[json-glib] Support for surrogate pairs in json string.

[Emmanuele Bassi <ebassi src gnome org>]

commit fcd07918d3ed2b31b047900da9d2fed23dddf7da
Author: Eiichi Sato <eiichi for public gmail com>
Date:   Sat Apr 10 01:52:10 2010 +0900

    Support for surrogate pairs in json string.
    
    https://bugzilla.gnome.org/show_bug.cgi?id=615799
    
    Signed-off-by: Emmanuele Bassi <ebassi linux intel com>

 json-glib/json-scanner.c |   16 +++++++++++++++-
 1 files changed, 15 insertions(+), 1 deletions(-)
---
diff --git a/json-glib/json-scanner.c b/json-glib/json-scanner.c
index 9cd0938..6eca2ba 100644
--- a/json-glib/json-scanner.c
+++ b/json-glib/json-scanner.c
@@ -892,7 +892,7 @@ json_scanner_get_unichar (JsonScanner *scanner,
         break;
     }
 
-  g_assert (g_unichar_validate (uchar));
+  g_assert (g_unichar_validate (uchar) || g_unichar_type (uchar) == G_UNICODE_SURROGATE);
 
   return uchar;
 }
@@ -1439,6 +1439,20 @@ json_scanner_get_token_ll (JsonScanner *scanner,
                               gunichar ucs;
 
                               ucs = json_scanner_get_unichar (scanner, line_p, position_p);
+
+                              if (g_unichar_type (ucs) == G_UNICODE_SURROGATE)
+                                {
+                                  /* read next surrogate */
+                                  if ('\\' == json_scanner_get_char (scanner, line_p, position_p)
+                                      && 'u' == json_scanner_get_char (scanner, line_p, position_p))
+                                    {
+                                      gunichar ucs_lo = json_scanner_get_unichar (scanner, line_p, position_p);
+                                      g_assert (g_unichar_type (ucs_lo) == G_UNICODE_SURROGATE);
+                                      ucs = (((ucs & 0x3ff) << 10) | (ucs_lo & 0x3ff)) + 0x10000;
+                                    }
+                                }
+
+                              g_assert (g_unichar_validate (ucs));
                               gstring = g_string_append_unichar (gstring, ucs);
                             }
                           break;