diff options
author | drbrain <drbrain@b2dd03c8-39d4-4d8f-98ff-823fe69b080e> | 2010-12-20 03:22:49 +0000 |
---|---|---|
committer | drbrain <drbrain@b2dd03c8-39d4-4d8f-98ff-823fe69b080e> | 2010-12-20 03:22:49 +0000 |
commit | 2ef9c50c6e405717d06362787c4549ca4f1c6485 (patch) | |
tree | ee99486567461dd5796f3d6edcc9e204187f2666 /lib/rdoc/encoding.rb | |
parent | d7effd506f5b91a636f2e6452ef1946b923007c7 (diff) | |
download | ruby-2ef9c50c6e405717d06362787c4549ca4f1c6485.tar.gz |
Import RDoc 3
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@30249 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
Diffstat (limited to 'lib/rdoc/encoding.rb')
-rw-r--r-- | lib/rdoc/encoding.rb | 79 |
1 files changed, 79 insertions, 0 deletions
diff --git a/lib/rdoc/encoding.rb b/lib/rdoc/encoding.rb new file mode 100644 index 0000000000..4f0779881c --- /dev/null +++ b/lib/rdoc/encoding.rb @@ -0,0 +1,79 @@ +require 'rdoc' + +## +# This class is a wrapper around File IO and Encoding that helps RDoc load +# files and convert them to the correct encoding. + +module RDoc::Encoding + + ## + # Reads the contents of +filename+ and handles any encoding directives in + # the file. + # + # The content will be converted to the +encoding+. If the file cannot be + # converted a warning will be printed and nil will be returned. + + def self.read_file filename, encoding + content = open filename, "rb" do |f| f.read end + + utf8 = content.sub!(/\A\xef\xbb\xbf/, '') + + RDoc::Encoding.set_encoding content + + if Object.const_defined? :Encoding then + encoding ||= Encoding.default_external + orig_encoding = content.encoding + + if utf8 then + content.force_encoding Encoding::UTF_8 + content.encode! encoding + else + # assume the content is in our output encoding + content.force_encoding encoding + end + + unless content.valid_encoding? then + # revert and try to transcode + content.force_encoding orig_encoding + content.encode! encoding + end + + unless content.valid_encoding? then + warn "unable to convert #{filename} to #{encoding}, skipping" + content = nil + end + end + + content + rescue ArgumentError => e + raise unless e.message =~ /unknown encoding name - (.*)/ + warn "unknown encoding name \"#{$1}\" for #{filename}, skipping" + nil + rescue Encoding::UndefinedConversionError => e + warn "unable to convert #{e.message} for #{filename}, skipping" + nil + rescue Errno::EISDIR, Errno::ENOENT + nil + end + + ## + # Sets the encoding of +string+ based on the magic comment + + def self.set_encoding string + return unless Object.const_defined? :Encoding + + first_line = string[/\A(?:#!.*\n)?.*\n/] + + name = case first_line + when /^<\?xml[^?]*encoding=(["'])(.*?)\1/ then $2 + when /\b(?:en)?coding[=:]\s*([^\s;]+)/i then $1 + else return + end + + enc = Encoding.find name + string.force_encoding enc if enc + end + +end + + |