uri-whatwg_parser 0.1.4 → 0.1.6

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 2061d819a10365a97f389d9cb70325c87b75c390c4145897ed640c3e4cc85f70
-  data.tar.gz: 905ec3bdd174039fc91ce433f84a079fe12ff93f9ac999f420b08760b3fcee53
+  metadata.gz: 712c7079461310fdab82e79bea4491ce91a4a8b26827a338334a0e161520a8c0
+  data.tar.gz: 85871841fdee54a61c942134398bfb85e2e89f798144921464058e4edd2ea8fb
 SHA512:
-  metadata.gz: fc4a0d83fb46e959feb1f7990fa1f119428764447c6d4d455ec5287d5823b19eec845ef4b8912ceeffd6233266b3142859a837c7a26e47e20315d64bb5660006
-  data.tar.gz: bd36808311ae2330cd897ba824219d82cff0497d75531441b3679def9ff5ea1969297ccfcbb8cd02b83909a2f7a5c0de7f0d505bb79e1117494544dd3d4b3cfa
+  metadata.gz: 42b046952ea2817c3db80edcc69bafbf86822c9b88f23837b4decb9d9693b0044f68b9a74c920707e16ee28121f2c03d6992df300227b32cf62a3995e7af7ee5
+  data.tar.gz: aa9c3d27934eec97f11f579cc2b194b5b88f0d69da5c224cb4ebf6fd13878c8a493a7e3391f7f985674e602629da303364aa49c7ed9b88f0b101b66c304db6d2

data/CHANGELOG.md

@@ -1,3 +1,12 @@
+## 0.1.6
+
+* Fix `opaque` part parsing
+* Correctly join multiple paths
+
+## 0.1.5
+
+* Remove `ipaddr` dependency
+
 ## 0.1.4
 
 * Support passing `encoding` option

data/README.md

@@ -2,9 +2,7 @@
 
 Ruby implementation of the [WHATWG URL Living Standard](https://url.spec.whatwg.org/).
 
-The latest revision that this package implements of the standard is ([24 March 2025](https://url.spec.whatwg.org/commit-snapshots/cc8b776b89a6d92b5cc74581c8d90450d3c1e762/)).
-
-NOTE: Some features haven't been implemented yet. Please see the TODO for details.
+The latest revision that this package implements of the standard is ([18 August 2025](https://url.spec.whatwg.org/commit-snapshots/05a5d834deba31622390ee05a3dcbc22496b7bb5/)).
 
 ## Installation
 

data/lib/uri/whatwg_parser/host_parser.rb

@@ -1,7 +1,6 @@
 # frozen_string_literal: true
 
 require "uri/idna"
-require "ipaddr"
 require_relative "parser_helper"
 
 class URI::WhatwgParser
@@ -66,26 +65,128 @@ class URI::WhatwgParser
       output.join(".")
     end
 
-    def parse_ipv6(host)
-      addr = IPAddr.new(host)
-      # NOTE: URL Standard doesn't support `zone_id`.
-      raise ParseError, "invalid IPv6 format" unless addr.zone_id.nil?
-      formatted_addr = (("%.32x" % addr).gsub(/.{4}(?!$)/, '\&:'))
-      "[#{compress_ipv6(formatted_addr)}]"
-    rescue IPAddr::InvalidAddressError
-      raise ParseError, "invalid IPv6 format"
+    def parse_ipv6(input)
+      input = input[1..-2] if input.start_with?("[") && input.end_with?("]")
+      address = Array.new(8, 0)
+      piece_index = 0
+      compress = nil
+      chars = input.chars
+      i = 0
+
+      if chars[i] == ":"
+        raise ParseError, "invalid IPv6 format" unless chars[i + 1] == ":"
+        i += 2
+        compress = piece_index
+      end
+
+      while i < chars.length
+        raise ParseError, "invalid IPv6 format" if piece_index == 8
+
+        if chars[i] == ":"
+          raise ParseError, "invalid IPv6 format" if compress
+          i += 1
+          compress = piece_index
+          next
+        end
+
+        value = 0
+        length = 0
+        while length < 4 && i < chars.length && chars[i].match?(/[0-9A-Fa-f]/)
+          value = value * 16 + chars[i].to_i(16)
+          i += 1
+          length += 1
+        end
+
+        if chars[i] == "."
+          # IPv4-mapped address must be valid and complete, no trailing dot
+          ipv4_piece = chars[i - length, chars.length - (i - length)].join
+          parts = ipv4_piece.split(".")
+          if parts.length != 4 || parts.any? { |p| p.empty? } || ipv4_piece.end_with?(".")
+            raise ParseError, "invalid IPv6 format"
+          end
+
+          ipv4 = parse_ipv4(ipv4_piece)
+          address[piece_index] = (ipv4 >> 16) & 0xFFFF
+          address[piece_index + 1] = ipv4 & 0xFFFF
+          piece_index += 2
+          i = chars.length
+          break
+        end
+
+        raise ParseError, "invalid IPv6 format" if length == 0
+
+        address[piece_index] = value
+        piece_index += 1
+
+        if i < chars.length
+          if chars[i] == ":"
+            i += 1
+          elsif chars[i] != nil
+            raise ParseError, "invalid IPv6 format"
+          end
+        end
+      end
+
+      if compress
+        swaps = piece_index - compress
+        (0...swaps).each do |j|
+          address[7 - j] = address[compress + swaps - 1 - j]
+          address[compress + swaps - 1 - j] = 0
+        end
+      elsif piece_index != 8
+        raise ParseError, "invalid IPv6 format"
+      end
+
+      compress_ipv6(address)
     end
 
-    def compress_ipv6(ip)
-      output = ip.split(":").map { |term| term.sub(/\A0+/, "").empty? ? "0" : term.sub(/\A0+/, "") }.join(":")
-      zeros = output.scan(/:?(?:0+:?){2,}/)
+    def compress_ipv6(address)
+      # Find the longest run of zeros for '::' compression
+      best_base = nil
+      best_len = 0
+      base = nil
+      len = 0
+
+      8.times do |idx|
+        if address[idx] == 0
+          base = idx if base.nil?
+          len += 1
+        else
+          if len > best_len
+            best_base = base
+            best_len = len
+          end
+          base = nil
+          len = 0
+        end
+      end
+
+      if len > best_len
+        best_base = base
+        best_len = len
+      end
+
+      # Only compress if the run is at least two 0s
+      if best_len < 2
+        best_base = nil
+      end
 
-      unless zeros.empty?
-        max = zeros.max_by { |item| item.gsub(":", "").length }
-        output = output.sub(max, "::")
+      # Build the string with '::' for the longest zero run
+      result = []
+      idx = 0
+      while idx < 8
+        if best_base == idx
+          result << "" if idx == 0
+          result << ""
+          idx += best_len
+          result << "" if idx == 8
+          next
+        end
+        result << address[idx].to_s(16)
+        idx += 1
       end
 
-      output
+      "[#{result.join(":").gsub(/:{3,}/, "::")}]"
     end
 
     def parse_opaque_host(host)
@@ -126,7 +227,7 @@ class URI::WhatwgParser
       validation_error = false
       r = 10
 
-      if str.size >= 2 && (str.start_with?("0x") || str.start_with?("0X"))
+      if str.size >= 2 && str.start_with?("0x", "0X")
         validation_error  = true
         str = str[2..-1]
         r = 16

data/lib/uri/whatwg_parser/version.rb

@@ -2,6 +2,6 @@
 
 module URI
   class WhatwgParser
-    VERSION = "0.1.4"
+    VERSION = "0.1.6"
   end
 end

data/lib/uri/whatwg_parser.rb

@@ -126,22 +126,22 @@ module URI
           @state = :path_or_authority_state
           @pos += 1
         else
-          @parse_result[:path] = nil
+          @parse_result[:opaque] = ""
           @state = :opaque_path_state
         end
       else
         @buffer.clear
-        @pos -= 1
+        @pos = -1
         @state = :no_scheme_state
       end
     end
 
     def no_scheme_state(c)
-      raise ParseError, "scheme is missing" if @base.nil? || !@base[:opaque].nil? && c != "#"
+      raise ParseError, "scheme is missing" if @base.nil? || (!@base[:opaque].nil? && c != "#")
 
       if !@base[:opaque].nil? && c == "#"
         @parse_result[:scheme] = @base[:scheme]
-        @parse_result[:path] = @base[:path]
+        @paths = @base_paths
         @parse_result[:query] = @base[:query]
         @parse_result[:fragment] = nil
         @state = :fragment_state
@@ -180,10 +180,11 @@ module URI
       elsif special_url? && c == "\\"
         @state = :relative_slash_state
       else
+
         @parse_result[:userinfo] = @base[:userinfo]
         @parse_result[:host] = @base[:host]
         @parse_result[:port] = @base[:port]
-        @parse_result[:path] = @base[:path]
+        @paths = @base_paths
         @parse_result[:query] = @base[:query]
 
         if c == "?"
@@ -396,7 +397,6 @@ module URI
       @paths ||= []
 
       if (c.nil? || c == "/") || (special_url? && c == "\/") || (c == "?" || c == "#")
-
         if double_dot_path_segments?(@buffer)
           shorten_url_path
           if c != "/" || (special_url? && c == "\/")
@@ -408,6 +408,7 @@ module URI
           if @parse_result[:scheme] == "file" && @paths.empty? && windows_drive_letter?(@buffer)
             @buffer[1] = ":"
           end
+
           @paths << @buffer
         end
 
@@ -433,13 +434,13 @@ module URI
         @parse_result[:fragment] = nil
         @state = :fragment_state
       elsif c == " "
-        if rest.start_with?("?") || rest.start_with?("#")
-          @parse_result[:path] = @parse_result[:path].to_s + "%20"
+        if rest.start_with?("?", "#")
+          @parse_result[:opaque] += "%20"
         else
-          @parse_result[:path] = @parse_result[:path].to_s + " "
+          @parse_result[:opaque] += " "
         end
       elsif !c.nil?
-        @parse_result[:path] = @parse_result[:path].to_s + percent_encode(c, C0_CONTROL_PERCENT_ENCODE_SET, @encoding)
+        @parse_result[:opaque] += percent_encode(c, C0_CONTROL_PERCENT_ENCODE_SET, @encoding)
       end
     end
 
@@ -488,10 +489,10 @@ module URI
     end
 
     def shorten_url_path
-      return if @parse_result[:path]&.empty?
+      return if @paths.nil?
 
-      return true if @parse_result[:scheme] == "file" && @parse_result[:path]&.length == 1 && normalized_windows_drive_letter?(@parse_result[:path])
-      @parse_result[:path]&.chomp!
+      return true if @parse_result[:scheme] == "file" && @paths.length == 1 && normalized_windows_drive_letter?(@paths.first)
+      @paths.pop
     end
 
     def rest

metadata

@@ -1,13 +1,13 @@
 --- !ruby/object:Gem::Specification
 name: uri-whatwg_parser
 version: !ruby/object:Gem::Version
-  version: 0.1.4
+  version: 0.1.6
 platform: ruby
 authors:
 - Yuji Yaginuma
 bindir: exe
 cert_chain: []
-date: 2025-06-12 00:00:00.000000000 Z
+date: 1980-01-02 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: uri
@@ -86,7 +86,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.6.2
+rubygems_version: 3.6.7
 specification_version: 4
 summary: Ruby implementation of the WHATWG URL Living Standard
 test_files: []