Even when processing a valid file (bulgarian), I am getting this error:

Processing file: "Localizable.strings"

/usr/local/bin/rubustrings:24:in gsub': invalid byte sequence in UTF-8 (ArgumentError) from /usr/local/bin/rubustrings:24:inremove_comments_and_empty_lines' from /usr/local/bin/rubustrings:80:in validate_localizable_string_file' from /usr/local/bin/rubustrings:94:inblock in

Devanshus:kplms devanshus$ sudo rubustrings Localizable.strings /System/Library/Frameworks/Ruby.framework/Versions/2.3/usr/lib/ruby/2.3.0/rubygems/core_ext/kernel_require.rb:55:in require': cannot load such file -- colored (LoadError) from /System/Library/Frameworks/Ruby.framework/Versions/2.3/usr/lib/ruby/2.3.0/rubygems/core_ext/kernel_require.rb:55:inrequire' from /Library/Ruby/Gems/2.3.0/gems/rubustrings-0.0.3/lib/rubustrings/action.rb:2:in <top (required)>' from /System/Library/Frameworks/Ruby.framework/Versions/2.3/usr/lib/ruby/2.3.0/rubygems/core_ext/kernel_require.rb:55:inrequire' from /System/Library/Frameworks/Ruby.framework/Versions/2.3/usr/lib/ruby/2.3.0/rubygems/core_ext/kernel_require.rb:55:in require' from /Library/Ruby/Gems/2.3.0/gems/rubustrings-0.0.3/lib/rubustrings.rb:2:in<top (required)>' from /System/Library/Frameworks/Ruby.framework/Versions/2.3/usr/lib/ruby/2.3.0/rubygems/core_ext/kernel_require.rb:127:in require' from /System/Library/Frameworks/Ruby.framework/Versions/2.3/usr/lib/ruby/2.3.0/rubygems/core_ext/kernel_require.rb:127:inrescue in require' from /System/Library/Frameworks/Ruby.framework/Versions/2.3/usr/lib/ruby/2.3.0/rubygems/core_ext/kernel_require.rb:40:in require' from /Library/Ruby/Gems/2.3.0/gems/rubustrings-0.0.3/bin/rubustrings:3:in<top (required)>' from /usr/local/bin/rubustrings:22:in load' from /usr/local/bin/rubustrings:22:in

The validation incorrectly fails when the argument position in the translation string is changed like so:

"%@, %@" = "%2$@ %1$@";

Example: en: "1, Infinite Loop" de: "Infinite Loop 1"

I just tried to fix this, but I'm a complete regex dead-loss :(

Correct issue #4 by expanding the validate_special_characters method. Now it will extract the parameter, length, and type fields from the format strings and does basic checking for format strings with a specified parameter. So something like

"%@, %d" = "%2$d %1$@";

will now be considered valid. It also checks to make sure that there are no specified parameter arguments in the key, and that if a parameter argument is specified somewhere in the value then it's used everywhere.

I included a stable sort to handle the case where no parameters are used and I still needed to preserve their original ordering.

There is a slight conflict with this pull request and the other I made with regards to the tests: the included tests pass on the master branch but with the xcode changes on the other pull request there is a slight change around the way color strings are generated (specifically the coloring formatting surrounding the newline characters, which doesn't affect the visible output but still causes a test fail). For whichever pull request gets included first I'll make sure to update the other to fix the issue.

I tried to run it over my file and it says:

./rubustrings.rb:17:inread': ""d" on UTF-16 (Encoding::InvalidByteSequenceError) from ./rubustrings.rb:17:in open_and_read_file' from ./rubustrings.rb:79:invalidate_localizable_string_file' from ./rubustrings.rb:94:in block in <main>' from ./rubustrings.rb:92:ineach' from ./rubustrings.rb:92:in <main>'

Support UTF-8 in addition to UTF-16. Apple documentation suggests UTF-16 is preferred for strings files, but I know a lot of projects use UTF-8 (including Apple's own ResearchKit)

This change adds a new -xcode flag that could be used to generate xcode-friendly output. This means that if rubustrings is used with this flag as part of a custom build rule or build phase shell script, it will let xcode know the file and line with the error, allowing the developer to double-click on the error to see it in the Xcode file editor.

It does make the code not quite as elegant: this is also my first time using Ruby so I'm sure there are things I have done that could be done much better.

There are a couple edge cases where I know this will not work correctly: Mainly if there is a comment that spans multiple lines with text after the closing tag e.g.

/* multi-line
   comment */"KEY" = "VALUE";

Although I'm not sure if that's a valid use-case to be concerned with.

Essentially the modifications preprocess the file by adding the line number to the beginning of every line, removes whitespace and comments, and then uses that line number when it generates error output.

in Localizable.strings, there is a line like this. "Turn off airplane mode or use Wi-Fi to access %@." = "Отключить режим "в самолете" или использовать Wi-Fi для доступа к %@.";

within the string, quotation marks are not escaped, wrong translation, while rubustring reports it is ok. plutil could catch the error.

$ plutil -lint test.strings plutil[60246:655609] CFPropertyListCreateFromXMLData(): Old-style plist parser: missing semicolon in dictionary on line 1. Parsing will be abandoned. Break on _CFPropertyListMissingSemicolon to debug. test.strings: Unexpected character " at line 1029

@dcordero if a localizable.string file is empty, maybe it should just a warning, not an error ? thanks. https://github.com/dcordero/Rubustrings/blob/7852877e82af8e4457cdfa4ac8e70badca846fa8/lib/rubustrings/action.rb#L44