I've got a text file with names like this:
Variable1[30]
extraVar22[22]
Test42Exists[104]

how do i get just the array name in each case
e.g.:
N="Variable1[30]"
echo "$N"|sed <what goes here>"
or
echo "$N"|eval match "$N" \(<what goes here>\)"

the output would be Variable1 (dropping the [30])

It can be sed or bash, or whatever, I'm using this in a bash script.
I was playing with echo 'eval match "$N" \( stuff here \)'
but i couldnt get it to work right, array names can be a mix of upper and
lower case, numbers (and possibly underscores).
Thanks
Eric

"Eric" <Scorpus@gordinator.org> wrote in message
news:7O6dnSa6OeSKQYrVnZ2dnUVZ_jqdnZ2d@comcast.com. ..
> I've got a text file with names like this:
> Variable1[30]
> extraVar22[22]
> Test42Exists[104]
>
> how do i get just the array name in each case
> e.g.:
> N="Variable1[30]"
> echo "$N"|sed <what goes here>"
> or
> echo "$N"|eval match "$N" \(<what goes here>\)"
>
> the output would be Variable1 (dropping the [30])
>
> It can be sed or bash, or whatever, I'm using this in a bash script.
> I was playing with echo 'eval match "$N" \( stuff here \)'
> but i couldnt get it to work right, array names can be a mix of upper and
> lower case, numbers (and possibly underscores).
> Thanks
> Eric

Did you try a cut?
echo "$N" | cut -d"[" -f1
or just cut -d"[" -f1 yourfile

On Wednesday 30 April 2008 04:40, Eric wrote:

> I've got a text file with names like this:
> Variable1[30]
> extraVar22[22]
> Test42Exists[104]
>
> how do i get just the array name in each case
> e.g.:
> N="Variable1[30]"
> echo "$N"|sed <what goes here>"
> or
> echo "$N"|eval match "$N" \(<what goes here>\)"
>
> the output would be Variable1 (dropping the [30])
>
> It can be sed or bash, or whatever, I'm using this in a bash script.
> I was playing with echo 'eval match "$N" \( stuff here \)'
> but i couldnt get it to work right, array names can be a mix of upper and
> lower case, numbers (and possibly underscores).

Plenty of ways to do that, in addition to cut, which was already suggested.

$ printf "%s\n" 'Test42Exists[104]' | awk -F'[' '{print $1}'
Test42Exists

$ printf "%s\n" 'Test42Exists[104]' | sed 's/\[[^]]*\]$//'
Test42Exists

$ expr 'Test42Exists[104]' : '\([^[]*\)'
Test42Exists
$ expr match 'Test42Exists[104]' '\([^[]*\)'
Test42Exists

$ var='Test42Exists[104]'
$ printf "%s\n" "${var%\[*\]}"
Test42Exists

And, since you use bash:

$ var='Test42Exists[104]'
$ printf "%s\n" "${var/%\[*\]/}"
Test42Exists

--
D.

On 30 Apr., 04:40, Eric <Scor...@gordinator.org> wrote:
> I've got a text file with names like this:
> Variable1[30]
> extraVar22[22]
> Test42Exists[104]
>
> how do i get just the array name in each case
> e.g.:
> N="Variable1[30]"
> echo "$N"|sed <what goes here>"
> or
> echo "$N"|eval match "$N" \(<what goes here>\)"
>
> the output would be Variable1 (dropping the [30])
>
> It can be sed or bash, or whatever, I'm using this in a bash script.
> I was playing with echo 'eval match "$N" \( stuff here \)'
> but i couldnt get it to work right, array names can be a mix of upper and
> lower case, numbers (and possibly underscores).
> Thanks
> Eric

N="Variable1[30]"
echo ${N%%[[]*}


Janis

On Wednesday 30 April 2008 10:06, Janis wrote:

> N="Variable1[30]"
> echo ${N%%[[]*}

Why not just

echo ${N%[*}

then?

--
All the commands are tested with bash and GNU tools, so they may use
nonstandard features. I try to mention when something is nonstandard (if
I'm aware of that), but I may miss something. Corrections are welcome.

On 30 Apr., 10:13, pk <p...@pk.invalid> wrote:
> On Wednesday 30 April 2008 10:06, Janis wrote:
>
> > N="Variable1[30]"
> > echo ${N%%[[]*}
>
> Why not just
>
> echo ${N%[*}
>
> then?

$ sh -c 'N="Variable1[30]"; echo ${N%%[*}'
Variable1[30]
$ sh -c 'N="Variable1[30]"; echo ${N%%[[]*}'
Variable1

More reliable (with other shells).

Janis

2008-04-30, 10:13(+02), pk:
> On Wednesday 30 April 2008 10:06, Janis wrote:
>
>> N="Variable1[30]"
>> echo ${N%%[[]*}
>
> Why not just
>
> echo ${N%[*}
[...]

$ zsh -c 'echo ${N%[*}'
zsh:1: bad pattern: [*
$ ash -c 'echo ${N%[*}'
Variable1[30]
$ ksh93 -c 'echo ${N%[*}'
Variable1[30]
$ pdksh -c 'echo ${N%[*}'
Variable1[30]
$ bash -c 'echo ${N%[*}'
Variable1

From SUSv3:
An expression containing a '[' that is not preceded by a
backslash and is not part of a bracket expression produces
undefined results.

That's in the RE section, but the globbing spec refers to REs
for the bracket expression.

--
Stéphane

On 30 Apr., 10:34, Janis <janis_papanag...@hotmail.com> wrote:
> On 30 Apr., 10:13, pk <p...@pk.invalid> wrote:
>
> > On Wednesday 30 April 2008 10:06, Janis wrote:
>
> > > N="Variable1[30]"
> > > echo ${N%%[[]*}
>
> > Why not just
>
> > echo ${N%[*}
>
> > then?
>
> $ sh -c 'N="Variable1[30]"; echo ${N%%[*}'
> Variable1[30]
> $ sh -c 'N="Variable1[30]"; echo ${N%%[[]*}'
> Variable1

Ah, you had two questions in your posting.

$ sh -c 'N="Variable1[30][40]"; echo ${N%[[]*}'
Variable1[30]
$ sh -c 'N="Variable1[30][40]"; echo ${N%%[[]*}'
Variable1

I used double %% because (depending on the parsed
language) arrays can possibly have multiple dimensions.

>
> More reliable (with other shells).
>
> Janis

On Wednesday 30 April 2008 10:39, Janis wrote:

> I used double %% because (depending on the parsed
> language) arrays can possibly have multiple dimensions.

I did not think about that, since the OP only posted examples of
single-dimension arrays.

Thanks!

--
All the commands are tested with bash and GNU tools, so they may use
nonstandard features. I try to mention when something is nonstandard (if
I'm aware of that), but I may miss something. Corrections are welcome.

On Wednesday 30 April 2008 10:36, Stephane CHAZELAS wrote:

> From SUSv3:
> An expression containing a '[' that is not preceded by a
> backslash and is not part of a bracket expression produces
> undefined results.

Ah ok, I just stumbled upon another bash peculiarity.
Thanks!

--
All the commands are tested with bash and GNU tools, so they may use
nonstandard features. I try to mention when something is nonstandard (if
I'm aware of that), but I may miss something. Corrections are welcome.

Eric wrote:

> I've got a text file with names like this:
> Variable1[30]
> extraVar22[22]
> Test42Exists[104]
>
> how do i get just the array name in each case
> e.g.:
> N="Variable1[30]"
> echo "$N"|sed <what goes here>"
> or
> echo "$N"|eval match "$N" \(<what goes here>\)"
>
> the output would be Variable1 (dropping the [30])
>
> It can be sed or bash, or whatever, I'm using this in a bash script.
> I was playing with echo 'eval match "$N" \( stuff here \)'
> but i couldnt get it to work right, array names can be a mix of upper and
> lower case, numbers (and possibly underscores).
> Thanks
> Eric
Thanks all, I finally came up with this by experiment:
arrName=$(echo `expr match "$N" '\([A-Za-z0−9]*\)'`)
but I'll probably replace it with something from replies to my post,
your suggestions seem more to the point.
Thanks,
Eric

On Thursday 1 May 2008 07:36, Eric wrote:

> Thanks all, I finally came up with this by experiment:
> arrName=$(echo `expr match "$N" '\([A-Za-z0−9]*\)'`)

arrName=$(expr match "$N" '\([A-Za-z0−9]*\)')

Consider using [[:alnum:]] instead of [A-Za-z0−9]. Finally, underscores
("_") and possibly other characters are probably allowed in names.

--
D.

Eric wrote:
> Eric wrote:
>
>
>>I've got a text file with names like this:
>> Variable1[30]
>> extraVar22[22]
>> Test42Exists[104]
>>
>>how do i get just the array name in each case
>>e.g.:
>> N="Variable1[30]"
>> echo "$N"|sed <what goes here>"
>>or
>> echo "$N"|eval match "$N" \(<what goes here>\)"
>>
>>the output would be Variable1 (dropping the [30])
>>
>>It can be sed or bash, or whatever, I'm using this in a bash script.
>>I was playing with echo 'eval match "$N" \( stuff here \)'
>>but i couldnt get it to work right, array names can be a mix of upper and
>>lower case, numbers (and possibly underscores).
>>Thanks
>>Eric
>
> Thanks all, I finally came up with this by experiment:
> arrName=$(echo `expr match "$N" '\([A-Za-z0−9]*\)'`)

Why this double program substitution $(echo `...`) instead of using
just one?

arrName=$(expr match "$N" '\([A-Za-z0-9]*\)')

Though I'd use something simpler than that, anyway.

> but I'll probably replace it with something from replies to my post,
> your suggestions seem more to the point.
> Thanks,
> Eric