REXML Trees

Synced with latest HTML5lib.
Added preliminary support (currently disabled) for sanitizing REXML trees.
This commit is contained in:
Jacques Distler 2007-06-05 16:34:49 -05:00
parent 4dd70af5ae
commit bd8ba1f4b1
28 changed files with 1317 additions and 112 deletions

View file

@ -21,3 +21,53 @@ rescue LoadError
end
end
end
module HTML5lib
module TestSupport
def self.startswith?(a, b)
b[0... a.length] == a
end
def self.parseTestcase(data)
innerHTML = nil
input = []
output = []
errors = []
currentList = input
data.split(/\n/).each do |line|
if !line.empty? and !startswith?("#errors", line) and
!startswith?("#document", line) and
!startswith?("#data", line) and
!startswith?("#document-fragment", line)
if currentList == output and startswith?("|", line)
currentList.push(line[2..-1])
else
currentList.push(line)
end
elsif line == "#errors"
currentList = errors
elsif line == "#document" or startswith?("#document-fragment", line)
if startswith?("#document-fragment", line)
innerHTML = line[19..-1]
raise AssertionError unless innerHTML
end
currentList = output
end
end
return innerHTML, input.join("\n"), output.join("\n"), errors
end
# convert the output of str(document) to the format used in the testcases
def convertTreeDump(treedump)
treedump.split(/\n/)[1..-1].map { |line| (line.length > 2 and line[0] == ?|) ? line[3..-1] : line }.join("\n")
end
def sortattrs(output)
output.gsub(/^(\s+)\w+=.*(\n\1\w+=.*)+/) do |match|
match.split("\n").sort.join("\n")
end
end
end
end