Ruby + htmlfile(WIN32OLE) + JavaScript-XPath で XPath

htmlfile-xpath.rb

# Ruby + htmlfile + JavaScript-XPath で XPath
#
# Example:
#
#    > ruby htmlfile-xpath.rb http://gist.github.com/gists "//div[@class='info']/span/a"
#    gist: 497491
#    gist: 497481
#    gist: 497469
#          :
#
# See:
#   http://d.hatena.ne.jp/amachang/20071112/1194856493
#   http://d.hatena.ne.jp/mobitan/20090829/1251542456#20090829f10

require "open-uri"
require "win32ole"

$jsxpath_uri = "http://svn.coderepos.org/share/lang/javascript/javascript-xpath/trunk/release/javascript-xpath-latest-cmp.js"

if ARGV.size != 2
  $stderr.puts "Usage: uri xpath"
  exit 1
end

uri, xpath = *ARGV
dom = WIN32OLE.new("htmlfile")
dom.Write(open(uri){|f| f.read })
dom.parentWindow.eval(open($jsxpath_uri){|f| f.read })

items = dom.evaluate(xpath, dom, nil, 7, nil)
len = items.snapshotLength
(0...len).each do |i|
  item = items.snapshotItem(i)
  puts item.innerHTML
end