Merge pull request #2222 from sparklemotion/nokogumbo-functional-inte…

…gration

get Nokogumbo's test suite green for basic cases

.github/workflows/windows.yml

@@ -0,0 +1,50 @@
+# this is a work in progress!
+name: windows
+on:
+  push:
+    branches:
+      - main
+  pull_request:
+    types: [opened, synchronize]
+    branches:
+      - '*'
+
+jobs:
+  windows:
+    name: "windows, sys: ${{ matrix.sys }}, ${{ matrix.ruby }}"
+
+    env:
+      MAKEFLAGS: -j2
+
+    runs-on: windows-latest
+
+    strategy:
+      fail-fast: false
+      matrix:
+        sys: [ enable, disable ]
+        ruby: [ "2.5", "2.6", "2.7", "3.0", "mingw" ]
+
+    steps:
+      - name: configure git crlf on windows
+        run: |
+          git config --system core.autocrlf false
+          git config --system core.eol lf
+      - name: checkout
+        uses: actions/checkout@v2
+      - name: load Ruby and bundle install
+        uses: MSP-Greg/setup-ruby-pkgs@v1
+        with:
+          ruby-version: ${{ matrix.ruby }}
+          mingw: libxml2 libxslt
+          bundler-cache: true
+      - uses: actions/cache@v2
+        if: matrix.sys == 'disable'
+        with:
+          path: ports/archives
+          key: ${{ matrix.os }}-${{ matrix.ruby }}-tarballs-${{ hashFiles('**/dependencies.yml') }}
+          restore-keys: ${{ matrix.os }}-${{ matrix.ruby }}-tarballs-
+      - name: bundle exec rake compile
+        run: |
+          bundle exec rake compile -- --${{ matrix.sys }}-system-libraries
+      - name: bundle exec rake test
+        run:  bundle exec rake test

.gitignore

@@ -9,6 +9,8 @@
 /gems/
 /lib/nokogiri/**/nokogiri.bundle
 /lib/nokogiri/**/nokogiri.so
+/lib/nokogumbo/**/nokogumbo.bundle
+/lib/nokogumbo/**/nokogumbo.so
 /lib/nokogiri/nokogiri.jar
 /pkg/
 /ports/

CHANGELOG.md

@@ -6,6 +6,11 @@ Nokogiri follows [Semantic Versioning](https://semver.org/), please see the [REA
 
 ## next / unreleased
 
+### Dependencies
+
+* [MRI] Upgrade mini_portile2 dependency from `~> 2.5.0` to `~> 2.5.1`.
+
+
 ### Changed
 
 * Introduce `Nokogiri::XML::ParseOptions::DEFAULT_XSLT` which adds the libxslt-preferred options of `NOENT | DTDLOAD | DTDATTR | NOCDATA` to `ParseOptions::DEFAULT_XML`.

ext/java/nokogiri/XmlDocument.java

@@ -443,7 +443,7 @@ private static class DocumentBuilderFactoryHolder
       return new_root;
     }
     if (!(new_root instanceof XmlNode)) {
-        throw context.runtime.newArgumentError("expected Nokogiri::XML::Node but received " + new_root.getType());
+      throw context.runtime.newArgumentError("expected Nokogiri::XML::Node but received " + new_root.getType());
     }
     XmlNode newRoot = asXmlNode(context, new_root);
 

ext/nokogiri/extconf.rb

@@ -14,7 +14,7 @@
 
 # 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 = "~> 2.5.0"
+REQUIRED_MINI_PORTILE_VERSION = "~> 2.5.1"
 REQUIRED_PKG_CONFIG_VERSION = "~> 1.1"
 
 # Keep track of what versions of what libraries we build against
@@ -402,7 +402,7 @@ def process_recipe(name, version, static_p, cross_p)
   require 'mini_portile2'
   message("Using mini_portile version #{MiniPortile::VERSION}\n")
 
-  if name != "libxml2" && name != "libxslt"
+  unless ["libxml2", "libxslt"].include?(name)
     OTHER_LIBRARY_VERSIONS[name] = version
   end
 
@@ -486,7 +486,7 @@ def process_recipe(name, version, static_p, cross_p)
         end
       end
 
-      message(<<~EOM)
+      message(<<~EOM) if name != "libgumbo"
 
         The Nokogiri maintainers intend to provide timely security updates, but if
         this is a concern for you and want to use your OS/distro system library
@@ -498,7 +498,7 @@ def process_recipe(name, version, static_p, cross_p)
       EOM
 
       message(<<~EOM) if name == 'libxml2'
-        Note, however, that nokogiri cannot guarantee compatiblity with every
+        Note, however, that nokogiri cannot guarantee compatibility with every
         version of libxml2 that may be provided by OS/package vendors.
 
       EOM
@@ -868,6 +868,56 @@ def compile
   ensure_func("exsltFuncRegister", "libexslt/exslt.h")
 end
 
+libgumbo_recipe = process_recipe("libgumbo", "1.0.0-nokogiri", static_p, cross_build_p) do |recipe|
+  recipe.configure_options = []
+
+  class << recipe
+    def downloaded?
+      true
+    end
+
+    def extract
+      target = File.join(tmp_path, "gumbo-parser")
+      output "Copying gumbo-parser files into #{target}..."
+      FileUtils.mkdir_p target
+      FileUtils.cp Dir.glob(File.join(PACKAGE_ROOT_DIR, "gumbo-parser/src/*")), target
+    end
+
+    def configured?
+      true
+    end
+
+    def install
+      lib_dir = File.join(port_path, "lib")
+      inc_dir = File.join(port_path, "include")
+      FileUtils.mkdir_p([lib_dir, inc_dir])
+      FileUtils.cp File.join(work_path, "libgumbo.a"), lib_dir
+      FileUtils.cp Dir.glob(File.join(work_path, "*.h")), inc_dir
+    end
+
+    def compile
+      cflags = concat_flags(ENV["CFLAGS"], "-fPIC", "-g")
+
+      env = {"CC" => gcc_cmd, "CFLAGS" => cflags}
+      if config_cross_build?
+        if host =~ /darwin/
+          env["AR"] = "#{host}-libtool"
+          env["ARFLAGS"] = "-o"
+        else
+          env["AR"] = "#{host}-ar"
+        end
+        env["RANLIB"] = "#{host}-ranlib"
+      end
+
+      execute("compile", make_cmd, {env: env})
+    end
+  end
+end
+append_cppflags("-I#{File.join(libgumbo_recipe.path, "include")}")
+$libs = $libs + " " + File.join(libgumbo_recipe.path, "lib", "libgumbo.a")
+$LIBPATH = $LIBPATH | [File.join(libgumbo_recipe.path, "lib")]
+ensure_func("gumbo_parse_with_options", "gumbo.h")
+
 have_func('xmlHasFeature') || abort("xmlHasFeature() is missing.") # introduced in libxml 2.6.21
 have_func('xmlFirstElementChild') # introduced in libxml 2.7.3
 have_func('xmlRelaxNGSetParserStructuredErrors') # introduced in libxml 2.6.24