UTF-8 string support in core

define MRB_UTF8_STRING (in mrbconf.h) to enable UTF-8 support.

Author: matz

include/mrbconf.h

@@ -26,6 +26,9 @@
 /* represent mrb_value as a word (natural unit of data for the processor) */
 //#define MRB_WORD_BOXING
 
+/* string class to handle UTF-8 encoding */
+//#define MRB_UTF8_STRING
+
 /* argv max size in mrb_funcall */
 //#define MRB_FUNCALL_ARGC_MAX 16
 

mrbgems/mruby-string-ext/mrblib/string.rb

@@ -310,4 +310,30 @@ def upto(other_str, excl=false, &block)
       return self if excl && str == other_str
     end
   end
+
+  def chars(&block)
+    if block_given?
+      self.split('').map do |i|
+        block.call(i)
+      end
+      self
+    else
+      self.split('')
+    end
+  end
+  alias each_char chars
+
+  def codepoints(&block)
+    len = self.size
+
+    if block_given?
+      self.split('').map do|x|
+        block.call(x.ord)
+      end
+      self
+    else
+      self.split('').map{|x| x.ord}
+    end
+  end
+  alias each_codepoint codepoints
 end

mrbgems/mruby-string-ext/src/string.c

@@ -245,6 +245,51 @@ mrb_str_chr(mrb_state *mrb, mrb_value self)
   return mrb_str_substr(mrb, self, 0, 1);
 }
 
+static mrb_value
+mrb_fixnum_chr(mrb_state *mrb, mrb_value num)
+{
+  mrb_int cp = mrb_fixnum(num);
+#ifdef MRB_UTF8_STRING
+  char utf8[4];
+  mrb_int len;
+
+  if (cp < 0 || 0x10FFFF < cp) {
+    mrb_raisef(mrb, E_RANGE_ERROR, "%S out of char range", num);
+  }
+  if (cp < 0x80) {
+    utf8[0] = (char)cp;
+    len = 1;
+  }
+  else if (cp < 0x800) {
+    utf8[0] = (char)(0xC0 | (cp >> 6));
+    utf8[1] = (char)(0x80 | (cp & 0x3F));
+    len = 2;
+  }
+  else if (cp < 0x10000) {
+    utf8[0] = (char)(0xE0 |  (cp >> 12));
+    utf8[1] = (char)(0x80 | ((cp >>  6) & 0x3F));
+    utf8[2] = (char)(0x80 | ( cp        & 0x3F));
+    len = 3;
+  }
+  else {
+    utf8[0] = (char)(0xF0 |  (cp >> 18));
+    utf8[1] = (char)(0x80 | ((cp >> 12) & 0x3F));
+    utf8[2] = (char)(0x80 | ((cp >>  6) & 0x3F));
+    utf8[3] = (char)(0x80 | ( cp        & 0x3F));
+    len = 4;
+  }
+  return mrb_str_new(mrb, utf8, len);
+#else
+  char c;
+
+  if (cp < 0 || 0xff < cp) {
+    mrb_raisef(mrb, E_RANGE_ERROR, "%S out of char range", num);
+  }
+  c = (char)cp;
+  return mrb_str_new(mrb, &c, 1);
+#endif
+}
+
 /*
  *  call-seq:
  *     string.lines    ->  array of string
@@ -422,6 +467,72 @@ mrb_str_prepend(mrb_state *mrb, mrb_value self)
   return self;
 }
 
+#ifdef MRB_UTF8_STRING
+static const char utf8len_codepage_zero[256] =
+{
+  1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+  1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+  1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+  1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+  0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+  0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+  2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,
+  3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,4,4,4,4,4,0,0,0,0,0,0,0,0,0,0,0,
+};
+
+static mrb_int
+utf8code(unsigned char* p)
+{
+  mrb_int len;
+
+  if (p[0] < 0x80)
+    return p[0];
+
+  len = utf8len_codepage_zero[p[0]];
+  if (len > 1 && (p[1] & 0xc0) == 0x80) {
+    if (len == 2)
+      return ((p[0] & 0x1f) << 6) + (p[1] & 0x3f);
+    if ((p[2] & 0xc0) == 0x80) {
+      if (len == 3)
+        return ((p[0] & 0x0f) << 12) + ((p[1] & 0x3f) << 6)
+          + (p[2] & 0x3f);
+      if ((p[3] & 0xc0) == 0x80) {
+        if (len == 4)
+          return ((p[0] & 0x07) << 18) + ((p[1] & 0x3f) << 12)
+            + ((p[2] & 0x3f) << 6) + (p[3] & 0x3f);
+        if ((p[4] & 0xc0) == 0x80) {
+          if (len == 5)
+            return ((p[0] & 0x03) << 24) + ((p[1] & 0x3f) << 18)
+              + ((p[2] & 0x3f) << 12) + ((p[3] & 0x3f) << 6)
+              + (p[4] & 0x3f);
+          if ((p[5] & 0xc0) == 0x80 && len == 6)
+            return ((p[0] & 0x01) << 30) + ((p[1] & 0x3f) << 24)
+              + ((p[2] & 0x3f) << 18) + ((p[3] & 0x3f) << 12)
+              + ((p[4] & 0x3f) << 6) + (p[5] & 0x3f);
+        }
+      }
+    }
+  }
+  return p[0];
+}
+
+static mrb_value
+mrb_str_ord(mrb_state* mrb, mrb_value str)
+{
+  if (RSTRING_LEN(str) == 0)
+    mrb_raise(mrb, E_ARGUMENT_ERROR, "empty string");
+  return mrb_fixnum_value(utf8code((unsigned char*) RSTRING_PTR(str)));
+}
+#else
+static mrb_value
+mrb_str_ord(mrb_state* mrb, mrb_value str)
+{
+  if (RSTRING_LEN(str) == 0)
+    mrb_raise(mrb, E_ARGUMENT_ERROR, "empty string");
+  return mrb_fixnum_value(RSTRING_PTR(str)[0]);
+}
+#endif
+
 void
 mrb_mruby_string_ext_gem_init(mrb_state* mrb)
 {
@@ -446,6 +557,9 @@ mrb_mruby_string_ext_gem_init(mrb_state* mrb)
   mrb_define_method(mrb, s, "prepend",         mrb_str_prepend,         MRB_ARGS_REQ(1));
   mrb_alias_method(mrb, s, mrb_intern_lit(mrb, "next"), mrb_intern_lit(mrb, "succ"));
   mrb_alias_method(mrb, s, mrb_intern_lit(mrb, "next!"), mrb_intern_lit(mrb, "succ!"));
+  mrb_define_method(mrb, s, "ord", mrb_str_ord, MRB_ARGS_NONE());
+
+  mrb_define_method(mrb, mrb->fixnum_class, "chr", mrb_fixnum_chr, MRB_ARGS_NONE());
 }
 
 void

mrbgems/mruby-string-ext/test/string.rb

@@ -1,6 +1,8 @@
 ##
 # String(Ext) Test
 
+UTF8STRING = ("\343\201\202".size == 1)
+
 assert('String#getbyte') do
   str1 = "hello"
   bytes1 = [104, 101, 108, 108, 111]
@@ -180,6 +182,8 @@ def o.to_str
 
 assert('String#chr') do
   assert_equal "a", "abcde".chr
+  # test Fixnum#chr as well
+  assert_equal "a", 97.chr
 end
 
 assert('String#lines') do
@@ -374,8 +378,8 @@ def o.to_str
   assert_equal "-b-", a
   a = "-z-"; a.succ!
   assert_equal "-aa-", a
-  a = "あa"; a.succ!
-  assert_equal "あb", a
+  a = "あb"; a.succ!
+  assert_equal "あc", a
   a = "あaz"; a.succ!
   assert_equal "あba", a
 end
@@ -471,3 +475,96 @@ def o.to_str
   })
   assert_equal(2, count)
 end
+
+assert('String#ord') do
+  got = "hello!".split('').map {|x| x.ord}
+  expect = [104, 101, 108, 108, 111, 33]
+  assert_equal expect, got
+end
+
+assert('String#ord(UTF-8)') do
+  got = "こんにちは世界!".split('').map {|x| x.ord}
+  expect = [0x3053,0x3093,0x306b,0x3061,0x306f,0x4e16,0x754c,0x21]
+  assert_equal expect, got
+end if UTF8STRING
+
+assert('String#chr') do
+  assert_equal "h", "hello!".chr
+end
+assert('String#chr(UTF-8)') do
+  assert_equal "こ", "こんにちは世界!".chr
+end if UTF8STRING
+
+assert('String#chars') do
+  expect = ["h", "e", "l", "l", "o", "!"]
+  assert_equal expect, "hello!".chars
+  s = ""
+  "hello!".chars do |x|
+    s += x
+  end
+  assert_equal "hello!", s
+end
+
+assert('String#chars(UTF-8)') do
+  expect = ['こ', 'ん', 'に', 'ち', 'は', '世', '界', '!']
+  assert_equal expect, "こんにちは世界!".chars
+  s = ""
+  "こんにちは世界!".chars do |x|
+    s += x
+  end
+  assert_equal "こんにちは世界!", s
+end if UTF8STRING
+
+assert('String#each_char') do
+  s = ""
+  "hello!".each_char do |x|
+    s += x
+  end
+  assert_equal "hello!", s
+end
+
+assert('String#each_char(UTF-8)') do
+  s = ""
+  "こんにちは世界!".each_char do |x|
+    s += x
+  end
+  assert_equal "こんにちは世界!", s
+end if UTF8STRING
+
+assert('String#codepoints') do
+  expect = [104, 101, 108, 108, 111, 33]
+  assert_equal expect, "hello!".codepoints
+  cp = []
+  "hello!".codepoints do |x|
+    cp << x
+  end
+  assert_equal expect, cp
+end
+
+assert('String#codepoints(UTF-8)') do
+  expect = [12371, 12435, 12395, 12385, 12399, 19990, 30028, 33]
+  assert_equal expect, "こんにちは世界!".codepoints
+  cp = []
+  "こんにちは世界!".codepoints do |x|
+    cp << x
+  end
+  assert_equal expect, cp
+end if UTF8STRING
+
+assert('String#each_codepoint') do
+  expect = [104, 101, 108, 108, 111, 33]
+  cp = []
+  "hello!".each_codepoint do |x|
+    cp << x
+  end
+  assert_equal expect, cp
+end
+
+assert('String#each_codepoint(UTF-8)') do
+  expect = [12371, 12435, 12395, 12385, 12399, 19990, 30028, 33]
+  cp = []
+  "こんにちは世界!".each_codepoint do |x|
+    cp << x
+  end
+  assert_equal expect, cp
+end if UTF8STRING