Binary scan and NULL-char
Esa Heikkinen
I would want to scan binary string by following commands:
set point_file_fileID [open $point_file_full_name r]
fconfigure $point_file_fileID -translation binary -encoding binary
set line [read $point_file_fileID 52]
set ret [binary scan $line "iiiissica21a5" \
PtSWLongitude \
PtSWLatitude \
PtNELongitude \
PtNELatitude \
PtLine \
PtIndex \
PtInt \
PtType \
PtComName \
PtShort ]
The problem is that there are NULL-character (0x00) in data regarding to
PtComName and PtShort.
This NULL-character can be in various positions. I do not want anything
after NULL-character.
How can i detect NULL-character correctly and how can i discard characters
after NULL-character ?
I think that code above should work, but it does not.
---
Esa
Helmut Giese
<esa.he...@insta.fi> wrote:
>How can i detect NULL-character correctly and how can i discard characters
>after NULL-character ?
Hi Esa,
you can detect a NULL-character via
set pos [string first \x00 $line]
and then delimit your string via
if { $pos =! -1 } {
set line [string range $line 0 [incr pos -1]]
}
HTH
Helmut Giese
Arjen Markus
not the C convention). But to discard the first NUL and anything after
that, you can use:
set PtComName [string range $PtComName 0 [expr {[string first "\0"
$PtComName]-1}]]
Regards,
Arjen
suchenwi
NUL:
64 % regsub {(.+\x00).+} "foo-\x00-bar" {\1}
foo-
Arjen Markus
character.
Regards,
Arjen
Esa Heikkinen
It works :) Thank you
---
Esa
"Helmut Giese" <hgi...@ratiosoft.com> wrote in message
news:43e75aa6...@News.Individual.DE...
Donal K. Fellows
> How can i detect NULL-character correctly and how can i discard characters
> after NULL-character ?
Use the 'A' format to scan the strings and not the 'a' format.
Donal.
Andreas Leitgeb
In the given context, this probably isn't it, since A${n} might
consume fewer than $n bytes from input, thus offsetting the start
of all following elements.
Donal K. Fellows
> In the given context, this probably isn't it, since A${n} might
> consume fewer than $n bytes from input, thus offsetting the start
> of all following elements.
Actually, I tested before making that comment, and it works. Witness:
% set s [binary format a10c abc 100]; string length $s
11
% binary scan $s a10c x y
2
% string length $x
10
% string length $y
3
% set y
100
% unset x y
% binary scan $s A10c x y
2
% set x
abc
% string length $x
3
% set y
100
As you can see, the 'A' specifier is exactly right. :-)
Donal.
Andreas Leitgeb
>> In the given context, this probably isn't it, since A${n} might
>> consume fewer than $n bytes from input, thus offsetting the start
>> of all following elements.
> Actually, I tested before making that comment, and it works. Witness:
ok, ok, I misread the doku (which is actually very clear on that, saying
that the *resulting* string is stripped - I don't know how I could ever
misunderstand it. I guess I'd have used it a couple of times already if
it were not for the misunderstanding)
> As you can see, the 'A' specifier is exactly right. :-)
Not exactly :-) The OP asked for removal of anything starting
with the first \0 not for a [string trim ... " \0"]-effect.
Even if we assume that \0's appear only towards the end,
it will still also strip leading and trailing blanks,
which the OP didn't ask for.
Donal K. Fellows
>Even if we assume that \0's appear only towards the end,
>it will still also strip leading and trailing blanks,
>which the OP didn't ask for.
Bet it's what they really want though, even if they don't know it yet.
;-)
Donal.
Uwe Klein
should work as well
uwe