nokogiri 1.6.8.rc3-x86-mingw32 → 1.6.8-x86-mingw32

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: c0ec7810f8915eb3f4f7cf25ffeb07289f58a155
-  data.tar.gz: 85bd7afc38bb90c7aa8024725dc6aefe6cb7adbf
+  metadata.gz: e99debe6aeb9a5efe23829ceb2c509b78bccf7fe
+  data.tar.gz: 0a87cbb2324b73854dedde53e063097482f1e56c
 SHA512:
-  metadata.gz: 03ac7a0e499a19b4b17a1a1e390d75598048e11812bf66f4bf163f4df9e03fd277a8f9ae812b561bba21c86b5bcad97759013507d64d6b9c21a10f8d232da035
-  data.tar.gz: 422ce41dd6c84832ae3ea48438530beb62315ce05f5b095e8f2e7291f341d554a999f205b0ff8d410125c792f02c9c95c24e9d1c1f5f2b8f85fb97f958ece5fc
+  metadata.gz: d9ca761b92ec08e70d04df86803f4df5a44cdcfcf858564e8513c989dfe7431a9dc2af8d591d8733bb0854799c607cc1768a2d385fe2980e29d7b8ae6b648352
+  data.tar.gz: f50a0d73cc8f249412b337d8ac0ee3353990554d1a216fc4a5401231e7faaf74adc27143425875e9e104a1527d124597aa0981e22d7461234081ee8814be1cac

data/.travis.yml

@@ -10,8 +10,9 @@ rvm:
 - 2.2.3
 - 2.3.0
 - jruby-19mode
-- jruby-1.7.23
+- jruby
 - jruby-9.0.4.0
+- jruby-head
 - rbx-19mode
 - rbx-2
 
@@ -24,13 +25,14 @@ matrix:
   # https://github.com/travis-ci/travis-ci/issues/5361
   - os: osx
     rvm: 2.3.0
+  - rvm: jruby-head
   exclude:
   - os: osx
     rvm: ruby-1.9.2
   - os: osx
     rvm: jruby-19mode
   - os: osx
-    rvm: jruby-1.7.23
+    rvm: jruby
   - os: osx
     rvm: jruby-9.0.4.0
   - os: osx
@@ -42,7 +44,7 @@ notifications:
   irc:
     channels:
       - "chat.freenode.net#nokogiri"
-    on_success: change
+    on_success: always
     on_failure: always
     template:
       - "%{repository} (%{branch}:%{commit} by %{author}): %{message} (%{build_url})"

data/CHANGELOG.rdoc

@@ -1,10 +1,24 @@
 === 1.6.8 / unreleased
 
-==== Security Note
+==== Security Notes
+
+[MRI] Bundled libxml2 is upgraded to 2.9.4, which fixes many security issues. Many of these had previously been patched in the vendored libxml 2.9.2 in the 1.6.7.x branch, but some are newer.
+
+See these libxml2 email posts for more:
+
+* https://mail.gnome.org/archives/xml/2015-November/msg00012.html
+* https://mail.gnome.org/archives/xml/2016-May/msg00023.html
+
+For a more detailed analysis, you may care to read Canonical's take on these security issues:
+
+* http://www.ubuntu.com/usn/usn-2994-1
+
+
+[MRI] Bundled libxslt is upgraded to 1.1.29, which fixes a security issue as well as many long-known outstanding bugs, some features, some portability improvements, and general cleanup.
+
+See this libxslt email post for more:
 
-* (MRI) Bundled Libxml2 is upgraded to 2.9.3, which fixes a series of security issues.
-  These security issues had been patched in the vendored libxml 2.9.2 in the 1.6.7.x branch.
-  See https://mail.gnome.org/archives/xml/2015-November/msg00012.html for more.
+* https://mail.gnome.org/archives/xslt/2016-May/msg00004.html
 
 
 ==== Features
@@ -16,19 +30,28 @@ Several changes were made to improve performance:
 * Use Symbol#to_proc where we weren't previously. (#1296) (Thanks, Bruno Sutic!)
 * XML::DTD#each uses implicit block calls. (Thanks, @glaucocustodio!)
 * Fall back to the `pkg-config` gem if we're having trouble finding the system libxml2. This should help many FreeBSD users. (#1417)
+* Set document encoding appropriately even on blank document. (#1043) (Thanks, @batter!)
 
 
 ==== Bug Fixes
 
+* [JRuby] fix slow add_child (#692)
 * [JRuby] fix load errors when deploying to JRuby/Torquebox (#1114) (Thanks, @atambo and @jvshahid!)
 * [JRuby] fix NPE when inspecting nodes returned by NodeSet#drop (#1042) (Thanks, @mkristian!)
 * [JRuby] fix nil attriubte node's namespace in reader (#1327) (Thanks, @codekitchen!)
 * [JRuby] fix Nokogiri munging unicode characters that require more than 2 bytes (#1113) (Thanks, @mkristian!)
+* [JRuby] allow unlinking an unparented node (#1112, #1152) (Thanks, @esse!)
+* [JRuby] allow Fragment parsing on a frozen string (#444, #1077)
+* [JRuby] HTML `style` tags are no longer encoded (#1316) (Thanks, @tbeauvais!)
 * [MRI] fix assertion failure while accessing attribute node's namespace in reader (#843) (Thanks, @2potatocakes!)
 * [MRI] fix issue with GCing namespace nodes returned in an xpath query. (#1155)
 * [MRI] Ensure C strings are null-terminated. (#1381)
 * [MRI] Ensure Rubygems is loaded before using mini_portile2 at installation. (#1393, #1411) (Thanks, @JonRowe!)
 * [MRI] Handling another edge case where the `libxml-ruby` gem's global callbacks were smashing the heap. (#1426). (Thanks to @bbergstrom for providing an isolated test case!)
+* [MRI] Ensure encodings are passed to Sax::Parser xmldecl callback. (#844)
+* [MRI] Ensure default ns prefix is applied correctly when reparenting nodes to another document. (#391) (Thanks, @ylecuyer!)
+* [MRI] Ensure Reader handles non-existent attributes as expected. (#1254) (Thanks, @ccutrer!)
+* [MRI] Cleanup around namespace handling when reparenting nodes. (#1332, #1333, #1444) (Thanks, @cuttrer and @bradleybeddoes!)
 * unescape special characters in CSS queries (#1303) (Thanks, @twalpole!)
 * consistently handle empty documents (#1349)
 * Update to mini_portile2 2.1.0 to address whitespace-handling during patching. (#1402)

data/CONTRIBUTING.md

@@ -1,4 +1,4 @@
-# Issue and Pull Request Guidlines
+# Issue and Pull Request Guidelines
 
 Thank you for getting involved in making Nokogiri better!
 

data/Gemfile

@@ -8,16 +8,16 @@ gem "mini_portile2", "~>2.1.0"
 gem "pkg-config", "~>1.1.7"
 
 gem "rdoc", "~>4.0", :group => [:development, :test]
-gem "hoe-bundler", ">=1.1", :group => [:development, :test]
+gem "hoe-bundler", "~>1.2.0", :group => [:development, :test]
 gem "hoe-debugging", "~>1.2.1", :group => [:development, :test]
-gem "hoe-gemspec", ">=1.0", :group => [:development, :test]
-gem "hoe-git", ">=1.4", :group => [:development, :test]
-gem "minitest", "~>2.2.2", :group => [:development, :test]
-gem "rake", ">=0.9", :group => [:development, :test]
+gem "hoe-gemspec", "~>1.0.0", :group => [:development, :test]
+gem "hoe-git", "~>1.6.0", :group => [:development, :test]
+gem "minitest", "~>5.8.4", :group => [:development, :test]
+gem "rake", "~>10.5.0", :group => [:development, :test]
 gem "rake-compiler", "~>0.9.2", :group => [:development, :test]
 gem "rake-compiler-dock", "~>0.5.1", :group => [:development, :test]
-gem "racc", ">=1.4.6", :group => [:development, :test], :platform => :ruby
-gem "rexical", ">=1.0.5", :group => [:development, :test], :platform => :ruby
+gem "racc", "~>1.4.14", :group => [:development, :test], :platform => :ruby
+gem "rexical", "~>1.0.5", :group => [:development, :test], :platform => :ruby
 gem "hoe", "~>3.14", :group => [:development, :test]
 
 # vim: syntax=ruby

data/Manifest.txt

@@ -319,7 +319,6 @@ test/test_css_cache.rb
 test/test_encoding_handler.rb
 test/test_memory_leak.rb
 test/test_nokogiri.rb
-test/test_reader.rb
 test/test_soap4r_sax.rb
 test/test_xslt_transforms.rb
 test/xml/node/test_save_options.rb
@@ -351,6 +350,7 @@ test/xml/test_node_reparenting.rb
 test/xml/test_node_set.rb
 test/xml/test_parse_options.rb
 test/xml/test_processing_instruction.rb
+test/xml/test_reader.rb
 test/xml/test_reader_encoding.rb
 test/xml/test_relax_ng.rb
 test/xml/test_schema.rb

data/README.md

@@ -22,9 +22,6 @@ Nokogiri (鋸) is an HTML, XML, SAX, and Reader parser.  Among
 Nokogiri's many features is the ability to search documents via XPath
 or CSS3 selectors.
 
-XML is like violence - if it doesn’t solve your problems, you are not
-using enough of it.
-
 
 ## Features
 

data/Rakefile

@@ -8,7 +8,6 @@ Hoe.plugin :debugging
 Hoe.plugin :git
 Hoe.plugin :gemspec
 Hoe.plugin :bundler
-Hoe.add_include_dirs '.'
 
 GENERATED_PARSER    = "lib/nokogiri/css/parser.rb"
 GENERATED_TOKENIZER = "lib/nokogiri/css/tokenizer.rb"
@@ -134,16 +133,16 @@ HOE = Hoe.spec 'nokogiri' do
   end
 
   self.extra_dev_deps += [
-    ["hoe-bundler",        ">= 1.1"],
+    ["hoe-bundler",        "~> 1.2.0"],
     ["hoe-debugging",      "~> 1.2.1"],
-    ["hoe-gemspec",        ">= 1.0"],
-    ["hoe-git",            ">= 1.4"],
-    ["minitest",           "~> 2.2.2"],
-    ["rake",               ">= 0.9"],
+    ["hoe-gemspec",        "~> 1.0.0"],
+    ["hoe-git",            "~> 1.6.0"],
+    ["minitest",           "~> 5.8.4"],
+    ["rake",               "~> 10.5.0"],
     ["rake-compiler",      "~> 0.9.2"],
     ["rake-compiler-dock", "~> 0.5.1"],
-    ["racc",               ">= 1.4.6"],
-    ["rexical",            ">= 1.0.5"]
+    ["racc",               "~> 1.4.14"],
+    ["rexical",            "~> 1.0.5"]
   ]
 
   if java?
@@ -207,7 +206,7 @@ else
 
   task gem_build_path do
     %w[libxml2 libxslt].each do |lib|
-      version = dependencies[lib]
+      version = dependencies[lib]["version"]
       archive = File.join("ports", "archives", "#{lib}-#{version}.tar.gz")
       add_file_to_gem archive
       patchesdir = File.join("patches", lib)
@@ -232,7 +231,7 @@ else
       ext.cross_platform = CROSS_RUBIES.map(&:platform).uniq
       ext.cross_config_options << "--enable-cross-build"
       ext.cross_compiling do |spec|
-        libs = dependencies.map { |name, version| "#{name}-#{version}" }.join(', ')
+        libs = dependencies.map { |name, dep| "#{name}-#{dep["version"]}" }.join(', ')
 
         spec.required_ruby_version = [
           '>= 1.9.2',
@@ -293,7 +292,7 @@ task :debug do
   ENV['CFLAGS'] += " -DDEBUG"
 end
 
-require 'tasks/test'
+require File.join File.dirname(__FILE__), 'tasks/test'
 
 task :java_debug do
   ENV['JRUBY_OPTS'] = "#{ENV['JRUBY_OPTS']} --debug --dev"

data/dependencies.yml

@@ -1,4 +1,29 @@
-libxml2: "2.9.3"
-libxslt: "1.1.28"
-zlib: "1.2.8"
-libiconv: "1.14"
+libxml2:
+  version: "2.9.4"
+  md5: "ae249165c173b1ff386ee8ad676815f5" # manually confirmed via `gpg --verify`
+  # gpg: Signature made Mon 23 May 2016 04:02:13 AM EDT using DSA key ID DE95BC1F
+  # gpg: Good signature from "Daniel Veillard (Red Hat work email) <veillard@redhat.com>"
+  # gpg:                 aka "Daniel Veillard <Daniel.Veillard@w3.org>"
+  # 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
+
+libxslt:
+  version: "1.1.29"
+  md5: "a129d3c44c022de3b9dcf6d6f288d72e"
+  # gpg: Signature made Mon 23 May 2016 09:58:52 PM EDT using DSA key ID DE95BC1F
+  # gpg: Good signature from "Daniel Veillard (Red Hat work email) <veillard@redhat.com>"
+  # gpg:                 aka "Daniel Veillard <Daniel.Veillard@w3.org>"
+  # 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
+
+zlib:
+  version: "1.2.8"
+  md5: "44d667c142d7cda120332623eab69f40"
+
+libiconv:
+  version: "1.14"
+  md5: "e34509b1623cec449dfeb73d7ce9c6c6"
+  # gpg: Signature made Sun 07 Aug 2011 01:58:18 PM EDT using DSA key ID F059B1D1
+  # gpg: BAD signature from "Bruno Haible (Open Source Development) <bruno@clisp.org>"

data/ext/nokogiri/extconf.rb

@@ -8,6 +8,42 @@ ROOT = File.expand_path(File.join(File.dirname(__FILE__), '..', '..'))
 #
 # functions
 #
+def windows?
+  RbConfig::CONFIG['target_os'] =~ /mingw32|mswin/
+end
+
+def solaris?
+  RbConfig::CONFIG['target_os'] =~ /solaris/
+end
+
+def darwin?
+  RbConfig::CONFIG['target_os'] =~ /darwin/
+end
+
+def nix?
+  ! (windows? || solaris? || darwin?)
+end
+
+def sh_export_path path
+  # because libxslt 1.1.29 configure.in uses AC_PATH_TOOL which treats ":"
+  # as a $PATH separator, we need to convert windows paths from
+  #
+  #   C:/path/to/foo
+  #
+  # to
+  #
+  #   /C/path/to/foo
+  #
+  # which is sh-compatible, in order to find things properly during
+  # configuration
+  if windows?
+    match = Regexp.new("^([A-Z]):(/.*)").match(path)
+    if match && match.length == 3
+      return File.join("/", match[1], match[2])
+    end
+  end
+  path
+end
 
 def do_help
   print <<HELP
@@ -364,7 +400,10 @@ when arg_config('--clean')
 end
 
 RbConfig::MAKEFILE_CONFIG['CC'] = ENV['CC'] if ENV['CC']
+# use same c compiler for libxml and libxslt
+ENV['CC'] = RbConfig::MAKEFILE_CONFIG['CC']
 
+# TODO: deprecate MacRuby: https://github.com/sparklemotion/nokogiri/issues/1474
 if defined?(RUBY_ENGINE) && RUBY_ENGINE == 'macruby'
   $LIBRUBYARG_STATIC.gsub!(/-static/, '')
 end
@@ -374,17 +413,20 @@ $LIBS << " #{ENV["LIBS"]}"
 # Read CFLAGS from ENV and make sure compiling works.
 add_cflags(ENV["CFLAGS"])
 
-case RbConfig::CONFIG['target_os']
-when 'mingw32', /mswin/
-  windows_p = true
+if windows?
   $CFLAGS << " -DXP_WIN -DXP_WIN32 -DUSE_INCLUDED_VASPRINTF"
-when /solaris/
+end
+
+if solaris?
   $CFLAGS << " -DUSE_INCLUDED_VASPRINTF"
-when /darwin/
-  darwin_p = true
+end
+
+if darwin?
   # Let Apple LLVM/clang 5.1 ignore unknown compiler flags
   add_cflags("-Wno-error=unused-command-line-argument-hard-error-in-future")
-else
+end
+
+if nix?
   $CFLAGS << " -g -DXP_UNIX"
 end
 
@@ -436,9 +478,12 @@ else
   dependencies = YAML.load_file(File.join(ROOT, "dependencies.yml"))
 
   cross_build_p = enable_config("cross-build")
-  if cross_build_p || windows_p
-    zlib_recipe = process_recipe("zlib", dependencies["zlib"], static_p, cross_build_p) do |recipe|
-      recipe.files = ["http://zlib.net/#{recipe.name}-#{recipe.version}.tar.gz"]
+  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/#{recipe.name}-#{recipe.version}.tar.gz",
+          md5: dependencies["zlib"]["md5"]
+        }]
       class << recipe
         attr_accessor :cross_build_p
 
@@ -472,8 +517,11 @@ else
       recipe.cross_build_p = cross_build_p
     end
 
-    libiconv_recipe = process_recipe("libiconv", dependencies["libiconv"], static_p, cross_build_p) do |recipe|
-      recipe.files = ["http://ftp.gnu.org/pub/gnu/libiconv/#{recipe.name}-#{recipe.version}.tar.gz"]
+    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",
+          md5: dependencies["libiconv"]["md5"]
+        }]
       recipe.configure_options += [
         "CPPFLAGS=-Wall",
         "CFLAGS=-O2 -g",
@@ -482,7 +530,7 @@ else
       ]
     end
   else
-    if darwin_p && !have_header('iconv.h')
+    if darwin? && !have_header('iconv.h')
       abort <<'EOM'.chomp
 -----
 The file "iconv.h" is missing in your build environment,
@@ -498,17 +546,21 @@ EOM
     end
   end
 
-  unless windows_p
+  unless windows?
     preserving_globals {
       have_library('z', 'gzdopen', 'zlib.h')
     } or abort 'zlib is missing; necessary for building libxml2'
   end
 
-  libxml2_recipe = process_recipe("libxml2", dependencies["libxml2"], static_p, cross_build_p) do |recipe|
-    recipe.files = ["http://xmlsoft.org/sources/#{recipe.name}-#{recipe.version}.tar.gz"]
+  libxml2_recipe = process_recipe("libxml2", dependencies["libxml2"]["version"], static_p, cross_build_p) do |recipe|
+    recipe.files = [{
+        url: "http://xmlsoft.org/sources/#{recipe.name}-#{recipe.version}.tar.gz",
+        md5: dependencies["libxml2"]["md5"]
+      }]
     recipe.configure_options += [
       "--without-python",
       "--without-readline",
+      *(zlib_recipe ? ["--with-zlib=#{zlib_recipe.path}", "CFLAGS=-I#{zlib_recipe.path}/include"] : []),
       *(libiconv_recipe ? "--with-iconv=#{libiconv_recipe.path}" : iconv_configure_flags),
       "--with-c14n",
       "--with-debug",
@@ -516,13 +568,16 @@ EOM
     ]
   end
 
-  libxslt_recipe = process_recipe("libxslt", dependencies["libxslt"], static_p, cross_build_p) do |recipe|
-    recipe.files = ["http://xmlsoft.org/sources/#{recipe.name}-#{recipe.version}.tar.gz"]
+  libxslt_recipe = process_recipe("libxslt", dependencies["libxslt"]["version"], static_p, cross_build_p) do |recipe|
+    recipe.files = [{
+        url: "http://xmlsoft.org/sources/#{recipe.name}-#{recipe.version}.tar.gz",
+        md5: dependencies["libxslt"]["md5"]
+      }]
     recipe.configure_options += [
       "--without-python",
       "--without-crypto",
       "--with-debug",
-      "--with-libxml-prefix=#{libxml2_recipe.path}"
+      "--with-libxml-prefix=#{sh_export_path(libxml2_recipe.path)}"
     ]
   end
 

data/ext/nokogiri/xml_node.c

@@ -35,13 +35,13 @@ static void relink_namespace(xmlNodePtr reparented)
   xmlNsPtr ns;
 
   if (reparented->type != XML_ATTRIBUTE_NODE &&
-      reparented->type != XML_ELEMENT_NODE) return;
+      reparented->type != XML_ELEMENT_NODE) { return; }
 
   if (reparented->ns == NULL || reparented->ns->prefix == NULL) {
     name = xmlSplitQName2(reparented->name, &prefix);
 
     if(reparented->type == XML_ATTRIBUTE_NODE) {
-      if (prefix == NULL || strcmp((char*)prefix, XMLNS_PREFIX) == 0) return;
+      if (prefix == NULL || strcmp((char*)prefix, XMLNS_PREFIX) == 0) { return; }
     }
 
     ns = xmlSearchNs(reparented->doc, reparented, prefix);
@@ -57,18 +57,19 @@ static void relink_namespace(xmlNodePtr reparented)
   }
 
   /* Avoid segv when relinking against unlinked nodes. */
-  if (reparented->type != XML_ELEMENT_NODE || !reparented->parent) return;
+  if (reparented->type != XML_ELEMENT_NODE || !reparented->parent) { return; }
 
   /* Make sure that our reparented node has the correct namespaces */
-  if(!reparented->ns && reparented->doc != (xmlDocPtr)reparented->parent)
+  if (!reparented->ns && reparented->doc != (xmlDocPtr)reparented->parent) {
     xmlSetNs(reparented, reparented->parent->ns);
+  }
 
   /* Search our parents for an existing definition */
-  if(reparented->nsDef) {
+  if (reparented->nsDef) {
     xmlNsPtr curr = reparented->nsDef;
     xmlNsPtr prev = NULL;
 
-    while(curr) {
+    while (curr) {
       xmlNsPtr ns = xmlSearchNsByHref(
           reparented->doc,
           reparented->parent,
@@ -76,7 +77,7 @@ static void relink_namespace(xmlNodePtr reparented)
       );
       /* If we find the namespace is already declared, remove it from this
        * definition list. */
-      if(ns && ns != curr) {
+      if (ns && ns != curr && xmlStrEqual(ns->prefix, curr->prefix)) {
         if (prev) {
           prev->next = curr->next;
         } else {
@@ -92,12 +93,12 @@ static void relink_namespace(xmlNodePtr reparented)
 
   /* Only walk all children if there actually is a namespace we need to */
   /* reparent. */
-  if(NULL == reparented->ns) return;
+  if (NULL == reparented->ns) { return; }
 
   /* When a node gets reparented, walk it's children to make sure that */
   /* their namespaces are reparented as well. */
   child = reparented->children;
-  while(NULL != child) {
+  while (NULL != child) {
     relink_namespace(child);
     child = child->next;
   }
@@ -140,6 +141,7 @@ static VALUE reparent_node_with(VALUE pivot_obj, VALUE reparentee_obj, pivot_rep
 {
   VALUE reparented_obj ;
   xmlNodePtr reparentee, pivot, reparented, next_text, new_next_text, parent ;
+  int original_ns_prefix_is_default = 0 ;
 
   if(!rb_obj_is_kind_of(reparentee_obj, cNokogiriXmlNode))
     rb_raise(rb_eArgError, "node must be a Nokogiri::XML::Node");
@@ -250,10 +252,21 @@ ok:
        */
       reparentee->_private = NULL ;
     }
+
+    if (reparentee->ns != NULL && reparentee->ns->prefix == NULL) {
+      original_ns_prefix_is_default = 1;
+    }
+
     nokogiri_root_node(reparentee);
+
     if (!(reparentee = xmlDocCopyNode(reparentee, pivot->doc, 1))) {
       rb_raise(rb_eRuntimeError, "Could not reparent node (xmlDocCopyNode)");
     }
+
+    if (original_ns_prefix_is_default && reparentee->ns != NULL && reparentee->ns->prefix != NULL) {
+      /* issue #391, where new node's prefix may become the string "default" */
+      reparentee->ns->prefix = NULL;
+    }
   }
 
   if (prf != xmlAddPrevSibling && prf != xmlAddNextSibling