This one works with non-terminated strings that we may need to handle.
It also validates that the sequence is valid UTF-8 so it results in
the same behaviour as the version that is used when converting to
codes (syncing the enumeration and drawing behaviour).
correctly (or at all...)
* Use the common 0xfffd character as a substitute for invalid characters
* Corrected comment
This function is used by the app_server to convert the UTF8 strings to char
codes to feed FreeType. Using a non space substitute character at least for
now, as it makes it more obvious where invalid characters are present.
I tested this change with some UTF8 test files and it seems to work well.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24175 a95241bf-73f2-0310-859d-f6bbb57e9c96
breaking out of the loop too early, and would therefore miss the last
character.
* Replaced UTF8CountChars() with a short and safe version as well.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24173 a95241bf-73f2-0310-859d-f6bbb57e9c96
of stuff. I think the trouble is that you keep decrementing charCount while
still inside the same UTF8 glyph. I cannot really tell, I am hoping you don't
mind me reverting to the old version, since the new one is only three lines,
so it should be easy to correct and re-commit.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24172 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Unified UTF8CountChars and UTF8ToLength and removed the latter
* Rewrote UTF8CountBytes to use the more safe algorithm from UTF8CountChars
* Removed the unsafe count_utf8_bytes() function
This should fix bug #839. Marcus can you please review?
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@19624 a95241bf-73f2-0310-859d-f6bbb57e9c96
