-sanitize utf8 api a bit

3 files changed, 32 insertions, 21 deletions

diff --git a/src/util/crypto_hash.c b/src/util/crypto_hash.c
index dcd46e5f9..f516f5474 100644
--- a/src/util/crypto_hash.c
+++ b/src/util/crypto_hash.c
@@ -76,7 +76,8 @@ GNUNET_CRYPTO_hash_from_string2 (const char *enc,
   char upper_enc[enclen];
   char *up_ptr = upper_enc;
 
-  GNUNET_STRINGS_utf8_toupper (enc, up_ptr);
+  if (GNUNET_OK != GNUNET_STRINGS_utf8_toupper (enc, up_ptr))
+    return GNUNET_SYSERR;
 
   return GNUNET_STRINGS_string_to_data (upper_enc, enclen,
                                         (unsigned char *) result,
diff --git a/src/util/strings.c b/src/util/strings.c
index c63870e42..75e45ba6d 100644
--- a/src/util/strings.c
+++ b/src/util/strings.c
@@ -390,7 +390,7 @@ GNUNET_STRINGS_conv (const char *input,
   ret[encoded_string_length] = '\0';
   free (encoded_string);
   return ret;
-fail:
+  fail:
   LOG (GNUNET_ERROR_TYPE_WARNING,
        _ ("Character sets requested were `%s'->`%s'\n"),
        "UTF-8",
@@ -426,24 +426,28 @@ GNUNET_STRINGS_from_utf8 (const char *input,
 }
 
 
-void
-GNUNET_STRINGS_utf8_normalize (const char *input,
-                               char *output)
+char *
+GNUNET_STRINGS_utf8_normalize (const char *input)
 {
   uint8_t *tmp;
   size_t len;
-
-  tmp = u8_normalize (UNINORM_NFKC,
-                      (uint8_t *) input,
-                      strlen ((char*) input),
-                      NULL,
-                      &len);
+  char *output;
+  tmp = u8_tolower ((uint8_t *) input,
+                    strlen ((char *) input),
+                    NULL,
+                    UNINORM_NFKC,
+                    NULL,
+                    &len);
+  if (NULL == tmp)
+    return NULL;
+  output = GNUNET_malloc (len + 1);
   GNUNET_memcpy (output, tmp, len);
   output[len] = '\0';
   free (tmp);
+  return output;
 }
 
-void
+enum GNUNET_GenericReturnValue
 GNUNET_STRINGS_utf8_tolower (const char *input,
                              char *output)
 {
@@ -456,13 +460,16 @@ GNUNET_STRINGS_utf8_tolower (const char *input,
                        UNINORM_NFD,
                        NULL,
                        &len);
+  if (NULL == tmp_in)
+    return GNUNET_SYSERR;
   GNUNET_memcpy (output, tmp_in, len);
   output[len] = '\0';
-  free (tmp_in);
+  GNUNET_free (tmp_in);
+  return GNUNET_OK;
 }
 
 
-void
+enum GNUNET_GenericReturnValue
 GNUNET_STRINGS_utf8_toupper (const char *input,
                              char *output)
 {
@@ -475,9 +482,13 @@ GNUNET_STRINGS_utf8_toupper (const char *input,
                        UNINORM_NFD,
                        NULL,
                        &len);
+  if (NULL == tmp_in)
+    return GNUNET_SYSERR;
+  /* 0-terminator does not fit */
   GNUNET_memcpy (output, tmp_in, len);
   output[len] = '\0';
-  free (tmp_in);
+  GNUNET_free (tmp_in);
+  return GNUNET_OK;
 }
 
 
@@ -1604,7 +1615,7 @@ GNUNET_STRINGS_base64_encode (const void *in,
   char *opt;
 
   ret = 0;
-  GNUNET_assert (len < SIZE_MAX / 4 * 3 );
+  GNUNET_assert (len < SIZE_MAX / 4 * 3);
   opt = GNUNET_malloc (2 + (len * 4 / 3) + 8);
   for (size_t i = 0; i < len; ++i)
   {
@@ -1741,7 +1752,7 @@ GNUNET_STRINGS_base64_decode (const char *data,
       output[ret++] = c;
     }
   }
-END:
+  END:
   *out = output;
   return ret;
 }
diff --git a/src/util/test_strings.c b/src/util/test_strings.c
index 1ecd31464..cccffcaf5 100644
--- a/src/util/test_strings.c
+++ b/src/util/test_strings.c
@@ -114,10 +114,9 @@ main (int argc, char *argv[])
   /* Normalization */
   r = "q\u0307\u0323"; /* Non-canonical order */
 
-  GNUNET_STRINGS_utf8_normalize (r,
-                                 buf);
-  GNUNET_assert (0 == strcmp ("q\u0323\u0307", buf));
-
+  b = GNUNET_STRINGS_utf8_normalize (r);
+  GNUNET_assert (0 == strcmp ("q\u0323\u0307", b));
+  GNUNET_free (b);
   b = GNUNET_STRINGS_to_utf8 ("TEST", 4, "ASCII");
   WANT ("TEST", b);