[ID 20000724.003] Documentation changes for perllocale.pod

[p5sagit/p5-mst-13.2.git] / utf8.c

diff --git a/utf8.c b/utf8.c
index 9bb89a4..95f457f 100644 (file)
--- a/utf8.c
+++ b/utf8.c
@@ -134,6 +134,30 @@ Perl_is_utf8_char(pTHX_ U8 *s)
     return len;
 }
 
+/*
+=for apidoc Am|bool_utf8_string|U8 *s|STRLEN len
+
+Returns true if first C<len> bytes of the given string form valid a UTF8
+string, false otherwise.
+
+=cut
+*/
+
+bool 
+Perl_is_utf8_string(pTHX_ U8 *s, STRLEN len)
+{
+    U8* x=s;
+    U8* send=s+len;
+    int c;
+    while (x < send) {
+        c = is_utf8_char(x);
+        x += c;
+        if (!c || x > send)
+            return 0;
+    }
+    return 1;
+}
+
 UV
 Perl_utf8_to_uv(pTHX_ U8* s, I32* retlen)
 {
@@ -296,13 +320,20 @@ Perl_bytes_to_utf8(pTHX_ U8* s, STRLEN *len)
     return dst;
 }
 
-/* XXX NOTHING CALLS THE FOLLOWING TWO ROUTINES YET!!! */
 /*
  * Convert native or reversed UTF-16 to UTF-8.
  *
  * Destination must be pre-extended to 3/2 source.  Do not use in-place.
  * We optimize for native, for obvious reasons. */
 
+/* There are several problems with utf16_to_utf8().
+ * (1) U16 is not necessarily *exactly* two bytes.
+ * (2) Secondly, no check is made for odd length.
+ * (3) Thirdly, the "Malformed UTF-16 surrogate" should probably be
+ *     a hard error (and it should be listed in perldiag).
+ * (4) The tests (in comp/t/require.t) are a joke: the UTF16 BOM
+ *     really ought to be followed by valid UTF16 characters.
+ * --Mike Guy */
 U8*
 Perl_utf16_to_utf8(pTHX_ U16* p, U8* d, I32 bytelen)
 {