UTF-8: More generic sanitization function.

I could now directly use it in Gigamail's parser.

diff --git a/ucw/unicode.h b/ucw/unicode.h
index c47d436a137a134ec4e6c269060cc9585fa91dce..8966167c15e671ec7baa7e13e6795052e2a1450c 100644 (file)
--- a/ucw/unicode.h
+++ b/ucw/unicode.h
@@ -282,7 +282,7 @@ static inline void *utf16_be_get(const void *p, uint *uu)
 }
 
 // Internal helper for functions below
-static inline uint unicode_sanitize_char_internal(uint u, uint limit, uint allowed_ctl)
+static inline uint unicode_sanitize_char_internal(uint u, uint limit, uint allowed_ctl, uint repl)
 {
   if (u - 0x20 >= 0x60)                        // Otherwise printable ASCII 0x20..0x7f (optimized for this case)
     {
@@ -301,7 +301,7 @@ static inline uint unicode_sanitize_char_internal(uint u, uint limit, uint allow
       else if (u < 0xa0)
        {                               // 0x80..0x9f Latin-1 control chars
 bad:
-         u = UNI_REPLACEMENT;
+         u = repl;
        }
     }
   return u;
@@ -314,7 +314,7 @@ bad:
  **/
 static inline uint unicode_sanitize_char(uint u)
 {
-  return unicode_sanitize_char_internal(u, 0x10000, 1U<<'\t');
+  return unicode_sanitize_char_internal(u, 0x10000, 1U<<'\t', UNI_REPLACEMENT);
 }
 
 /**
@@ -324,7 +324,7 @@ static inline uint unicode_sanitize_char(uint u)
  **/
 static inline uint unicode_full_sanitize_char(uint u)
 {
-  return unicode_sanitize_char_internal(u, 0x110000, 1U<<'\t');
+  return unicode_sanitize_char_internal(u, 0x110000, 1U<<'\t', UNI_REPLACEMENT);
 }
 
 /* unicode-utf8.c */