source/app/models/version.rb - epf/org.eclipse.epf.wiki - Git at Google

 require 'ftools'
 require 'xhtmldiff'

 # More information:
 # * {EPF Wiki Data model}[link:files/doc/DATAMODEL.html]
 #
 #--######################################################################
 # Copyright (c) 2006 Logica
 #
 # All rights reserved. This program and the accompanying materials
 # are made available under the terms of the Eclipse Public License v1.0
 # which accompanies this distribution, and is available at
 # http://www.eclipse.org/legal/epl-v10.html
 #
 # Contributors:
 #
 # Onno van der Straaten:: initial implementation
 #++######################################################################
 # {Copyright (c) 2006 Logica}[link:files/COPYRIGHT.html]

 class Version < ActiveRecord::Base

   belongs_to  :user
   belongs_to  :wiki, :foreign_key => 'wiki_id' # TODO not necessary?
   belongs_to  :page
   has_one     :checkout
   belongs_to  :source_version,        :class_name => 'Version', :foreign_key => 'version_id'
   belongs_to  :reviewer,              :class_name => 'User',    :foreign_key => 'reviewer_id'
   has_many    :child_versions,        :class_name => 'Version', :foreign_key => 'version_id'
   has_many    :comments

   attr_accessor :type_filter # used for selection filter

   # NOTE: it is not possible to use 'update' below, this is reserved by ActiveRecord.
   # If you use it, the record won't save
   belongs_to  :baseline_update, :class_name => 'Update', :foreign_key => 'update_id'

   validates_presence_of :user, :wiki, :page

   BODY_PATTERN = /<body.*<\/body>/m
   LINKS_PATTERN = /<a.*?href=".*?".*?<\/a>/
   HREF_PATTERN = /href="(.*?)"/

   DIFF_STYLE =   ["<style type=\"text/css\">",
 "  .diffdel,del.diffmod, .diffdel *, del.diffmod * {",
 "  background-color:#FAA;",
 "  text-decoration:line-through;",
 "  }",
 "  .diffins,ins.diffmod, .diffins *, ins.diffmod * {",
 "  background-color:#AFA;",
 "  text-decoration:underline;",
 "  }",
 "  .diffmod, .diffmod * {",
 "  border: 2px solid #FE0;",
 "}",
 "</style>"].join("\n")

   def previous_version
     case self.version
     when 0 then nil
     when nil then self.page.current_version
     else Version.find(:first, :conditions => ['page_id=? and version=?', self.page_id, self.version - 1])
     end
   end

   # TODO this doesn't add value
   def baseversion
     return Version.find(:first, :conditions => ['wiki_id = ? and page_id=? and version=0',
     self.wiki_id, self.page_id])
   end

   def current_version
     return self.page.current_version
   end

   # return the latest version of page in a site
   def self.find_latest_version(site, page)
     return Version.find(:first ,:order => 'version DESC', :conditions => ["page_id=? and wiki_id =?", page.id, site.id])
   end

   # return the latest version based on a version
   # TODO use page.last_version
   def latest_version
     return Version.find(:first ,:order => 'version DESC', :conditions => ["page_id=? and wiki_id =?", self.page, self.wiki])
   end

   # Returns relative path of a version in folder 'public',
   # the column rel_path stores the relatieve path in a Site folder
     def rel_path_root
     return self.path.gsub("#{ENV['EPFWIKI_ROOT_DIR']}#{ENV['EPFWIKI_PUBLIC_FOLDER']}/", '')
   end

   # use #tidy to tidy the file using HTML Tidy.
   # This also removes any empty lines that may have been
   # created by PinEdit.
   def tidy
     logger.info("Tidying file " + path)
     tidy_file(path)
   end


   # method #html read or writes the HTML of a version from or to the version file.
   # When writing HTML Tidy is used to cleanup the file
   def html
     IO.readlines(self.path).join
   end

   # Create the tmp diff source file, clean it using Tidy and prepare for use with XHTMLDiff
   def html4diff(h = nil)
     logger.info("Create tmp diff source file #{self.path_to_tmp_diff_html} using #{self.path}")
     File.copy(self.path, self.path_to_tmp_diff_html)
     tidy_file(self.path_to_tmp_diff_html)
     h = IO.readlines(self.path_to_tmp_diff_html).join if h.nil?
     h = BODY_PATTERN.match(h)[0]
     h = h.gsub(Page::TREEBROWSER_PLACEHOLDER, '').gsub(Page::TREEBROWSER_PATTERN, '')
     h = h.gsub(Page::SHIM_TAG_PATTERN, Page::SHIM_TAG) # TODO remove? This should not be necessary, see html
     h = h.gsub(/<!-- epfwiki.*end -->/m, '') # TODO remove?
     h = h.gsub('</body>','').gsub(/<body.*>/, '')

     # If the html originates from TinyMCE it contains extra tbody tags.
     # TinyMCE adds those and this is good XHTML but if
     # we strip those so we get better diff results
     # Because without it, compare with the baseversion (not modified by TinyMCE) will
     # result false positives (changes)
     h = h.gsub(/<[\/]{0,1}(tbody){1}[.]*>/, '')

     # TODO Workaround for Tidy adding title attributes which causes false changes
     #--
     # The question mark makes it lazy, so it won't match for instance
     # title="Org"></a><span class="overview" it will match
     # title="Org" only
     #++
     h = h.gsub(/title="(.*?)"/, '') # WE REMOVE ALL TITLE ATTRIBUTES FOR BETTER DIFFS
     logger.debug("html4diff #{self.path}: ")
     #logger.debug(h)
     file = File.new(self.path_to_tmp_diff_html, 'w')
     file.puts(h)
     file.close
     h
   end

   # Create diff results using XHTMLDiff
   def xhtmldiff(from_version)
     content_from = "<div>\n" + from_version.html4diff + "\n</div>"
     content_to = "<div>\n" + self.html4diff + "\n</div>"

     diff_doc = REXML::Document.new
     diff_doc << (div = REXML::Element.new 'div')
     hd = XHTMLDiff.new(div)

     parsed_from_content = REXML::HashableElementDelegator.new(REXML::XPath.first(REXML::Document.new(content_from), '/div'))
     parsed_to_content = REXML::HashableElementDelegator.new(REXML::XPath.first(REXML::Document.new(content_to), '/div'))
     Diff::LCS.traverse_balanced(parsed_from_content, parsed_to_content , hd)
     diffs = ''
     diff_doc.write(diffs, -1, true, true)
     diffs
   end

   # Create diff file using #xhtmldiff. The file is generated if doesn't exist of if one of the versions is checked out.
   def xhtmldiffpage(from_version, force = false)
     p = path_to_diff(from_version)
       if !File.exists?(p) || !from_version.checkout.nil? || !self.checkout.nil? || force
       logger.info("Generating xhtmldiffpage #{p}")
       diffs = xhtmldiff(from_version)
       h = page.html
       body_tag = Page::BODY_TAG_PATTERN.match(h)[0]
       h = h.gsub(BODY_PATTERN,body_tag + diffs + '</body>')
       h = h.gsub('</head>',DIFF_STYLE + '</head>')
       h = h.gsub(Page::PAGE_HEAD_SNIPPET_PATTERN, '')
       file = File.new(p, 'w')
       file.puts(h)
       file.close
     end
   end

   # Absolute path to the diff file and paths to the intermediate (tidied, cleaned) HTML files
   def path_to_diff(from_version)
     return "#{ENV['EPFWIKI_ROOT_DIR']}#{ENV['EPFWIKI_PUBLIC_FOLDER']}/#{relpath_to_diff(from_version)}"
   end

   # Relative path to the diff file
   def relpath_to_diff(from_version)
     return '/' + self.wiki.rel_path + '/' + self.page.rel_path + "_EPFWIKI_DIFF_V#{from_version.version}_V#{self.version}.html"
   end

   # Path intermediate (tidied, cleaned) HTML files, prepared for XHTMLDiff
   def path_to_tmp_diff_html
     File.makedirs(ENV['EPFWIKI_DIFFS_PATH']) unless File.exists?(ENV['EPFWIKI_DIFFS_PATH'])
     "#{ENV['EPFWIKI_DIFFS_PATH']}#{self.id.to_s}.html"
   end

   # Compares hrefs to determine new or added
   #--
   # Depends on covert_urls settings of TinyMCE
   #++
   # TODO implement
   #def diff_links(from_version)
   #  links = self.html.scan(LINKS_PATTERN)
   #  links_from = from_version.html.scan(LINKS_PATTERN)
   #  hrefs = links.collect {|link | HREF_PATTERN.match(link)[0]}
   #  hrefs_from = links_from.collect {|link | HREF_PATTERN.match(link)[0]}
   #  new_removed = [hrefs - hrefs_from, hrefs_from - hrefs]
   #  new_removed
   #end

   # TODO implement
   #def diff_img(from_version)
   #end

   #def diff_area_links
   #end

   def template?
     self.wiki.title == 'Templates'
   end

   def base_version?
     !user_version?
   end

   def user_version?
     baseline_process_id.nil?
   end

   # #version_text returns version number, site title and baseline,
   # for example <tt>1 from OpenUP(OUP_20060721)</tt>
   def version_text
     if self.base_version?
       return self.version.to_s + ' from ' + self.wiki.title + ' (' + self.baseline_process.title+ ')'
     else
       return self.version.to_s + ' from ' + self.wiki.title + ' (' + self.user.name + ')'
     end
   end

   def before_destroy
     logger.debug("Before Destroy: deleting file #{self.path}")
     File.delete(self.path)
   end

   # #uma_type_descr is used to retrieve the brief description for a template, see app/views/pages/new
   def uma_type_descr
     logger.debug("Returning uma_type_description from: #{self.inspect}")
     if self.note.blank? || self.note == 'Automatically created'
       match = /<table class="overviewTable".*?<td valign="top">(.*?)<\/td>/m.match(self.html)
       if match
         self.note = match[1]
         self.save!
         logger.info("Match found: #{self.note}")
       else
         logger.debug('No match')
       end
     end
     return self.note
   end

 end
	require 'ftools'
	require 'xhtmldiff'

	# More information:
	# * {EPF Wiki Data model}[link:files/doc/DATAMODEL.html]
	#
	#--######################################################################
	# Copyright (c) 2006 Logica
	#
	# All rights reserved. This program and the accompanying materials
	# are made available under the terms of the Eclipse Public License v1.0
	# which accompanies this distribution, and is available at
	# http://www.eclipse.org/legal/epl-v10.html
	#
	# Contributors:
	#
	# Onno van der Straaten:: initial implementation
	#++######################################################################
	# {Copyright (c) 2006 Logica}[link:files/COPYRIGHT.html]

	class Version < ActiveRecord::Base

	belongs_to :user
	belongs_to :wiki, :foreign_key => 'wiki_id' # TODO not necessary?
	belongs_to :page
	has_one :checkout
	belongs_to :source_version, :class_name => 'Version', :foreign_key => 'version_id'
	belongs_to :reviewer, :class_name => 'User', :foreign_key => 'reviewer_id'
	has_many :child_versions, :class_name => 'Version', :foreign_key => 'version_id'
	has_many :comments

	attr_accessor :type_filter # used for selection filter

	# NOTE: it is not possible to use 'update' below, this is reserved by ActiveRecord.
	# If you use it, the record won't save
	belongs_to :baseline_update, :class_name => 'Update', :foreign_key => 'update_id'

	validates_presence_of :user, :wiki, :page

	BODY_PATTERN = /<body.*<\/body>/m
	LINKS_PATTERN = /<a.?href=".?".*?<\/a>/
	HREF_PATTERN = /href="(.*?)"/

	DIFF_STYLE = ["<style type=\"text/css\">",
	" .diffdel,del.diffmod, .diffdel , del.diffmod {",
	" background-color:#FAA;",
	" text-decoration:line-through;",
	" }",
	" .diffins,ins.diffmod, .diffins , ins.diffmod {",
	" background-color:#AFA;",
	" text-decoration:underline;",
	" }",
	" .diffmod, .diffmod * {",
	" border: 2px solid #FE0;",
	"}",
	"</style>"].join("\n")

	def previous_version
	case self.version
	when 0 then nil
	when nil then self.page.current_version
	else Version.find(:first, :conditions => ['page_id=? and version=?', self.page_id, self.version - 1])
	end
	end

	# TODO this doesn't add value
	def baseversion
	return Version.find(:first, :conditions => ['wiki_id = ? and page_id=? and version=0',
	self.wiki_id, self.page_id])
	end

	def current_version
	return self.page.current_version
	end

	# return the latest version of page in a site
	def self.find_latest_version(site, page)
	return Version.find(:first ,:order => 'version DESC', :conditions => ["page_id=? and wiki_id =?", page.id, site.id])
	end

	# return the latest version based on a version
	# TODO use page.last_version
	def latest_version
	return Version.find(:first ,:order => 'version DESC', :conditions => ["page_id=? and wiki_id =?", self.page, self.wiki])
	end

	# Returns relative path of a version in folder 'public',
	# the column rel_path stores the relatieve path in a Site folder
	def rel_path_root
	return self.path.gsub("#{ENV['EPFWIKI_ROOT_DIR']}#{ENV['EPFWIKI_PUBLIC_FOLDER']}/", '')
	end

	# use #tidy to tidy the file using HTML Tidy.
	# This also removes any empty lines that may have been
	# created by PinEdit.
	def tidy
	logger.info("Tidying file " + path)
	tidy_file(path)
	end


	# method #html read or writes the HTML of a version from or to the version file.
	# When writing HTML Tidy is used to cleanup the file
	def html
	IO.readlines(self.path).join
	end

	# Create the tmp diff source file, clean it using Tidy and prepare for use with XHTMLDiff
	def html4diff(h = nil)
	logger.info("Create tmp diff source file #{self.path_to_tmp_diff_html} using #{self.path}")
	File.copy(self.path, self.path_to_tmp_diff_html)
	tidy_file(self.path_to_tmp_diff_html)
	h = IO.readlines(self.path_to_tmp_diff_html).join if h.nil?
	h = BODY_PATTERN.match(h)[0]
	h = h.gsub(Page::TREEBROWSER_PLACEHOLDER, '').gsub(Page::TREEBROWSER_PATTERN, '')
	h = h.gsub(Page::SHIM_TAG_PATTERN, Page::SHIM_TAG) # TODO remove? This should not be necessary, see html
	h = h.gsub(/<!-- epfwiki.*end -->/m, '') # TODO remove?
	h = h.gsub('</body>','').gsub(/<body.*>/, '')

	# If the html originates from TinyMCE it contains extra tbody tags.
	# TinyMCE adds those and this is good XHTML but if
	# we strip those so we get better diff results
	# Because without it, compare with the baseversion (not modified by TinyMCE) will
	# result false positives (changes)
	h = h.gsub(/<[\/]{0,1}(tbody){1}[.]*>/, '')

	# TODO Workaround for Tidy adding title attributes which causes false changes
	#--
	# The question mark makes it lazy, so it won't match for instance
	# title="Org"></a><span class="overview" it will match
	# title="Org" only
	#++
	h = h.gsub(/title="(.*?)"/, '') # WE REMOVE ALL TITLE ATTRIBUTES FOR BETTER DIFFS
	logger.debug("html4diff #{self.path}: ")
	#logger.debug(h)
	file = File.new(self.path_to_tmp_diff_html, 'w')
	file.puts(h)
	file.close
	h
	end

	# Create diff results using XHTMLDiff
	def xhtmldiff(from_version)
	content_from = "<div>\n" + from_version.html4diff + "\n</div>"
	content_to = "<div>\n" + self.html4diff + "\n</div>"

	diff_doc = REXML::Document.new
	diff_doc << (div = REXML::Element.new 'div')
	hd = XHTMLDiff.new(div)

	parsed_from_content = REXML::HashableElementDelegator.new(REXML::XPath.first(REXML::Document.new(content_from), '/div'))
	parsed_to_content = REXML::HashableElementDelegator.new(REXML::XPath.first(REXML::Document.new(content_to), '/div'))
	Diff::LCS.traverse_balanced(parsed_from_content, parsed_to_content , hd)
	diffs = ''
	diff_doc.write(diffs, -1, true, true)
	diffs
	end

	# Create diff file using #xhtmldiff. The file is generated if doesn't exist of if one of the versions is checked out.
	def xhtmldiffpage(from_version, force = false)
	p = path_to_diff(from_version)
	if !File.exists?(p) \|\| !from_version.checkout.nil? \|\| !self.checkout.nil? \|\| force
	logger.info("Generating xhtmldiffpage #{p}")
	diffs = xhtmldiff(from_version)
	h = page.html
	body_tag = Page::BODY_TAG_PATTERN.match(h)[0]
	h = h.gsub(BODY_PATTERN,body_tag + diffs + '</body>')
	h = h.gsub('</head>',DIFF_STYLE + '</head>')
	h = h.gsub(Page::PAGE_HEAD_SNIPPET_PATTERN, '')
	file = File.new(p, 'w')
	file.puts(h)
	file.close
	end
	end

	# Absolute path to the diff file and paths to the intermediate (tidied, cleaned) HTML files
	def path_to_diff(from_version)
	return "#{ENV['EPFWIKI_ROOT_DIR']}#{ENV['EPFWIKI_PUBLIC_FOLDER']}/#{relpath_to_diff(from_version)}"
	end

	# Relative path to the diff file
	def relpath_to_diff(from_version)
	return '/' + self.wiki.rel_path + '/' + self.page.rel_path + "_EPFWIKI_DIFF_V#{from_version.version}_V#{self.version}.html"
	end

	# Path intermediate (tidied, cleaned) HTML files, prepared for XHTMLDiff
	def path_to_tmp_diff_html
	File.makedirs(ENV['EPFWIKI_DIFFS_PATH']) unless File.exists?(ENV['EPFWIKI_DIFFS_PATH'])
	"#{ENV['EPFWIKI_DIFFS_PATH']}#{self.id.to_s}.html"
	end

	# Compares hrefs to determine new or added
	#--
	# Depends on covert_urls settings of TinyMCE
	#++
	# TODO implement
	#def diff_links(from_version)
	# links = self.html.scan(LINKS_PATTERN)
	# links_from = from_version.html.scan(LINKS_PATTERN)
	# hrefs = links.collect {\|link \| HREF_PATTERN.match(link)[0]}
	# hrefs_from = links_from.collect {\|link \| HREF_PATTERN.match(link)[0]}
	# new_removed = [hrefs - hrefs_from, hrefs_from - hrefs]
	# new_removed
	#end

	# TODO implement
	#def diff_img(from_version)
	#end

	#def diff_area_links
	#end

	def template?
	self.wiki.title == 'Templates'
	end

	def base_version?
	!user_version?
	end

	def user_version?
	baseline_process_id.nil?
	end

	# #version_text returns version number, site title and baseline,
	# for example <tt>1 from OpenUP(OUP_20060721)</tt>
	def version_text
	if self.base_version?
	return self.version.to_s + ' from ' + self.wiki.title + ' (' + self.baseline_process.title+ ')'
	else
	return self.version.to_s + ' from ' + self.wiki.title + ' (' + self.user.name + ')'
	end
	end

	def before_destroy
	logger.debug("Before Destroy: deleting file #{self.path}")
	File.delete(self.path)
	end

	# #uma_type_descr is used to retrieve the brief description for a template, see app/views/pages/new
	def uma_type_descr
	logger.debug("Returning uma_type_description from: #{self.inspect}")
	if self.note.blank? \|\| self.note == 'Automatically created'
	match = /<table class="overviewTable".?<td valign="top">(.?)<\/td>/m.match(self.html)
	if match
	self.note = match[1]
	self.save!
	logger.info("Match found: #{self.note}")
	else
	logger.debug('No match')
	end
	end
	return self.note
	end

	end