nokogiri 1.4.7-x86-mingw32 → 1.5.0-x86-mingw32

data/lib/nokogiri/xml/sax/document.rb

@@ -8,7 +8,7 @@ module Nokogiri
     # The basic way a SAX style parser works is by creating a parser,
     # telling the parser about the events we're interested in, then giving
     # the parser some XML to process.  The parser will notify you when
-    # it encounters events you said you would like to know about.
+    # it encounters events your said you would like to know about.
     #
     # To register for events, you simply subclass Nokogiri::XML::SAX::Document,
     # and implement the methods for which you would like notification.
@@ -36,7 +36,7 @@ module Nokogiri
     #   parser.parse(File.read(ARGV[0], 'rb'))
     #
     # Now my document handler will be called when each node starts, and when
-    # the document ends.  To see what kinds of events are available, take
+    # then document ends.  To see what kinds of events are available, take
     # a look at Nokogiri::XML::SAX::Document.
     #
     # Two SAX parsers for XML are available, a parser that reads from a string

data/lib/nokogiri/xslt/stylesheet.rb

@@ -6,7 +6,7 @@ module Nokogiri
     # an XML::Document with a Stylesheet:
     #
     #   doc   = Nokogiri::XML(File.read('some_file.xml'))
-    #   xslt  = Nokogir::XSLT(File.read('some_transformer.xslt'))
+    #   xslt  = Nokogiri::XSLT(File.read('some_transformer.xslt'))
     #
     #   puts xslt.transform(doc)
     #

data/nokogiri_help_responses.md

@@ -0,0 +1,40 @@
+# not-enough-information
+
+Hello!
+
+Thanks for asking this question! However, without more information,
+Team Nokogiri cannot reproduce your issue, and so we cannot offer much
+help.
+
+Please provide us with:
+
+* A self-contained script (one that we can run without modification,
+  and preferably without making external network connections).
+
+* Please note that you need to include the XML/HTML that you are
+  operating on.
+
+* The output of `nokogiri -v`, which will provide details about your
+  platform and versions of ruby, libxml2 and nokogiri.
+
+For more information about requesting help or reporting bugs, please
+take a look at http://bit.ly/nokohelp
+
+Thank you so much!
+
+# not-a-bug
+
+Hello!
+
+Thanks for asking this question! Your request for help will not go
+unanswered!
+
+However, Nokogiri's Github Issues is reserved for reporting bugs or
+submitting patches. If you ask your question on the mailing list, Team
+Nokogiri promises someone will provide you with an answer in a timely
+manner.
+
+If you'd like to read up on Team Nokogiri's rationale for this policy,
+please go to http://bit.ly/nokohelp.
+
+Thank you so much for understanding! And thank you for using Nokogiri.

data/tasks/cross_compile.rb

@@ -1,173 +1,152 @@
 require 'rake/extensioncompiler'
 HOST = Rake::ExtensionCompiler.mingw_host
-TARGET = 'i386-pc-mingw32'
 
-ZLIB    = 'zlib-1.2.5'
-ICONV   = 'libiconv-1.13.1'
-LIBXML  = 'libxml2-2.7.7'
-LIBXSLT = 'libxslt-1.1.26'
-RAKE_COMPILER_PKGCONFIG = File.expand_path(File.join(Dir.pwd, "tmp/cross/lib/pkgconfig/"))
+require 'mini_portile'
+$recipes = {}
+$recipes[:zlib]     = MiniPortile.new "zlib", "1.2.5"
+$recipes[:libiconv] = MiniPortile.new "libiconv", "1.13.1"
+$recipes[:libxml2]  = MiniPortile.new "libxml2", "2.7.7"
+$recipes[:libxslt]  = MiniPortile.new "libxslt", "1.1.26"
+$recipes.each { |_, recipe| recipe.host = HOST }
 
-### Build zlib ###
-file "tmp/cross/download/#{ZLIB}" do |t|
-  FileUtils.mkdir_p('tmp/cross/download')
-
-  file = ZLIB
-  url  = "http://zlib.net/#{file}.tar.gz"
-
-  Dir.chdir('tmp/cross/download') do
-    sh "wget #{url} || curl -O #{url}"
-    sh "tar zxvf #{file}.tar.gz"
-  end
-
-  Dir.chdir t.name do
-    mk = File.read('win32/Makefile.gcc')
-    File.open('win32/Makefile.gcc', 'wb') do |f|
-      f.puts "BINARY_PATH = #{CROSS_DIR}/bin"
-      f.puts "LIBRARY_PATH = #{CROSS_DIR}/lib"
-      f.puts "INCLUDE_PATH = #{CROSS_DIR}/include"
-
-      f.puts mk.sub(/^PREFIX\s*=\s*$/, "PREFIX = #{HOST}-") #.
-        #sub(/^SHARED_MODE=0$/, 'SHARED_MODE=1').
-        #sub(/^IMPLIB\s*=.*$/, 'IMPLIB=libz.dll.a')
-    end
-  end
-end
-
-file 'tmp/cross/lib/libz.a' => "tmp/cross/download/#{ZLIB}" do |t|
-  Dir.chdir t.prerequisites.first do
-    sh 'make -f win32/Makefile.gcc'
-    sh 'make -f win32/Makefile.gcc install'
-  end
-end
-### End build zlib ###
-
-### Build iconv ###
-file "tmp/cross/download/#{ICONV}" do |t|
-  FileUtils.mkdir_p('tmp/cross/download')
-
-  file = ICONV
-  url  = "http://ftp.gnu.org/pub/gnu/libiconv/#{file}.tar.gz"
-
-  Dir.chdir('tmp/cross/download') do
-    sh "wget #{url} || curl -O #{url}"
-    sh "tar zxvf #{file}.tar.gz"
-  end
-
-  Dir.chdir t.name do
-    sh "./configure --disable-shared --enable-static --host=#{HOST} --target=#{TARGET} --prefix=#{CROSS_DIR} CPPFLAGS='-mno-cygwin -Wall' CFLAGS='-mno-cygwin -O2 -g' CXXFLAGS='-mno-cygwin -O2 -g' LDFLAGS=-mno-cygwin"
-  end
-end
-
-file 'tmp/cross/bin/iconv.exe' => "tmp/cross/download/#{ICONV}" do |t|
-  Dir.chdir t.prerequisites.first do
-    sh 'make'
-    sh 'make install'
-  end
-end
-### End build iconv ###
-
-### Build libxml2 ###
-file "tmp/cross/download/#{LIBXML}" do |t|
-  FileUtils.mkdir_p('tmp/cross/download')
-
-  file = LIBXML
-  url  = "ftp://ftp.xmlsoft.org/libxml2/#{file}.tar.gz"
-
-  Dir.chdir('tmp/cross/download') do
-    sh "wget #{url} || curl -O #{url}"
-    sh "tar zxvf #{file}.tar.gz"
-  end
-
-  Dir.chdir t.name do
-    sh "CFLAGS='-DIN_LIBXML' ./configure --host=#{HOST} --target=#{TARGET} --enable-static --disable-shared --prefix=#{CROSS_DIR} --with-zlib=#{CROSS_DIR} --with-iconv=#{CROSS_DIR} --without-python --without-readline"
+file "lib/nokogiri/nokogiri.rb" do
+  File.open("lib/nokogiri/nokogiri.rb", 'wb') do |f|
+    f.write %Q{require "nokogiri/\#{RUBY_VERSION.sub(/\\.\\d+$/, '')}/nokogiri"\n}
   end
 end
 
-file 'tmp/cross/bin/xml2-config' => "tmp/cross/download/#{LIBXML}" do |t|
-  Dir.chdir t.prerequisites.first do
-    sh 'make LDFLAGS="-avoid-version"'
-    sh 'make install'
-  end
-end
-### End build libxml2 ###
-
-### Build libxslt ###
-file "tmp/cross/download/#{LIBXSLT}" do |t|
-  FileUtils.mkdir_p('tmp/cross/download')
-
-  file = LIBXSLT
-  url  = "ftp://ftp.xmlsoft.org/libxml2/#{file}.tar.gz"
-
-  Dir.chdir('tmp/cross/download') do
-    sh "wget #{url} || curl -O #{url}"
-    sh "tar zxvf #{file}.tar.gz"
-  end
-
-  Dir.chdir t.name do
-    sh "CFLAGS='-DIN_LIBXML' ./configure --host=#{HOST} --target=#{TARGET} --enable-static --disable-shared --prefix=#{CROSS_DIR} --with-libxml-prefix=#{CROSS_DIR} --without-python --without-crypto"
-  end
-end
+namespace :cross do
+  task :zlib do
+    recipe = $recipes[:zlib]
+    recipe.files = ["http://zlib.net/#{recipe.name}-#{recipe.version}.tar.gz"]
+    class << recipe
+      def configure
+        Dir.chdir work_path do
+          mk = File.read 'win32/Makefile.gcc'
+          File.open 'win32/Makefile.gcc', 'wb' do |f|
+            f.puts "BINARY_PATH = #{CROSS_DIR}/bin"
+            f.puts "LIBRARY_PATH = #{CROSS_DIR}/lib"
+            f.puts "INCLUDE_PATH = #{CROSS_DIR}/include"
+            f.puts mk.sub(/^PREFIX\s*=\s*$/, "PREFIX = #{HOST}-")
+          end
+        end
+      end
+
+      def configured?
+        Dir.chdir work_path do
+          !! (File.read('win32/Makefile.gcc') =~ /^BINARY_PATH/)
+        end
+      end
+
+      def compile
+        execute "compile", "make -f win32/Makefile.gcc"
+      end
+
+      def install
+        execute "install", "make -f win32/Makefile.gcc install"
+      end
+    end
 
-file 'tmp/cross/bin/xslt-config' => "tmp/cross/download/#{LIBXSLT}" do |t|
-  Dir.chdir t.prerequisites.first do
-    sh 'make LDFLAGS="-avoid-version"'
-    sh 'make install'
-  end
-end
-### End build libxslt ###
+    checkpoint = "#{CROSS_DIR}/#{recipe.name}-#{recipe.version}.installed"
+    unless File.exist?(checkpoint)
+      recipe.cook
+      touch checkpoint
+    end
+    recipe.activate
+  end
+
+  task :libiconv do
+    recipe = $recipes[:libiconv]
+    recipe.files = ["http://ftp.gnu.org/pub/gnu/libiconv/#{recipe.name}-#{recipe.version}.tar.gz"]
+    recipe.configure_options = [
+      "--host=#{HOST}",
+      "--enable-static",
+      "--disable-shared",
+      "CPPFLAGS='-mno-cygwin -Wall'",
+      "CFLAGS='-mno-cygwin -O2 -g'",
+      "CXXFLAGS='-mno-cygwin -O2 -g'",
+      "LDFLAGS=-mno-cygwin"
+    ]
+
+    checkpoint = "#{CROSS_DIR}/#{recipe.name}-#{recipe.version}.installed"
+    unless File.exist?(checkpoint)
+      recipe.cook
+      touch checkpoint
+    end
+    recipe.activate
+  end
+
+  task :libxml2 => ["cross:zlib", "cross:libiconv"] do
+    recipe = $recipes[:libxml2]
+    recipe.files = ["ftp://ftp.xmlsoft.org/libxml2/#{recipe.name}-#{recipe.version}.tar.gz"]
+    recipe.configure_options = [
+      "--host=#{HOST}",
+      "--enable-static",
+      "--disable-shared",
+      "--with-zlib=#{CROSS_DIR}",
+      "--with-iconv=#{$recipes[:libiconv].path}",
+      "--without-python",
+      "--without-readline",
+      "CFLAGS='-DIN_LIBXML'"
+    ]
+    class << recipe
+      def download
+        Dir.chdir archives_path do
+          @files.each do |url|
+            sh "wget #{url} || curl -O #{url}"
+          end
+        end
+      end
+    end
 
-file 'lib/nokogiri/nokogiri.rb' => 'cross:check' do
-  File.open("lib/#{HOE.name}/#{HOE.name}.rb", 'wb') do |f|
-    f.write <<-eoruby
-require "#{HOE.name}/\#{RUBY_VERSION.sub(/\\.\\d+$/, '')}/#{HOE.name}"
-    eoruby
-  end
-end
+    checkpoint = "#{CROSS_DIR}/#{recipe.name}-#{recipe.version}.installed"
+    unless File.exist?(checkpoint)
+      recipe.cook
+      touch checkpoint
+    end
+    recipe.activate
+  end
+
+  task :libxslt => ['cross:libxml2'] do
+    recipe = $recipes[:libxslt]
+    recipe.files = ["ftp://ftp.xmlsoft.org/libxml2/#{recipe.name}-#{recipe.version}.tar.gz"]
+    recipe.configure_options = [
+      "--host=#{HOST}",
+      "--enable-static",
+      "--disable-shared",
+      "--with-libxml-prefix=#{$recipes[:libxml2].path}",
+      "--without-python",
+      "--without-crypto",
+      "CFLAGS='-DIN_LIBXML'"
+    ]
+    class << recipe
+      def download
+        Dir.chdir archives_path do
+          @files.each do |url|
+            sh "wget #{url} || curl -O #{url}"
+          end
+        end
+      end
+    end
 
-namespace :cross do
-  task :iconv   => 'tmp/cross/bin/iconv.exe'
-  task :zlib    => 'tmp/cross/lib/libz.a'
-  task :libxml2 => ['cross:zlib', 'cross:iconv', 'tmp/cross/bin/xml2-config']
-  task :libxslt => ['cross:libxml2', 'tmp/cross/bin/xslt-config']
-
-  task :check => ["cross:libxslt"] do
-    unless File.directory?(RAKE_COMPILER_PKGCONFIG)
-      raise RuntimeError.new("looks like rake-compiler changed where pkgconfig info is kept. (#{RAKE_COMPILER_PKGCONFIG})")
+    checkpoint = "#{CROSS_DIR}/#{recipe.name}-#{recipe.version}.installed"
+    unless File.exist?(checkpoint)
+      recipe.cook
+      touch checkpoint
     end
+    recipe.activate
   end
 
   task :file_list do
-    HOE.spec.extensions = []
-    HOE.spec.files += Dir["lib/#{HOE.name}/#{HOE.name}.rb"]
-    HOE.spec.files += Dir["lib/#{HOE.name}/1.{8,9}/#{HOE.name}.so"]
+    HOE.spec.files += Dir["lib/nokogiri/nokogiri.rb"]
+    HOE.spec.files += Dir["lib/nokogiri/1.{8,9}/nokogiri.so"]
   end
 end
 
-CLOBBER.include("lib/nokogiri/nokogiri.{so,dylib,rb,bundle}")
-CLOBBER.include("lib/nokogiri/1.{8,9}")
-
-if Rake::Task.task_defined?(:cross)
-  Rake::Task[:cross].prerequisites << "lib/nokogiri/nokogiri.rb"
-  Rake::Task[:cross].prerequisites << "cross:file_list"
-end
+HOE.clean_globs += [
+  "#{CROSS_DIR}/*.installed",
+  "#{CROSS_DIR}/#{HOST}",
+  "tmp/#{HOST}",
+]
 
-desc "build a windows gem without all the ceremony."
-task "gem:windows" do
-  rake_compiler_config = YAML.load_file("#{ENV['HOME']}/.rake-compiler/config.yml")
-
-  # check that rake-compiler config contains the right patchlevels of 1.8.6 and 1.9.1. see #279.
-  ["1.8.6-p383", "1.9.1-p243"].each do |version|
-    majmin, patchlevel = version.split("-")
-    rbconfig = "rbconfig-#{majmin}"
-    unless rake_compiler_config.key?(rbconfig) && rake_compiler_config[rbconfig] =~ /-#{patchlevel}/
-      raise "rake-compiler '#{rbconfig}' not #{patchlevel}. try running 'env --unset=HOST rake-compiler cross-ruby VERSION=#{version}'"
-    end
-  end
-
-  # verify that --export-all is in the 1.9.1 rbconfig. see #279,#374,#375.
-  rbconfig_191 = rake_compiler_config["rbconfig-1.9.1"]
-  raise "rbconfig #{rbconfig_191} needs --export-all in its DLDFLAGS value" if File.read(rbconfig_191).grep(/CONFIG\["DLDFLAGS"\].*--export-all/).empty?
-
-  system("env PKG_CONFIG_PATH=#{RAKE_COMPILER_PKGCONFIG} RUBY_CC_VERSION=1.8.6:1.9.1 rake cross native gem") || raise("build failed!")
-end
+task :cross => ["cross:libxslt", "lib/nokogiri/nokogiri.rb", "cross:file_list"]

data/tasks/nokogiri.org.rb

@@ -0,0 +1,18 @@
+namespace :docs do
+  desc "generate HTML docs for nokogiri.org"
+  task :website do
+    title = "#{HOE.name}-#{HOE.version} Documentation"
+
+    options = []
+    options << "--main=#{HOE.readme_file}"
+    options << '--format=activerecord'
+    options << '--threads=1'
+    options << "--title=#{title.inspect}"
+
+    options += HOE.spec.require_paths
+    options += HOE.spec.extra_rdoc_files
+    require 'rdoc/rdoc'
+    ENV['RAILS_ROOT'] ||= File.expand_path(File.join('..', 'nokogiri_ws'))
+    RDoc::RDoc.new.document options
+  end
+end

data/tasks/test.rb

@@ -64,7 +64,7 @@ namespace :test do
       end
 
       Dir.chdir File.join(MULTI_XML, 'build', filename) do
-        system "./configure --prefix=#{install_dir}"
+        system "./configure --without-http --prefix=#{install_dir}"
         system "make && make install"
       end
     end
@@ -73,7 +73,7 @@ namespace :test do
     libxslt = Dir[File.join(MULTI_XML, 'install', 'libxslt*')].first
 
     directories = ENV['MULTIXML2_DIR'] ? [ENV['MULTIXML2_DIR']] : Dir[File.join(MULTI_XML, 'install', '*')]
-    directories.sort.reverse.each do |xml2_version|
+    directories.sort.reverse_each do |xml2_version|
       next unless xml2_version =~ /libxml2/
       extopts = "--with-xml2-include=#{xml2_version}/include/libxml2 --with-xml2-lib=#{xml2_version}/lib --with-xslt-dir=#{libxslt} --with-iconv-dir=/usr"
       cmd = "#{$0} clean test EXTOPTS='#{extopts}' LD_LIBRARY_PATH='#{xml2_version}/lib'"

data/test/files/metacharset.html

@@ -0,0 +1,10 @@
+<!DOCTYPE HTML>
+<html>
+<head>
+<title>$B$?$3>F$-2>LL(B</title>
+<meta charset="iso-2022-jp" />
+</head>
+<body>
+<p>$B0-$$;v$r9=A[Cf!#(B</p>
+</body>
+</html>

data/test/files/noencoding.html

@@ -0,0 +1,47 @@
+<!DOCTYPE HTML>
+<html>
+<head>
+<title>I have no encoding declaration.</title>
+</head>
+<body>
+<p>I want one.</p>
+<p>I really want one.</p>
+<p>I really really want one.</p>
+<p>I really really really want one.</p>
+<p>I really really really really want one.</p>
+<p>I really really really really really want one.</p>
+<p>I really really really really really really want one.</p>
+<p>I really really really really really really really want one.</p>
+<p>I really really really really really really really really want one.</p>
+<p>I really really really really really really really really really want one.</p>
+<p>I really really really really really really really really really really want one.</p>
+<p>I really really really really really really really really really really really want one.</p>
+<p>I really really really really really really really really really really really really want one.</p>
+<p>I really really really really really really really really really really really really really want one.</p>
+<p>I really really really really really really really really really really really really really really want one.</p>
+<p>I really really really really really really really really really really really really really really really want one.</p>
+<p>I really really really really really really really really really really really really really really really really want one.</p>
+<p>I really really really really really really really really really really really really really really really really really want one.</p>
+<p>I really really really really really really really really really really really really really really really really really really want one.</p>
+<p>I really really really really really really really really really really really really really really really really really really really want one.</p>
+<p>I really really really really really really really really really really really really really really really really really really really really want one.</p>
+<p>I really really really really really really really really really really really really really really really really really really really really really want one.</p>
+<p>I really really really really really really really really really really really really really really really really really really really really really really want one.</p>
+<p>I really really really really really really really really really really really really really really really really really really really really really really really want one.</p>
+<p>I really really really really really really really really really really really really really really really really really really really really really really really really want one.</p>
+<p>I really really really really really really really really really really really really really really really really really really really really really really really really really want one.</p>
+<p>I really really really really really really really really really really really really really really really really really really really really really really really really really really want one.</p>
+<p>I really really really really really really really really really really really really really really really really really really really really really really really really really really really want one.</p>
+<p>I really really really really really really really really really really really really really really really really really really really really really really really really really really really really want one.</p>
+<p>I really really really really really really really really really really really really really really really really really really really really really really really really really really really really really want one.</p>
+<p>I really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really want one.</p>
+<p>I really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really want one.</p>
+<p>I really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really want one.</p>
+<p>I really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really want one.</p>
+<p>I really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really want one.</p>
+<p>I really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really want one.</p>
+<p>I really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really want one.</p>
+<p>I really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really want one.</p>
+<p>I really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really really want one.</p>
+</body>
+</html>