display()
+            
+        elif action in ["scroll", "step"]:
+            # Relative movement
+            direction = int(args[1])
+            # Handle the units parameter properly - it might be a string "units"
+            if len(args) > 2:
+                try:
+                    units = int(args[2])
+                except ValueError:
+                    # If args[2] is "units" string, default to 1
+                    units = 1
+            else:
+                units = 1
+            
+            if action == "scroll":
+                # Line-based scrolling
+                self.scroll_by_lines(direction * units)
+            elif action == "step":
+                # Page-based scrolling
+                self.scroll_by_pages(direction * units)
+    
+    def update_scrollbar(self):
+        """Update scrollbar position and size based on viewport state"""
+        if not self.current_viewport:
+            self.v_scrollbar.set(0, 1)
+            return
+        
+        scroll_info = self.current_viewport.get_scroll_info()
+        content_height = scroll_info['content_size'][1]
+        viewport_height = scroll_info['viewport_size'][1]
+        current_offset = scroll_info['offset'][1]
+        
+        if content_height <= viewport_height:
+            # No scrolling needed
+            self.v_scrollbar.set(0, 1)
+        else:
+            # Calculate scrollbar position and size
+            top_fraction = current_offset / content_height
+            bottom_fraction = (current_offset + viewport_height) / content_height
+            self.v_scrollbar.set(top_fraction, bottom_fraction)
+        
+        # Update scroll info display
+        progress = scroll_info['scroll_progress_y']
+        self.scroll_info_var.set(f"Scroll: {progress:.1%} ({current_offset}/{content_height - viewport_height})")
+    
+    def on_mouse_wheel(self, event):
+        """Handle mouse wheel scrolling"""
+        if not self.current_viewport:
+            return
+        
+        # Determine scroll direction and amount
+        if event.num == 4 or event.delta > 0:
+            # Scroll up
+            self.scroll_by_lines(-3)
+        elif event.num == 5 or event.delta < 0:
+            # Scroll down
+            self.scroll_by_lines(3)
+    
+    def scroll_by_lines(self, lines: int):
+        """Scroll by a number of lines"""
+        if not self.current_viewport:
+            return
+        
+        delta_y = lines * self.scroll_speed
+        self.current_viewport.scroll_by(0, delta_y)
+        self.update_viewport_display()
+    
+    def scroll_by_pages(self, pages: int):
+        """Scroll by a number of pages"""
+        if not self.current_viewport:
+            return
+        
+        viewport_height = self.current_viewport.viewport_size[1]
+        delta_y = int(pages * viewport_height * self.page_scroll_ratio)
+        self.current_viewport.scroll_by(0, delta_y)
+        self.update_viewport_display()
+    
+    def page_up(self, event=None):
+        """Scroll up by one page"""
+        self.scroll_by_pages(-1)
+    
+    def page_down(self, event=None):
+        """Scroll down by one page"""
+        self.scroll_by_pages(1)
+    
+    def scroll_to_top(self, event=None):
+        """Scroll to the top of the document"""
+        if not self.current_viewport:
+            return
+        
+        self.current_viewport.scroll_to_top()
+        self.update_viewport_display()
+    
+    def scroll_to_bottom(self, event=None):
+        """Scroll to the bottom of the document"""
+        if not self.current_viewport:
+            return
+        
+        self.current_viewport.scroll_to_bottom()
+        self.update_viewport_display()
+    
+    def update_viewport_display(self):
+        """Update the canvas display with current viewport content"""
+        if not self.current_viewport:
+            return
+        
+        try:
+            # Render the current viewport
+            viewport_img = self.current_viewport.render()
+            
+            # Convert to PhotoImage for Tkinter
+            self.photo = ImageTk.PhotoImage(viewport_img)
+            
+            # Clear canvas and display image
+            self.canvas.delete("all")
+            self.canvas.create_image(0, 0, anchor=tk.NW, image=self.photo)
+            
+            # Update scrollbar
+            self.update_scrollbar()
+            
+        except Exception as e:
+            self.status_var.set(f"Error rendering viewport: {str(e)}")
+    
+    def on_drag(self, event):
+        """Handle mouse dragging for text selection"""
+        canvas_x = self.canvas.canvasx(event.x)
+        canvas_y = self.canvas.canvasy(event.y)
+        
+        if not self.is_selecting:
+            # Start selection
+            self.is_selecting = True
+            self.selection_start = (canvas_x, canvas_y)
+            self.selection_end = (canvas_x, canvas_y)
+        else:
+            # Update selection end
+            self.selection_end = (canvas_x, canvas_y)
+        
+        # Update visual selection
+        self.update_selection_visual()
+        
+        # Update status
+        self.status_var.set("Selecting text...")
+    
+    def on_release(self, event):
+        """Handle mouse release to complete text selection"""
+        if self.is_selecting:
+            canvas_x = self.canvas.canvasx(event.x)
+            canvas_y = self.canvas.canvasy(event.y)
+            self.selection_end = (canvas_x, canvas_y)
+            
+            # Extract selected text
+            self.extract_selected_text()
+            
+            # Update status
+            if self.selected_text:
+                self.status_var.set(f"Selected: {len(self.selected_text)} characters")
+            else:
+                self.status_var.set("No text selected")
+                self.clear_selection()
+    
+    def update_selection_visual(self):
+        """Update the visual representation of text selection"""
+        # Remove existing selection overlay
+        if self.selection_overlay:
+            self.canvas.delete(self.selection_overlay)
+        
+        if self.selection_start and self.selection_end:
+            # Create selection rectangle
+            x1, y1 = self.selection_start
+            x2, y2 = self.selection_end
+            
+            # Ensure proper coordinates (top-left to bottom-right)
+            left = min(x1, x2)
+            top = min(y1, y2)
+            right = max(x1, x2)
+            bottom = max(y1, y2)
+            
+            # Draw selection rectangle with transparency effect
+            self.selection_overlay = self.canvas.create_rectangle(
+                left, top, right, bottom,
+                fill='blue', stipple='gray50', outline='blue', width=1
+            )
+    
+    def extract_selected_text(self):
+        """Extract text that falls within the selection area"""
+        if not self.selection_start or not self.selection_end or not self.current_viewport:
+            self.selected_text = ""
+            return
+        
+        # Get selection bounds in viewport coordinates
+        x1, y1 = self.selection_start
+        x2, y2 = self.selection_end
+        left = min(x1, x2)
+        top = min(y1, y2)
+        right = max(x1, x2)
+        bottom = max(y1, y2)
+        
+        # Convert to content coordinates
+        viewport_offset = self.current_viewport.viewport_offset
+        content_left = left + viewport_offset[0]
+        content_top = top + viewport_offset[1]
+        content_right = right + viewport_offset[0]
+        content_bottom = bottom + viewport_offset[1]
+        
+        # Extract text elements in selection area
+        selected_elements = []
+        visible_elements = self.current_viewport.get_visible_elements()
+        
+        for element, visible_origin, visible_size, clip_info in visible_elements:
+            # Check if element intersects with selection
+            elem_left = visible_origin[0]
+            elem_top = visible_origin[1]
+            elem_right = elem_left + visible_size[0]
+            elem_bottom = elem_top + visible_size[1]
+            
+            if (elem_left < right and elem_right > left and 
+                elem_top < bottom and elem_bottom > top):
+                
+                # Extract text from element
+                if hasattr(element, '_text'):
+                    text_content = element._text
+                    if text_content.strip():
+                        selected_elements.append((text_content.strip(), elem_left, elem_top))
+        
+        # Sort by position (top to bottom, left to right)
+        selected_elements.sort(key=lambda x: (x[2], x[1]))
+        
+        # Combine text
+        self.selected_text = " ".join([element[0] for element in selected_elements])
+    
+    def copy_selection(self, event=None):
+        """Copy selected text to clipboard"""
+        if self.selected_text:
+            try:
+                pyperclip.copy(self.selected_text)
+                self.status_var.set(f"Copied {len(self.selected_text)} characters to clipboard")
+            except Exception as e:
+                self.status_var.set(f"Error copying to clipboard: {str(e)}")
+        else:
+            self.status_var.set("No text selected to copy")
+    
+    def select_all(self, event=None):
+        """Select all text on the page"""
+        if not self.current_viewport:
+            return
+        
+        # Set selection to entire viewport area
+        viewport_size = self.current_viewport.viewport_size
+        
+        self.selection_start = (0, 0)
+        self.selection_end = viewport_size
+        self.is_selecting = True
+        
+        # Extract all visible text
+        self.extract_selected_text()
+        
+        # Update visual
+        self.update_selection_visual()
+        
+        if self.selected_text:
+            self.status_var.set(f"Selected all visible text: {len(self.selected_text)} characters")
+        else:
+            self.status_var.set("No text found to select")
+    
+    def clear_selection(self):
+        """Clear the current text selection"""
+        self.selection_start = None
+        self.selection_end = None
+        self.is_selecting = False
+        self.selected_text = ""
+        
+        # Remove visual selection
+        if self.selection_overlay:
+            self.canvas.delete(self.selection_overlay)
+            self.selection_overlay = None
+        
+        self.status_var.set("Selection cleared")
+    
+    def load_default_page(self):
+        """Load a default welcome page"""
+        html_content = """
+        <html>
+        <head><title>pyWebLayout Browser with Viewport - Welcome</title></head>
+        <body>
+            <h1>Welcome to pyWebLayout Browser with Viewport System</h1>
+            <p>This enhanced browser uses the new viewport system for efficient scrolling through large documents.</p>
+            
+            <h2>New Viewport Features:</h2>
+            <ul>
+                <li><b>Efficient Rendering:</b> Only visible content is rendered</li>
+                <li><b>Smooth Scrolling:</b> Mouse wheel, keyboard, and scrollbar support</li>
+                <li><b>Large Document Support:</b> Handle documents of any size</li>
+                <li><b>Memory Efficient:</b> Low memory usage even for huge pages</li>
+            </ul>
+            
+            <h2>Scrolling Controls:</h2>
+            <p><b>Mouse Wheel:</b> Scroll up and down</p>
+            <p><b>Page Up/Down:</b> Scroll by viewport height</p>
+            <p><b>Home/End:</b> Jump to top/bottom</p>
+            <p><b>Arrow Keys:</b> Scroll line by line</p>
+            <p><b>Scrollbar:</b> Click and drag for precise positioning</p>
+            
+            <h2>Text Selection:</h2>
+            <p>Click and drag to select text, then use <b>Ctrl+C</b> to copy</p>
+            <p>Use <b>Ctrl+A</b> to select all visible text</p>
+            
+            <h3>Try scrolling with different methods!</h3>
+            <p>This page demonstrates the viewport system. All the content above and below is efficiently managed.</p>
+            
+            <h2>Sample Content for Scrolling</h2>
+            <p>Here's some additional content to demonstrate scrolling capabilities:</p>
+            
+            <h3>Lorem Ipsum</h3>
+            <p>Lorem ipsum dolor sit amet, consectetur adipiscing elit. Sed do eiusmod tempor incididunt ut labore et dolore magna aliqua.</p>
+            <p>Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat.</p>
+            <p>Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur.</p>
+            
+            <h3>More Sample Text</h3>
+            <p>Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum.</p>
+            <p>Sed ut perspiciatis unde omnis iste natus error sit voluptatem accusantium doloremque laudantium.</p>
+            <p>Totam rem aperiam, eaque ipsa quae ab illo inventore veritatis et quasi architecto beatae vitae dicta sunt explicabo.</p>
+            
+            <h3>Even More Content</h3>
+            <p>Nemo enim ipsam voluptatem quia voluptas sit aspernatur aut odit aut fugit, sed quia consequuntur magni dolores eos.</p>
+            <p>Qui ratione voluptatem sequi nesciunt. Neque porro quisquam est, qui dolorem ipsum quia dolor sit amet.</p>
+            <p>Consectetur, adipisci velit, sed quia non numquam eius modi tempora incidunt ut labore et dolore magnam aliquam quaerat voluptatem.</p>
+            
+            <h2>Technical Details</h2>
+            <p>The viewport system works by:</p>
+            <ul>
+                <li>Creating a large content container that can hold any amount of content</li>
+                <li>Providing a viewport window that shows only a portion of the content</li>
+                <li>Efficiently calculating which elements are visible</li>
+                <li>Rendering only the visible elements</li>
+                <li>Supporting smooth scrolling through the content</li>
+            </ul>
+            
+            <p>This allows handling of very large documents without performance issues.</p>
+            
+            <h2>Load Your Own Content</h2>
+            <p>Use the "Open File" button to load local HTML files, or enter a URL in the address bar.</p>
+        </body>
+        </html>
+        """
+        
+        # Get current canvas size for viewport
+        self.root.update_idletasks()
+        canvas_width = max(800, self.canvas.winfo_width())
+        canvas_height = max(600, self.canvas.winfo_height())
+        
+        parser = HTMLViewportAdapter()
+        self.current_viewport = parser.parse_html_string(html_content, viewport_size=(canvas_width, canvas_height))
+        
+        # Update window title
+        if hasattr(self.current_viewport, 'title'):
+            self.root.title(f"pyWebLayout Browser - {self.current_viewport.title}")
+        
+        self.update_viewport_display()
+        self.status_var.set("Welcome page loaded with viewport system")
+    
+    def navigate_to_url(self, event=None):
+        """Navigate to the URL in the address bar"""
+        url = self.url_var.get().strip()
+        if not url:
+            return
+            
+        self.status_var.set(f"Loading {url}...")
+        self.root.update()
+        
+        # Get current canvas size for viewport
+        self.root.update_idletasks()
+        canvas_width = max(800, self.canvas.winfo_width())
+        canvas_height = max(600, self.canvas.winfo_height())
+        
+        try:
+            parser = HTMLViewportAdapter()
+            
+            if url.startswith(('http://', 'https://')):
+                # Web URL
+                response = requests.get(url, timeout=10)
+                response.raise_for_status()
+                html_content = response.text
+                
+                self.current_viewport = parser.parse_html_string(html_content, url, (canvas_width, canvas_height))
+                
+            elif os.path.isfile(url):
+                # Local file
+                self.current_viewport = parser.parse_html_file(url, (canvas_width, canvas_height))
+                
+            else:
+                # Try to treat as a local file path
+                if not url.startswith('file://'):
+                    url = 'file://' + os.path.abspath(url)
+                
+                file_path = url.replace('file://', '')
+                if os.path.isfile(file_path):
+                    self.current_viewport = parser.parse_html_file(file_path, (canvas_width, canvas_height))
+                else:
+                    raise FileNotFoundError(f"File not found: {file_path}")
+            
+            # Update window title
+            if hasattr(self.current_viewport, 'title'):
+                self.root.title(f"pyWebLayout Browser - {self.current_viewport.title}")
+            
+            # Add to history
+            self.add_to_history(url)
+            self.update_viewport_display()
+            self.status_var.set(f"Loaded {url}")
+            
+        except Exception as e:
+            self.status_var.set(f"Error loading {url}: {str(e)}")
+            messagebox.showerror("Error", f"Failed to load {url}:\n{str(e)}")
+    
+    def open_file(self):
+        """Open a local HTML file"""
+        file_path = filedialog.askopenfilename(
+            title="Open HTML File",
+            filetypes=[("HTML files", "*.html *.htm"), ("All files", "*.*")]
+        )
+        
+        if file_path:
+            self.url_var.set(file_path)
+            self.navigate_to_url()
+    
+    def on_click(self, event):
+        """Handle mouse clicks on the canvas"""
+        if not self.current_viewport:
+            return
+        
+        # Convert canvas coordinates to viewport coordinates
+        canvas_x = self.canvas.canvasx(event.x)
+        canvas_y = self.canvas.canvasy(event.y)
+        
+        # Use viewport hit testing
+        hit_element = self.current_viewport.hit_test((canvas_x, canvas_y))
+        
+        if hit_element and hasattr(hit_element, '_callback'):
+            # Handle clickable elements
+            try:
+                result = hit_element._callback()
+                if result:
+                    self.status_var.set(result)
+                    
+                    # For external links, open in system browser
+                    if hasattr(hit_element, '_link') and hit_element._link.link_type == LinkType.EXTERNAL:
+                        webbrowser.open(hit_element._link.location)
+            except Exception as e:
+                self.status_var.set(f"Click error: {str(e)}")
+    
+    def on_mouse_move(self, event):
+        """Handle mouse movement for hover effects"""
+        if not self.current_viewport:
+            return
+        
+        # Convert canvas coordinates to viewport coordinates
+        canvas_x = self.canvas.canvasx(event.x)
+        canvas_y = self.canvas.canvasy(event.y)
+        
+        # Check if mouse is over any clickable element
+        hit_element = self.current_viewport.hit_test((canvas_x, canvas_y))
+        
+        if hit_element and hasattr(hit_element, '_callback'):
+            self.canvas.configure(cursor="hand2")
+        else:
+            self.canvas.configure(cursor="arrow")
+    
+    def add_to_history(self, url):
+        """Add URL to navigation history"""
+        # Remove any forward history
+        self.history = self.history[:self.history_index + 1]
+        
+        # Add new URL
+        self.history.append(url)
+        self.history_index = len(self.history) - 1
+        
+        # Update navigation buttons
+        self.update_nav_buttons()
+    
+    def update_nav_buttons(self):
+        """Update the state of navigation buttons"""
+        self.back_btn.configure(state=tk.NORMAL if self.history_index > 0 else tk.DISABLED)
+        self.forward_btn.configure(state=tk.NORMAL if self.history_index < len(self.history) - 1 else tk.DISABLED)
+    
+    def go_back(self):
+        """Navigate back in history"""
+        if self.history_index > 0:
+            self.history_index -= 1
+            url = self.history[self.history_index]
+            self.url_var.set(url)
+            self.navigate_to_url()
+    
+    def go_forward(self):
+        """Navigate forward in history"""
+        if self.history_index < len(self.history) - 1:
+            self.history_index += 1
+            url = self.history[self.history_index]
+            self.url_var.set(url)
+            self.navigate_to_url()
+    
+    def refresh(self):
+        """Refresh the current page"""
+        current_url = self.url_var.get()
+        if current_url:
+            self.navigate_to_url()
+        else:
+            self.load_default_page()
+    
+    def run(self):
+        """Start the browser"""
+        self.root.mainloop()
+
+
+def main():
+    """Main function to run the enhanced browser"""
+    print("Starting pyWebLayout HTML Browser with Viewport System...")
+    
+    try:
+        browser = ViewportBrowserWindow()
+        browser.run()
+    except Exception as e:
+        print(f"Error starting browser: {e}")
+        import traceback
+        traceback.print_exc()
+
+
+if __name__ == "__main__":
+    main()

pyWebLayout/concrete/__init__.py

@@ -3,3 +3,4 @@ from .page import Container, Page
 from .text import Text, Line
 from .functional import RenderableLink, RenderableButton, RenderableForm, RenderableFormField
 from .image import RenderableImage
+from .viewport import Viewport, ScrollablePageContent

pyWebLayout/concrete/page.py

@@ -542,7 +542,7 @@ class Page(Container):
                         break
             
             # Add the line if it has any words
-            if len(line.renderable_words) > 0:
+            if len(line._text_objects) > 0:
                 lines.append(line)
                 line_y_offset += line_height
             else:

pyWebLayout/concrete/text.py

@@ -144,11 +144,11 @@ class JustifyAlignmentHandler(AlignmentHandler):
         
         if num_spaces > 0:
             projected_spacing = available_space // num_spaces
-            # Be more conservative about hyphenation - only suggest it if spacing would be very large
-            # Use a higher threshold to avoid unnecessary hyphenation
-            max_acceptable_spacing = spacing * 3  # Allow up to 3x normal spacing before hyphenating
-            # Also ensure we have a minimum threshold to avoid hyphenating for tiny improvements
-            min_threshold_for_hyphenation = spacing + 10  # At least 10 pixels above min spacing
+            # Be much more conservative about hyphenation - only suggest it if spacing would be extremely large
+            # Increase the threshold significantly to avoid mid-sentence hyphenation
+            max_acceptable_spacing = spacing * 5  # Allow up to 5x normal spacing before hyphenating
+            # Increase minimum threshold to make hyphenation much less likely
+            min_threshold_for_hyphenation = spacing + 20  # At least 20 pixels above min spacing
             return projected_spacing > max(max_acceptable_spacing, min_threshold_for_hyphenation)
             
         return False
@@ -402,6 +402,54 @@ class Line(Box):
         """Set the next line in sequence"""
         self._next = line
 
+    def _calculate_available_width(self, font: Font) -> int:
+        """Calculate available width for adding a word."""
+        min_spacing = self._spacing[0]
+        spacing_needed = min_spacing if self._text_objects else 0
+        safety_margin = self._get_safety_margin(font)
+        return int(self._size[0] - self._current_width - spacing_needed - safety_margin)
+    
+    def _get_safety_margin(self, font: Font) -> int:
+        """Calculate safety margin to prevent text cropping."""
+        return max(1, int(font.font_size * 0.05))  # 5% of font size
+    
+    def _fits_with_normal_spacing(self, word_width: int, available_width: int, font: Font) -> bool:
+        """Check if word fits with normal spacing."""
+        if word_width > available_width:
+            return False
+        
+        # Check if alignment handler suggests hyphenation anyway
+        should_hyphenate = self._alignment_handler.should_try_hyphenation(
+            self._text_objects, word_width, available_width, self._spacing[0], font)
+        return not should_hyphenate
+    
+    def _add_word_with_normal_spacing(self, text: str, font: Font, word_width: int) -> None:
+        """Add word to line with normal spacing."""
+        spacing_needed = self._spacing[0] if self._text_objects else 0
+        
+        text_obj = Text(text, font)
+        text_obj.add_to_line(self)
+        self._text_objects.append(text_obj)
+        
+        self._current_width += spacing_needed + word_width
+        return None
+    
+    def _try_hyphenation(self, text: str, font: Font, available_width: int) -> Union[str, None]:
+        """Try hyphenation to fit part of the word."""
+        spacing_needed = self._spacing[0] if self._text_objects else 0
+        safety_margin = self._get_safety_margin(font)
+        return self._try_hyphenation_or_fit(text, font, available_width, spacing_needed, safety_margin)
+    
+    def _handle_word_overflow(self, text: str, font: Font, available_width: int) -> str:
+        """Handle case where word doesn't fit."""
+        if self._text_objects:
+            # Line already has words, move this word to the next line
+            return text
+        else:
+            # Empty line with word that's too long - force fit as last resort
+            safety_margin = self._get_safety_margin(font)
+            return self._force_fit_long_word(text, font, available_width + safety_margin)
+
     def _try_reduced_spacing_fit(self, text: str, font: Font, word_width: int, safety_margin: int) -> Union[None, str]:
         """
         Try to fit the word by reducing spacing between existing words.
@@ -490,14 +538,7 @@ class Line(Box):
 
     def add_word(self, text: str, font: Optional[Font] = None) -> Union[None, str]:
         """
-        Add a word to this line as a Text object using intelligent word fitting strategies.
-        
-        This method implements a comprehensive word fitting algorithm that:
-        1. First tries to fit the word with normal spacing
-        2. If that fails, tries reducing spacing to minimize gaps
-        3. Uses hyphenation when beneficial for spacing quality
-        4. Falls back to moving the word to the next line
-        5. As a last resort, force-fits long words
+        Add a word to this line using intelligent word fitting strategies.
         
         Args:
             text: The text content of the word
@@ -509,63 +550,26 @@ class Line(Box):
         if not font:
             font = self._font
         
-        # Create a Text object to measure the word
-        text_obj = Text(text, font)
-        word_width = text_obj.width
+        available_width = self._calculate_available_width(font)
+        word_width = Text(text, font).width
         
-        # If this is the first word, no spacing is needed
-        min_spacing, max_spacing = self._spacing
-        spacing_needed = min_spacing if self._text_objects else 0
+        # Strategy 1: Try normal spacing first
+        if self._fits_with_normal_spacing(word_width, available_width, font):
+            return self._add_word_with_normal_spacing(text, font, word_width)
         
-        # Add a small margin to prevent edge cases where words appear to fit but get cropped
-        safety_margin = max(1, int(font.font_size * 0.05))  # 5% of font size as safety margin
-        
-        # Check if word fits in the line with safety margin
-        available_width = self._size[0] - self._current_width - spacing_needed - safety_margin
-        
-        # Strategy 1: Try to fit with normal spacing
-        if word_width <= available_width:
-            # Check if alignment handler suggests hyphenation for better spacing quality
-            should_hyphenate = self._alignment_handler.should_try_hyphenation(
-                self._text_objects, word_width, available_width, min_spacing, font)
-            
-            if not should_hyphenate:
-                # Word fits with normal spacing and no hyphenation needed - add it
-                text_obj.add_to_line(self)
-                self._text_objects.append(text_obj)
-                self._current_width += spacing_needed + word_width
-                return None
-            else:
-                # Word fits but hyphenation might improve spacing - try it
-                hyphen_result = self._try_hyphenation_or_fit(text, font, available_width, spacing_needed, safety_margin)
-                if hyphen_result is None:
-                    # Hyphenation worked and improved spacing
-                    return None
-                # If hyphenation didn't work or didn't improve things, fall through to add the whole word
-                text_obj.add_to_line(self)
-                self._text_objects.append(text_obj)
-                self._current_width += spacing_needed + word_width
+        # Strategy 2: Try reduced spacing
+        if self._text_objects:
+            result = self._try_reduced_spacing_fit(text, font, word_width, self._get_safety_margin(font))
+            if result is None:
                 return None
         
-        # Strategy 2: Try reducing spacing to maximize fit
-        if self._text_objects and word_width > available_width:
-            reduced_spacing_result = self._try_reduced_spacing_fit(text, font, word_width, safety_margin)
-            if reduced_spacing_result is None:
-                # Word fitted by reducing spacing
-                return None
-        
-        # Strategy 3: Try hyphenation to fit part of the word
-        hyphen_result = self._try_hyphenation_or_fit(text, font, available_width, spacing_needed, safety_margin)
-        if hyphen_result != text:  # Some progress was made with hyphenation
+        # Strategy 3: Try hyphenation
+        hyphen_result = self._try_hyphenation(text, font, available_width)
+        if hyphen_result != text:
             return hyphen_result
         
-        # Strategy 4: Word doesn't fit and no hyphenation helped
-        if self._text_objects:
-            # Line already has words, move this word to the next line
-            return text
-        else:
-            # Empty line with word that's too long - force fit as last resort
-            return self._force_fit_long_word(text, font, available_width + safety_margin)
+        # Strategy 4: Handle overflow
+        return self._handle_word_overflow(text, font, available_width)
     
     def _try_hyphenation_or_fit(self, text: str, font: Font, available_width: int, 
                                spacing_needed: int, safety_margin: int) -> Union[None, str]:

pyWebLayout/concrete/viewport.py

@@ -0,0 +1,461 @@
+from typing import List, Tuple, Optional, Dict, Any
+import numpy as np
+from PIL import Image
+
+from pyWebLayout.core.base import Renderable, Layoutable
+from .box import Box
+from .page import Container
+from pyWebLayout.style.layout import Alignment
+
+
+class Viewport(Box, Layoutable):
+    """
+    A viewport that provides a movable window into a larger content area.
+    
+    This class allows you to have a large layout containing many elements,
+    but only render the portion that's currently visible in the viewport.
+    This enables efficient scrolling and memory usage for large documents.
+    """
+    
+    def __init__(self, viewport_size: Tuple[int, int], content_size: Optional[Tuple[int, int]] = None,
+                 origin=(0, 0), callback=None, sheet=None, mode='RGBA',
+                 background_color=(255, 255, 255)):
+        """
+        Initialize a viewport.
+        
+        Args:
+            viewport_size: The size of the visible viewport window (width, height)
+            content_size: The total size of the content area (None for auto-sizing)
+            origin: The origin of the viewport in its parent container
+            callback: Optional callback function
+            sheet: Optional image sheet
+            mode: Image mode
+            background_color: Background color for the viewport
+        """
+        super().__init__(origin, viewport_size, callback, sheet, mode)
+        
+        self._viewport_size = np.array(viewport_size)
+        self._content_size = np.array(content_size) if content_size else None
+        self._background_color = background_color
+        
+        # Viewport position within the content (scroll offset)
+        self._viewport_offset = np.array([0, 0])
+        
+        # Content container that holds all the actual content
+        self._content_container = Container(
+            origin=(0, 0),
+            size=content_size or viewport_size,
+            direction='vertical',
+            spacing=0,
+            padding=(0, 0, 0, 0)
+        )
+        
+        # Cached content bounds for optimization
+        self._content_bounds_cache = None
+        self._cache_dirty = True
+    
+    @property
+    def viewport_size(self) -> Tuple[int, int]:
+        """Get the viewport size"""
+        return tuple(self._viewport_size)
+    
+    @property
+    def content_size(self) -> Tuple[int, int]:
+        """Get the total content size"""
+        if self._content_size is not None:
+            return tuple(self._content_size)
+        else:
+            # Auto-calculate from content
+            self._update_content_size()
+            return tuple(self._content_size)
+    
+    @property
+    def viewport_offset(self) -> Tuple[int, int]:
+        """Get the current viewport offset (scroll position)"""
+        return tuple(self._viewport_offset)
+    
+    @property
+    def max_scroll_x(self) -> int:
+        """Get the maximum horizontal scroll position"""
+        content_w, content_h = self.content_size
+        viewport_w, viewport_h = self.viewport_size
+        return max(0, content_w - viewport_w)
+    
+    @property
+    def max_scroll_y(self) -> int:
+        """Get the maximum vertical scroll position"""
+        content_w, content_h = self.content_size
+        viewport_w, viewport_h = self.viewport_size
+        return max(0, content_h - viewport_h)
+    
+    def add_content(self, renderable: Renderable) -> 'Viewport':
+        """Add content to the viewport's content area"""
+        self._content_container.add_child(renderable)
+        self._cache_dirty = True
+        return self
+    
+    def clear_content(self) -> 'Viewport':
+        """Clear all content from the viewport"""
+        self._content_container._children.clear()
+        self._cache_dirty = True
+        return self
+    
+    def set_content_size(self, size: Tuple[int, int]) -> 'Viewport':
+        """Set the total content size explicitly"""
+        self._content_size = np.array(size)
+        self._content_container._size = self._content_size
+        self._cache_dirty = True
+        return self
+    
+    def _update_content_size(self):
+        """Auto-calculate content size from children"""
+        if not self._content_container._children:
+            self._content_size = self._viewport_size.copy()
+            return
+        
+        # Layout children to get their positions
+        self._content_container.layout()
+        
+        # Find the bounds of all children
+        max_x = 0
+        max_y = 0
+        
+        for child in self._content_container._children:
+            if hasattr(child, '_origin') and hasattr(child, '_size'):
+                child_origin = np.array(child._origin)
+                child_size = np.array(child._size)
+                child_end = child_origin + child_size
+                
+                max_x = max(max_x, child_end[0])
+                max_y = max(max_y, child_end[1])
+        
+        # Ensure content size is at least as large as viewport
+        self._content_size = np.array([
+            max(max_x, self._viewport_size[0]),
+            max(max_y, self._viewport_size[1])
+        ])
+        
+        self._content_container._size = self._content_size
+    
+    def _get_content_bounds(self) -> List[Tuple]:
+        """Get bounds of all content elements for efficient intersection testing"""
+        if not self._cache_dirty and self._content_bounds_cache is not None:
+            return self._content_bounds_cache
+        
+        bounds = []
+        self._collect_element_bounds(self._content_container, np.array([0, 0]), bounds)
+        
+        self._content_bounds_cache = bounds
+        self._cache_dirty = False
+        return bounds
+    
+    def _collect_element_bounds(self, container, offset: np.ndarray, bounds: List[Tuple]):
+        """Recursively collect bounds of all renderable elements"""
+        if not hasattr(container, '_children'):
+            return
+        
+        for child in container._children:
+            if hasattr(child, '_origin') and hasattr(child, '_size'):
+                child_origin = np.array(child._origin)
+                child_size = np.array(child._size)
+                
+                # Calculate absolute position
+                abs_origin = offset + child_origin
+                abs_end = abs_origin + child_size
+                
+                # Store bounds info
+                bounds.append((child, abs_origin, abs_end, child_size))
+                
+                # Recursively process children
+                if hasattr(child, '_children'):
+                    self._collect_element_bounds(child, abs_origin, bounds)
+    
+    def scroll_to(self, x: int, y: int) -> 'Viewport':
+        """
+        Scroll the viewport to a specific position.
+        
+        Args:
+            x: Horizontal scroll position
+            y: Vertical scroll position
+            
+        Returns:
+            Self for method chaining
+        """
+        # Clamp scroll position to valid range
+        max_x = self.max_scroll_x
+        max_y = self.max_scroll_y
+        
+        self._viewport_offset[0] = max(0, min(x, max_x))
+        self._viewport_offset[1] = max(0, min(y, max_y))
+        
+        return self
+    
+    def scroll_by(self, dx: int, dy: int) -> 'Viewport':
+        """
+        Scroll the viewport by a relative amount.
+        
+        Args:
+            dx: Horizontal scroll delta
+            dy: Vertical scroll delta
+            
+        Returns:
+            Self for method chaining
+        """
+        current_x, current_y = self.viewport_offset
+        return self.scroll_to(current_x + dx, current_y + dy)
+    
+    def scroll_to_top(self) -> 'Viewport':
+        """Scroll to the top of the content"""
+        return self.scroll_to(self._viewport_offset[0], 0)
+    
+    def scroll_to_bottom(self) -> 'Viewport':
+        """Scroll to the bottom of the content"""
+        return self.scroll_to(self._viewport_offset[0], self.max_scroll_y)
+    
+    def scroll_page_up(self) -> 'Viewport':
+        """Scroll up by one viewport height"""
+        return self.scroll_by(0, -self._viewport_size[1])
+    
+    def scroll_page_down(self) -> 'Viewport':
+        """Scroll down by one viewport height"""
+        return self.scroll_by(0, self._viewport_size[1])
+    
+    def scroll_line_up(self, line_height: int = 20) -> 'Viewport':
+        """Scroll up by one line"""
+        return self.scroll_by(0, -line_height)
+    
+    def scroll_line_down(self, line_height: int = 20) -> 'Viewport':
+        """Scroll down by one line"""
+        return self.scroll_by(0, line_height)
+    
+    def get_visible_elements(self) -> List[Tuple]:
+        """
+        Get all elements that are currently visible in the viewport.
+        
+        Returns:
+            List of tuples (element, visible_origin, visible_size) for each visible element
+        """
+        # Define viewport bounds
+        viewport_left = self._viewport_offset[0]
+        viewport_top = self._viewport_offset[1]
+        viewport_right = viewport_left + self._viewport_size[0]
+        viewport_bottom = viewport_top + self._viewport_size[1]
+        
+        visible_elements = []
+        content_bounds = self._get_content_bounds()
+        
+        for element, abs_origin, abs_end, element_size in content_bounds:
+            # Check if element intersects with viewport
+            if (abs_origin[0] < viewport_right and abs_end[0] > viewport_left and
+                abs_origin[1] < viewport_bottom and abs_end[1] > viewport_top):
+                
+                # Calculate visible portion of the element
+                visible_left = max(abs_origin[0], viewport_left)
+                visible_top = max(abs_origin[1], viewport_top)
+                visible_right = min(abs_end[0], viewport_right)
+                visible_bottom = min(abs_end[1], viewport_bottom)
+                
+                # Calculate visible origin relative to viewport
+                visible_origin = np.array([
+                    visible_left - viewport_left,
+                    visible_top - viewport_top
+                ])
+                
+                # Calculate visible size
+                visible_size = np.array([
+                    visible_right - visible_left,
+                    visible_bottom - visible_top
+                ])
+                
+                # Calculate clipping info for the element
+                element_clip_x = visible_left - abs_origin[0]
+                element_clip_y = visible_top - abs_origin[1]
+                element_clip_w = visible_size[0]
+                element_clip_h = visible_size[1]
+                
+                visible_elements.append((
+                    element, 
+                    visible_origin, 
+                    visible_size,
+                    (element_clip_x, element_clip_y, element_clip_w, element_clip_h)
+                ))
+        
+        return visible_elements
+    
+    def layout(self):
+        """Layout the content within the viewport"""
+        # Update content size if needed
+        if self._content_size is None:
+            self._update_content_size()
+        
+        # Layout all content
+        self._content_container.layout()
+        self._cache_dirty = True
+    
+    def render(self) -> Image.Image:
+        """
+        Render only the visible portion of the content.
+        
+        Returns:
+            A PIL Image containing the rendered viewport
+        """
+        # Ensure content is laid out
+        self.layout()
+        
+        # Create viewport canvas
+        canvas = Image.new(self._mode, tuple(self._viewport_size), self._background_color)
+        
+        # Get visible elements
+        visible_elements = self.get_visible_elements()
+        
+        # Render each visible element
+        for element, visible_origin, visible_size, clip_info in visible_elements:
+            try:
+                # Render the full element
+                element_img = element.render()
+                
+                # Extract the visible portion
+                clip_x, clip_y, clip_w, clip_h = clip_info
+                if clip_x >= 0 and clip_y >= 0 and clip_w > 0 and clip_h > 0:
+                    # Ensure clipping bounds are within element image
+                    elem_w, elem_h = element_img.size
+                    clip_x = min(clip_x, elem_w)
+                    clip_y = min(clip_y, elem_h)
+                    clip_w = min(clip_w, elem_w - clip_x)
+                    clip_h = min(clip_h, elem_h - clip_y)
+                    
+                    if clip_w > 0 and clip_h > 0:
+                        # Crop the visible portion
+                        visible_img = element_img.crop((
+                            clip_x, clip_y, 
+                            clip_x + clip_w, clip_y + clip_h
+                        ))
+                        
+                        # Paste onto viewport canvas
+                        paste_pos = tuple(visible_origin.astype(int))
+                        if visible_img.mode == 'RGBA' and canvas.mode == 'RGBA':
+                            canvas.paste(visible_img, paste_pos, visible_img)
+                        else:
+                            canvas.paste(visible_img, paste_pos)
+                            
+            except Exception as e:
+                # Skip elements that fail to render
+                continue
+        
+        return canvas
+    
+    def hit_test(self, point: Tuple[int, int]) -> Optional[Renderable]:
+        """
+        Find the topmost element at the given viewport coordinates.
+        
+        Args:
+            point: Coordinates within the viewport
+            
+        Returns:
+            The element at the given point, or None
+        """
+        viewport_x, viewport_y = point
+        
+        # Convert viewport coordinates to content coordinates
+        content_x = viewport_x + self._viewport_offset[0]
+        content_y = viewport_y + self._viewport_offset[1]
+        
+        # Find elements at this point (in reverse order for top-to-bottom hit testing)
+        content_bounds = self._get_content_bounds()
+        
+        for element, abs_origin, abs_end, element_size in reversed(content_bounds):
+            if (abs_origin[0] <= content_x < abs_end[0] and
+                abs_origin[1] <= content_y < abs_end[1]):
+                return element
+        
+        return None
+    
+    def get_scroll_info(self) -> Dict[str, Any]:
+        """
+        Get information about the current scroll state.
+        
+        Returns:
+            Dictionary with scroll information
+        """
+        content_w, content_h = self.content_size
+        viewport_w, viewport_h = self.viewport_size
+        offset_x, offset_y = self.viewport_offset
+        
+        return {
+            'content_size': (content_w, content_h),
+            'viewport_size': (viewport_w, viewport_h),
+            'offset': (offset_x, offset_y),
+            'max_scroll': (self.max_scroll_x, self.max_scroll_y),
+            'scroll_progress_x': offset_x / max(1, self.max_scroll_x) if self.max_scroll_x > 0 else 0,
+            'scroll_progress_y': offset_y / max(1, self.max_scroll_y) if self.max_scroll_y > 0 else 0,
+            'can_scroll_up': offset_y > 0,
+            'can_scroll_down': offset_y < self.max_scroll_y,
+            'can_scroll_left': offset_x > 0,
+            'can_scroll_right': offset_x < self.max_scroll_x
+        }
+
+
+class ScrollablePageContent(Container):
+    """
+    A specialized container for page content that's designed to work with viewports.
+    This extends the regular Page functionality but allows for much larger content areas.
+    """
+    
+    def __init__(self, content_width: int = 800, initial_height: int = 1000, 
+                 direction='vertical', spacing=10, padding=(0, 0, 0, 0)):
+        """
+        Initialize scrollable page content.
+        
+        Args:
+            content_width: Width of the content area
+            initial_height: Initial height (will grow as content is added)
+            direction: Layout direction
+            spacing: Spacing between elements
+            padding: Padding around content (no padding to avoid viewport clipping issues)
+        """
+        super().__init__(
+            origin=(0, 0),
+            size=(content_width, initial_height),
+            direction=direction,
+            spacing=spacing,
+            padding=padding  # No padding to avoid any positioning issues with viewport
+        )
+        
+        self._content_width = content_width
+        self._auto_height = True
+    
+    def add_child(self, child: Renderable):
+        """Add a child and update content height if needed"""
+        super().add_child(child)
+        
+        if self._auto_height:
+            self._update_content_height()
+        
+        return self
+    
+    def _update_content_height(self):
+        """Update the content height based on children"""
+        if not self._children:
+            return
+        
+        # Layout children to get accurate positions
+        super().layout()
+        
+        # Find the bottom-most child
+        max_bottom = 0
+        for child in self._children:
+            if hasattr(child, '_origin') and hasattr(child, '_size'):
+                child_bottom = child._origin[1] + child._size[1]
+                max_bottom = max(max_bottom, child_bottom)
+        
+        # Add some bottom padding
+        new_height = max_bottom + self._padding[2] + self._spacing
+        
+        # Update size if needed
+        if new_height > self._size[1]:
+            self._size = np.array([self._content_width, new_height])
+    
+    def get_content_height(self) -> int:
+        """Get the total content height"""
+        self._update_content_height()
+        return self._size[1]

tests/test_line_splitting_bug.py

@@ -102,6 +102,83 @@ class TestLineSplittingBug(unittest.TestCase):
         self.assertEqual(len(line.text_objects), 1)
         self.assertEqual(line.text_objects[0].text, "short")
 
+    def test_conservative_justified_hyphenation(self):
+        """Test that justified alignment is more conservative about mid-sentence hyphenation"""
+        font = Font(font_path=None, font_size=12, colour=(0, 0, 0))
+        line = Line((5, 15), (0, 0), (200, 20), font, halign=Alignment.JUSTIFY)
+        
+        with patch('pyWebLayout.abstract.inline.pyphen') as mock_pyphen_module:
+            mock_dic = Mock()
+            mock_pyphen_module.Pyphen.return_value = mock_dic
+            mock_dic.inserted.return_value = "test-word"
+            
+            # Add words that should fit without hyphenation
+            result1 = line.add_word("This")
+            result2 = line.add_word("should")
+            result3 = line.add_word("testword")  # Should NOT be hyphenated with conservative settings
+            
+            self.assertIsNone(result1)
+            self.assertIsNone(result2)
+            self.assertIsNone(result3)  # Should fit without hyphenation
+            self.assertEqual(len(line.text_objects), 3)
+            self.assertEqual([obj.text for obj in line.text_objects], ["This", "should", "testword"])
+
+    def test_helper_methods_exist(self):
+        """Test that refactored helper methods exist and work"""
+        font = Font(font_path=None, font_size=12, colour=(0, 0, 0))
+        line = Line((5, 10), (0, 0), (200, 20), font)
+        
+        # Test helper methods exist and return reasonable values
+        available_width = line._calculate_available_width(font)
+        self.assertIsInstance(available_width, int)
+        self.assertGreater(available_width, 0)
+        
+        safety_margin = line._get_safety_margin(font)
+        self.assertIsInstance(safety_margin, int)
+        self.assertGreaterEqual(safety_margin, 1)
+        
+        fits = line._fits_with_normal_spacing(50, 100, font)
+        self.assertIsInstance(fits, bool)
+
+    def test_no_cropping_with_safety_margin(self):
+        """Test that safety margin prevents text cropping"""
+        font = Font(font_path=None, font_size=12, colour=(0, 0, 0))
+        
+        # Create a line that's just barely wide enough
+        line = Line((2, 5), (0, 0), (80, 20), font)
+        
+        # Add words that should fit with safety margin
+        result1 = line.add_word("test")
+        result2 = line.add_word("word")
+        
+        self.assertIsNone(result1)
+        self.assertIsNone(result2)
+        
+        # Verify both words were added
+        self.assertEqual(len(line.text_objects), 2)
+        self.assertEqual([obj.text for obj in line.text_objects], ["test", "word"])
+
+    def test_modular_word_fitting_strategies(self):
+        """Test that word fitting strategies work in proper order"""
+        font = Font(font_path=None, font_size=12, colour=(0, 0, 0))
+        line = Line((5, 10), (0, 0), (80, 20), font)  # Narrower line to force overflow
+        
+        # Test normal spacing strategy
+        result1 = line.add_word("short")
+        self.assertIsNone(result1)
+        
+        # Test that we can add multiple words
+        result2 = line.add_word("words")
+        self.assertIsNone(result2)
+        
+        # Test overflow handling with a definitely too-long word
+        result3 = line.add_word("verylongwordthatdefinitelywontfitinnarrowline")
+        self.assertIsNotNone(result3)  # Should return overflow
+        
+        # Line should have the first two words only
+        self.assertEqual(len(line.text_objects), 2)
+        self.assertEqual([obj.text for obj in line.text_objects], ["short", "words"])
+
 
 def demonstrate_bug():
     """Demonstrate the bug with a practical example"""