xcop 0.9.0 → 0.10.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: f920205808498fe938096e3c94b78366b740bd435c80a96e1c6b414609db2580
-  data.tar.gz: 99ceffd495685cd59a26788dc8c1f8ef384d105ffc80967e0e12443639585829
+  metadata.gz: fa89e4f1fcf4bd98059669b6fdd5dfae445a9c6fc43bbee41486f974b8d7af6e
+  data.tar.gz: 32f664bba8e5dc4450919c6af6f19146a80d97ef9bc5db595cbd4800afd8d9e2
 SHA512:
-  metadata.gz: 8859eebb530bbad420b2645b4971fcb2c272a244957de2d6575e932a9ba516ecd6112bd6369da69a61ac2ad08cdba228ea45bf92427be69b713d885c16620f75
-  data.tar.gz: b38404a73d36189719654efbf9f1f3c0e12fdbeab9cd8d7e207224e2470e0de8d77ad72dc4b955fc82ebe12f19ab6b585ddc658672186d7f19170cad08e0c3ac
+  metadata.gz: 9fc34e6689bb4f67f3f14c764f2443764daf856b6b4c5be763b07bc9b4ab0003b3841fdb38c3120478dfc9e1f93d23462c8021c4d189ec9a4cf69a60c588200d
+  data.tar.gz: 71fae53f841ad1cb2ff31e3ca741d4d167643099daad5d270aac8aaac86732b3851c659d18974a707d51f01f4fae1eb7742324681d3ce7aca3b295ddd38776a2

data/Gemfile

@@ -11,6 +11,7 @@ gem 'minitest', '~>6.0', require: false
 gem 'minitest-reporters', '~>1.7', require: false
 gem 'rake', '~>13.2', require: false
 gem 'rubocop', '~>1.75', require: false
+gem 'rubocop-elegant', '~>0.5', require: false
 gem 'rubocop-minitest', '~>0.38', require: false
 gem 'rubocop-performance', '~>1.25', require: false
 gem 'rubocop-rake', '~>0.7', require: false

data/Gemfile.lock

@@ -13,9 +13,9 @@ GEM
     ansi (1.6.0)
     ast (2.4.3)
     base64 (0.3.0)
-    bigdecimal (4.1.1)
+    bigdecimal (4.1.2)
     builder (3.3.0)
-    cucumber (11.0.0)
+    cucumber (11.1.1)
       base64 (~> 0.2)
       builder (~> 3.2)
       cucumber-ci-environment (> 9, < 12)
@@ -32,9 +32,9 @@ GEM
       cucumber-gherkin (> 36, < 40)
       cucumber-messages (> 31, < 33)
       cucumber-tag-expressions (> 6, < 9)
-    cucumber-cucumber-expressions (19.0.0)
+    cucumber-cucumber-expressions (19.0.1)
       bigdecimal
-    cucumber-gherkin (39.0.0)
+    cucumber-gherkin (39.1.0)
       cucumber-messages (>= 31, < 33)
     cucumber-html-formatter (23.1.0)
       cucumber-messages (> 23, < 33)
@@ -45,7 +45,7 @@ GEM
     docile (1.4.1)
     drb (2.2.3)
     ellipsized (0.3.0)
-    erb (6.0.4)
+    erb (6.0.5)
     ffi (1.17.4-aarch64-linux-gnu)
     ffi (1.17.4-aarch64-linux-musl)
     ffi (1.17.4-arm-linux-gnu)
@@ -55,8 +55,8 @@ GEM
     ffi (1.17.4-x86_64-darwin)
     ffi (1.17.4-x86_64-linux-gnu)
     ffi (1.17.4-x86_64-linux-musl)
-    json (2.19.3)
-    language_server-protocol (3.17.0.5)
+    json (2.20.0)
+    language_server-protocol (3.17.0.6)
     lint_roller (1.1.0)
     logger (1.7.0)
     loog (0.8.0)
@@ -64,7 +64,7 @@ GEM
       logger (~> 1.0)
     memoist3 (1.0.0)
     mini_mime (1.1.5)
-    minitest (6.0.5)
+    minitest (6.0.6)
       drb (~> 2.0)
       prism (~> 1.5)
     minitest-reporters (1.8.0)
@@ -73,25 +73,25 @@ GEM
       minitest (>= 5.0, < 7)
       ruby-progressbar
     multi_test (1.1.0)
-    nokogiri (1.19.0-aarch64-linux-gnu)
+    nokogiri (1.19.3-aarch64-linux-gnu)
       racc (~> 1.4)
-    nokogiri (1.19.0-aarch64-linux-musl)
+    nokogiri (1.19.3-aarch64-linux-musl)
       racc (~> 1.4)
-    nokogiri (1.19.0-arm-linux-gnu)
+    nokogiri (1.19.3-arm-linux-gnu)
       racc (~> 1.4)
-    nokogiri (1.19.0-arm-linux-musl)
+    nokogiri (1.19.3-arm-linux-musl)
       racc (~> 1.4)
-    nokogiri (1.19.0-arm64-darwin)
+    nokogiri (1.19.3-arm64-darwin)
       racc (~> 1.4)
-    nokogiri (1.19.0-x64-mingw-ucrt)
+    nokogiri (1.19.3-x64-mingw-ucrt)
       racc (~> 1.4)
-    nokogiri (1.19.0-x86_64-darwin)
+    nokogiri (1.19.3-x86_64-darwin)
       racc (~> 1.4)
-    nokogiri (1.19.0-x86_64-linux-gnu)
+    nokogiri (1.19.3-x86_64-linux-gnu)
       racc (~> 1.4)
-    nokogiri (1.19.0-x86_64-linux-musl)
+    nokogiri (1.19.3-x86_64-linux-musl)
       racc (~> 1.4)
-    parallel (1.27.0)
+    parallel (1.28.0)
     parser (3.3.11.1)
       ast (~> 2.4.1)
       racc
@@ -99,9 +99,9 @@ GEM
     racc (1.8.1)
     rainbow (3.1.1)
     rake (13.4.2)
-    regexp_parser (2.11.3)
+    regexp_parser (2.12.0)
     rexml (3.4.4)
-    rubocop (1.86.1)
+    rubocop (1.88.2)
       json (~> 2.3)
       language_server-protocol (~> 3.17.0.2)
       lint_roller (~> 1.1.0)
@@ -115,6 +115,9 @@ GEM
     rubocop-ast (1.49.1)
       parser (>= 3.3.7.2)
       prism (~> 1.7)
+    rubocop-elegant (0.7.1)
+      lint_roller (~> 1.1)
+      rubocop (~> 1.75)
     rubocop-minitest (0.39.1)
       lint_roller (~> 1.1)
       rubocop (>= 1.75.0, < 2.0)
@@ -148,7 +151,7 @@ GEM
       unicode-emoji (~> 4.1)
     unicode-emoji (4.2.0)
     win32ole (1.9.3)
-    yard (0.9.43)
+    yard (0.9.44)
 
 PLATFORMS
   aarch64-linux-gnu
@@ -170,6 +173,7 @@ DEPENDENCIES
   minitest-reporters (~> 1.7)
   rake (~> 13.2)
   rubocop (~> 1.75)
+  rubocop-elegant (~> 0.5)
   rubocop-minitest (~> 0.38)
   rubocop-performance (~> 1.25)
   rubocop-rake (~> 0.7)

data/README.md

@@ -122,7 +122,7 @@ To customize the files pattern, use the `files` parameter:
 ## How to use in Maven `pom.xml`?
 
 You can integrate it with the help of
-[maven-antrun-plugin](http://maven.apache.org/plugins/maven-antrun-plugin/):
+[maven-antrun-plugin](https://maven.apache.org/plugins/maven-antrun-plugin/):
 
 ```xml
 <project>

data/Rakefile

@@ -1,9 +1,9 @@
 # SPDX-FileCopyrightText: Copyright (c) 2017-2026 Yegor Bugayenko
 # SPDX-License-Identifier: MIT
 
-require 'rubygems'
 require 'rake'
 require 'rake/clean'
+require 'rubygems'
 
 def name
   @name ||= File.basename(Dir['*.gemspec'].first, '.*')

data/bin/xcop

@@ -65,8 +65,8 @@ if opts[:exclude]
 end
 
 if opts.fix?
-  Xcop::CLI.new(files).fix do |f|
-    puts "#{f} fixed" unless opts.quiet?
+  Xcop::CLI.new(files).fix do |f, status|
+    puts "#{f} #{status}" unless opts.quiet?
   end
 else
   begin

data/features/comments.feature

@@ -0,0 +1,65 @@
+# SPDX-FileCopyrightText: Copyright (c) 2017-2026 Yegor Bugayenko
+# SPDX-License-Identifier: MIT
+Feature: Comment Formatting
+  As an author of XML I want XCOP to enforce
+  a single canonical layout for XML comments
+
+  Scenario: Messy single-line comment is rejected
+    Given I have a "single.xml" file with content:
+    """
+    <?xml version="1.0"?>
+    <a>
+      <!--   text   -->
+    </a>
+
+    """
+    When I run bin/xcop with "single.xml"
+    Then Exit code is not zero
+
+  Scenario: Messy single-line comment is fixed
+    Given I have a "single.xml" file with content:
+    """
+    <?xml version="1.0"?>
+    <a>
+      <!--   text   -->
+    </a>
+
+    """
+    When I run bin/xcop with "--fix single.xml"
+    Then Exit code is zero
+    And I run bin/xcop with "single.xml"
+    Then Exit code is zero
+    And Stdout contains "single.xml looks good"
+
+  Scenario: Over-indented multi-line comment is rejected
+    Given I have a "multi.xml" file with content:
+    """
+    <?xml version="1.0"?>
+    <a>
+      <!--
+          line1
+          line2
+      -->
+    </a>
+
+    """
+    When I run bin/xcop with "multi.xml"
+    Then Exit code is not zero
+
+  Scenario: Over-indented multi-line comment is fixed
+    Given I have a "multi.xml" file with content:
+    """
+    <?xml version="1.0"?>
+    <a>
+      <!--
+          line1
+          line2
+      -->
+    </a>
+
+    """
+    When I run bin/xcop with "--fix multi.xml"
+    Then Exit code is zero
+    And I run bin/xcop with "multi.xml"
+    Then Exit code is zero
+    And Stdout contains "multi.xml looks good"

data/features/step_definitions/steps.rb

@@ -1,20 +1,21 @@
 # SPDX-FileCopyrightText: Copyright (c) 2017-2026 Yegor Bugayenko
 # SPDX-License-Identifier: MIT
 
+require 'English'
 require 'nokogiri'
-require 'tmpdir'
 require 'slop'
-require 'English'
+require 'tmpdir'
 
 Before do
   @cwd = Dir.pwd
   @dir = Dir.mktmpdir('test')
   FileUtils.mkdir_p(@dir)
   Dir.chdir(@dir)
-  @opts = Slop.parse ['-v', '-s', @dir] do |o|
-    o.bool '-v', '--verbose'
-    o.string '-s', '--source'
-  end
+  @opts =
+    Slop.parse(['-v', '-s', @dir]) do |o|
+      o.bool('-v', '--verbose')
+      o.string('-s', '--source')
+    end
 end
 
 After do
@@ -34,19 +35,19 @@ When(%r{^I run bin/xcop with "([^"]*)"$}) do |arg|
 end
 
 Then(/^Stdout contains "([^"]*)"$/) do |txt|
-  raise "STDOUT doesn't contain '#{txt}':\n#{@stdout}" unless @stdout.include?(txt)
+  raise(StandardError, "STDOUT doesn't contain '#{txt}':\n#{@stdout}") unless @stdout.include?(txt)
 end
 
 Then(/^Stdout is empty$/) do
-  raise "STDOUT is not empty:\n#{@stdout}" unless @stdout == ''
+  raise(StandardError, "STDOUT is not empty:\n#{@stdout}") unless @stdout == ''
 end
 
 Then(/^Exit code is zero$/) do
-  raise "Non-zero exit #{@exitstatus}:\n#{@stdout}" unless @exitstatus.zero?
+  raise(StandardError, "Non-zero exit #{@exitstatus}:\n#{@stdout}") unless @exitstatus.zero?
 end
 
 Then(/^Exit code is not zero$/) do
-  raise 'Zero exit code' if @exitstatus.zero?
+  raise(StandardError, 'Zero exit code') if @exitstatus.zero?
 end
 
 When(/^I run bash with "([^"]*)"$/) do |text|

data/lib/xcop/cli.rb

@@ -1,18 +1,17 @@
 # SPDX-FileCopyrightText: Copyright (c) 2017-2026 Yegor Bugayenko
 # SPDX-License-Identifier: MIT
 
-require 'nokogiri'
 require 'differ'
+require 'nokogiri'
 require 'rainbow'
-require_relative 'version'
 require_relative 'document'
+require_relative 'version'
 
 # Command line interface.
 # Author:: Yegor Bugayenko (yegor256@gmail.com)
 # Copyright:: Copyright (c) 2017-2026 Yegor Bugayenko
 # License:: MIT
 class Xcop::CLI
-  # Extensions recognized when a directory is passed as input.
   EXTENSIONS = %w[xml xsd xhtml xsl html].freeze
 
   def initialize(files, nocolor: false)
@@ -30,18 +29,25 @@ class Xcop::CLI
       doc = Xcop::Document.new(f)
       diff = doc.diff(nocolor: @nocolor)
       unless diff.empty?
-        puts diff
-        raise "Invalid XML formatting in #{f}"
+        puts(diff)
+        raise(StandardError, "Invalid XML formatting in #{f}")
+      end
+      errors = doc.validate
+      unless errors.empty?
+        puts(errors.join("\n"))
+        raise(StandardError, "XSD validation failed in #{f}")
       end
       yield(f) if block_given?
     end
   end
 
-  # Fix them all.
+  # Fix them all. The block, when given, receives the file path and a
+  # status symbol that is +:fixed+ when the file was rewritten and
+  # +:untouched+ when the file was already canonical.
   def fix
     @files.each do |f|
-      Xcop::Document.new(f).fix
-      yield(f) if block_given?
+      status = Xcop::Document.new(f).fix
+      yield(f, status) if block_given?
     end
   end
 end

data/lib/xcop/document.rb

@@ -1,9 +1,10 @@
 # SPDX-FileCopyrightText: Copyright (c) 2017-2026 Yegor Bugayenko
 # SPDX-License-Identifier: MIT
 
-require 'nokogiri'
 require 'differ'
+require 'nokogiri'
 require 'rainbow'
+require 'set'
 require_relative 'version'
 
 # One document.
@@ -19,21 +20,127 @@ class Xcop::Document
 
   # Return the difference, if any (empty string if everything is clean).
   def diff(nocolor: false)
-    xml = Nokogiri::XML(File.open(@path), &:noblanks)
-    ideal = xml.to_xml(indent: 2)
-    now = File.read(@path)
-    differ(ideal, now, nocolor: nocolor)
+    differ(ideal, File.read(@path), nocolor: nocolor)
   end
 
-  # Fixes the document.
+  # Fixes the document, returning +:fixed+ when the file content
+  # changed on disk and +:untouched+ when the file was already
+  # canonical.
   def fix
-    xml = Nokogiri::XML(File.open(@path), &:noblanks)
-    ideal = xml.to_xml(indent: 2)
-    File.write(@path, ideal)
+    canonical = ideal
+    return :untouched if canonical == File.read(@path)
+    File.write(@path, canonical)
+    :fixed
+  end
+
+  # Validates the document against its declared XSD schema, if any.
+  # Returns an array of error message strings (empty when valid or no
+  # schema declared); a declared but missing schema is itself an error.
+  def validate
+    xml = Nokogiri::XML(File.read(@path))
+    xsd = schema(xml)
+    return [] unless xsd
+    return ["schema file is absent at #{xsd}"] unless File.exist?(xsd)
+    Nokogiri::XML::Schema(File.read(xsd)).validate(xml).map(&:message)
   end
 
+  XSI_NS = 'http://www.w3.org/2001/XMLSchema-instance'.freeze
+
   private
 
+  # The canonical, well-formatted version of the document.
+  def ideal
+    xml = Nokogiri::XML(File.read(@path), &:noblanks)
+    reshape(xml)
+    text = xml.to_xml(indent: 2)
+    unused(xml).each do |prefix|
+      text =
+        if prefix.nil?
+          text.gsub(/\s+xmlns="[^"]*"/, '')
+        else
+          text.gsub(/\s+xmlns:#{Regexp.escape(prefix)}="[^"]*"/, '')
+        end
+    end
+    Nokogiri::XML(text, &:noblanks).to_xml(indent: 2)
+  end
+
+  # Rewrites every comment node into one of two canonical shapes so that
+  # +diff+ and +--fix+ can report and correct sloppy comment layout. A
+  # comment whose text has no end-of-line character is collapsed onto a
+  # single line as +<!-- text -->+. A comment whose text spans several
+  # lines is re-emitted with +<!--+ and +-->+ each alone on their own
+  # line and every body line trimmed to the comment's own indent.
+  def reshape(xml)
+    xml.traverse do |node|
+      next unless node.comment?
+      node.native_content = canonical(node)
+    end
+  end
+
+  # The canonical inner text (the part between +<!--+ and +-->+) for a
+  # single comment node, indented to match the node's depth.
+  def canonical(node)
+    raw = node.content
+    body = raw.strip
+    return " #{body} " unless raw.include?("\n")
+    indent = '  ' * depth(node)
+    body = body.split("\n").map { |line| line.strip.empty? ? '' : indent + line.strip }.join("\n")
+    "\n#{body}\n#{indent}"
+  end
+
+  # The number of ancestor elements of +node+, which equals the number
+  # of two-space indentation levels that +to_xml+ places before it.
+  def depth(node)
+    levels = 0
+    parent = node.parent
+    while parent && parent.element?
+      levels += 1
+      parent = parent.parent
+    end
+    levels
+  end
+
+  # Returns the set of namespace prefixes that are declared in +xml+
+  # but never referenced by any element or attribute. A +nil+ entry in
+  # the set represents the default namespace.
+  #
+  # A prefix counts as "used" not only when it appears as an element or
+  # attribute QName, but also when it is referenced from an attribute
+  # value as a +prefix:+ token. XSLT and XPath routinely mention a
+  # prefix only inside +select+, +test+, +match+, +as+, +use+ and
+  # similar attributes (e.g. +as="xs:integer"+ or +select="eo:foo()"+),
+  # so stripping such a "declared but not QName-used" prefix would break
+  # the stylesheet. See #161.
+  def unused(xml)
+    used = Set.new
+    declared = Set.new
+    xml.traverse do |node|
+      next unless node.is_a?(Nokogiri::XML::Element)
+      used << node.namespace.prefix if node.namespace
+      node.attribute_nodes.each do |attr|
+        used << attr.namespace.prefix if attr.namespace
+        attr.value.scan(/([A-Za-z_][\w.-]*):/) { |m| used << m.first }
+      end
+      node.namespace_definitions.each { |ns| declared << ns.prefix }
+    end
+    declared - used
+  end
+
+  def schema(xml)
+    root = xml.root
+    return unless root
+    plain = root.at_xpath('@xsi:noNamespaceSchemaLocation', 'xsi' => XSI_NS)
+    mapped = root.at_xpath('@xsi:schemaLocation', 'xsi' => XSI_NS)
+    location =
+      if plain
+        plain.value
+      elsif mapped
+        mapped.value.split.last
+      end
+    return unless location
+    File.expand_path(location, File.dirname(@path))
+  end
+
   def differ(ideal, fact, nocolor: false)
     return '' if ideal == fact
     if nocolor

data/lib/xcop/rake_task.rb

@@ -18,9 +18,9 @@ class Xcop::RakeTask < Rake::TaskLib
     @includes = %w[xml xsd xhtml xsl html].map { |e| "**/*.#{e}" }
     @excludes = []
     @quiet = false
-    desc 'Run Xcop' unless ::Rake.application.last_description
+    desc('Run Xcop') unless ::Rake.application.last_description
     task(name, *args) do |_, task_args|
-      RakeFileUtils.send(:verbose, true) do
+      RakeFileUtils.__send__(:verbose, true) do
         yield(*[self, task_args].slice(0, task_block.arity)) if block_given?
         run
       end
@@ -30,22 +30,23 @@ class Xcop::RakeTask < Rake::TaskLib
   private
 
   def run
-    puts 'Running xcop...' unless @quiet
+    puts('Running xcop...') unless @quiet
     bad = Dir.glob(@excludes)
     good = Dir.glob(@includes).reject { |f| bad.include?(f) }
-    puts "Inspecting #{pluralize(good.length, 'file')}..." unless @quiet
+    puts("Inspecting #{pluralize(good.length, 'file')}...") unless @quiet
     begin
       Xcop::CLI.new(good).run do
-        print Rainbow('.').green unless @quiet
+        print(Rainbow('.').green) unless @quiet
       end
     rescue StandardError => e
-      puts e.message
+      puts(e.message)
       abort(e.message)
     end
     return if @quiet
-    puts \
+    puts(
       "\n#{pluralize(good.length, 'file')} checked, " \
       "everything looks #{Rainbow('pretty').green}"
+    )
   end
 
   def pluralize(num, text)

data/lib/xcop/version.rb

@@ -6,5 +6,5 @@
 # Copyright:: Copyright (c) 2017-2026 Yegor Bugayenko
 # License:: MIT
 module Xcop
-  VERSION = '0.9.0'.freeze
+  VERSION = '0.10.1'.freeze
 end

data/xcop.gemspec

@@ -8,7 +8,7 @@ $LOAD_PATH.unshift(lib) unless $LOAD_PATH.include?(lib)
 require_relative 'lib/xcop/version'
 
 Gem::Specification.new do |s|
-  s.required_rubygems_version = Gem::Requirement.new('>= 0') if s.respond_to? :required_rubygems_version=
+  s.required_rubygems_version = Gem::Requirement.new('>= 0') if s.respond_to?(:required_rubygems_version=)
   s.required_ruby_version = '>= 2.2'
   s.name = 'xcop'
   s.version = Xcop::VERSION
@@ -23,8 +23,8 @@ Gem::Specification.new do |s|
   s.executables = s.files.grep(%r{^bin/}) { |f| File.basename(f) }
   s.rdoc_options = ['--charset=UTF-8']
   s.extra_rdoc_files = ['README.md', 'LICENSE.txt']
-  s.add_dependency 'differ', '~>0.1.2'
-  s.add_dependency 'nokogiri', '~>1.8'
-  s.add_dependency 'rainbow', '~>3.0'
-  s.add_dependency 'slop', '~>4.4'
+  s.add_dependency('differ', '~>0.1.2')
+  s.add_dependency('nokogiri', '~>1.8')
+  s.add_dependency('rainbow', '~>3.0')
+  s.add_dependency('slop', '~>4.4')
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: xcop
 version: !ruby/object:Gem::Version
-  version: 0.9.0
+  version: 0.10.1
 platform: ruby
 authors:
 - Yegor Bugayenko
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2026-04-24 00:00:00.000000000 Z
+date: 2026-07-17 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: differ
@@ -85,6 +85,7 @@ files:
 - bin/xcop
 - cucumber.yml
 - features/cli.feature
+- features/comments.feature
 - features/gem_package.feature
 - features/rake.feature
 - features/step_definitions/steps.rb