pagecapture.rb

#!/usr/local/bin/macruby
framework 'Cocoa'
framework 'WebKit'

class VirtualBrowser
  attr_reader :view, :file_name
  
  def initialize
    NSApplication.sharedApplication.delegate = self
    @view     = WebView.alloc.initWithFrame([0, 0, 1024, 768])
    window    = NSWindow.alloc.initWithContentRect([0, 0, 1024, 768],
                                                styleMask:NSBorderlessWindowMask, 
                                                backing:NSBackingStoreBuffered, 
                                                defer:false)

    window.contentView = view
    setup_view_prefs
    view.frameLoadDelegate = self
    @captured = false
  end
  
  def fetch(url, file_name='capture.pdf')
    @file_name = file_name
    page_url = NSURL.URLWithString(url)
    view.mainFrame.loadRequest NSURLRequest.requestWithURL(page_url)
    until @captured
      NSRunLoop.currentRunLoop.runUntilDate NSDate.date
    end
  end
  
  def setup_view_prefs
    view.mediaStyle = 'screen'
    view.customUserAgent = 'Mozilla/5.0 (Macintosh; U; Intel Mac OS X 10_6_2; en-us) AppleWebKit/531.21.8 (KHTML, like Gecko) Version/4.0.4 Safari/531.21.10'
    view.preferences.autosaves = false
    view.preferences.shouldPrintBackgrounds = true
    view.preferences.javaScriptCanOpenWindowsAutomatically = false
    view.preferences.allowsAnimatedImages = false
    view.mainFrame.frameView.allowsScrolling = false
  end
  
  def webView(view, didFinishLoadForFrame:frame)
    save
  end

  def webView(view, didFailLoadWithError:error, forFrame:frame)
    raise "Failed to take snapshot: #{error.localizedDescription}"
    quit
  end

  def webView(view, didFailProvisionalLoadWithError:error, forFrame:frame)
    raise "Failed to take snapshot: #{error.localizedDescription}"
    quit
  end
  
  def quit
    NSApplication.sharedApplication.terminate(nil)
  end
  
  def save
    filepath = File.expand_path(@file_name)
    puts "saving #{filepath}"
    doc_view  = setup_docview
    doc_view.dataWithPDFInsideRect(doc_view.bounds).writeToFile(filepath, atomically:true)
    doc_view.unlockFocus
    @captured = true
    `open #{filepath}`
  end
  
  def setup_docview
    doc_view = view.mainFrame.frameView.documentView
    doc_view.window.contentSize = [doc_view.bounds.size.width, doc_view.bounds.size.height]
    doc_view.frame = view.bounds
    doc_view.needsDisplay = true
    doc_view.lockFocus
    doc_view
  end
  
end

url       = ARGV.shift || "http://rubyinside.com/"
file_name = ARGV.shift || "rubyinside.pdf"
VirtualBrowser.new.fetch(url, file_name)