bash expansion help

[Christopher Friedt]

Hi Paul,

I'm very familiar w/ sed, and I would gladly use it. However, I don't 
know enough about sed to use these types of regular expressions w/ it. I 
usually just attempt regexp stuff w/ range matching and whitespace.

In this particular case, I would have needed to use some extension that 
had memory as well (for previous matches).

I think something along the line of

(1){a,b,c...} => 1a 1b 1c

where a b c could have any number of the same patterns, would have done 
the trick... I'm not really sure how to code that w/ gnu or posix 
regular expressions... Any suggestions? It's probably pretty straight 
forward.

Luckily, the expansion that I needed was exactly the same type of 
expansion used in bash for filenames. Except in this case, the items 
didn't actually represent filenames. For some reason I couldn't make 
bash do the proper expansion with ${}.

so Steven's suggestion of

echo echo $var | bash

was exactly the solution I was looking for.

I can't actually use perl in this case either, because my task is going 
to need to be run w/ a bare minimum of dependencies, and especially disk 
space. Bash for me was about as large as I wanted to get.

Thanks for the reply in any case though ;-)

Do you know the right way to key in the type of grep regexp mentioned 
above? I'll give it another shot - it would tidy up my code a bit anyway.

~/Chris

Paul King wrote:
> [ Charset UTF-8 unsupported, converting... ]
>> Hi everyone,
>>
>> I'm trying to find some method to expand this variable $arm_opts:
>>
>> "GENERIC_ARM ARM{610,710,720T,920T,922T,926T,_{SA110,SA1100,XSCALE}}"
>>
>> to this:
>>
>> "GENERIC_ARM ARM610 ARM710 ARM720T ARM920T ARM922T ARM926T ARM_SA110 
>> ARM_SA1100 ARM_XSCALE"
> 
> Looks easy enough to parse in Perl. Any reason to insist on Bash?
> 
> In BASH, I find it easier to parse through a list if you use space
> delimiters. I couldn't get it to work with my standard method. I
> suggest, if BASH is insisted upon and Perl must be avoided, that you use
> sed/awk (probably sed more than awk).
> 
> In case you don't know (many seem to think sed is out of "fahsion"), sed
> is a parser, like perl (sed stands for Stream editor). THere are some
> similarities in syntax, but sed is not a language, like Perl is.
> 
> Paul King
> 
>> in bash.
>>
>> Any tips ?
>>
>>
>> Thanks,
>>
>> ~/Chris
>> --
>> The Toronto Linux Users Group.      Meetings: http://gtalug.org/
>> TLUG requests: Linux topics, No HTML, wrap text below 80 columns
>> How to UNSUBSCRIBE: http://gtalug.org/wiki/Mailing_lists
>>
> --
> The Toronto Linux Users Group.      Meetings: http://gtalug.org/
> TLUG requests: Linux topics, No HTML, wrap text below 80 columns
> How to UNSUBSCRIBE: http://gtalug.org/wiki/Mailing_lists
> 
> 
--
The Toronto Linux Users Group.      Meetings: http://gtalug.org/
TLUG requests: Linux topics, No HTML, wrap text below 80 columns
How to UNSUBSCRIBE: http://gtalug.org/wiki/Mailing_lists