oj 1.0.4 → 1.0.5

data/README.md

@@ -24,9 +24,9 @@ A fast JSON parser and Object marshaller as a Ruby gem.
 
 ## <a name="release">Release Notes</a>
 
-### Release 1.0.4
+### Release 1.0.5
 
- - Fixed bug that did not allow symbols as keys in :compat mode.
+ - Added :ascii_only options for dumping JSON where all high-bit characters are encoded as escaped sequences.
 
 ## <a name="description">Description</a>
 

data/ext/oj/dump.c

@@ -94,45 +94,50 @@ static int	dump_attr_cb(ID key, VALUE value, Out out);
 static void	dump_obj_attrs(VALUE obj, int with_class, slot_t id, int depth, Out out);
 
 static void     grow(Out out, size_t len);
-static int      is_json_friendly(const u_char *str, size_t len);
 static size_t	json_friendly_size(const u_char *str, size_t len);
+static size_t	ascii_friendly_size(const u_char *str, size_t len);
 
 
 static const char	hex_chars[17] = "0123456789abcdef";
 
 static char     json_friendly_chars[256] = "\
-uuuuuuuuxxxuxxuuuuuuuuuuuuuuuuuu\
-ooxooooooooooooxoooooooooooooooo\
-ooooooooooooooooooooooooooooxooo\
-ooooooooooooooooooooooooooooooou\
-uuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuu\
-uuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuu\
-uuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuu\
-uuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuu";
-
-inline static int
-is_json_friendly(const u_char *str, size_t len) {
+66666666222622666666666666666666\
+11211111111111121111111111111111\
+11111111111111111111111111112111\
+11111111111111111111111111111111\
+11111111111111111111111111111111\
+11111111111111111111111111111111\
+11111111111111111111111111111111\
+11111111111111111111111111111111";
+
+static char     ascii_friendly_chars[256] = "\
+66666666222622666666666666666666\
+11211111111111121111111111111111\
+11111111111111111111111111112111\
+11111111111111111111111111111116\
+66666666666666666666666666666666\
+66666666666666666666666666666666\
+66666666666666666666666666666666\
+66666666666666666666666666666666";
+
+inline static size_t
+json_friendly_size(const u_char *str, size_t len) {
+    size_t	size = 0;
+
     for (; 0 < len; str++, len--) {
-        if ('o' != json_friendly_chars[*str]) {
-            return 0;
-        }
+	size += json_friendly_chars[*str];
     }
-    return 1;
+    return size - len * (size_t)'0';
 }
 
 inline static size_t
-json_friendly_size(const u_char *str, size_t len) {
-    int		cnt = 0;
-    
+ascii_friendly_size(const u_char *str, size_t len) {
+    size_t	size = 0;
+
     for (; 0 < len; str++, len--) {
-	switch (json_friendly_chars[*str]) {
-	case 'o':	cnt++;		break;
-	case 'x':	cnt += 2;	break;
-	case 'u':	cnt += 6;	break;
-	default:			break;
-	}
+	size += ascii_friendly_chars[*str];
     }
-    return cnt;
+    return size - len * (size_t)'0';
 }
 
 inline static void
@@ -341,8 +346,16 @@ dump_float(VALUE obj, Out out) {
 
 static void
 dump_cstr(const char *str, size_t cnt, int is_sym, int escape1, Out out) {
-    size_t	size = json_friendly_size((u_char*)str, cnt);
-    
+    size_t	size;
+    char	*cmap;
+
+    if (Yes == out->opts->ascii_only) {
+	cmap = ascii_friendly_chars;
+	size = ascii_friendly_size((u_char*)str, cnt);
+    } else {
+	cmap = json_friendly_chars;
+	size = json_friendly_size((u_char*)str, cnt);
+    }
     if (out->end - out->cur <= (long)size + 10) { // extra 10 for escaped first char, quotes, and sym
 	grow(out, size + 10);
     }
@@ -371,11 +384,11 @@ dump_cstr(const char *str, size_t cnt, int is_sym, int escape1, Out out) {
 	    *out->cur++ = ':';
 	}
 	for (; 0 < cnt; cnt--, str++) {
-	    switch (json_friendly_chars[(u_char)*str]) {
-	    case 'o':
+	    switch (cmap[(u_char)*str]) {
+	    case '1':
 		*out->cur++ = *str;
 		break;
-	    case 'x':
+	    case '2':
 		*out->cur++ = '\\';
 		switch (*str) {
 		case '\b':	*out->cur++ = 'b';	break;
@@ -386,7 +399,7 @@ dump_cstr(const char *str, size_t cnt, int is_sym, int escape1, Out out) {
 		default:	*out->cur++ = *str;	break;
 		}
 		break;
-	    case 'u':
+	    case '6':
 		*out->cur++ = '\\';
 		*out->cur++ = 'u';
 		if ((u_char)*str <= 0x7F) {

data/ext/oj/oj.c

@@ -66,6 +66,7 @@ VALUE	oj_time_class;
 
 VALUE	oj_slash_string;
 
+static VALUE	ascii_only_sym;
 static VALUE	auto_define_sym;
 static VALUE	circular_sym;
 static VALUE	compat_sym;
@@ -86,6 +87,7 @@ struct _Options	oj_default_options = {
     No,			// circular
     Yes,		// auto_define
     No,			// sym_key
+    No,			// ascii_only
     ObjectMode,		// mode
 };
 
@@ -109,6 +111,7 @@ get_def_opts(VALUE self) {
     rb_hash_aset(opts, indent_sym, INT2FIX(oj_default_options.indent));
     rb_hash_aset(opts, circular_sym, (Yes == oj_default_options.circular) ? Qtrue : ((No == oj_default_options.circular) ? Qfalse : Qnil));
     rb_hash_aset(opts, auto_define_sym, (Yes == oj_default_options.auto_define) ? Qtrue : ((No == oj_default_options.auto_define) ? Qfalse : Qnil));
+    rb_hash_aset(opts, ascii_only_sym, (Yes == oj_default_options.ascii_only) ? Qtrue : ((No == oj_default_options.ascii_only) ? Qfalse : Qnil));
     rb_hash_aset(opts, symbol_keys_sym, (Yes == oj_default_options.sym_key) ? Qtrue : ((No == oj_default_options.sym_key) ? Qfalse : Qnil));
     switch (oj_default_options.mode) {
     case StrictMode:	rb_hash_aset(opts, mode_sym, strict_sym);	break;
@@ -129,6 +132,7 @@ get_def_opts(VALUE self) {
  * @param [true|false|nil] :circular support circular references while dumping
  * @param [true|false|nil] :auto_define automatically define classes if they do not exist
  * @param [true|false|nil] :symbol_keys convert hash keys to symbols
+ * @param [true|false|nil] :ascii_only encode all high-bit characters as escaped sequences if true
  * @param [:object|:strict|:compat|:null] load and dump mode to use for JSON
  *        :strict raises an exception when a non-supported Object is
  *        encountered. :compat attempts to extract variable values from an
@@ -144,6 +148,7 @@ set_def_opts(VALUE self, VALUE opts) {
         { circular_sym, &oj_default_options.circular },
         { auto_define_sym, &oj_default_options.auto_define },
         { symbol_keys_sym, &oj_default_options.sym_key },
+        { ascii_only_sym, &oj_default_options.ascii_only },
         { Qnil, 0 }
     };
     YesNoOpt    o;
@@ -182,19 +187,18 @@ set_def_opts(VALUE self, VALUE opts) {
     }
 
     for (o = ynos; 0 != o->attr; o++) {
-	if (Qtrue != rb_funcall(opts, rb_intern("has_key?"), 1, mode_sym)) {
+	if (Qtrue != rb_funcall(opts, rb_intern("has_key?"), 1, o->sym)) {
 	    continue;
 	}
-        v = rb_hash_lookup(opts, o->sym);
-	if (Qnil == v) {
-            *o->attr = NotSet;
-        } else if (Qtrue == v) {
-            *o->attr = Yes;
-        } else if (Qfalse == v) {
-            *o->attr = No;
-        } else {
-            rb_raise(rb_eArgError, "%s must be true, false, or nil.\n", StringValuePtr(o->sym));
-        }
+        if (Qnil != (v = rb_hash_lookup(opts, o->sym))) {
+	    if (Qtrue == v) {
+		*o->attr = Yes;
+	    } else if (Qfalse == v) {
+		*o->attr = No;
+	    } else {
+		rb_raise(rb_eArgError, "%s must be true, false, or nil.\n", rb_id2name(SYM2ID(o->sym)));
+	    }
+	}
     }
     return Qnil;
 }
@@ -205,6 +209,7 @@ parse_options(VALUE ropts, Options copts) {
         { circular_sym, &copts->circular },
         { auto_define_sym, &copts->auto_define },
         { symbol_keys_sym, &copts->sym_key },
+        { ascii_only_sym, &copts->ascii_only },
         { Qnil, 0 }
     };
     YesNoOpt    o;
@@ -239,13 +244,12 @@ parse_options(VALUE ropts, Options copts) {
         }
         for (o = ynos; 0 != o->attr; o++) {
             if (Qnil != (v = rb_hash_lookup(ropts, o->sym))) {
-
                 if (Qtrue == v) {
                     *o->attr = Yes;
                 } else if (Qfalse == v) {
                     *o->attr = No;
                 } else {
-                    rb_raise(rb_eArgError, "%s must be true or false.\n", StringValuePtr(o->sym));
+                    rb_raise(rb_eArgError, "%s must be true or false.\n", rb_id2name(SYM2ID(o->sym)));
                 }
             }
         }
@@ -431,6 +435,7 @@ void Init_oj() {
     oj_date_class = rb_const_get(rb_cObject, rb_intern("Date"));
     oj_stringio_class = rb_const_get(rb_cObject, rb_intern("StringIO"));
 
+    ascii_only_sym = ID2SYM(rb_intern("ascii_only"));	rb_ary_push(keep, ascii_only_sym);
     auto_define_sym = ID2SYM(rb_intern("auto_define"));	rb_ary_push(keep, auto_define_sym);
     circular_sym = ID2SYM(rb_intern("circular"));	rb_ary_push(keep, circular_sym);
     compat_sym = ID2SYM(rb_intern("compat"));		rb_ary_push(keep, compat_sym);

data/ext/oj/oj.h

@@ -85,6 +85,7 @@ typedef struct _Options {
     char        circular;	// YesNo
     char        auto_define;	// YesNo
     char        sym_key;	// YesNo
+    char        ascii_only;	// YesNo
     char        mode;		// Mode
 } *Options;
 

data/lib/oj/version.rb

@@ -1,5 +1,5 @@
 
 module Oj
   # Current version of the module. 
-  VERSION = '1.0.4'
+  VERSION = '1.0.5'
 end

data/test/tests.rb

@@ -72,6 +72,7 @@ class Juice < ::Test::Unit::TestCase
                    :circular=>false,
                    :auto_define=>true,
                    :symbol_keys=>false,
+                   :ascii_only=>false,
                    :mode=>:object}, opts)
   end
 
@@ -82,13 +83,15 @@ class Juice < ::Test::Unit::TestCase
       :circular=>false,
       :auto_define=>true,
       :symbol_keys=>false,
+      :ascii_only=>false,
       :mode=>:object}
     o2 = {
       :encoding=>"UTF-8",
       :indent=>4,
       :circular=>true,
       :auto_define=>false,
-      :symbol_keys=>false,
+      :symbol_keys=>true,
+      :ascii_only=>true,
       :mode=>:compat}
     o3 = { :indent => 4 }
     Oj.default_options = o2
@@ -143,7 +146,9 @@ class Juice < ::Test::Unit::TestCase
   def test_encode
     Oj.default_options = { :encoding => 'UTF-8' }
     dump_and_load("ぴーたー", false)
-    Oj.default_options = { :encoding => nil }
+    Oj.default_options = { :ascii_only => true }
+    dump_and_load("ぴーたー", false)
+    Oj.default_options = { :encoding => nil, :ascii_only => false }
   end
 
   def test_array

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: oj
 version: !ruby/object:Gem::Version
-  version: 1.0.4
+  version: 1.0.5
   prerelease: 
 platform: ruby
 authors:
@@ -9,7 +9,7 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2012-03-16 00:00:00.000000000 Z
+date: 2012-03-17 00:00:00.000000000 Z
 dependencies: []
 description: ! 'The fastest JSON parser and object serializer. '
 email: peter@ohler.com