mirror of
https://github.com/ruby/ruby.git
synced 2022-11-09 12:17:21 -05:00
* lib/rexml/encoding.rb (REXML::Encoding#encoding=): store @encoding
a String which means the name of the encoding. this partially revert r29646. * lib/rexml/document.rb: follow above. * lib/rexml/output.rb: ditto. * lib/rexml/parsers/baseparser.rb: ditto. * lib/rexml/source.rb: ditto. * lib/rexml/xmldecl.rb: ditto. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@31008 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
This commit is contained in:
parent
cddcffb8f9
commit
f25ff846f6
11 changed files with 114 additions and 147 deletions
|
@ -1,8 +1,9 @@
|
|||
module REXML
|
||||
module Encoding
|
||||
# ID ---> Encoding object
|
||||
# ID ---> Encoding name
|
||||
attr_reader :encoding
|
||||
def encoding=(encoding)
|
||||
encoding = encoding.name if encoding.is_a?(Encoding)
|
||||
if encoding.is_a?(String)
|
||||
original_encoding = encoding
|
||||
encoding = find_encoding(encoding)
|
||||
|
@ -11,35 +12,25 @@ module REXML
|
|||
end
|
||||
end
|
||||
return false if defined?(@encoding) and encoding == @encoding
|
||||
if encoding and encoding != ::Encoding::UTF_8
|
||||
@encoding = encoding
|
||||
if encoding
|
||||
@encoding = encoding.upcase
|
||||
else
|
||||
@encoding = ::Encoding::UTF_8
|
||||
@encoding = 'UTF-8'
|
||||
end
|
||||
true
|
||||
end
|
||||
|
||||
def check_encoding(xml)
|
||||
# We have to recognize UTF-16, LSB UTF-16, and UTF-8
|
||||
# We have to recognize UTF-16BE, UTF-16LE, and UTF-8
|
||||
if xml[0, 2] == "\xfe\xff"
|
||||
xml[0, 2] = ""
|
||||
::Encoding::UTF_16BE
|
||||
return 'UTF-16BE'
|
||||
elsif xml[0, 2] == "\xff\xfe"
|
||||
xml[0, 2] = ""
|
||||
::Encoding::UTF_16LE
|
||||
else
|
||||
if /\A\s*<\?xml\s+version\s*=\s*(['"]).*?\1
|
||||
\s+encoding\s*=\s*(["'])(.*?)\2/mx =~ xml
|
||||
encoding_name = $3
|
||||
if /\Autf-16\z/i =~ encoding_name
|
||||
::Encoding::UTF_16BE
|
||||
else
|
||||
find_encoding(encoding_name)
|
||||
end
|
||||
else
|
||||
::Encoding::UTF_8
|
||||
end
|
||||
return 'UTF-16LE'
|
||||
end
|
||||
xml =~ /^\s*<\?xml\s+version\s*=\s*(['"]).*?\1\s+encoding\s*=\s*(["'])(.*?)\2/m
|
||||
return $3 ? $3.upcase : 'UTF-8'
|
||||
end
|
||||
|
||||
def encode(string)
|
||||
|
@ -53,14 +44,19 @@ module REXML
|
|||
private
|
||||
def find_encoding(name)
|
||||
case name
|
||||
when "UTF-16"
|
||||
name = "UTF-16BE"
|
||||
when /\Ashift-jis\z/i
|
||||
name = "Shift_JIS"
|
||||
return "SHIFT_JIS"
|
||||
when /\ACP-(\d+)\z/
|
||||
name = "CP#{$1}"
|
||||
when /\AUTF-8\z/i
|
||||
return name
|
||||
end
|
||||
::Encoding.find(name)
|
||||
begin
|
||||
::Encoding::Converter.search_convpath(name, 'UTF-8')
|
||||
rescue ::Encoding::ConverterNotFoundError
|
||||
return nil
|
||||
end
|
||||
name
|
||||
end
|
||||
end
|
||||
end
|
||||
|
|
Loading…
Add table
Add a link
Reference in a new issue