add a tool to convert UTF-8 to UTF8-MAC

.gitignore

@@ -5,6 +5,7 @@ s3config.yaml
 
 bin/llc
 bin/ruby
+bin/nfd
 bin/ios/deploy
 bin/ios/sim
 bin/osx/sim

Rakefile

@@ -85,7 +85,7 @@ desc "Install"
 task :install do
   public_binaries = ['./bin/motion']
   binaries = public_binaries.dup.concat(['./bin/ios/deploy', './bin/ios/sim',
-    './bin/osx/sim', './bin/llc', './bin/ruby', './bin/ctags',
+    './bin/osx/sim', './bin/llc', './bin/ruby', './bin/ctags', './bin/nfd',
     'lib/yard/bin/yard', 'lib/yard/bin/yardoc', 'lib/yard/bin/yri', './lldb/lldb.py'])
   data = ['./NEWS']
   data.concat(Dir.glob('./lib/**/*', File::FNM_DOTMATCH) - ['./lib/Rakefile'])

bin/Rakefile

@@ -1,7 +1,7 @@
 verbose(true)
 
 task :default => :all
-task :all => [:files, :deploy, :sim]
+task :all => [:files, :deploy, :sim, :nfd]
 
 STRIP = "/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/strip"
 
@@ -33,6 +33,15 @@ task :sim do
   end
 end
 
-task :clean do
-  %w{ruby llc ios/deploy ios/sim osx/sim}.each { |path| rm_rf(path) }
+task :nfd do
+  bin = 'nfd'
+  if !File.exist?(bin) or File.mtime('nfd.m') > File.mtime(bin)
+    sh "/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/clang -mmacosx-version-min=10.6 -isysroot /Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX10.7.sdk -I./src -std=c99 -Wall -O3 nfd.m -o \"#{bin}\" -framework Foundation -I."
+    sh "#{STRIP} -x \"#{bin}\""
+  end
+end
+
+
+task :clean do
+  %w{ruby llc ios/deploy ios/sim osx/sim ndf}.each { |path| rm_rf(path) }
 end

bin/nfd.m

@@ -0,0 +1,25 @@
+// Convert UTF-8 from Normalization Form C (NFC) to Normalization Form D (NFD)
+// This tool outputs the escaped Ruby string.
+//   ex) ./nfd "a" #=> "\x61"
+#import <Foundation/Foundation.h>
+
+int main(int argc, char* argv[])
+{
+    [[NSAutoreleasePool alloc] init];
+    if (argc <= 0) {
+	exit(0);
+    }
+
+    const char *cstr = argv[1];
+    NSString *str = [[[NSString alloc] initWithCString:cstr encoding:NSUTF8StringEncoding] autorelease];
+    NSString *str_nfd = [str decomposedStringWithCanonicalMapping];
+
+    // output the string
+    const char *string = [str_nfd UTF8String];
+    const long len = strlen(string);
+    printf("\"");
+    for (int i = 0; i < len; i++) {
+	printf("\\x%X", (unsigned char)string[i]);
+    }
+    printf("\"");
+}