rexml 3.2.6 → 3.4.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 2583ae302aa5e698f0887a689c416e5debe0533ac472a9f96fce6a8912040fd8
-  data.tar.gz: b0ffa6301fd899969a78e060ccaeafebfc2169e3c63ff499ebc6170468866475
+  metadata.gz: 582bb5339257c81f2ce9c076155c01d7adfe8fb169c09bc7f5f489f6a76bca80
+  data.tar.gz: 160de8899d8d1f995bafca23631e9e4ab928ebbffa21684e3b61dad805a6187b
 SHA512:
-  metadata.gz: f63fb0b84ef51e790cc6310244f2106d8c47ec9a00687c58c743afda82b60be9986d503c6f56f947db06f6758707facccd03405c4d1009376e856080aa26d0e4
-  data.tar.gz: db62bea7391837a7ab4cfc5cb5a412ed4deb8d232653ca66d93a323a5a76383eed520cd4ced5b20204f29b04e84678791cd6f807195868f5d4a5e519a73d2aaf
+  metadata.gz: e2b095792523f54301e8a6af2f1682a9ad24d92cdd5d94c9e6088b27520e3c03b68fe06061b6ff2fd96b001b9cb947c57e4095244d83206a83fc2a1829dd4243
+  data.tar.gz: 4f335d2b1e58c1da233c3f0a0588def502c8cb2660633e0e06b4d0930bbcedcaae36b52dc550923704b4525d94a1011f4b5f4e87a81e5d689cce24ee89210a23

data/NEWS.md

@@ -1,5 +1,404 @@
 # News
 
+## 3.4.0 - 2024-12-15 {#version-3-4-0}
+
+### Improvement
+
+  * Improved performance.
+    * GH-216
+    * Patch by NAITOH Jun
+
+  * JRuby: Improved parse performance.
+    * GH-219
+    * Patch by João Duarte
+
+  * Added support for reusing pull parser.
+    * GH-214
+    * GH-220
+    * Patch by Dmitry Pogrebnoy
+
+  * Improved error handling when source is `IO`.
+    * GH-221
+    * Patch by NAITOH Jun
+
+### Thanks
+
+  * NAITOH Jun
+
+  * João Duarte
+
+  * Dmitry Pogrebnoy
+
+## 3.3.9 - 2024-10-24 {#version-3-3-9}
+
+### Improvements
+
+  * Improved performance.
+    * GH-210
+    * Patch by NAITOH Jun.
+
+### Fixes
+
+  * Fixed a parse bug for text only invalid XML.
+    * GH-215
+    * Patch by NAITOH Jun.
+
+  * Fixed a parse bug that `&#0x...;` is accepted as a character
+    reference.
+
+### Thanks
+
+  * NAITOH Jun
+
+## 3.3.8 - 2024-09-29 {#version-3-3-8}
+
+### Improvements
+
+  * SAX2: Improve parse performance.
+    * GH-207
+    * Patch by NAITOH Jun.
+
+### Fixes
+
+  * Fixed a bug that unexpected attribute namespace conflict error for
+    the predefined "xml" namespace is reported.
+    * GH-208
+    * Patch by KITAITI Makoto
+
+### Thanks
+
+  * NAITOH Jun
+
+  * KITAITI Makoto
+
+## 3.3.7 - 2024-09-04 {#version-3-3-7}
+
+### Improvements
+
+  * Added local entity expansion limit methods
+    * GH-192
+    * GH-202
+    * Reported by takuya kodama.
+    * Patch by NAITOH Jun.
+
+  * Removed explicit strscan dependency
+    * GH-204
+    * Patch by Bo Anderson.
+
+### Thanks
+
+  * takuya kodama
+
+  * NAITOH Jun
+
+  * Bo Anderson
+
+## 3.3.6 - 2024-08-22 {#version-3-3-6}
+
+### Improvements
+
+  * Removed duplicated entity expansions for performance.
+    * GH-194
+    * Patch by Viktor Ivarsson.
+
+  * Improved namespace conflicted attribute check performance. It was
+    too slow for deep elements.
+    * Reported by l33thaxor.
+
+### Fixes
+
+  * Fixed a bug that default entity expansions are counted for
+    security check. Default entity expansions should not be counted
+    because they don't have a security risk.
+    * GH-198
+    * GH-199
+    * Patch Viktor Ivarsson
+
+  * Fixed a parser bug that parameter entity references in internal
+    subsets are expanded. It's not allowed in the XML specification.
+    * GH-191
+    * Patch by NAITOH Jun.
+
+  * Fixed a stream parser bug that user-defined entity references in
+    text aren't expanded.
+    * GH-200
+    * Patch by NAITOH Jun.
+
+### Thanks
+
+  * Viktor Ivarsson
+
+  * NAITOH Jun
+
+  * l33thaxor
+
+## 3.3.5 - 2024-08-12 {#version-3-3-5}
+
+### Fixes
+
+  * Fixed a bug that `REXML::Security.entity_expansion_text_limit`
+    check has wrong text size calculation in SAX and pull parsers.
+    * GH-193
+    * GH-195
+    * Reported by Viktor Ivarsson.
+    * Patch by NAITOH Jun.
+
+### Thanks
+
+  * Viktor Ivarsson
+
+  * NAITOH Jun
+
+## 3.3.4 - 2024-08-01 {#version-3-3-4}
+
+### Fixes
+
+  * Fixed a bug that `REXML::Security` isn't defined when
+    `REXML::Parsers::StreamParser` is used and
+    `rexml/parsers/streamparser` is only required.
+    * GH-189
+    * Patch by takuya kodama.
+
+### Thanks
+
+  * takuya kodama
+
+## 3.3.3 - 2024-08-01 {#version-3-3-3}
+
+### Improvements
+
+  * Added support for detecting invalid XML that has unsupported
+    content before root element
+    * GH-184
+    * Patch by NAITOH Jun.
+
+  * Added support for `REXML::Security.entity_expansion_limit=` and
+    `REXML::Security.entity_expansion_text_limit=` in SAX2 and pull
+    parsers
+    * GH-187
+    * Patch by NAITOH Jun.
+
+  * Added more tests for invalid XMLs.
+    * GH-183
+    * Patch by Watson.
+
+  * Added more performance tests.
+    * Patch by Watson.
+
+  * Improved parse performance.
+    * GH-186
+    * Patch by tomoya ishida.
+
+### Thanks
+
+  * NAITOH Jun
+
+  * Watson
+
+  * tomoya ishida
+
+## 3.3.2 - 2024-07-16 {#version-3-3-2}
+
+### Improvements
+
+  * Improved parse performance.
+    * GH-160
+    * Patch by NAITOH Jun.
+
+  * Improved parse performance.
+    * GH-169
+    * GH-170
+    * GH-171
+    * GH-172
+    * GH-173
+    * GH-174
+    * GH-175
+    * GH-176
+    * GH-177
+    * Patch by Watson.
+
+  * Added support for raising a parse exception when an XML has extra
+    content after the root element.
+    * GH-161
+    * Patch by NAITOH Jun.
+
+  * Added support for raising a parse exception when an XML
+    declaration exists in wrong position.
+    * GH-162
+    * Patch by NAITOH Jun.
+
+  * Removed needless a space after XML declaration in pretty print mode.
+    * GH-164
+    * Patch by NAITOH Jun.
+
+  * Stopped to emit `:text` event after the root element.
+    * GH-167
+    * Patch by NAITOH Jun.
+
+### Fixes
+
+  * Fixed a bug that SAX2 parser doesn't expand predefined entities for
+    `characters` callback.
+    * GH-168
+    * Patch by NAITOH Jun.
+
+### Thanks
+
+  * NAITOH Jun
+
+  * Watson
+
+## 3.3.1 - 2024-06-25 {#version-3-3-1}
+
+### Improvements
+
+  * Added support for detecting malformed top-level comments.
+    * GH-145
+    * Patch by Hiroya Fujinami.
+
+  * Improved `REXML::Element#attribute` performance.
+    * GH-146
+    * Patch by Hiroya Fujinami.
+
+  * Added support for detecting malformed `<!-->` comments.
+    * GH-147
+    * Patch by Hiroya Fujinami.
+
+  * Added support for detecting unclosed `DOCTYPE`.
+    * GH-152
+    * Patch by Hiroya Fujinami.
+
+  * Added `changlog_uri` metadata to gemspec.
+    * GH-156
+    * Patch by fynsta.
+
+  * Improved parse performance.
+    * GH-157
+    * GH-158
+    * Patch by NAITOH Jun.
+
+### Fixes
+
+  * Fixed a bug that large XML can't be parsed.
+    * GH-154
+    * Patch by NAITOH Jun.
+
+  * Fixed a bug that private constants are visible.
+    * GH-155
+    * Patch by NAITOH Jun.
+
+### Thanks
+
+  * Hiroya Fujinami
+
+  * NAITOH Jun
+
+  * fynsta
+
+## 3.3.0 - 2024-06-11 {#version-3-3-0}
+
+### Improvements
+
+  * Added support for strscan 0.7.0 installed with Ruby 2.6.
+    * GH-142
+    * Reported by Fernando Trigoso.
+
+### Thanks
+
+  * Fernando Trigoso
+
+## 3.2.9 - 2024-06-09 {#version-3-2-9}
+
+### Improvements
+
+  * Added support for old strscan.
+    * GH-132
+    * Reported by Adam.
+
+  * Improved attribute value parse performance.
+    * GH-135
+    * Patch by NAITOH Jun.
+
+  * Improved `REXML::Node#each_recursive` performance.
+    * GH-134
+    * GH-139
+    * Patch by Hiroya Fujinami.
+
+  * Improved text parse performance.
+    * Reported by mprogrammer.
+
+### Thanks
+
+  * Adam
+  * NAITOH Jun
+  * Hiroya Fujinami
+  * mprogrammer
+
+## 3.2.8 - 2024-05-16 {#version-3-2-8}
+
+### Fixes
+
+  * Suppressed a warning
+
+## 3.2.7 - 2024-05-16 {#version-3-2-7}
+
+### Improvements
+
+  * Improve parse performance by using `StringScanner`.
+
+    * GH-106
+    * GH-107
+    * GH-108
+    * GH-109
+    * GH-112
+    * GH-113
+    * GH-114
+    * GH-115
+    * GH-116
+    * GH-117
+    * GH-118
+    * GH-119
+    * GH-121
+
+    * Patch by NAITOH Jun.
+
+  * Improved parse performance when an attribute has many `<`s.
+
+    * GH-126
+
+### Fixes
+
+  * XPath: Fixed a bug of `normalize_space(array)`.
+
+    * GH-110
+    * GH-111
+
+    * Patch by flatisland.
+
+  * XPath: Fixed a bug that wrong position is used with nested path.
+
+    * GH-110
+    * GH-122
+
+    * Reported by jcavalieri.
+    * Patch by NAITOH Jun.
+
+  * Fixed a bug that an exception message can't be generated for
+    invalid encoding XML.
+
+    * GH-29
+    * GH-123
+
+    * Reported by DuKewu.
+    * Patch by NAITOH Jun.
+
+### Thanks
+
+  * NAITOH Jun
+  * flatisland
+  * jcavalieri
+  * DuKewu
+
 ## 3.2.6 - 2023-07-27 {#version-3-2-6}
 
 ### Improvements

data/lib/rexml/attribute.rb

@@ -148,8 +148,9 @@ module REXML
     # have been expanded to their values
     def value
       return @unnormalized if @unnormalized
-      @unnormalized = Text::unnormalize( @normalized, doctype )
-      @unnormalized
+
+      @unnormalized = Text::unnormalize(@normalized, doctype,
+                                        entity_expansion_text_limit: @element&.document&.entity_expansion_text_limit)
     end
 
     # The normalized value of this attribute.  That is, the attribute with

data/lib/rexml/document.rb

@@ -91,6 +91,8 @@ module REXML
     #
     def initialize( source = nil, context = {} )
       @entity_expansion_count = 0
+      @entity_expansion_limit = Security.entity_expansion_limit
+      @entity_expansion_text_limit = Security.entity_expansion_text_limit
       super()
       @context = context
       return if source.nil?
@@ -431,10 +433,12 @@ module REXML
     end
 
     attr_reader :entity_expansion_count
+    attr_writer :entity_expansion_limit
+    attr_accessor :entity_expansion_text_limit
 
     def record_entity_expansion
       @entity_expansion_count += 1
-      if @entity_expansion_count > Security.entity_expansion_limit
+      if @entity_expansion_count > @entity_expansion_limit
         raise "number of entity expansions exceeded, processing aborted."
       end
     end

data/lib/rexml/element.rb

@@ -7,14 +7,6 @@ require_relative "xpath"
 require_relative "parseexception"
 
 module REXML
-  # An implementation note about namespaces:
-  # As we parse, when we find namespaces we put them in a hash and assign
-  # them a unique ID.  We then convert the namespace prefix for the node
-  # to the unique ID.  This makes namespace lookup much faster for the
-  # cost of extra memory use.  We save the namespace prefix for the
-  # context node and convert it back when we write it.
-  @@namespaces = {}
-
   # An \REXML::Element object represents an XML element.
   #
   # An element:
@@ -449,9 +441,14 @@ module REXML
     # Related: #root_node, #document.
     #
     def root
-      return elements[1] if self.kind_of? Document
-      return self if parent.kind_of? Document or parent.nil?
-      return parent.root
+      target = self
+      while target
+        return target.elements[1] if target.kind_of? Document
+        parent = target.parent
+        return target if parent.kind_of? Document or parent.nil?
+        target = parent
+      end
+      nil
     end
 
     # :call-seq:
@@ -627,8 +624,12 @@ module REXML
       else
         prefix = "xmlns:#{prefix}" unless prefix[0,5] == 'xmlns'
       end
-      ns = attributes[ prefix ]
-      ns = parent.namespace(prefix) if ns.nil? and parent
+      ns = nil
+      target = self
+      while ns.nil? and target
+        ns = target.attributes[prefix]
+        target = target.parent
+      end
       ns = '' if ns.nil? and prefix == 'xmlns'
       return ns
     end
@@ -1284,16 +1285,11 @@ module REXML
     #   document.root.attribute("x", "a") # => a:x='a:x'
     #
     def attribute( name, namespace=nil )
-      prefix = nil
-      if namespaces.respond_to? :key
-        prefix = namespaces.key(namespace) if namespace
-      else
-        prefix = namespaces.index(namespace) if namespace
-      end
+      prefix = namespaces.key(namespace) if namespace
       prefix = nil if prefix == 'xmlns'
 
       ret_val =
-        attributes.get_attribute( "#{prefix ? prefix + ':' : ''}#{name}" )
+        attributes.get_attribute( prefix ? "#{prefix}:#{name}" : name )
 
       return ret_val unless ret_val.nil?
       return nil if prefix.nil?
@@ -2388,17 +2384,6 @@ module REXML
       elsif old_attr.kind_of? Hash
         old_attr[value.prefix] = value
       elsif old_attr.prefix != value.prefix
-        # Check for conflicting namespaces
-        if value.prefix != "xmlns" and old_attr.prefix != "xmlns"
-          old_namespace = old_attr.namespace
-          new_namespace = value.namespace
-          if old_namespace == new_namespace
-            raise ParseException.new(
-                    "Namespace conflict in adding attribute \"#{value.name}\": "+
-                    "Prefix \"#{old_attr.prefix}\" = \"#{old_namespace}\" and "+
-                    "prefix \"#{value.prefix}\" = \"#{new_namespace}\"")
-          end
-        end
         store value.name, {old_attr.prefix => old_attr,
                            value.prefix    => value}
       else

data/lib/rexml/entity.rb

@@ -12,6 +12,7 @@ module REXML
     EXTERNALID = "(?:(?:(SYSTEM)\\s+#{SYSTEMLITERAL})|(?:(PUBLIC)\\s+#{PUBIDLITERAL}\\s+#{SYSTEMLITERAL}))"
     NDATADECL = "\\s+NDATA\\s+#{NAME}"
     PEREFERENCE = "%#{NAME};"
+    PEREFERENCE_RE = /#{PEREFERENCE}/um
     ENTITYVALUE = %Q{((?:"(?:[^%&"]|#{PEREFERENCE}|#{REFERENCE})*")|(?:'([^%&']|#{PEREFERENCE}|#{REFERENCE})*'))}
     PEDEF = "(?:#{ENTITYVALUE}|#{EXTERNALID})"
     ENTITYDEF = "(?:#{ENTITYVALUE}|(?:#{EXTERNALID}(#{NDATADECL})?))"
@@ -19,7 +20,7 @@ module REXML
     GEDECL = "<!ENTITY\\s+#{NAME}\\s+#{ENTITYDEF}\\s*>"
     ENTITYDECL = /\s*(?:#{GEDECL})|(?:#{PEDECL})/um
 
-    attr_reader :name, :external, :ref, :ndata, :pubid
+    attr_reader :name, :external, :ref, :ndata, :pubid, :value
 
     # Create a new entity.  Simple entities can be constructed by passing a
     # name, value to the constructor; this creates a generic, plain entity
@@ -68,14 +69,14 @@ module REXML
     end
 
     # Evaluates to the unnormalized value of this entity; that is, replacing
-    # all entities -- both %ent; and &ent; entities.  This differs from
-    # +value()+ in that +value+ only replaces %ent; entities.
+    # &ent; entities.
     def unnormalized
-      document.record_entity_expansion unless document.nil?
-      v = value()
-      return nil if v.nil?
-      @unnormalized = Text::unnormalize(v, parent)
-      @unnormalized
+      document&.record_entity_expansion
+
+      return nil if @value.nil?
+
+      @unnormalized = Text::unnormalize(@value, parent,
+                                        entity_expansion_text_limit: document&.entity_expansion_text_limit)
     end
 
     #once :unnormalized
@@ -121,46 +122,6 @@ module REXML
       write rv
       rv
     end
-
-    PEREFERENCE_RE = /#{PEREFERENCE}/um
-    # Returns the value of this entity.  At the moment, only internal entities
-    # are processed.  If the value contains internal references (IE,
-    # %blah;), those are replaced with their values.  IE, if the doctype
-    # contains:
-    #  <!ENTITY % foo "bar">
-    #  <!ENTITY yada "nanoo %foo; nanoo>
-    # then:
-    #  doctype.entity('yada').value   #-> "nanoo bar nanoo"
-    def value
-      @resolved_value ||= resolve_value
-    end
-
-    def parent=(other)
-      @resolved_value = nil
-      super
-    end
-
-    private
-    def resolve_value
-      return nil if @value.nil?
-      return @value unless @value.match?(PEREFERENCE_RE)
-
-      matches = @value.scan(PEREFERENCE_RE)
-      rv = @value.clone
-      if @parent
-        sum = 0
-        matches.each do |entity_reference|
-          entity_value = @parent.entity( entity_reference[0] )
-          if sum + entity_value.bytesize > Security.entity_expansion_text_limit
-            raise "entity expansion has grown too large"
-          else
-            sum += entity_value.bytesize
-          end
-          rv.gsub!( /%#{entity_reference.join};/um, entity_value )
-        end
-      end
-      rv
-    end
   end
 
   # This is a set of entity constants -- the ones defined in the XML

data/lib/rexml/formatters/pretty.rb

@@ -111,7 +111,7 @@ module REXML
         # itself, then we don't need a carriage return... which makes this
         # logic more complex.
         node.children.each { |child|
-          next if child == node.children[-1] and child.instance_of?(Text)
+          next if child.instance_of?(Text)
           unless child == node.children[0] or child.instance_of?(Text) or
             (child == node.children[1] and !node.children[0].writethis)
             output << "\n"

data/lib/rexml/functions.rb

@@ -262,11 +262,10 @@ module REXML
       string(string).length
     end
 
-    # UNTESTED
     def Functions::normalize_space( string=nil )
       string = string(@@context[:node]) if string.nil?
       if string.kind_of? Array
-        string.collect{|x| string.to_s.strip.gsub(/\s+/um, ' ') if string}
+        string.collect{|x| x.to_s.strip.gsub(/\s+/um, ' ') if x}
       else
         string.to_s.strip.gsub(/\s+/um, ' ')
       end

data/lib/rexml/node.rb

@@ -52,10 +52,14 @@ module REXML
 
     # Visit all subnodes of +self+ recursively
     def each_recursive(&block) # :yields: node
-      self.elements.each {|node|
-        block.call(node)
-        node.each_recursive(&block)
-      }
+      stack = []
+      each { |child| stack.unshift child if child.node_type == :element }
+      until stack.empty?
+        child = stack.pop
+        yield child
+        n = stack.size
+        child.each { |grandchild| stack.insert n, grandchild if grandchild.node_type == :element }
+      end
     end
 
     # Find (and return) first subnode (recursively) for which the block

data/lib/rexml/parseexception.rb

@@ -29,6 +29,7 @@ module REXML
         err << "\nLine: #{line}\n"
         err << "Position: #{position}\n"
         err << "Last 80 unconsumed characters:\n"
+        err.force_encoding("ASCII-8BIT")
         err << @source.buffer[0..80].force_encoding("ASCII-8BIT").gsub(/\n/, ' ')
       end