Is there any gem which adds # encoding: UTF-8 to each Ruby file automatically?
Or is there any other way to prevent from the invalid multibyte char (US-ASCII) error in the entire Ruby on Rails project (not in a single class only)?
-
3This isn't what you asked for, but for what it's worth some text editors (e.g. emacs) automatically insert "#encoding: UTF-8" at the top when you save a ruby file containing UTF-8.sepp2k– sepp2k2011-01-26 13:02:46 +00:00Commented Jan 26, 2011 at 13:02
Add a comment
|
6 Answers
Upgrade to Ruby 2.0, as it makes UTF-8 the default encoding, removing the need for magic comments.
1 Comment
Automatico
@Subimage I believe most legacy code should work in the newer versions of ruby, the other way is harder.
Try magic_encoding gem, it can insert uft-8 magic comment to all ruby files in your app.
[EDIT] Having switched to SublimeText now I use auto-encoding-for-ruby plugin.
4 Comments
marcus erronius
note that this won't be a dependency, it's a tool that edits all the files for you. you can uninstall afterwords if you want
Zakwan
This is Atom version for auto-encoding-for-ruby
Magne
@Mirko Did you get the auto-encoding-for-ruby plugin to work with Sublime Text 3 ?
Magne
@Mirko Hm, I'm also on ruby 2.x, but I was having problems with running certain scripts in the irb console for some reason.
Vim:
:args **/*.ruby
:set hidden
:argdo norm! O# encoding: UTF-8
:wqa
2 Comments
William Denniss
Thanks! Great suggestion. For me, vim crashed (segfault) on the
:wqa command during the write, resulting in some written files and a bunch of .swp. So I ended up using :wa, then :q which worked fine. Of course, change .ruby to .rb if the latter is your ruby extension.
Benoit
@William Denniss: Another one:
argdo 0put ='#encoding: UTF-8'If you're using Sublime Text 2, you can use a plugin that automatically includes encoding declaration when needed: https://github.com/elomarns/auto-encoding-for-ruby.
Comments
How about just running a script?
#!/usr/bin/env ruby1.9.1
require 'find'
fixfile = []
Find.find('.') do |path|
next unless /\.rb$/.match(path);
File.open(path) do |file|
count = 0;
type = :lib
file.each do |line|
if count == 0 and /#!/.match(line)
type = :script
end
if /utf/.match(line)
break
end
if (count += 1) > 10 then
fixfile.push path:path, type:type
break
end
end
if file.eof?
fixfile.push path:path, type:type
end
end
end
fixfile.each do |info|
path = info[:path]
backuppath = path + '~'
type = info[:type]
begin
File.delete(backuppath) if File.exist?(backuppath)
File.link(path, backuppath)
rescue Errno::ENOENT => x
puts "could not make backup file '#{backuppath}' for '#{ path }': #{$!}"
raise
end
begin
inputfile = File.open(backuppath, 'r')
File.unlink(path)
File.open(path, 'w') do |outputfile|
if type == :script
line = inputfile.readline
outputfile.write line
end
outputfile.write "# encoding: utf-8\n"
inputfile.each do |line|
outputfile.write line
end
inputfile.close
outputfile.close
end
rescue => x
puts "error: #{x} #{$!}"
exit
end
To make it automatic add this to your Rakefile.
You could run file -bi #{path} and look for charset=utf-8 if you only want to update files that have utf-8 chars.
Comments
Adding a # encoding: UTF-8 to each Ruby file automatically makes only sense, when your files are really stored in UTF-8.
If your files are encoded CP850 (AFAIK default in Windows) and you use Non-ASCII characters, you replace invalid multibyte char (US-ASCII) with invalid multibyte char (UTF-8).
I would prefer a manual modification and check of each file, if it is really UTF-8.
