#!/usr/bin/env ruby ############################################################################### # test.rb # Run this script with `help` for more information (or examine this file.) # # This script replaces the previous tools-cmd and tools-sh shell scripts, # as a single, cross-platform means of providing regression testing support # for HTML Tidy. ############################################################################### require 'rubygems' # Ensure bundler is used. require 'bundler/setup' # Make sure the Gemfile is recognized. require 'logger' # Log output simplified. require 'fileutils' # File utilities. require 'thor' # Thor provides robust command line parameter parsing. require 'tty-editor' # Cross-platform opening of files in default editor. ################################################################################ # module TidyRegressionTestModule # This module encapsulates module-level variables, utilities, logging, # and the CLI handling class. ############################################################################### #noinspection RubyTooManyInstanceVariablesInspection module TidyRegressionTestModule ########################################################### # Setup # Change these variables in case directories are changed. # Note that DIR_TIDY can be changed with input options. ########################################################### DIR_TEST_SETS = File.expand_path('cases') EXE_TIDY = File.expand_path(File.join('..', 'build', 'cmake', 'tidy')) VERSION_FILE = File.join(File.expand_path('..', File.dirname(__FILE__)), 'version.txt') ########################################################### # The shared logger. # This is a module-level logger, shared amongst everyone # in a singleton-like fashion. ########################################################### SHARED_LOGGER = Logger.new(STDOUT) SHARED_LOGGER.level = Logger::ERROR SHARED_LOGGER.datetime_format = '%Y-%m-%d %H:%M:%S' WINDOWS_OS = (/cygwin|mswin|mingw|bccwin|wince|emx/ =~ RUBY_PLATFORM) ############################################################################# # class TidyTestSet # This class contains information a single test set. ############################################################################# class TidyTestSet attr_reader :test_set_name attr_accessor :dir_test_set attr_reader :dir_expects attr_reader :dir_results attr_reader :case_paths attr_reader :case_names attr_accessor :path_tidy_exe ######################################################### # initialize ######################################################### def initialize( with_directory:, path_to_tidy: ) @dir_expects = nil @dir_results = nil @case_paths = nil self.dir_test_set = with_directory self.path_tidy_exe = path_to_tidy end ######################################################### # property test_set_name ######################################################### def test_set_name @test_set_name end ######################################################### # property dir_test_set ######################################################### def dir_test_set @dir_test_set end def dir_test_set=( test_set_directory ) @test_set_name = File.basename(test_set_directory)[/(.*)-cases/,1] candidate_expects = File.join(File.dirname(test_set_directory), "#{@test_set_name}-expects") if File.exist?(candidate_expects) @dir_test_set = test_set_directory else SHARED_LOGGER.warn("Skipping test set #{test_set_directory} because expectations directory #{candidate_expects} does not exist.") @dir_test_set = nil end end ######################################################### # property dir_expects ######################################################### def dir_expects unless @dir_expects @dir_expects = File.join(File.dirname(self.dir_test_set), "#{@test_set_name}-expects") end @dir_expects end ######################################################### # property dir_results ######################################################### def dir_results unless @dir_results @dir_results = File.join(File.dirname(self.dir_test_set), "#{@test_set_name}-results") end @dir_results end ######################################################### # property case_paths ######################################################### def case_paths unless @case_paths @case_paths = [] Dir.glob("#{self.dir_test_set}/*.{html,xml,xhtml,htm}").sort.each do |html_file| SHARED_LOGGER.info("Adding case path #{html_file}") case_name = File.basename( html_file[/(.*)@/,1] ) candidate_expects_html = File.join( self.dir_expects, "#{case_name}#{File.extname(html_file)}" ) candidate_expects_output = File.join( self.dir_expects, "#{case_name}.txt" ) #noinspection RubyNilAnalysis -- this is a false positive, because we've explicitly made this an array. @case_paths.push(html_file) unless File.exist?(candidate_expects_html) && File.exist?(candidate_expects_output) SHARED_LOGGER.warn("Note: #{html_file} is missing some expectations files in #{self.dir_expects}.") end end end @case_paths end ######################################################### # property case_names ######################################################### def case_names #noinspection RubyNilAnalysis -- cannot be nil at this point. self.case_paths.map { |case_path| File.basename(case_path)[/^case-(.*)@.*$/,1] } end ######################################################### # property path_tidy_exe ######################################################### def path_tidy_exe @path_tidy_exe end def path_tidy_exe=( value ) unless File.exist?( value ) raise StandardError.new "The path #{value} to tidy doesn’t exist." end unless File.executable?( value ) raise StandardError.new "The path #{value} to tidy is there, but isn’t executable." end @path_tidy_exe = value end end # class TidyTestSet ############################################################################# # class TidyTestCase # This class contains information a single test case. ############################################################################# class TidyTestCase attr_accessor :path_file attr_reader :name_test_set attr_reader :name_case attr_reader :path_config attr_reader :expects_exit_code attr_reader :dir_test_sets attr_reader :dir_expects attr_reader :dir_results attr_reader :path_expects_html attr_reader :path_expects_output attr_reader :path_result_html attr_reader :path_result_output attr_accessor :path_tidy_exe attr_reader :test_passed attr_reader :test_report attr_reader :test_exit_status ######################################################### # initialize ######################################################### def initialize( with_file:, path_to_tidy: ) @path_file = nil @name_test_set = nil @name_case = nil @path_config = nil @expects_exit_code = nil @dir_test_sets = nil @dir_expects = nil @dir_results = nil @path_expects_html = nil @path_expects_output = nil @path_result_html = nil @path_result_output = nil @path_tidy_exe = nil @test_report = nil @test_exit_status = nil @test_complete = false @test_diff_html = nil @test_diff_html_status = nil @test_diff_output = nil @test_diff_output_status = nil self.path_file = with_file self.path_tidy_exe = path_to_tidy end # initialize ############################################################################# # Property Accessors ############################################################################# ######################################################### # property path_file ######################################################### def path_file @path_file end def path_file=( value ) if File.exist?(File.expand_path(value)) @path_file = value else raise StandardError.new "File #{value} doesn’t exist." end valid_extensions = %w(.html .xhtml .htm .xml) unless valid_extensions.include?(File.extname(value).downcase) raise StandardError.new "File #{value} doesn’t have a correct extension (one of #{valid_extensions})." end @path_file = value end ######################################################### # property name_test_set # This is the enclosing directory name, without the # -cases suffix. ######################################################### def name_test_set unless @name_test_set @name_test_set = File.basename( File.dirname(self.path_file)[/(.*)-cases/,1] ) end @name_test_set end ######################################################### # property name_case ######################################################### def name_case unless @name_case @name_case = File.basename(self.path_file)[/case-(.*)@/, 1] end @name_case end ######################################################### # property path_config # The config file is either the same as the test case # file (without the exit code metadata) and the .conf # extension, or the config_default.conf if not found. ######################################################### def path_config unless @path_config directory = File.dirname(self.path_file) candidate = File.join( directory, "case-#{name_case}.conf" ) unless File.exist?(candidate) candidate = File.join( directory, 'config_default.conf') unless File.exist?(candidate) raise StandardError.new "Case #{name_case} doesn’t have a configuration file." end end @path_config = candidate end @path_config end ######################################################### # property expects_exit_code # This is the metadata between the @ and extension. ######################################################### def expects_exit_code unless @expects_exit_code @expects_exit_code = File.basename(self.path_file)[/case-.*@(.)./, 1] end @expects_exit_code end ######################################################### # property dir_test_sets # This is the directory containing the test sets. ######################################################### def dir_test_sets unless @dir_test_sets @dir_test_sets = File.dirname( File.dirname(self.path_file) ) end @dir_test_sets end ######################################################### # property dir_expects # This is the directory within the test sets that # contains the expectations files. ######################################################### def dir_expects unless @dir_expects @dir_expects = File.join( self.dir_test_sets, "#{self.name_test_set}-expects" ) unless File.exist?(@dir_expects) raise StandardError.new "The expectations directory #{dir_expects} doesn’t seem to exist." end end @dir_expects end ######################################################### # property dir_results # This is the in-source directory where results will # be generated. ######################################################### def dir_results unless @dir_results @dir_results = File.join( self.dir_test_sets, "#{self.name_test_set}-results" ) unless File.exist?(@dir_results) FileUtils.mkdir_p(@dir_results) unless File.exist?(@dir_results) raise StandardError.new "The expectations directory #{@dir_results} doesn’t seem to exist and couldn’t be created." end end end @dir_results end ######################################################### # property path_expects_html # A file in the expectations directory with the same # name as the HTML file (without @ metadata). ######################################################### def path_expects_html unless @path_expects_html basename = File.basename(self.path_file)[/(.*)@/,1] + File.extname(self.path_file) @path_expects_html = File.join( self.dir_expects, basename ) unless File.exist?(@path_expects_html) SHARED_LOGGER.warn("The expectations file #{@path_expects_html} doesn’t seem to exist.") end end @path_expects_html end ######################################################### # property path_expects_output # A file in the expectations directory with the same # name as the HTML file (without @ metadata), but # of type .txt. ######################################################### def path_expects_output unless @path_expects_output basename = File.basename(self.path_file)[/(.*)@/,1] + '.txt' @path_expects_output = File.join( self.dir_expects, basename ) unless File.exist?(@path_expects_output) SHARED_LOGGER.warn("The expectations file #{@path_expects_output} doesn’t seem to exist.") end end @path_expects_output end ######################################################### # property path_result_html # A file in the results directory with the same # name as the HTML file (without @ metadata). This # file doesn't necessarily exist until the test is # executed. ######################################################### def path_result_html unless @path_result_html basename = File.basename(self.path_file)[/(.*)@/,1] + File.extname(self.path_file) @path_result_html = File.join( self.dir_results, basename ) end @path_result_html end ######################################################### # property path_result_output # A file in the results directory with the same # name as the HTML file (without @ metadata), but # of type .txt. This file doesn't necessarily exist # until the test is executed. ######################################################### def path_result_output unless @path_result_output basename = File.basename(self.path_file)[/(.*)@/,1] + '.txt' @path_result_output = File.join( self.dir_results, basename ) end @path_result_output end ######################################################### # property path_tidy_exe ######################################################### def path_tidy_exe @path_tidy_exe end def path_tidy_exe=( value ) unless File.exist?( value ) raise StandardError.new "The path #{value} to tidy doesn’t exist." end unless File.executable?( value ) raise StandardError.new "The path #{value} to tidy is there, but isn’t executable." end @path_tidy_exe = value end ######################################################### # property @test_passed ######################################################### def test_passed unless @test_complete self.perform_test end @test_exit_status == self.expects_exit_code.to_i && @test_diff_html_status == 0 && @test_diff_output_status == 0 end ######################################################### # property @test_report ######################################################### def test_report unless @test_complete self.perform_test end if self.test_passed result = "PASSED | status #{@test_exit_status} expected #{self.expects_exit_code} | #{self.path_file}\n" else result = WINDOWS_OS ? "\n++++++++++++++++++++\n" : "\n⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓\n" result += "FAILED | status #{@test_exit_status} expected #{self.expects_exit_code} | #{self.path_file}\n" result += "\n#{@test_diff_html}" unless @test_diff_html_status == 0 result += "\n#{@test_diff_output}" unless @test_diff_output_status == 0 result += WINDOWS_OS ? "^^^^^^^^^^^^^^^^^^^^\n\n" : "⇑⇑⇑⇑⇑⇑⇑⇑⇑⇑⇑⇑⇑⇑⇑⇑⇑⇑⇑⇑\n\n" end result end ######################################################### # property @test_exit_status ######################################################### def test_exit_status @test_exit_status end ############################################################################# # Protected Methods ############################################################################# protected ######################################################### # perform_test ######################################################### def perform_test #-------------------------------------------------- # Remove existing output files, if any. #-------------------------------------------------- File.delete(self.path_result_html) if File.exist?(self.path_result_html) File.delete(self.path_result_output) if File.exist?(self.path_result_output) #-------------------------------------------------- # Run HTML Tidy on the input file. #-------------------------------------------------- shell_command = %Q<"#{self.path_tidy_exe}" -lang en_us -f "#{self.path_result_output}" -config "#{self.path_config}" --tidy-mark no -o "#{self.path_result_html}" "#{self.path_file}"> SHARED_LOGGER.info("Performing: #{shell_command}") %x(#{shell_command}) @test_exit_status = $?.exitstatus SHARED_LOGGER.info("Tidy exited with status: #{@test_exit_status}") #-------------------------------------------------- # Get the diffs for the html files. #-------------------------------------------------- diff_cmd = WINDOWS_OS ? 'fc /l' : 'diff -ua' if File.exist?(self.path_expects_html) shell_command = %Q<#{diff_cmd} "#{possible_windows_shell_path(self.path_expects_html)}" "#{possible_windows_shell_path(self.path_result_html)}"> SHARED_LOGGER.info("Performing: #{shell_command}") @test_diff_html = %x( #{shell_command} ) @test_diff_html_status = $?.exitstatus SHARED_LOGGER.info("diff/fc exited with status: #{@test_diff_html_status}") else # Although the HTML -expects file is missing, that might be by design, because # some cases are Tidy failures and don't produce output. If the HTML -results # is also missing, then this is actually a match, so we'll give it a pass. if File.exist?(self.path_result_html) @test_diff_html = "diff: Expectations file #{self.path_expects_html} is missing!" else @test_diff_html = "" @test_diff_html_status = 0 end end #-------------------------------------------------- # Get the diffs for the output files. #-------------------------------------------------- if File.exist?(self.path_expects_output) shell_command = %Q<#{diff_cmd} "#{possible_windows_shell_path(self.path_expects_output)}" "#{possible_windows_shell_path(self.path_result_output)}"> SHARED_LOGGER.info("Performing: #{shell_command}") @test_diff_output = %x( #{shell_command} ) @test_diff_output_status = $?.exitstatus SHARED_LOGGER.info("diff/fc exited with status: #{@test_diff_output_status}") else @test_diff_output = "diff: Expectations file #{self.path_expects_output} is missing!" end @test_complete = true end ########################################################### # Sometimes Windows still rejects normal file path # specifiers, such as the fc command, which thinks that # the file isn't found when using forward slashes. ########################################################### def possible_windows_shell_path(unix_path) if WINDOWS_OS unix_path.gsub(File::SEPARATOR, File::ALT_SEPARATOR || File::SEPARATOR) else unix_path end end end # class TidyTestCase ############################################################################# # class TidyRegressionTestCLI # This class provides handlers for CLI parameters. Because it's such a # simple application, it also serves as our Application. ############################################################################# class TidyRegressionTestCLI < Thor class_option :verbose, :type => :boolean, :desc => 'Provides verbose debug output.', :aliases => '-v' class_option :debug, :type => :boolean, :desc => 'Provides really verbose debug output.', :aliases => '-d' class_option :tidy, :type => :string, :desc => "Specify an alternate HTML Tidy binary instead of the default at #{EXE_TIDY}.", :aliases => '-t' ######################################################### # Property declarations ######################################################### attr_accessor :tidy_path attr_reader :all_test_sets # array of all test sets. ######################################################### # Initialize ######################################################### def initialize(args = nil, options = nil, config = nil) super set_options end ######################################################### # exit_on_failure? ######################################################### def self.exit_on_failure? true end ######################################################### # help # Override the default help in order to better describe # what we're doing. ######################################################### def help(*args) if args.count == 0 puts <<~HEREDOC This script (#{File.basename($0)}) runs HTML Tidy’s regression tests. You can instruct it to run all test sets, a single test set, or even a single case. Complete Help: -------------- HEREDOC end super end # help ######################################################### # test # See long_desc ######################################################### desc 'test', 'Runs all test cases in all test sets.' option :omitpassed, :type => :boolean, :desc => 'Omits PASSED test cases to keep output quieter.', :aliases => '-q' long_desc <<-LONG_DESC Runs all test cases in all test sets. Any directory in #{DIR_TEST_SETS} ending with '-cases' and having a matching '-expects' counterpart is assumed to be a test set. This command will exit with a non-zero exit code if any of tests do no exit with the expected exit code, or if any of the output produced does not match the expectation in '-expects'. It will also fail if no test cases can be found. Summary data will be printed to stdout, as well as any diffs in the event of a regression test failure. Use case: this is the general test you should perform when deciding whether or not to accept changes to Tidy's source code. LONG_DESC def test total_tested = 0 total_passed = 0 self.all_test_sets.each do |test_set| test_set.case_paths.each do |case_file| test_run = TidyTestCase.new(with_file: case_file, path_to_tidy: self.tidy_path) puts test_run.test_report unless test_run.test_passed && options[:omitpassed] total_tested += 1 total_passed += 1 if test_run.test_passed end end self.print_report_footer( total_tested, total_passed ) raise Thor::Error, "Automated Testing Failed." if total_passed < total_tested end # test ######################################################### # only # See long_desc ######################################################### desc 'only ', 'Only runs all of the test cases in the test set given.' option :omitpassed, :type => :boolean, :desc => 'Omits PASSED test cases to keep output quieter.', :aliases => '-q' long_desc <<-LONG_DESC Runs all test cases in the given test set. Any directory in #{DIR_TEST_SETS} ending with '-cases' and having a matching '-expects' counterpart is assumed to be a test set. The name of the test set does not include the '-cases' suffix, and so do not provide it. This command will exit with a non-zero exit code if any of tests do no exit with the expected exit code, or if any of the output produced does not match the expectation in '-expects'. It will also fail if no test cases can be found. Summary data will be printed to stdout, as well as any diffs in the event of a regression test failure. Use case: Use this command if you want to speed up testing, especially if you already know that other test sets have already been tested successfully. LONG_DESC def only( test_set_name ) total_tested = 0 total_passed = 0 test_set = all_test_sets.select { |test_case| test_case.test_set_name == test_set_name }.first if test_set test_set.case_paths.each do |case_file| test_run = TidyTestCase.new(with_file: case_file, path_to_tidy: self.tidy_path) puts test_run.test_report unless test_run.test_passed && options[:omitpassed] total_tested += 1 total_passed += 1 if test_run.test_passed end self.print_report_footer( total_tested, total_passed ) else puts "Test set #{test_set_name} could not be used, found, or does not meet conventions correctly." end raise Thor::Error, "Automated Testing Failed." if total_passed < total_tested end # only ######################################################### # case # See long_desc ######################################################### desc 'case ', 'Only tests a single case.' long_desc <<-LONG_DESC Run only the specific test case. If the same case name exists in multiple test sets, then they will be tested in the order that they are found. Don’t give your cases the same name if you do not wish to experience this behavior. Any directory in #{DIR_TEST_SETS} ending with '-cases' and having a matching '-expects' counterpart is assumed to be a test set. Do not provide any prefix or suffix parts of the case name. This command will exit with a non-zero exit code if the test does not exit with the expected exit code, or if any of the output produced does not match the expectation in '-expects'. It will also fail if the test cases cannot be found. Summary data will be printed to stdout, as well as any diffs in the event of a regression test failure. Use case: Use this command if you want to speed up testing, especially if you're adding a new test case. LONG_DESC def case( case_name ) total_tested = 0 total_passed = 0 self.all_test_sets.each do |test_set| test_set.case_paths.select {|case_path| File.basename(case_path)[/^case-(.*)@.*$/,1] == case_name }.each do |case_file| test_run = TidyTestCase.new(with_file: case_file, path_to_tidy: self.tidy_path) puts test_run.test_report unless test_run.test_passed && options[:omitpassed] total_tested += 1 total_passed += 1 if test_run.test_passed end end self.print_report_footer( total_tested, total_passed ) raise Thor::Error, "Automated Testing Failed." if total_passed < total_tested end # case ######################################################### # qo # See long_desc ######################################################### desc 'qo ', 'Quick-Opens the case file, the -expects, and -results files.' long_desc <<-LONG_DESC Quick-Opens the specific test case in your environment‘s default text editor, as well as the corresponding -expects file and -results files if present. Additionally, it will print the diff commands for your operating system pointing to the html files and the output files, respectively. This makes it easy to copy and paste if you would like to see a diff view. Use case: Use this command if a test case fails, and you want to view all of the relevant files in your text editor for comparison. LONG_DESC def qo( case_name ) if WINDOWS_OS diff_cmd = 'fc -l' editor = TTY::Editor.new else diff_cmd = 'diff -ua' # This means of instantiation is temporary until # tty-editor is patched to trust the program = File.basename(ENV['EDITOR']) editor = TTY::Editor.new(command: program ) end self.all_test_sets.each do |test_set| test_set.case_paths.select {|case_path| File.basename(case_path)[/^case-(.*)@.*$/,1] == case_name }.each do |case_file| # This will *not* run the tests and overwrite the files; that would only occur # if we try to pull a test report. This makes it simple to get our paths. test_run = TidyTestCase.new(with_file: case_file, path_to_tidy: self.tidy_path) editor.open("#{test_run.path_file}") if File.exist?(test_run.path_file) editor.open("#{test_run.path_config}") if File.exist?(test_run.path_config) editor.open("#{test_run.path_expects_html}") if File.exist?(test_run.path_expects_html) editor.open("#{test_run.path_expects_output}") if File.exist?(test_run.path_expects_output) editor.open("#{test_run.path_result_html}") if File.exist?(test_run.path_result_html) editor.open("#{test_run.path_result_output}") if File.exist?(test_run.path_result_output) puts <<~HEREDOC #{diff_cmd} #{test_run.path_expects_html} #{test_run.path_result_html} #{diff_cmd} #{test_run.path_expects_output} #{test_run.path_result_output} HEREDOC end end end # qo protected ######################################################### # set_options # Handles command line options. ######################################################### def set_options SHARED_LOGGER.level = Logger::WARN if options[:verbose] SHARED_LOGGER.level = Logger::DEBUG if options[:debug] self.tidy_path = options[:tidy] ? File.expand_path(options[:tidy]) : EXE_TIDY end # set_options ######################################################### # print_report_footer ######################################################### def print_report_footer( total_tested, total_passed ) tidy_version = %x( #{self.tidy_path} --version ) tidy_version = tidy_version[/^.*(\d+\.\d+\.\d+).*/,1] cases_version = File.open(VERSION_FILE).first.chomp puts puts "Note that no valid cases were found." if total_tested < 1 puts "Ran #{total_tested} tests, of which #{total_passed} passed and #{total_tested - total_passed} failed." puts "Test conducted with HTML Tidy #{tidy_version} using test sets for version #{cases_version}." end ######################################################### # property tidy_path ######################################################### def tidy_path @tidy_path end def tidy_path=( value ) value += '.exe' if WINDOWS_OS && File.extname(value) != '.exe' unless File.exist?( value ) raise StandardError.new "The path #{value} to tidy doesn’t exist." end @tidy_path = value end ######################################################### # property all_test_sets ######################################################### def all_test_sets unless @all_test_sets @all_test_sets = [] Dir.glob("#{DIR_TEST_SETS}/*-cases").select { |entry| File.directory? entry }.sort.each do |test_set_directory| new_test_set = TidyTestSet.new(with_directory: test_set_directory, path_to_tidy: self.tidy_path) unless new_test_set.dir_test_set.nil? @all_test_sets.push(new_test_set) end end end @all_test_sets end end # TidyRegressionTestCLI end # TidyRegressionTestModule ########################################################### # Main ########################################################### TidyRegressionTestModule::TidyRegressionTestCLI.start(ARGV)