From 5640f805d471b2da7f54fca7842f7e700154a9b6 Mon Sep 17 00:00:00 2001
From: Laurent Sansonetti <lsansonetti@apple.com>
Date: Tue, 7 Dec 2010 08:47:56 +0000
Subject: [PATCH] Dir.glob: convert paths to unicode normalization form C

git-svn-id: http://svn.macosforge.org/repository/ruby/MacRuby/trunk@4991 23306eb0-4c56-4727-a40e-e92c0eb68959
---
 dir.c      |  5 ++++-
 encoding.h |  2 ++
 string.c   | 19 +++++++++++++++++++
 3 files changed, 25 insertions(+), 1 deletion(-)

diff --git a/dir.c b/dir.c
index 85e04c23c..409ccbd2c 100644
--- a/dir.c
+++ b/dir.c
@@ -15,6 +15,7 @@
 #include "ruby/node.h"
 #include "ruby/util.h"
 #include "vm.h"
+#include "encoding.h"
 
 #include <sys/types.h>
 #include <sys/stat.h>
@@ -1391,7 +1392,9 @@ rb_glob(const char *path, void (*func)(const char *, VALUE), VALUE arg)
 static void
 push_pattern(const char *path, VALUE ary)
 {
-    rb_ary_push(ary, rb_tainted_str_new2(path));
+    VALUE str = rstr_new_path(path);
+    OBJ_TAINT(str);
+    rb_ary_push(ary, str);
 }
 
 int
diff --git a/encoding.h b/encoding.h
index f7ca2894f..f31a7a955 100644
--- a/encoding.h
+++ b/encoding.h
@@ -368,6 +368,8 @@ str_check_ascii_compatible(VALUE str)
     }
 }
 
+VALUE rstr_new_path(const char *path);
+
 #if defined(__cplusplus)
 } // extern "C"
 #endif
diff --git a/string.c b/string.c
index a1c11e248..869e7bfbb 100644
--- a/string.c
+++ b/string.c
@@ -6724,3 +6724,22 @@ rb_str_modify(VALUE obj)
 	rb_raise(rb_eArgError, "can't modify NSString");
     }
 }
+
+VALUE
+rstr_new_path(const char *path)
+{
+    // XXX this should be rewritten using ICU (utrans.h?) to avoid creating
+    // these 2 temporary CFStrings.
+    CFStringRef tmp = CFStringCreateWithFileSystemRepresentation(NULL, path); 
+    assert(tmp != NULL);
+
+    CFMutableStringRef tmp2 = CFStringCreateMutableCopy(NULL, 0, tmp); 
+    assert(tmp2 != NULL); 
+    CFRelease(tmp); 
+    CFStringNormalize(tmp2, kCFStringNormalizationFormC); 
+
+    rb_str_t *str = str_new_from_cfstring(tmp2);
+    CFRelease(tmp2);
+ 
+    return (VALUE)str;
+}