mirror/CMake
1
0
Fork 0
mirror of https://github.com/Kitware/CMake.git synced 2026-01-05 13:20:47 -06:00
Code Issues Packages Projects Releases Wiki Activity

cm_utf8: reject UTF-16 surrogate half codepoints

Browse Source
This commit is contained in:
Ben Boeckel
2019-03-14 11:01:27 -04:00
parent 7111873efd
commit a11e5e021b
2 changed files with 7 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
Source/cm_utf8.c
View File

@@ -71,6 +71,11 @@ const char* cm_utf8_decode_character(const char* first, const char* last,
return 0;
}
/* UTF-16 surrogate halves. */
if (0xD800 <= uc && uc <= 0xDFFF) {
return 0;
}
*pc = uc;
return first;
}

4
Tests/CMakeLib/testUTF8.cxx
View File

@@ -29,8 +29,6 @@ static test_utf8_entry const good_entry[] = {
{ 3, "\xED\x80\x80\x00", 0xD000 }, /* Valid 0xED prefixed codepoint. */
{ 4, "\xF4\x8F\xBF\xBF", 0x10FFFF }, /* Highest valid RFC codepoint. */
/* These are invalid according to the RFC, but accepted here. */
{ 3, "\xED\xA0\x80\x00", 0xD800 }, /* UTF-16 surrogate half. */
{ 3, "\xED\xBF\xBF\x00", 0xDFFF }, /* UTF-16 surrogate half. */
{ 4, "\xF4\x90\x80\x80", 0x110000 }, /* Lowest out-of-range codepoint. */
{ 4, "\xF5\x80\x80\x80",
0x140000 }, /* Prefix forces out-of-range codepoints. */
@@ -45,6 +43,8 @@ static test_utf8_char const bad_chars[] = {
"\xE0\x00\x00\x00", /* Missing continuation bytes. */
"\xE0\x80\x80\x00", /* Overlong encoding. */
"\xF0\x80\x80\x80", /* Overlong encoding. */
"\xED\xA0\x80\x00", /* UTF-16 surrogate half. */
"\xED\xBF\xBF\x00", /* UTF-16 surrogate half. */
{ 0, 0, 0, 0, 0 }
};