Implement surrogate pair handling using Bufs

Tested on Rakudo 2013.08-3-g7cb364b (Parrot & JVM)
moritz · Sep 8, 2013 · 60b3d80 · 60b3d80
1 parent 215b406
commit 60b3d80
Show file tree

Hide file tree

Showing 3 changed files with 10 additions and 5 deletions.
diff --git a/lib/JSON/Tiny.pm b/lib/JSON/Tiny.pm
@@ -26,7 +26,9 @@ multi to-json(Str:D  $d) {
     '"'
     ~ $d.trans(['"',  '\\',   "\b", "\f", "\n", "\r", "\t"]
             => ['\"', '\\\\', '\b', '\f', '\n', '\r', '\t'])\
-            .subst(/<-[\c32..\c126]>/, { ord(~$_).fmt('\u%04x') }, :g)
+            .subst(/<-[\c32..\c126]>/, {
+                $_.Str.encode('utf-16').values».fmt('\u%04x').join
+            }, :g)
     ~ '"'
 }
 multi to-json(Positional:D $d) {

diff --git a/lib/JSON/Tiny/Actions.pm b/lib/JSON/Tiny/Actions.pm
@@ -39,9 +39,8 @@ method value:sym<array>($/)  { make $<array>.ast }
 method str($/)               { make ~$/ }
 
 method str_escape($/) {
-    if $<xdigit> {
-        # make chr(:16($<xdigit>.join));  # preferred version of next line, but it doesn't work on Niecza yet
-        make chr(eval "0x" ~ $<xdigit>.join);
+    if $<utf16_codepoint> {
+        make utf16.new( $<utf16_codepoint>.map('0x'~*)».Int ).decode();
     } else {
         my %h = '\\' => "\\",
                 '/'  => "/",

diff --git a/lib/JSON/Tiny/Grammar.pm b/lib/JSON/Tiny/Grammar.pm
@@ -31,7 +31,11 @@ token str {
 }
 
 token str_escape {
-    <["\\/bfnrt]> | u <xdigit>**4
+    <["\\/bfnrt]> | 'u' <utf16_codepoint>+ % '\u'
+}
+
+token utf16_codepoint {
+    <.xdigit>**4
 }
 
 # vim: ft=perl6