Move new maven spec implementation into a new file

maven/lib/dependabot/maven/new_version.rb

@@ -0,0 +1,55 @@
+# typed: true
+# frozen_string_literal: true
+
+require "dependabot/maven/version_parser"
+require "dependabot/version"
+require "dependabot/utils"
+
+# See https://maven.apache.org/pom.html#Version_Order_Specification for details.
+
+module Dependabot
+  module Maven
+    class NewVersion < Dependabot::Version
+      PRERELEASE_QUALIFIERS = [
+        Dependabot::Maven::VersionParser::ALPHA,
+        Dependabot::Maven::VersionParser::BETA,
+        Dependabot::Maven::VersionParser::MILESTONE,
+        Dependabot::Maven::VersionParser::RC,
+        Dependabot::Maven::VersionParser::SNAPSHOT
+      ].freeze
+
+      attr_accessor :bucket
+
+      def self.correct?(version)
+        return false if version.nil? || version.empty?
+
+        Dependabot::Maven::VersionParser.parse(version.to_s).any?
+      end
+
+      def initialize(version)
+        @version_string = version.to_s
+        @bucket = Dependabot::Maven::VersionParser.parse(version.to_s)
+      end
+
+      def inspect
+        "#<#{self.class} #{@version_string}>"
+      end
+
+      def to_s
+        @version_string
+      end
+
+      def prerelease?
+        bucket.to_a.flatten.any? do |token|
+          token.is_a?(Integer) && token.negative?
+        end
+      end
+
+      def <=>(other)
+        bucket <=> other.bucket
+      end
+    end
+  end
+end
+
+Dependabot::Utils.register_version_class("maven", Dependabot::Maven::NewVersion)

maven/lib/dependabot/maven/version.rb

@@ -1,34 +1,47 @@
 # typed: true
 # frozen_string_literal: true
 
-require "dependabot/maven/version_parser"
 require "dependabot/version"
 require "dependabot/utils"
 
+# Java versions use dots and dashes when tokenising their versions.
+# Gem::Version converts a "-" to ".pre.", so we override the `to_s` method.
+#
 # See https://maven.apache.org/pom.html#Version_Order_Specification for details.
 
 module Dependabot
   module Maven
     class Version < Dependabot::Version
-      PRERELEASE_QUALIFIERS = [
-        Dependabot::Maven::VersionParser::ALPHA,
-        Dependabot::Maven::VersionParser::BETA,
-        Dependabot::Maven::VersionParser::MILESTONE,
-        Dependabot::Maven::VersionParser::RC,
-        Dependabot::Maven::VersionParser::SNAPSHOT
-      ].freeze
-
-      attr_accessor :bucket
+      NULL_VALUES = %w(0 final ga).freeze
+      PREFIXED_TOKEN_HIERARCHY = {
+        "." => { qualifier: 1, number: 4 },
+        "-" => { qualifier: 2, number: 3 },
+        "+" => { qualifier: 3, number: 2 }
+      }.freeze
+      NAMED_QUALIFIERS_HIERARCHY = {
+        "a" => 1, "alpha"     => 1,
+        "b" => 2, "beta"      => 2,
+        "m" => 3, "milestone" => 3,
+        "rc" => 4, "cr" => 4, "pr" => 4, "pre" => 4,
+        "snapshot" => 5, "dev" => 5,
+        "ga" => 6, "" => 6, "final" => 6,
+        "sp" => 7
+      }.freeze
+      VERSION_PATTERN =
+        "[0-9a-zA-Z]+" \
+        '(?>\.[0-9a-zA-Z]*)*' \
+        '([_\-\+][0-9A-Za-z_-]*(\.[0-9A-Za-z_-]*)*)?'
+      ANCHORED_VERSION_PATTERN = /\A\s*(#{VERSION_PATTERN})?\s*\z/
 
       def self.correct?(version)
-        return false if version.nil? || version.empty?
+        return false if version.nil?
 
-        Dependabot::Maven::VersionParser.parse(version.to_s).any?
+        version.to_s.match?(ANCHORED_VERSION_PATTERN)
       end
 
       def initialize(version)
         @version_string = version.to_s
-        @bucket = Dependabot::Maven::VersionParser.parse(version.to_s)
+        super(version.to_s.tr("_", "-"))
       end
 
       def inspect
@@ -40,13 +53,139 @@ def to_s
       end
 
       def prerelease?
-        bucket.to_a.flatten.any? do |token|
-          token.is_a?(Integer) && token.negative?
+        tokens.any? do |token|
+          next true if token == "eap"
+          next false unless NAMED_QUALIFIERS_HIERARCHY[token]
+
+          NAMED_QUALIFIERS_HIERARCHY[token] < 6
         end
       end
 
       def <=>(other)
-        bucket <=> other.bucket
+        version = stringify_version(@version_string)
+        version = fill_tokens(version)
+        version = trim_version(version)
+
+        other_version = stringify_version(other)
+        other_version = fill_tokens(other_version)
+        other_version = trim_version(other_version)
+
+        version, other_version = convert_dates(version, other_version)
+
+        prefixed_tokens = split_into_prefixed_tokens(version)
+        other_prefixed_tokens = split_into_prefixed_tokens(other_version)
+
+        prefixed_tokens, other_prefixed_tokens =
+          pad_for_comparison(prefixed_tokens, other_prefixed_tokens)
+
+        prefixed_tokens.count.times.each do |index|
+          comp = compare_prefixed_token(
+            prefix: prefixed_tokens[index][0],
+            token: prefixed_tokens[index][1..-1] || "",
+            other_prefix: other_prefixed_tokens[index][0],
+            other_token: other_prefixed_tokens[index][1..-1] || ""
+          )
+          return comp unless comp.zero?
+        end
+
+        0
+      end
+
+      private
+
+      def tokens
+        @tokens ||=
+          begin
+            version = @version_string.to_s.downcase
+            version = fill_tokens(version)
+            version = trim_version(version)
+            split_into_prefixed_tokens(version).map { |t| t[1..-1] }
+          end
+      end
+
+      def stringify_version(version)
+        version = version.to_s.downcase
+
+        # Not technically correct, but pragmatic
+        version.gsub(/^v(?=\d)/, "")
+      end
+
+      def fill_tokens(version)
+        # Add separators when transitioning from digits to characters
+        version = version.gsub(/(\d)([A-Za-z])/, '\1-\2')
+        version = version.gsub(/([A-Za-z])(\d)/, '\1-\2')
+
+        # Replace empty tokens with 0
+        version = version.gsub(/([\.\-])([\.\-])/, '\10\2')
+        version = version.gsub(/^([\.\-])/, '0\1')
+        version.gsub(/([\.\-])$/, '\10')
+      end
+
+      def trim_version(version)
+        version.split("-").filter_map do |v|
+          parts = v.split(".")
+          parts = parts[0..-2] while NULL_VALUES.include?(parts&.last)
+          parts&.join(".")
+        end.reject(&:empty?).join("-")
+      end
+
+      def convert_dates(version, other_version)
+        default = [version, other_version]
+        return default unless version.match?(/^\d{4}-?\d{2}-?\d{2}$/)
+        return default unless other_version.match?(/^\d{4}-?\d{2}-?\d{2}$/)
+
+        [version.delete("-"), other_version.delete("-")]
+      end
+
+      def split_into_prefixed_tokens(version)
+        ".#{version}".split(/(?=[\-\.\+])/)
+      end
+
+      def pad_for_comparison(prefixed_tokens, other_prefixed_tokens)
+        prefixed_tokens = prefixed_tokens.dup
+        other_prefixed_tokens = other_prefixed_tokens.dup
+
+        longest = [prefixed_tokens, other_prefixed_tokens].max_by(&:count)
+        shortest = [prefixed_tokens, other_prefixed_tokens].min_by(&:count)
+
+        longest.count.times do |index|
+          next unless shortest[index].nil?
+
+          shortest[index] = longest[index].start_with?(".") ? ".0" : "-"
+        end
+
+        [prefixed_tokens, other_prefixed_tokens]
+      end
+
+      def compare_prefixed_token(prefix:, token:, other_prefix:, other_token:)
+        token_type = token.match?(/^\d+$/) ? :number : :qualifier
+        other_token_type = other_token.match?(/^\d+$/) ? :number : :qualifier
+
+        hierarchy = PREFIXED_TOKEN_HIERARCHY.fetch(prefix).fetch(token_type)
+        other_hierarchy =
+          PREFIXED_TOKEN_HIERARCHY.fetch(other_prefix).fetch(other_token_type)
+
+        hierarchy_comparison = hierarchy <=> other_hierarchy
+        return hierarchy_comparison unless hierarchy_comparison.zero?
+
+        compare_token(token: token, other_token: other_token)
+      end
+
+      def compare_token(token:, other_token:)
+        if (token_hierarchy = NAMED_QUALIFIERS_HIERARCHY[token])
+          return -1 unless NAMED_QUALIFIERS_HIERARCHY[other_token]
+
+          return token_hierarchy <=> NAMED_QUALIFIERS_HIERARCHY[other_token]
+        end
+
+        return 1 if NAMED_QUALIFIERS_HIERARCHY[other_token]
+
+        if token.match?(/\A\d+\z/) && other_token.match?(/\A\d+\z/)
+          token = token.to_i
+          other_token = other_token.to_i
+        end
+
+        token <=> other_token
       end
     end
   end