diff options
Diffstat (limited to 'tests/gtests/blenlib/BLI_string_utf8_test.cc')
-rw-r--r-- | tests/gtests/blenlib/BLI_string_utf8_test.cc | 4 |
1 files changed, 2 insertions, 2 deletions
diff --git a/tests/gtests/blenlib/BLI_string_utf8_test.cc b/tests/gtests/blenlib/BLI_string_utf8_test.cc index 30dbccc97fe..89483d3d440 100644 --- a/tests/gtests/blenlib/BLI_string_utf8_test.cc +++ b/tests/gtests/blenlib/BLI_string_utf8_test.cc @@ -271,7 +271,7 @@ const char *utf8_invalid_tests[][3] = { // (to convert between UTF-16LE and UTF-16BE). // With such internal use of noncharacters, it may be desirable and safer to block those code points in // UTF-8 decoders, as they should never occur legitimately in incoming UTF-8 data, and could trigger -// unsafe behaviour in subsequent processing. +// unsafe behavior in subsequent processing. // // Particularly problematic noncharacters in 16-bit applications: {"5.3.1 U+FFFE = ef bf be = \"\xef\xbf\xbe\" |", @@ -281,7 +281,7 @@ const char *utf8_invalid_tests[][3] = { /* Fo now, we ignore those, they do not seem to be crucial anyway... */ // 5.3.3 U+FDD0 .. U+FDEF // 5.3.4 U+nFFFE U+nFFFF (for n = 1..10) - {NULL, NULL, NULL} + {NULL, NULL, NULL}, }; /* BLI_utf8_invalid_strip (and indirectly, BLI_utf8_invalid_byte). */ |