nokogiri 1.12.5 → 1.13.9

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 23ffbd6c1e4af09a26b9e6c9f4ecd6afda89115f024bf80e6b908b31e4d03e71
-  data.tar.gz: 7dd0bd80a3a48a11e581677e93e4645629260d82b05311d8d62da8185a417e1f
+  metadata.gz: ae3b5f6cb32585ea5caf7fdc4119cdbfb9fb2c1ef338e732a1a7e199a24a7eea
+  data.tar.gz: f95707623e5201074c68d49fb5cbaed6f15d8fd6c73e6a0f5cd3995da1c366ab
 SHA512:
-  metadata.gz: e54c3269141d29fa515d2bac02d5c02e2a2d54feb0e336b4d94d4bc32618653812fe308bacb55983293886903baffb14d4eb70032ed6b9b8482bba4530aeb78f
-  data.tar.gz: 5d76ebeffb4c1180c03eaf946ee5a2d13d80cc942aba1cae0b04f11a359b5adc5508fe8cbeb58c89cf4cc1d6232f6c6afad38264102282a314693e83fd05426d
+  metadata.gz: 0e2c6b7c3f537da076a47da4e53fb9e2aed0658f9342f6e5de307938d525cd24942cef46ee7f500c144b50d1379ddae669586ff83552ae604f1131955566d572
+  data.tar.gz: ef3f5af7bc323c13f4319703363ed6100aa9077fab48e34b2c6e3456b8d4a6d1587a222ac025ee8c714c9262557c820b7df2d1f9f264148f742e7a182b64e1d4

data/Gemfile

@@ -1,3 +1,5 @@
+# frozen_string_literal: true
+
 source "https://rubygems.org"
 
 gemspec

data/README.md

@@ -2,7 +2,7 @@
 
 # Nokogiri
 
-Nokogiri (鋸) makes it easy and painless to work with XML and HTML from Ruby. It provides a sensible, easy-to-understand API for [reading](https://nokogiri.org/tutorials/parsing_an_html_xml_document.html), writing, [modifying](https://nokogiri.org/tutorials/modifying_an_html_xml_document.html), and [querying](https://nokogiri.org/tutorials/searching_a_xml_html_document.html) documents. It is fast and standards-compliant by relying on native parsers like libxml2 (C) and xerces (Java).
+Nokogiri (鋸) makes it easy and painless to work with XML and HTML from Ruby. It provides a sensible, easy-to-understand API for [reading](https://nokogiri.org/tutorials/parsing_an_html_xml_document.html), writing, [modifying](https://nokogiri.org/tutorials/modifying_an_html_xml_document.html), and [querying](https://nokogiri.org/tutorials/searching_a_xml_html_document.html) documents. It is fast and standards-compliant by relying on native parsers like libxml2 (CRuby) and xerces (JRuby).
 
 ## Guiding Principles
 
@@ -30,7 +30,9 @@ Some guiding principles Nokogiri tries to follow:
 [![Appveyor CI](https://ci.appveyor.com/api/projects/status/xj2pqwvlxwuwgr06/branch/main?svg=true)](https://ci.appveyor.com/project/flavorjones/nokogiri/branch/main)
 
 [![Gem Version](https://badge.fury.io/rb/nokogiri.svg)](https://rubygems.org/gems/nokogiri)
-[![SemVer compatibility](https://api.dependabot.com/badges/compatibility_score?dependency-name=nokogiri&package-manager=bundler&version-scheme=semver)](https://dependabot.com/compatibility-score/?dependency-name=nokogiri&package-manager=bundler)
+[![SemVer compatibility](https://dependabot-badges.githubapp.com/badges/compatibility_score?dependency-name=nokogiri&package-manager=bundler&previous-version=1.11.7&new-version=1.12.5)](https://docs.github.com/en/code-security/supply-chain-security/managing-vulnerabilities-in-your-projects-dependencies/about-dependabot-security-updates#about-compatibility-scores)
+
+[![CII Best Practices](https://bestpractices.coreinfrastructure.org/projects/5344/badge)](https://bestpractices.coreinfrastructure.org/projects/5344)
 [![Tidelift dependencies](https://tidelift.com/badges/package/rubygems/nokogiri)](https://tidelift.com/subscription/pkg/rubygems-nokogiri?utm_source=rubygems-nokogiri&utm_medium=referral&utm_campaign=readme)
 
 
@@ -78,7 +80,7 @@ Full information and description of our security policy is in [`SECURITY.md`](SE
 
 ### Semantic Versioning Policy
 
-Nokogiri follows [Semantic Versioning](https://semver.org/) (since 2017 or so). [![Dependabot's SemVer compatibility score for Nokogiri](https://api.dependabot.com/badges/compatibility_score?dependency-name=nokogiri&package-manager=bundler&version-scheme=semver)](https://dependabot.com/compatibility-score/?dependency-name=nokogiri&package-manager=bundler)
+Nokogiri follows [Semantic Versioning](https://semver.org/) (since 2017 or so). [![Dependabot's SemVer compatibility score for Nokogiri](https://dependabot-badges.githubapp.com/badges/compatibility_score?dependency-name=nokogiri&package-manager=bundler&previous-version=1.11.7&new-version=1.12.5)](https://docs.github.com/en/code-security/supply-chain-security/managing-vulnerabilities-in-your-projects-dependencies/about-dependabot-security-updates#about-compatibility-scores)
 
 We bump `Major.Minor.Patch` versions following this guidance:
 
@@ -105,8 +107,8 @@ We bump `Major.Minor.Patch` versions following this guidance:
 
 Requirements:
 
-- Ruby >= 2.5
-- JRuby >= 9.2.0.0
+- Ruby >= 2.6
+- JRuby >= 9.3.0.0
 
 
 ### Native Gems: Faster, more reliable installation
@@ -115,12 +117,12 @@ Requirements:
 
 ### Supported Platforms
 
-As of v1.11.0, Nokogiri ships pre-compiled, "native" gems for the following platforms:
+Nokogiri ships pre-compiled, "native" gems for the following platforms:
 
 - Linux: `x86-linux` and `x86_64-linux` (req: `glibc >= 2.17`), including musl platforms like Alpine
 - Darwin/MacOS: `x86_64-darwin` and `arm64-darwin`
 - Windows: `x86-mingw32` and `x64-mingw32`
-- Java: any platform running JRuby 9.2 or higher
+- Java: any platform running JRuby 9.3 or higher
 
 To determine whether your system supports one of these gems, look at the output of `bundle platform` or `ruby -e 'puts Gem::Platform.local.to_s'`.
 

data/bin/nokogiri

@@ -1,61 +1,77 @@
 #!/usr/bin/env ruby
-require 'optparse'
-require 'open-uri'
-require 'uri'
-require 'rubygems'
-require 'nokogiri'
-autoload :IRB, 'irb'
+# frozen_string_literal: true
+
+require "optparse"
+require "open-uri"
+require "uri"
+require "rubygems"
+require "nokogiri"
+autoload :IRB, "irb"
 
 parse_class = Nokogiri
 encoding = nil
 
 # This module provides some tunables with the nokogiri CLI for use in
 # your ~/.nokogirirc.
-module Nokogiri::CLI
-  class << self
-    # Specify the console engine, defaulted to IRB.
-    #
-    # call-seq:
-    #   require 'pry'
-    #   Nokogiri::CLI.console = Pry
-    attr_writer :console
-
-    def console
-      case @console
-      when Symbol
-        Kernel.const_get(@console)
-      else
-        @console
+module Nokogiri
+  module CLI
+    class << self
+      # Specify the console engine, defaulted to IRB.
+      #
+      # call-seq:
+      #   require 'pry'
+      #   Nokogiri::CLI.console = Pry
+      attr_writer :console
+
+      def console
+        case @console
+        when Symbol
+          Kernel.const_get(@console)
+        else
+          @console
+        end
       end
+
+      attr_accessor :rcfile
     end
 
-    attr_accessor :rcfile
+    self.rcfile  = File.expand_path("~/.nokogirirc")
+    self.console = :IRB
   end
+end
 
-  self.rcfile  = File.expand_path('~/.nokogirirc')
-  self.console = :IRB
+def safe_read(uri_or_path)
+  uri = URI.parse(uri_or_path)
+  case uri
+  when URI::HTTP
+    uri.read
+  when URI::File
+    File.read(uri.path)
+  else
+    File.read(uri_or_path)
+  end
 end
 
 opts = OptionParser.new do |opts|
   opts.banner = "Nokogiri: an HTML, XML, SAX, and Reader parser"
-  opts.define_head "Usage: nokogiri <uri|path> [options]"
-  opts.separator ""
-  opts.separator "Examples:"
-  opts.separator "  nokogiri https://www.ruby-lang.org/"
-  opts.separator "  nokogiri ./public/index.html"
-  opts.separator "  curl -s http://www.nokogiri.org | nokogiri -e'p $_.css(\"h1\").length'"
-  opts.separator ""
-  opts.separator "Options:"
+  opts.define_head("Usage: nokogiri <uri|path> [options]")
+  opts.separator("")
+  opts.separator("Examples:")
+  opts.separator("  nokogiri https://www.ruby-lang.org/")
+  opts.separator("  nokogiri ./public/index.html")
+  opts.separator("  curl -s http://www.nokogiri.org | nokogiri -e'p $_.css(\"h1\").length'")
+  opts.separator("")
+  opts.separator("Options:")
 
   opts.on("--type type", "Parse as type: xml or html (default: auto)", [:xml, :html]) do |v|
-    parse_class = {:xml => Nokogiri::XML, :html => Nokogiri::HTML}[v]
+    parse_class = { xml: Nokogiri::XML, html: Nokogiri::HTML }[v]
   end
 
   opts.on("-C file", "Specifies initialization file to load (default #{Nokogiri::CLI.rcfile})") do |v|
     Nokogiri::CLI.rcfile = v
   end
 
-  opts.on("-E", "--encoding encoding", "Read as encoding (default: #{encoding || 'none'})") do |v|
+  opts.on("-E", "--encoding encoding", "Read as encoding (default: #{encoding || "none"})") do |v|
     encoding = v
   end
 
@@ -64,7 +80,7 @@ opts = OptionParser.new do |opts|
   end
 
   opts.on("--rng <uri|path>", "Validate using this rng file.") do |v|
-    @rng = open(v) {|f| Nokogiri::XML::RelaxNG(f)}
+    @rng = Nokogiri::XML::RelaxNG(safe_read(v))
   end
 
   opts.on_tail("-?", "--help", "Show this message") do
@@ -90,15 +106,10 @@ if File.file?(Nokogiri::CLI.rcfile)
   load Nokogiri::CLI.rcfile
 end
 
-if url || $stdin.tty?
-  case uri = (URI(url) rescue url)
-  when URI::HTTP
-    @doc = parse_class.parse(uri.read, url, encoding)
-  else
-    @doc = parse_class.parse(open(url).read, nil, encoding)
-  end
+@doc = if url || $stdin.tty?
+  parse_class.parse(safe_read(url), url, encoding)
 else
-  @doc = parse_class.parse($stdin, nil, encoding)
+  parse_class.parse($stdin, nil, encoding)
 end
 
 $_ = @doc
@@ -107,12 +118,14 @@ if @rng
   @rng.validate(@doc).each do |error|
     puts error.message
   end
-else
-  if @script
-    eval @script, binding, '<main>'
-  else
-    puts "Your document is stored in @doc..."
-    Nokogiri::CLI.console.start
+elsif @script
+  begin
+    eval(@script, binding, "<main>") # rubocop:disable Security/Eval
+  rescue Exception => e # rubocop:disable Lint/RescueException
+    warn("ERROR: Exception raised while evaluating '#{@script}'")
+    raise e
   end
+else
+  puts "Your document is stored in @doc..."
+  Nokogiri::CLI.console.start
 end
-

data/dependencies.yml

@@ -1,74 +1,23 @@
 libxml2:
-  version: "2.9.12"
-  sha256: "c8d6681e38c56f172892c85ddc0852e1fd4b53b4209e7f4ebf17f7e2eae71d92"
-  #  manually verified checksum:
-  #
-  #    $ gpg --verify libxml2-2.9.12.tar.gz.asc ports/archives/libxml2-2.9.12.tar.gz
-  #    gpg: Signature made Thu 13 May 2021 02:59:16 PM EDT
-  #    gpg:                using RSA key DB46681BB91ADCEA170FA2D415588B26596BEA5D
-  #    gpg: Good signature from "Daniel Veillard (Red Hat work email) <veillard@redhat.com>" [unknown]
-  #    gpg:                 aka "Daniel Veillard <Daniel.Veillard@w3.org>" [unknown]
-  #    gpg: WARNING: This key is not certified with a trusted signature!
-  #    gpg:          There is no indication that the signature belongs to the owner.
-  #    Primary key fingerprint: C744 15BA 7C9C 7F78 F02E  1DC3 4606 B8A5 DE95 BC1F
-  #         Subkey fingerprint: DB46 681B B91A DCEA 170F  A2D4 1558 8B26 596B EA5D
-  #
-  #  using this pgp signature:
-  #
-  #    -----BEGIN PGP SIGNATURE-----
-  #    
-  #    iQEzBAABCAAdFiEE20ZoG7ka3OoXD6LUFViLJllr6l0FAmCddwQACgkQFViLJllr
-  #    6l11LQgAioRTdfmcC+uK/7+6HPtF/3c5zkX6j8VGYuvFBwZ0jayqMRBAl++fcpjE
-  #    JUU/JKebSZ/KCYjzyeOWK/i3Gq77iqm3UbZFB85rqu4a5P3gmj/4STWVyAx0KU3z
-  #    G3jKqDhJOt7c0acXb5lh2DngfDa1dn/VGcQcIXsqplNxNr4ET7MnSJjZ3nlxYfW2
-  #    E5vWBdPCMUeXDBl6MjYvw9XnGGBLUAaEJWoFToG6jKmVf4GAd9nza20jj5dtbcJq
-  #    QEOaSDKDr+f9h2NS8haOhJ9vOpy52PdeGzaFlbRkXarGXuAr8kITgATVs8FAqcgv
-  #    MoVhmrO5r2hJf0dCM9fZoYqzpMfmNA==
-  #    =KfJ9
-  #    -----END PGP SIGNATURE-----
-  #
+  version: "2.10.3"
+  sha256: "5d2cc3d78bec3dbe212a9d7fa629ada25a7da928af432c93060ff5c17ee28a9c"
+  # sha-256 hash provided in https://download.gnome.org/sources/libxml2/2.10/libxml2-2.10.3.sha256sum
 
 libxslt:
-  version: "1.1.34"
-  sha256: "98b1bd46d6792925ad2dfe9a87452ea2adebf69dcb9919ffd55bf926a7f93f7f"
-  #  manually verified checksum:
-  #
-  #    $ gpg --verify ~/Downloads/libxslt-1.1.34.tar.gz.asc ports/archives/libxslt-1.1.34.tar.gz
-  #    gpg: Signature made Wed 30 Oct 2019 04:02:48 PM EDT
-  #    gpg:                using RSA key DB46681BB91ADCEA170FA2D415588B26596BEA5D
-  #    gpg: Good signature from "Daniel Veillard (Red Hat work email) <veillard@redhat.com>" [unknown]
-  #    gpg:                 aka "Daniel Veillard <Daniel.Veillard@w3.org>" [unknown]
-  #    gpg: WARNING: This key is not certified with a trusted signature!
-  #    gpg:          There is no indication that the signature belongs to the owner.
-  #    Primary key fingerprint: C744 15BA 7C9C 7F78 F02E  1DC3 4606 B8A5 DE95 BC1F
-  #         Subkey fingerprint: DB46 681B B91A DCEA 170F  A2D4 1558 8B26 596B EA5D
-  #
-  #  using this pgp signature:
-  #
-  #    -----BEGIN PGP SIGNATURE-----
-  #    
-  #    iQEzBAABCAAdFiEE20ZoG7ka3OoXD6LUFViLJllr6l0FAl257GgACgkQFViLJllr
-  #    6l2vVggAjJEHmASiS56SxhPOsGqbfBihM66gQFoIymQfMu2430N1GSTkLsfbkJO8
-  #    8yBX11NjzK/m9uxwshMW3rVCU7EpL3PUimN3reXdPiQj9hAOAWF1V3BZNevbQC2E
-  #    FCIraioukaidf8sjUG4/sGpK/gOcP/3hYoN0HUoBigCNJjDqhijxM3M3GJJtCASp
-  #    jL4CQbs2OmxW8ixOZbuWEESvFFHUgYRsdZjRVN+GRfSOvJjxypurmYwQ3RjO7JxL
-  #    2FY8qKQ+xpeID8NV8F5OUEvWBjk1QS133VTqBZNlONdnEtV/og6jNu5k0O/Kvhup
-  #    caR+8TMErOcLr9OgDklO6DoYyAsf9Q==
-  #    =g4i4
-  #    -----END PGP SIGNATURE-----
-  #
+  version: "1.1.37"
+  sha256: "3a4b27dc8027ccd6146725950336f1ec520928f320f144eb5fa7990ae6123ab4"
+  # sha-256 hash provided in https://download.gnome.org/sources/libxslt/1.1/libxslt-1.1.37.sha256sum
 
 zlib:
-  version: "1.2.11"
-  sha256: "c3e5e9fdd5004dcb542feda5ee4f0ff0744628baf8ed2dd5d66f8ca1197cb1a1"
+  version: "1.2.13"
+  sha256: "b3a24de97a8fdbc835b9833169501030b8977031bcb54b3b3ac13740f846ab30"
   # SHA-256 hash provided on http://zlib.net/
 
 libiconv:
-  version: "1.15"
-  sha256: "ccf536620a45458d26ba83887a983b96827001e92a13847b45e4925cc8913178"
-  # gpg: Signature made Fri Feb  3 00:38:12 2017 CET
+  version: "1.16"
+  sha256: "e6a1b1b589654277ee790cce3734f07876ac4ccfaecbee8afa0b649cf529cc04"
+  # gpg: Signature made Fri 26 Apr 2019 03:36:38 PM EDT
   # gpg:                using RSA key 4F494A942E4616C2
-  # gpg: Good signature from "Bruno Haible (Open Source Development) <bruno@clisp.org>" [unknown]
-  # gpg: WARNING: This key is not certified with a trusted signature!
-  # gpg:          There is no indication that the signature belongs to the owner.
+  # gpg: Good signature from "Bruno Haible (Open Source Development) <bruno@clisp.org>" [expired]
+  # gpg: Note: This key has expired!
   # Primary key fingerprint: 68D9 4D8A AEEA D48A E7DC  5B90 4F49 4A94 2E46 16C2

data/ext/nokogiri/extconf.rb

@@ -1,5 +1,8 @@
 # frozen_string_literal: true
-ENV["RC_ARCHS"] = "" if RUBY_PLATFORM =~ /darwin/
+
+# rubocop:disable Style/GlobalVars
+
+ENV["RC_ARCHS"] = "" if RUBY_PLATFORM.include?("darwin")
 
 require "mkmf"
 require "rbconfig"
@@ -12,16 +15,14 @@ PACKAGE_ROOT_DIR = File.expand_path(File.join(File.dirname(__FILE__), "..", ".."
 REQUIRED_LIBXML_VERSION = "2.6.21"
 RECOMMENDED_LIBXML_VERSION = "2.9.3"
 
-# The gem version constraint in the Rakefile is not respected at install time.
-# Keep this version in sync with the one in the Rakefile !
-REQUIRED_MINI_PORTILE_VERSION = "~> 2.6.1"
+REQUIRED_MINI_PORTILE_VERSION = "~> 2.8.0" # keep this version in sync with the one in the gemspec
 REQUIRED_PKG_CONFIG_VERSION = "~> 1.1"
 
 # Keep track of what versions of what libraries we build against
 OTHER_LIBRARY_VERSIONS = {}
 
 NOKOGIRI_HELP_MESSAGE = <<~HELP
-  USAGE: ruby #{$0} [options]
+  USAGE: ruby #{$PROGRAM_NAME} [options]
 
     Flags that are always valid:
 
@@ -175,23 +176,23 @@ def config_system_libraries?
 end
 
 def windows?
-  RbConfig::CONFIG["target_os"] =~ /mingw32|mswin/
+  RbConfig::CONFIG["target_os"].match?(/mingw|mswin/)
 end
 
 def solaris?
-  RbConfig::CONFIG["target_os"] =~ /solaris/
+  RbConfig::CONFIG["target_os"].include?("solaris")
 end
 
 def darwin?
-  RbConfig::CONFIG["target_os"] =~ /darwin/
+  RbConfig::CONFIG["target_os"].include?("darwin")
 end
 
 def openbsd?
-  RbConfig::CONFIG["target_os"] =~ /openbsd/
+  RbConfig::CONFIG["target_os"].include?("openbsd")
 end
 
 def aix?
-  RbConfig::CONFIG["target_os"] =~ /aix/
+  RbConfig::CONFIG["target_os"].include?("aix")
 end
 
 def nix?
@@ -210,6 +211,18 @@ def local_have_library(lib, func = nil, headers = nil)
   have_library(lib, func, headers) || have_library("lib#{lib}", func, headers)
 end
 
+def gnome_source
+  # As of 2022-02-20, some mirrors have expired SSL certificates. I'm able to retrieve from my home,
+  # but whatever host is resolved on the github actions workers see an expired cert.
+  #
+  # See https://github.com/sparklemotion/nokogiri/runs/5266206403?check_suite_focus=true
+  if ENV["NOKOGIRI_USE_CANONICAL_GNOME_SOURCE"]
+    "https://download.gnome.org"
+  else
+    "https://mirror.csclub.uwaterloo.ca/gnome" # old reliable
+  end
+end
+
 LOCAL_PACKAGE_RESPONSE = Object.new
 def LOCAL_PACKAGE_RESPONSE.%(package)
   package ? "yes: #{package}" : "no"
@@ -284,17 +297,16 @@ ensure
 end
 
 def abort_could_not_find_library(lib)
-  abort("-----\n#{caller[0]}\n#{lib} is missing. Please locate mkmf.log to investigate how it is failing.\n-----")
+  callers = caller(1..2).join("\n")
+  abort("-----\n#{callers}\n#{lib} is missing. Please locate mkmf.log to investigate how it is failing.\n-----")
 end
 
-def chdir_for_build
+def chdir_for_build(&block)
   # When using rake-compiler-dock on Windows, the underlying Virtualbox shared
   # folders don't support symlinks, but libiconv expects it for a build on
   # Linux. We work around this limitation by using the temp dir for cooking.
-  build_dir = ENV["RCD_HOST_RUBY_PLATFORM"].to_s =~ /mingw|mswin|cygwin/ ? "/tmp" : "."
-  Dir.chdir(build_dir) do
-    yield
-  end
+  build_dir = /mingw|mswin|cygwin/.match?(ENV["RCD_HOST_RUBY_PLATFORM"].to_s) ? "/tmp" : "."
+  Dir.chdir(build_dir, &block)
 end
 
 def sh_export_path(path)
@@ -370,6 +382,7 @@ def iconv_configure_flags
   ["iconv", "opt"].each do |target|
     config = preserving_globals { dir_config(target) }
     next unless config.any? && try_link_iconv("--with-#{target}-* flags") { dir_config(target) }
+
     idirs, ldirs = config.map do |dirs|
       Array(dirs).flat_map do |dir|
         dir.split(File::PATH_SEPARATOR)
@@ -402,9 +415,9 @@ def iconv_configure_flags
   abort_could_not_find_library("libiconv")
 end
 
-def process_recipe(name, version, static_p, cross_p, cacheable_p=true)
+def process_recipe(name, version, static_p, cross_p, cacheable_p = true)
   require "rubygems"
-  gem("mini_portile2", REQUIRED_MINI_PORTILE_VERSION)
+  gem("mini_portile2", REQUIRED_MINI_PORTILE_VERSION) # gemspec is not respected at install time
   require "mini_portile2"
   message("Using mini_portile version #{MiniPortile::VERSION}\n")
 
@@ -413,9 +426,13 @@ def process_recipe(name, version, static_p, cross_p, cacheable_p=true)
   end
 
   MiniPortile.new(name, version).tap do |recipe|
+    def recipe.port_path
+      "#{@target}/#{RUBY_PLATFORM}/#{@name}/#{@version}"
+    end
+
     recipe.target = File.join(PACKAGE_ROOT_DIR, "ports") if cacheable_p
-    # Prefer host_alias over host in order to use i586-mingw32msvc as
-    # correct compiler prefix for cross build, but use host if not set.
+    # Prefer host_alias over host in order to use the correct compiler prefix for cross build, but
+    # use host if not set.
     recipe.host = RbConfig::CONFIG["host_alias"].empty? ? RbConfig::CONFIG["host"] : RbConfig::CONFIG["host_alias"]
     recipe.configure_options << "--libdir=#{File.join(recipe.path, "lib")}"
 
@@ -462,7 +479,7 @@ def process_recipe(name, version, static_p, cross_p, cacheable_p=true)
     end
 
     if RbConfig::CONFIG["target_cpu"] == "universal"
-      %w[CFLAGS LDFLAGS].each do |key|
+      ["CFLAGS", "LDFLAGS"].each do |key|
         unless env[key].include?("-arch")
           env[key] = concat_flags(env[key], RbConfig::CONFIG["ARCH_FLAG"])
         end
@@ -473,7 +490,7 @@ def process_recipe(name, version, static_p, cross_p, cacheable_p=true)
       "#{key}=#{value.strip}"
     end
 
-    checkpoint = "#{recipe.target}/#{recipe.name}-#{recipe.version}-#{recipe.host}.installed"
+    checkpoint = "#{recipe.target}/#{recipe.name}-#{recipe.version}-#{RUBY_PLATFORM}.installed"
     if File.exist?(checkpoint) && !recipe.source_directory
       message("Building Nokogiri with a packaged version of #{name}-#{version}.\n")
     else
@@ -487,7 +504,7 @@ def process_recipe(name, version, static_p, cross_p, cacheable_p=true)
         message("The following patches are being applied:\n")
 
         recipe.patch_files.each do |patch|
-          message("  - %s\n" % File.basename(patch))
+          message(format("  - %s\n", File.basename(patch)))
         end
       end
 
@@ -508,6 +525,7 @@ def process_recipe(name, version, static_p, cross_p, cacheable_p=true)
 
       EOM
 
+      pp(recipe.files)
       chdir_for_build { recipe.cook }
       FileUtils.touch(checkpoint)
     end
@@ -656,7 +674,7 @@ else
   if cross_build_p || windows?
     zlib_recipe = process_recipe("zlib", dependencies["zlib"]["version"], static_p, cross_build_p) do |recipe|
       recipe.files = [{
-        url: "http://zlib.net/fossils/#{recipe.name}-#{recipe.version}.tar.gz",
+        url: "https://zlib.net/fossils/#{recipe.name}-#{recipe.version}.tar.gz",
         sha256: dependencies["zlib"]["sha256"],
       }]
       if windows?
@@ -694,16 +712,16 @@ else
       else
         class << recipe
           def configure
-            cflags = concat_flags(ENV["CFLAGS"], "-fPIC", "-g")
-            execute("configure",
-              ["env", "CHOST=#{host}", "CFLAGS=#{cflags}", "./configure", "--static", configure_prefix])
-          end
-
-          def compile
-            if host =~ /darwin/
-              execute("compile", "make AR=#{host}-libtool")
-            else
-              super
+            env = {}
+            env["CFLAGS"] = concat_flags(ENV["CFLAGS"], "-fPIC", "-g")
+            env["CHOST"] = host
+            execute("configure", ["./configure", "--static", configure_prefix], { env: env })
+            if darwin?
+              # needed as of zlib 1.2.13
+              Dir.chdir(work_path) do
+                makefile = File.read("Makefile").gsub(/^AR=.*$/, "AR=#{host}-libtool")
+                File.open("Makefile", "w") { |m| m.write(makefile) }
+              end
             end
           end
         end
@@ -714,10 +732,13 @@ else
       libiconv_recipe = process_recipe("libiconv", dependencies["libiconv"]["version"], static_p,
         cross_build_p) do |recipe|
         recipe.files = [{
-          url: "http://ftp.gnu.org/pub/gnu/libiconv/#{recipe.name}-#{recipe.version}.tar.gz",
+          url: "https://ftp.gnu.org/pub/gnu/libiconv/#{recipe.name}-#{recipe.version}.tar.gz",
           sha256: dependencies["libiconv"]["sha256"],
         }]
 
+        # The libiconv configure script doesn't accept "arm64" host string but "aarch64"
+        recipe.host = recipe.host.gsub("arm64-apple-darwin", "aarch64-apple-darwin")
+
         cflags = concat_flags(ENV["CFLAGS"], "-O2", "-U_FORTIFY_SOURCE", "-g")
 
         recipe.configure_options += [
@@ -741,12 +762,21 @@ else
       Tools" to open the developer site, download the installer for your OS
       version and run it.
       -----
-      EOM
+    EOM
   end
 
-  unless windows?
-    preserving_globals { local_have_library("z", "gzdopen", "zlib.h") } ||
-      abort("zlib is missing; necessary for building libxml2")
+  if zlib_recipe
+    append_cppflags("-I#{zlib_recipe.path}/include")
+    $LIBPATH = ["#{zlib_recipe.path}/lib"] | $LIBPATH
+    ensure_package_configuration(opt: "zlib", pc: "zlib", lib: "z",
+      headers: "zlib.h", func: "gzdopen")
+  end
+
+  if libiconv_recipe
+    append_cppflags("-I#{libiconv_recipe.path}/include")
+    $LIBPATH = ["#{libiconv_recipe.path}/lib"] | $LIBPATH
+    ensure_package_configuration(opt: "iconv", pc: "iconv", lib: "iconv",
+      headers: "iconv.h", func: "iconv_open")
   end
 
   libxml2_recipe = process_recipe("libxml2", dependencies["libxml2"]["version"], static_p, cross_build_p) do |recipe|
@@ -754,8 +784,9 @@ else
     if source_dir
       recipe.source_directory = source_dir
     else
+      minor_version = Gem::Version.new(recipe.version).segments.take(2).join(".")
       recipe.files = [{
-        url: "http://xmlsoft.org/sources/#{recipe.name}-#{recipe.version}.tar.gz",
+        url: "#{gnome_source}/sources/libxml2/#{minor_version}/#{recipe.name}-#{recipe.version}.tar.xz",
         sha256: dependencies["libxml2"]["sha256"],
       }]
       recipe.patch_files = Dir[File.join(PACKAGE_ROOT_DIR, "patches", "libxml2", "*.patch")].sort
@@ -765,7 +796,6 @@ else
 
     if zlib_recipe
       recipe.configure_options << "--with-zlib=#{zlib_recipe.path}"
-      cflags = concat_flags(cflags, "-I#{zlib_recipe.path}/include")
     end
 
     if libiconv_recipe
@@ -803,8 +833,9 @@ else
     if source_dir
       recipe.source_directory = source_dir
     else
+      minor_version = Gem::Version.new(recipe.version).segments.take(2).join(".")
       recipe.files = [{
-        url: "http://xmlsoft.org/sources/#{recipe.name}-#{recipe.version}.tar.gz",
+        url: "#{gnome_source}/sources/libxslt/#{minor_version}/#{recipe.name}-#{recipe.version}.tar.xz",
         sha256: dependencies["libxslt"]["sha256"],
       }]
       recipe.patch_files = Dir[File.join(PACKAGE_ROOT_DIR, "patches", "libxslt", "*.patch")].sort
@@ -816,6 +847,11 @@ else
       recipe.configure_options += ["RANLIB=/usr/bin/ranlib", "AR=/usr/bin/ar"]
     end
 
+    if windows?
+      cflags = concat_flags(cflags, "-ULIBXSLT_STATIC", "-DIN_LIBXSLT")
+      cflags = concat_flags(cflags, "-ULIBEXSLT_STATIC", "-DIN_LIBEXSLT")
+    end
+
     recipe.configure_options << if source_dir
       "--config-cache"
     else
@@ -834,9 +870,6 @@ else
   append_cppflags("-DNOKOGIRI_PACKAGED_LIBRARIES")
   append_cppflags("-DNOKOGIRI_PRECOMPILED_LIBRARIES") if cross_build_p
 
-  $LIBPATH = ["#{zlib_recipe.path}/lib"] | $LIBPATH if zlib_recipe
-  $LIBPATH = ["#{libiconv_recipe.path}/lib"] | $LIBPATH if libiconv_recipe
-
   $libs = $libs.shellsplit.tap do |libs|
     [libxml2_recipe, libxslt_recipe].each do |recipe|
       libname = recipe.name[/\Alib(.+)\z/, 1]
@@ -929,7 +962,7 @@ libgumbo_recipe = process_recipe("libgumbo", "1.0.0-nokogiri", static_p, cross_b
 
       env = { "CC" => gcc_cmd, "CFLAGS" => cflags }
       if config_cross_build?
-        if host =~ /darwin/
+        if /darwin/.match?(host)
           env["AR"] = "#{host}-libtool"
           env["ARFLAGS"] = "-o"
         else
@@ -953,6 +986,8 @@ have_func("xmlRelaxNGSetParserStructuredErrors") # introduced in libxml 2.6.24
 have_func("xmlRelaxNGSetValidStructuredErrors") # introduced in libxml 2.6.21
 have_func("xmlSchemaSetValidStructuredErrors") # introduced in libxml 2.6.23
 have_func("xmlSchemaSetParserStructuredErrors") # introduced in libxml 2.6.23
+have_func("rb_gc_location") # introduced in Ruby 2.7
+have_func("rb_category_warning") # introduced in Ruby 3.0
 
 have_func("vasprintf")
 

data/ext/nokogiri/gumbo.c

@@ -401,7 +401,7 @@ static xmlNodePtr
 extract_xml_node(VALUE node)
 {
   xmlNodePtr xml_node;
-  Data_Get_Struct(node, xmlNode, xml_node);
+  Noko_Node_Get_Struct(node, xmlNode, xml_node);
   return xml_node;
 }
 

data/ext/nokogiri/html4_sax_parser_context.c

@@ -19,9 +19,8 @@ parse_memory(VALUE klass, VALUE data, VALUE encoding)
 {
   htmlParserCtxtPtr ctxt;
 
-  if (NIL_P(data)) {
-    rb_raise(rb_eArgError, "data cannot be nil");
-  }
+  Check_Type(data, T_STRING);
+
   if (!(int)RSTRING_LEN(data)) {
     rb_raise(rb_eRuntimeError, "data cannot be empty");
   }