Crash on some multi-byte UTF-8 characters in wxSTC

[Paul K]

Neil:

I'm not sure if this is Scintilla problem or not as I can't test Scite on
OSX, but I opened this ticket for wxwidgets and would appreciate if you
could take a quick look to confirm that it's not Scintilla. I have all the
details (including the stack trace and the file it crashes on) in the
ticket: http://trac.wxwidgets.org/ticket/15621.

The short summary: when I load a file that includes some multi-byte
characters, for example, \240\157\128\128, it crashes during painting of
that character. wxSTC is using v3.2.1, but I didn't see any changes related
to that issue in the Scintilla history after that version. Please let me
know if you need any addtional information. Thank you.

Paul.

[Neil Hodgson]

I'm not sure if this is Scintilla problem or not as I can't test Scite on OSX,

It works fine with Scintilla and SciTE on OS X and Windows which displays the file as [xED][xB3][xB6][xED][xB2][xA2]

but I opened this ticket for wxwidgets and would appreciate if you could take a quick look to confirm that it's not Scintilla. I have all the details (including the stack trace and the file it crashes on) in the ticket: http://trac.wxwidgets.org/ticket/15621.

The text is "\xED\xB3\xB6\xED\xB2\xA2" which decodes into "\uDCF6\uDCA2". U+DCF6 and U+DCA2 are both trail surrogates which can not appear isolated. This text is not valid UTF-8.

The short summary: when I load a file that includes some multi-byte characters, for example, \240\157\128\128,

That isn't the contents of the sample file. A file containing U+1D000 as UTF-8 ("\xF0\x9D\x80\x80") opens fine in SciTE on both OS X and Windows.

Neil

[Paul K]

Neil:

Thank you for confirming it's not a Scintilla issue and for the details.
You are of course correct about the codes; I'm now not sure where I got
\240 code from as I rechecked and the codes indeed are the ones you showed.

Paul.