1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287
|
# encoding: utf-8
# frozen_string_literal: true
require "English"
require "helper"
if Nokogiri.uses_gumbo?
def parse_test(test_data)
test = { script: :both }
index = /(?:^#errors\n|\n#errors\n)/ =~ test_data
abort("Expected #errors in\n#{test_data}") if index.nil?
skip_amount = $LAST_MATCH_INFO[0].length
# Omit the final new line
test[:data] = test_data[0...index]
# Process the rest line by line
lines = test_data[index + skip_amount..-1].split("\n")
index = lines.find_index do |line|
line == "#document-fragment" ||
line == "#document" ||
line == "#script-off" ||
line == "#script-on" ||
line == "#new-errors"
end
abort("Expected #document") if index.nil?
test[:errors] = lines[0...index]
test[:new_errors] = []
if lines[index] == "#new-errors"
index += 1
until ["#document-fragment", "#document", "#script-off", "#script-on"].include?(lines[index])
test[:new_errors] << lines[index]
index += 1
end
end
if lines[index] == "#document-fragment"
test[:context] = lines[index + 1].chomp.split(" ", 2)
index += 2
end
abort("failed to find fragment: #{index}: #{lines[index]}") if test_data.include?("#document-fragment") && test[:context].nil?
if lines[index] =~ /#script-(on|off)/
test[:script] = $LAST_MATCH_INFO[1].to_sym
index += 1
end
abort("Expected #document, got #{lines[index]}") unless lines[index] == "#document"
index += 1
document = {
type: test[:context] ? :fragment : :document,
children: [],
}
open_nodes = [document]
while index < lines.length
abort("Expected '| ' but got #{lines[index]}") unless /^\| ( *)([^ ].*$)/ =~ lines[index]
depth = $LAST_MATCH_INFO[1].length
if depth.odd?
abort("Invalid nesting depth")
else
depth /= 2
end
abort("Too deep") if depth >= open_nodes.length
node = {}
node_text = $LAST_MATCH_INFO[2]
if node_text[0] == '"'
if node_text == '"' || node_text[-1] != '"'
loop do
index += 1
node_text << "\n" + lines[index]
break if node_text[-1] == '"'
end
end
node[:type] = :text
node[:contents] = node_text[1..-2]
elsif /^<!DOCTYPE ([^ >]*)(?: "([^"]*)" "(.*)")?>$/ =~ node_text
node[:type] = :doctype
node[:name] = $LAST_MATCH_INFO[1]
node[:public_id] = $LAST_MATCH_INFO[2].nil? || $LAST_MATCH_INFO[2].empty? ? nil : $LAST_MATCH_INFO[2]
node[:system_id] = $LAST_MATCH_INFO[3].nil? || $LAST_MATCH_INFO[3].empty? ? nil : $LAST_MATCH_INFO[3]
elsif node_text.start_with?("<!-- ")
loop do
break if lines[index].end_with?(" -->")
index += 1
node_text << "\n" + lines[index]
end
node[:type] = :comment
node[:contents] = node_text[5..-5]
elsif /^<(svg |math )?(.+)>$/ =~ node_text
node[:type] = :element
node[:ns] = $LAST_MATCH_INFO[1].nil? ? nil : $LAST_MATCH_INFO[1].rstrip
node[:tag] = $LAST_MATCH_INFO[2]
node[:attributes] = []
node[:children] = []
elsif /^([^ ]+ )?([^=]+)="(.*)"$/ =~ node_text
node[:type] = :attribute
node[:ns] = $LAST_MATCH_INFO[1].nil? ? nil : $LAST_MATCH_INFO[1].rstrip
node[:name] = $LAST_MATCH_INFO[2]
node[:value] = $LAST_MATCH_INFO[3]
elsif node_text == "content"
node[:type] = :template
else
abort("Unexpected node_text: #{node_text}")
end
if node[:type] == :attribute
abort("depth #{depth} != #{open_nodes.length}") unless depth == open_nodes.length - 1
abort("type :#{open_nodes[-1][:type]} != :element") unless open_nodes[-1][:type] == :element
abort("element has children") unless open_nodes[-1][:children].empty?
open_nodes[-1][:attributes] << node
elsif node[:type] == :template
abort("depth #{depth} != #{open_nodes.length}") unless depth == open_nodes.length - 1
abort("type :#{open_nodes[-1][:type]} != :element") unless open_nodes[-1][:type] == :element
abort("tag :#{open_nodes[-1][:tag]} != template") unless open_nodes[-1][:tag] == "template"
abort("template has children before the 'content'") unless open_nodes[-1][:children].empty?
# Hack. We want the children of this template node to be reparented as
# children of the template element.
# XXX: Template contents are _not_ supposed to be children of the
# template, but we currently mishandle this.
open_nodes << open_nodes[-1]
else
open_nodes[depth][:children] << node
open_nodes[depth + 1..-1] = []
if node[:type] == :element
open_nodes << node
end
end
index += 1
end
test[:document] = document
test
end
class TestHtml5TreeConstructionBase < Nokogiri::TestCase
def assert_equal_or_nil(exp, act)
if exp.nil?
assert_nil(act)
else
assert_equal(exp, act)
end
end
def compare_nodes(node, ng_node)
case ng_node.type
when Nokogiri::XML::Node::ELEMENT_NODE
assert_equal(:element, node[:type])
if node[:ns]
refute_nil(ng_node.namespace)
assert_equal(node[:ns], ng_node.namespace.prefix)
end
assert_equal(node[:tag], ng_node.name)
attributes = ng_node.attributes
assert_equal(node[:attributes].length, attributes.length)
node[:attributes].each do |attr|
value = if attr[:ns]
ng_node["#{attr[:ns]}:#{attr[:name]}"]
else
attributes[attr[:name]].value
end
assert_equal(attr[:value], value)
end
assert_equal(node[:children].length, ng_node.children.length,
"Element <#{node[:tag]}> has wrong number of children #{ng_node.children.map(&:name)} in #{@test[:data]}")
when Nokogiri::XML::Node::TEXT_NODE, Nokogiri::XML::Node::CDATA_SECTION_NODE
# We preserve the CDATA in the tree, but the tests represent it as text.
assert_equal(:text, node[:type])
assert_equal(node[:contents], ng_node.content)
when Nokogiri::XML::Node::COMMENT_NODE
assert_equal(:comment, node[:type])
assert_equal(node[:contents], ng_node.content)
when Nokogiri::XML::Node::HTML_DOCUMENT_NODE
assert_equal(:document, node[:type])
assert_equal(node[:children].length, ng_node.children.length)
when Nokogiri::XML::Node::DOCUMENT_FRAG_NODE
assert_equal(:fragment, node[:type])
assert_equal(node[:children].length, ng_node.children.length,
"Fragment node has wrong number of children #{ng_node.children.map(&:name)} in #{@test[:data]}")
when Nokogiri::XML::Node::DTD_NODE
assert_equal(:doctype, node[:type])
assert_equal(node[:name], ng_node.name)
assert_equal_or_nil(node[:public_id], ng_node.external_id)
assert_equal_or_nil(node[:system_id], ng_node.system_id)
else
flunk("Unknown node type #{ng_node.type} (expected #{node[:type]})")
end
end
def run_test
if @test[:context]
ctx = @test[:context].join(":")
doc = Nokogiri::HTML5::Document.new
doc = Nokogiri::HTML5::DocumentFragment.new(doc, @test[:data], ctx, max_errors: @test[:errors].length + 10)
else
doc = Nokogiri::HTML5.parse(@test[:data], max_errors: @test[:errors].length + 10)
end
# Walk the tree.
exp_nodes = [@test[:document]]
act_nodes = [doc]
children = [0]
compare_nodes(exp_nodes[0], doc)
while children.any?
child_index = children[-1]
exp = exp_nodes[-1]
act = act_nodes[-1]
if child_index == exp[:children].length
exp_nodes.pop
act_nodes.pop
children.pop
next
end
exp_child = exp[:children][child_index]
act_child = act.children[child_index]
compare_nodes(exp_child, act_child)
children[-1] = child_index + 1
next unless exp_child.key?(:children)
exp_nodes << exp_child
act_nodes << act_child
children << 0
end
# Test the errors.
assert_equal(@test[:errors].length, doc.errors.length, "Wrong number of errors for #{@test[:data]}")
# The new, standardized tokenizer errors live in @test[:new_errors]. Let's
# match each one to exactly one error in doc.errors. Unfortunately, the
# tests specify the column the error is detected, _not_ the column of the
# start of the problematic HTML (e.g., the start of a character reference
# or <![CDATA[) the way gumbo does. So check that Gumbo's column is no
# later than the error's column.
errors = doc.errors.map { |err| { line: err.line, column: err.column, code: err.str1 } }
errors.reject! { |err| err[:code] == "generic-parser" }
error_regex = /^\((?<line>\d+):(?<column>\d+)(?:-\d+:\d+)?\) (?<code>.*)$/
@test[:new_errors].each do |err|
assert_match(error_regex, err, "New error format does not match: #{mu_pp(err)}")
m = err.match(error_regex)
line = m[:line].to_i
column = m[:column].to_i
code = m[:code]
idx = errors.index do |e|
e[:line] == line &&
e[:code] == code &&
e[:column] <= column
end
# This error should be the first error in the list.
# refute_nil(idx, "Expected to find error #{code} at #{line}:#{column}")
assert_equal(0, idx, "Expected to find error #{code} at #{line}:#{column} in #{@test[:data]}")
errors.delete_at(idx)
end
end
end
tc_path = File.expand_path("../../html5lib-tests/tree-construction", __FILE__)
Dir[File.join(tc_path, "*.dat")].each do |path|
test_name = "TestHtml5TreeConstruction" + File.basename(path, ".dat")
.split(/[_-]/)
.map(&:capitalize)
.join("")
tests = []
File.open(path, "r", encoding: "UTF-8") do |f|
f.each("\n\n#data\n") do |test_data|
if test_data.start_with?("#data\n")
test_data = test_data[6..-1]
end
if test_data.end_with?("\n\n#data\n")
test_data = test_data[0..-9]
end
tests << parse_test(test_data)
end
end
klass = Class.new(TestHtml5TreeConstructionBase) do
tests.each_with_index do |test, index|
next if test[:script] == :on
define_method "test_#{index}".to_sym do
@test = test
@index = index
run_test
end
end
end
Object.const_set(test_name, klass)
end
end
|