#189 Text after tabs+comments gets borked

[Anonymous]

Hi,
when I run SciTE on Linux (Mandrake 9.2) with the
attached SciTEUser.properties file, any text appearing
after a comment that is preceded by a tab character in
a cpp/c/java etc. gets written twice, in different
colours, in differing places, on top of each other.
Eg:
int main(){$tab//text here gets borked

My SciTEUser.properties file attempts to recreate the
Emacs colour scheme. If for some obscure reason, I'm
defining two different rules that both get used to
display the comment, then surely the fact that that can
happen is a bug.

But it can't just be that, because where tab
character(s) should be another strange character(s)
shows up - rather like a squiggle.

This problem doesn't happen in Windows (XP).

Possibly related is the fact that whenever more then 2
forward slashes are used, they show up on a white
background, as opposed to the dark-slate-grey one that
I want.
Eg.
int main()///////////text here has a white background

But that might be by design, so I'm not going to file
that as a bug.

If you need more information, you can contact me at
rolandarnold
<at>
yahoo
<dot>
<the last three letters are guessable ;-)>

[Nobody/Anonymous]

The .SciTEUser.properties file that I have in my home directory

[Neil Hodgson]

Logged In: YES
user_id=12579

This problem was caused by the code.page=1251 setting. On
GTK+, code.page should only be set for UTF-8 or DBCS text,
not for Russian or Latin-1. However, this revealed an
underlying problem which was that the gdk_mbstowcs call
required \0 terminated input values so now DBCS strings are
copied into an extra buffer and \0 terminated. This may have
further implications as the length processed is restricted
to 10000 bytes. Another problem was that mblen can return -1
when the DBCS code page chosen does not match the text and a
byte is present that is not allowed in the code page. The -1
returned character length is now treated as 1 rather than as
a big unsigned number.
Three or more slashes is a documentation comment and if you
don't want to treat them differently assign them the same
style as ordinary comments.

[Nobody/Anonymous]

Logged In: NO

Similar things happen when using .ASM file editing, it just
changes some characters ( for example ";" and others
to "{" ... ) I am using Windows 2000.

It may be just that it changes those characters in edit
window, but when I do save, these changes get written to
the file :S

Thank, keep up the good work!

[Neil Hodgson]

Logged In: YES
user_id=12579

Are you using a non-English code page? One cause for
changing characters can be the case:u or case:l styles that
can display text in upper or lower case although that
shouldn't affect punctuation.