nokogiri 1.4.2 → 1.4.3

data/CHANGELOG.ja.rdoc

@@ -1,8 +1,28 @@
-=== 1.4.2
+=== 1.4.3 ２０１０年７月２８日
 
 * 新しい機能
 
-  * XML::Node#parse　定義されたコンテキストノードで、XML 又はHTMLのフレグメント
+  * XML::Reader#empty_element?  - 子の無いエレメントにtrueを返す　 #262
+  * Node#remove_namespaces! - 1.4.2では 名前空間のみを取り除いていたが、
+    1.4.3 では名前空間及び、名前空間宣言も取り除く　#294
+
+* バグの修正
+
+  * XML::NodeSet#{include?,delete,push} はXML::Namespaceを受入れる
+  * XML::Document#parse - 1.4.3より文書内の文脈を解析する機能を追加
+  * XML::DocumentFragment#inner_html= 文脈解析を共に実行する　#298, #281
+  * lib/nokogiri/css/parser.y  はCSSと疑似選別の両方を機能
+  * 演算によって近隣に存在する併合型ノードへの遊離問題の有無に関わらず、一切の
+    弊害なしにテキストノードの繰り返しが実行可能　 #283
+  * xmlFirstElementChild et al.による libxml2バージョンでの不適合性を修正 #303
+  * XML::Attr#add_namespace　（！）文字通りの機能実現!　 #252
+  * HTML::DocumentFragment が文字列に存在するエンコードを使用　#305
+
+=== 1.4.2 ２０１０年５月２２日
+
+* 新機能
+
+  * XML::Node#parse　定義されたコンテキストノードでXML 又はHTMLのフレグメント
     を解析する
   * XML::Node#namespacesが子ノードとその祖先ノード内で定義された全ての名前空間
     を返すようになった（以前は祖先ノードの名前空間は返されなかった）
@@ -30,18 +50,18 @@
   * XML::Node#xpath はNodeSetのオブジェクト以外のオブジェクトを返す　GH #208
   * XSLT::StyleSheet#transformはパラメーターのハッシュを受け入れる　GH #223
   * CSSのnot()の疑似セレクタの修正　 GH #205
-  * XML::Builder  はノード達が切り離されても破壊しない（vihaiの協力に感謝）
+  * XML::Builder  はノードらが切り離されても破壊しない（vihaiの協力に感謝）
     GH #228
   * SAX parser経由でエンコードを強制することが出来る　 Eugene Pimenovに感謝!
     GH #204
-  * XML::DocumentFragment はML::Node#parse を使用して子供を限定する
-  * XML Reader内のメモリリークを修正　 sdorさん、ありがとう! GH#244
+  * XML::DocumentFragment はML::Node#parse を使用して子を限定する
+  * XML Reader内のメモリリーク修正　 sdorさん、ありがとう! GH#244
 
 * ノート
 
-  * 今日４月１８日現在、Windows gems は libxml 2.7.6 とlibxslt
-    1.1.26にDLLsを正規装備しています。このリリース以前にも既にDLLsはlibxml
-    2.7.3 と libxslt 1.1.24に正規装備されています。
+  * 今日４月１８日現在、Windows gems は libxml 2.7.7 とlibxslt
+    1.1.26にDLLsを標準装備しています。このリリース以前にも既にDLLsはlibxml
+    2.7.3 と libxslt 1.1.24に標準装備済み。
 
 === 1.4.1 ２００９年１２月１０日
 

data/CHANGELOG.rdoc

@@ -1,4 +1,27 @@
-=== 1.4.2 (TBA)
+=== 1.4.3 / 2010/07/28
+
+* New Features
+
+  * XML::Reader#empty_element? returns true for empty elements. #262
+  * Node#remove_namespaces! now removes namespace *declarations* as well. #294
+  * NodeSet#at_xpath, NodeSet#at_css and NodeSet#> do what the corresponding
+    methods of Node do.
+
+* Bugfixes
+
+  * XML::NodeSet#{include?,delete,push} accept an XML::Namespace
+  * XML::Document#parse added for parsing in the context of a document
+  * XML::DocumentFragment#inner_html= works with contextual parsing! #298, #281
+  * lib/nokogiri/css/parser.y Combined CSS functions + pseudo selectors fixed
+  * Reparenting text nodes is safe, even when the operation frees adjacent merged nodes. #283
+  * Fixed libxml2 versionitis issue with xmlFirstElementChild et al. #303
+  * XML::Attr#add_namespace now works as expected. #252
+  * HTML::DocumentFragment uses the string's encoding. #305
+  * Fix the CSS3 selector translation rule for the general sibling combinator
+    (a.k.a. preceding selector) that incorrectly converted "E + F G" to
+    "//F//G[preceding-sibling::E]".
+
+=== 1.4.2 / 2010/05/22
 
 * New Features
 

data/Manifest.txt

@@ -44,6 +44,8 @@ ext/nokogiri/xml_entity_reference.c
 ext/nokogiri/xml_entity_reference.h
 ext/nokogiri/xml_io.c
 ext/nokogiri/xml_io.h
+ext/nokogiri/xml_libxml2_hacks.c
+ext/nokogiri/xml_libxml2_hacks.h
 ext/nokogiri/xml_namespace.c
 ext/nokogiri/xml_namespace.h
 ext/nokogiri/xml_node.c
@@ -157,7 +159,6 @@ lib/nokogiri/html/sax/parser.rb
 lib/nokogiri/html/sax/parser_context.rb
 lib/nokogiri/syntax_error.rb
 lib/nokogiri/version.rb
-lib/nokogiri/version_warning.rb
 lib/nokogiri/xml.rb
 lib/nokogiri/xml/attr.rb
 lib/nokogiri/xml/attribute_decl.rb

data/README.ja.rdoc

@@ -3,7 +3,7 @@
 * http://nokogiri.org/
 * http://github.com/tenderlove/nokogiri/wikis
 * http://github.com/tenderlove/nokogiri/tree/master
-* http://groups.google.com/group/nokogiri-talk
+* http://groups.google.com/group/nokogiri-list
 * http://github.com/tenderlove/nokogiri/issues
 
 == DESCRIPTION:

data/README.rdoc

@@ -81,9 +81,27 @@ The IRC channel is #nokogiri on freenode.
 
 == ENCODING:
 
-Strings are always stored as UTF-8 internally.  Methods that return text values
-will always return UTF-8 encoded strings.  Methods that return XML (like to_xml,
-to_html and inner_html) will return a string encoded like the source document.
+Strings are always stored as UTF-8 internally.  Methods that return
+text values will always return UTF-8 encoded strings.  Methods that
+return XML (like to_xml, to_html and inner_html) will return a string
+encoded like the source document.
+
+*WARNING*
+
+Some documents declare one particular encoding, but use a different
+one. So, which encoding should the parser choose?
+
+Remember that data is just a stream of bytes. Only us humans add
+meaning to that stream. Any particular set of bytes could be valid
+characters in multiple encodings, so detecting encoding with 100%
+accuracy is not possible. libxml2 does its best, but it can't be right
+100% of the time.
+
+If you want Nokogiri to handle the document encoding properly, your
+best bet is to explicitly set the encoding.  Here is an example of
+explicitly setting the encoding to EUC-JP on the parser:
+
+    doc = Nokogiri.XML('<foo><bar /><foo>', nil, 'EUC-JP')
 
 == INSTALL:
 
@@ -94,7 +112,7 @@ to_html and inner_html) will return a string encoded like the source document.
 Binary packages are available for:
 
 * SuSE[http://download.opensuse.org/repositories/devel:/languages:/ruby:/extensions/]
-* Fedora[https://admin.fedoraproject.org/pkgdb/packages/name/rubygem-nokogiri]
+* Fedora[http://s390.koji.fedoraproject.org/koji/packageinfo?packageID=6756]
 
 == DEVELOPMENT:
 

data/Rakefile

@@ -112,7 +112,7 @@ end
 
 file GENERATED_PARSER => "lib/nokogiri/css/parser.y" do |t|
   begin
-    racc = `which racc`.strip
+    racc = Config::CONFIG['target_os'] =~ /mswin32/ ? '' : `which racc`.strip
     racc = "#{::Config::CONFIG['bindir']}/racc" if racc.empty?
     sh "#{racc} -l -o #{t.name} #{t.prerequisites.first}"
   rescue
@@ -129,7 +129,11 @@ file GENERATED_TOKENIZER => "lib/nokogiri/css/tokenizer.rex" do |t|
 end
 
 require 'tasks/test'
-require 'tasks/cross_compile'
+begin
+  require 'tasks/cross_compile' unless java
+rescue RuntimeError => e
+  warn "WARNING: Could not perform some cross-compiling: #{e}"
+end
 
 desc "set environment variables to build and/or test with debug options"
 task :debug do

data/ext/nokogiri/extconf.rb

@@ -17,7 +17,7 @@ end
 $CFLAGS << " #{ENV["CFLAGS"]}"
 $LIBS << " #{ENV["LIBS"]}"
 
-if Config::CONFIG['target_os'] == 'mingw32'
+if Config::CONFIG['target_os'] == 'mingw32' || Config::CONFIG['target_os'] =~ /mswin32/
   $CFLAGS << " -DXP_WIN -DXP_WIN32 -DUSE_INCLUDED_VASPRINTF"
 elsif Config::CONFIG['target_os'] =~ /solaris/
   $CFLAGS << " -DUSE_INCLUDED_VASPRINTF"
@@ -29,58 +29,81 @@ if RbConfig::MAKEFILE_CONFIG['CC'] =~ /mingw/
   $CFLAGS << " -DIN_LIBXML"
 end
 
-$CFLAGS << " -O3 -Wall -Wcast-qual -Wwrite-strings -Wconversion -Wmissing-noreturn -Winline"
+if RbConfig::MAKEFILE_CONFIG['CC'] =~ /gcc/
+  $CFLAGS << " -O3 -Wall -Wcast-qual -Wwrite-strings -Wconversion -Wmissing-noreturn -Winline"
+end
+
+if Config::CONFIG['target_os'] =~ /mswin32/
+  lib_prefix = 'lib'
+
+  # There's no default include/lib dir on Windows. Let's just add the Ruby ones
+  # and resort on the search path specified by INCLUDE and LIB environment
+  # variables
+  HEADER_DIRS = [INCLUDEDIR]
+  LIB_DIRS = [LIBDIR]
+  XML2_HEADER_DIRS = [File.join(INCLUDEDIR, "libxml2"), INCLUDEDIR]
+
+else
+  lib_prefix = ''
+
+  HEADER_DIRS = [
+    # First search /opt/local for macports
+    '/opt/local/include',
 
-HEADER_DIRS = [
-  # First search /opt/local for macports
-  '/opt/local/include',
+    # Then search /usr/local for people that installed from source
+    '/usr/local/include',
 
-  # Then search /usr/local for people that installed from source
-  '/usr/local/include',
+    # Check the ruby install locations
+    INCLUDEDIR,
 
-  # Check the ruby install locations
-  INCLUDEDIR,
+    # Finally fall back to /usr
+    '/usr/include',
+    '/usr/include/libxml2',
+  ]
 
-  # Finally fall back to /usr
-  '/usr/include',
-  '/usr/include/libxml2',
-]
+  LIB_DIRS = [
+    # First search /opt/local for macports
+    '/opt/local/lib',
 
-LIB_DIRS = [
-  # First search /opt/local for macports
-  '/opt/local/lib',
+    # Then search /usr/local for people that installed from source
+    '/usr/local/lib',
 
-  # Then search /usr/local for people that installed from source
-  '/usr/local/lib',
+    # Check the ruby install locations
+    LIBDIR,
 
-  # Check the ruby install locations
-  LIBDIR,
+    # Finally fall back to /usr
+    '/usr/lib',
+  ]
 
-  # Finally fall back to /usr
-  '/usr/lib',
-]
+  XML2_HEADER_DIRS = [
+    '/opt/local/include/libxml2',
+    '/usr/local/include/libxml2',
+    File.join(INCLUDEDIR, "libxml2")
+  ] + HEADER_DIRS
+end
 
 dir_config('zlib', HEADER_DIRS, LIB_DIRS)
 dir_config('iconv', HEADER_DIRS, LIB_DIRS)
-dir_config('xml2', [
-           '/opt/local/include/libxml2',
-           '/usr/local/include/libxml2',
-           File.join(INCLUDEDIR, "libxml2")] + HEADER_DIRS, LIB_DIRS)
+dir_config('xml2', XML2_HEADER_DIRS, LIB_DIRS)
 dir_config('xslt', HEADER_DIRS, LIB_DIRS)
 
 def asplode(lib)
   abort "-----\n#{lib} is missing.  please visit http://nokogiri.org/tutorials/installing_nokogiri.html for help with installing dependencies.\n-----"
 end
 
-asplode "iconv"    unless find_header('iconv.h')
+# Use this with cross compiling
+# PKG_CONFIG_PATH=/Users/apatterson/git/nokogiri/tmp/cross/lib/pkgconfig/ \
+#   rake cross compile RUBY_CC_VERSION=1.9.1
+pkg_config('libxslt') if RUBY_PLATFORM =~ /mingw/
+
 asplode "libxml2"  unless find_header('libxml/parser.h')
 asplode "libxslt"  unless find_header('libxslt/xslt.h')
 asplode "libexslt" unless find_header('libexslt/exslt.h')
-asplode "zlib"     unless find_library('z', 'gzopen')
-asplode "libxml2"  unless find_library('xml2', 'xmlParseDoc')
-asplode "libxslt"  unless find_library('xslt', 'xsltParseStylesheetDoc')
-asplode "libexslt" unless find_library('exslt', 'exsltFuncRegister')
+asplode "libxml2"  unless find_library("#{lib_prefix}xml2", 'xmlParseDoc')
+asplode "libxslt"  unless find_library("#{lib_prefix}xslt", 'xsltParseStylesheetDoc')
+asplode "libexslt" unless find_library("#{lib_prefix}exslt", 'exsltFuncRegister')
 
+have_func 'xmlFirstElementChild'
 have_func('xmlRelaxNGSetParserStructuredErrors')
 have_func('xmlRelaxNGSetParserStructuredErrors')
 have_func('xmlRelaxNGSetValidStructuredErrors')

data/ext/nokogiri/nokogiri.h

@@ -77,6 +77,8 @@ int is_2_6_16(void) ;
 #define RBSTR_OR_QNIL(_str) \
   (_str ? NOKOGIRI_STR_NEW2(_str) : Qnil)
 
+#include <xml_libxml2_hacks.h>
+
 #include <xml_io.h>
 #include <xml_document.h>
 #include <html_entity_lookup.h>

data/ext/nokogiri/xml_document.c

@@ -46,6 +46,11 @@ static void recursively_remove_namespaces_from_node(xmlNodePtr node)
 
   for (child = node->children ; child ; child = child->next)
     recursively_remove_namespaces_from_node(child);
+
+  if (node->nsDef) {
+    xmlFreeNsList(node->nsDef);
+    node->nsDef = NULL;
+  }
 }
 
 /*

data/ext/nokogiri/xml_libxml2_hacks.c

@@ -0,0 +1,112 @@
+#ifndef HAVE_XMLFIRSTELEMENTCHILD
+#include <libxml/tree.h>
+/**
+ * xmlFirstElementChild:
+ * @parent: the parent node
+ *
+ * Finds the first child node of that element which is a Element node
+ * Note the handling of entities references is different than in
+ * the W3C DOM element traversal spec since we don't have back reference
+ * from entities content to entities references.
+ *
+ * Returns the first element child or NULL if not available
+ */
+xmlNodePtr
+xmlFirstElementChild(xmlNodePtr parent) {
+    xmlNodePtr cur = NULL;
+
+    if (parent == NULL)
+        return(NULL);
+    switch (parent->type) {
+        case XML_ELEMENT_NODE:
+        case XML_ENTITY_NODE:
+        case XML_DOCUMENT_NODE:
+        case XML_HTML_DOCUMENT_NODE:
+            cur = parent->children;
+            break;
+        default:
+            return(NULL);
+    }
+    while (cur != NULL) {
+        if (cur->type == XML_ELEMENT_NODE)
+            return(cur);
+        cur = cur->next;
+    }
+    return(NULL);
+}
+
+/**
+ * xmlNextElementSibling:
+ * @node: the current node
+ *
+ * Finds the first closest next sibling of the node which is an
+ * element node.
+ * Note the handling of entities references is different than in
+ * the W3C DOM element traversal spec since we don't have back reference
+ * from entities content to entities references.
+ *
+ * Returns the next element sibling or NULL if not available
+ */
+xmlNodePtr
+xmlNextElementSibling(xmlNodePtr node) {
+    if (node == NULL)
+        return(NULL);
+    switch (node->type) {
+        case XML_ELEMENT_NODE:
+        case XML_TEXT_NODE:
+        case XML_CDATA_SECTION_NODE:
+        case XML_ENTITY_REF_NODE:
+        case XML_ENTITY_NODE:
+        case XML_PI_NODE:
+        case XML_COMMENT_NODE:
+        case XML_DTD_NODE:
+        case XML_XINCLUDE_START:
+        case XML_XINCLUDE_END:
+            node = node->next;
+            break;
+        default:
+            return(NULL);
+    }
+    while (node != NULL) {
+        if (node->type == XML_ELEMENT_NODE)
+            return(node);
+        node = node->next;
+    }
+    return(NULL);
+}
+
+/**
+ * xmlLastElementChild:
+ * @parent: the parent node
+ *
+ * Finds the last child node of that element which is a Element node
+ * Note the handling of entities references is different than in
+ * the W3C DOM element traversal spec since we don't have back reference
+ * from entities content to entities references.
+ *
+ * Returns the last element child or NULL if not available
+ */
+xmlNodePtr
+xmlLastElementChild(xmlNodePtr parent) {
+    xmlNodePtr cur = NULL;
+
+    if (parent == NULL)
+        return(NULL);
+    switch (parent->type) {
+        case XML_ELEMENT_NODE:
+        case XML_ENTITY_NODE:
+        case XML_DOCUMENT_NODE:
+        case XML_HTML_DOCUMENT_NODE:
+            cur = parent->last;
+            break;
+        default:
+            return(NULL);
+    }
+    while (cur != NULL) {
+        if (cur->type == XML_ELEMENT_NODE)
+            return(cur);
+        cur = cur->prev;
+    }
+    return(NULL);
+}
+#endif

data/ext/nokogiri/xml_libxml2_hacks.h

@@ -0,0 +1,12 @@
+#ifndef HAVE_XMLFIRSTELEMENTCHILD
+
+#ifndef XML_LIBXML2_HACKS
+#define XML_LIBXML2_HACKS
+
+xmlNodePtr xmlFirstElementChild(xmlNodePtr parent);
+xmlNodePtr xmlNextElementSibling(xmlNodePtr node);
+xmlNodePtr xmlLastElementChild(xmlNodePtr parent);
+
+#endif
+
+#endif

data/ext/nokogiri/xml_node.c

@@ -103,7 +103,7 @@ static xmlNodePtr xmlReplaceNodeWrapper(xmlNodePtr pivot, xmlNodePtr new_node)
 static VALUE reparent_node_with(VALUE pivot_obj, VALUE reparentee_obj, pivot_reparentee_func prf)
 {
   VALUE reparented_obj ;
-  xmlNodePtr reparentee, pivot, reparented ;
+  xmlNodePtr reparentee, pivot, reparented, next_text, new_next_text ;
 
   if(!rb_obj_is_kind_of(reparentee_obj, cNokogiriXmlNode))
     rb_raise(rb_eArgError, "node must be a Nokogiri::XML::Node");
@@ -141,6 +141,34 @@ static VALUE reparent_node_with(VALUE pivot_obj, VALUE reparentee_obj, pivot_rep
     }
   }
 
+  if (reparentee->type == XML_TEXT_NODE && pivot->next && pivot->next->type == XML_TEXT_NODE) {
+    /*
+     *  libxml merges text nodes in a right-to-left fashion, meaning that if
+     *  there are two text nodes who would be adjacent, the right (or following,
+     *  or next) node will be merged into the left (or preceding, or previous)
+     *  node.
+     *
+     *  and by "merged" I mean the string contents will be concatenated onto the
+     *  left node's contents, and then the node will be freed.
+     *
+     *  which means that if we have a ruby object wrapped around the right node,
+     *  its memory would be freed out from under it.
+     *
+     *  so, we detect this edge case and unlink-and-root the text node before it gets
+     *  merged. then we dup the node and insert that duplicate back into the
+     *  document where the real node was.
+     *
+     *  yes, this is totally lame.
+     */
+    next_text     = pivot->next ;
+    new_next_text = xmlDocCopyNode(next_text, pivot->doc, 1) ;
+
+    xmlUnlinkNode(next_text);
+    NOKOGIRI_ROOT_NODE(next_text);
+
+    xmlAddNextSibling(pivot, new_next_text);
+  }
+
   /* TODO: I really want to remove this.  We shouldn't support 2.6.16 anymore */
   if ( reparentee->type == XML_TEXT_NODE && pivot->type == XML_TEXT_NODE && is_2_6_16() ) {
     /* work around a string-handling bug in libxml 2.6.16. we'd rather leak than segfault. */
@@ -1021,26 +1049,32 @@ static VALUE line(VALUE self)
  */
 static VALUE add_namespace_definition(VALUE self, VALUE prefix, VALUE href)
 {
-  xmlNodePtr node;
+  xmlNodePtr node, namespacee;
   xmlNsPtr ns;
 
   Data_Get_Struct(self, xmlNode, node);
+  namespacee = node ;
 
-  ns = xmlNewNs(
+  ns = xmlSearchNs(
+      node->doc,
       node,
-      (const xmlChar *)StringValuePtr(href),
       (const xmlChar *)(NIL_P(prefix) ? NULL : StringValuePtr(prefix))
   );
 
   if(!ns) {
-    ns = xmlSearchNs(
-        node->doc,
-        node,
+    if (node->type != XML_ELEMENT_NODE) {
+      namespacee = node->parent;
+    }
+    ns = xmlNewNs(
+        namespacee,
+        (const xmlChar *)StringValuePtr(href),
         (const xmlChar *)(NIL_P(prefix) ? NULL : StringValuePtr(prefix))
     );
   }
 
-  if(NIL_P(prefix)) xmlSetNs(node, ns);
+  if (!ns) return Qnil ;
+
+  if(NIL_P(prefix) || node != namespacee) xmlSetNs(node, ns);
 
   return Nokogiri_wrap_xml_namespace(node->doc, ns);
 }
@@ -1122,13 +1156,11 @@ static VALUE in_context(VALUE self, VALUE _str, VALUE _options)
   xmlNodePtr node;
   xmlNodePtr list;
   xmlNodeSetPtr set;
+  xmlParserErrors error;
   VALUE doc, err;
 
   Data_Get_Struct(self, xmlNode, node);
 
-  if(!node->parent)
-    rb_raise(rb_eRuntimeError, "no contextual parsing on unlinked nodes");
-
   doc = DOC_RUBY_OBJECT(node->doc);
   err = rb_iv_get(doc, "@errors");
 
@@ -1141,7 +1173,7 @@ static VALUE in_context(VALUE self, VALUE _str, VALUE _options)
   htmlHandleOmittedElem(0);
 #endif
 
-  xmlParseInNodeContext(
+  error = xmlParseInNodeContext(
       node,
       StringValuePtr(_str),
       (int)RSTRING_LEN(_str),
@@ -1154,6 +1186,20 @@ static VALUE in_context(VALUE self, VALUE _str, VALUE _options)
 
   xmlSetStructuredErrorFunc(NULL, NULL);
 
+  /* FIXME: This probably needs to handle more constants... */
+  switch(error) {
+    case XML_ERR_OK:
+      break;
+
+    case XML_ERR_INTERNAL_ERROR:
+    case XML_ERR_NO_MEMORY:
+      rb_raise(rb_eRuntimeError, "error parsing fragment (%d)", error);
+      break;
+
+    default:
+      break;
+  }
+
   set = xmlXPathNodeSetCreate(NULL);
 
   while(list) {