Re: Calling a function with implementation in awk

[alex xmb ratchev]

On Sat, Mar 18, 2023, 2:21 AM uzibalqa <uzibalqa@proton.me> wrote:

> I have a number of bash functions that consist in their entirety of awk
> commands,
>

u define those as vars
awkfuncs=( 'code1 ..' 'code2 ..' .. )
printf -v awkstr %s\\n "${awkfuncs[@]}"

because I do not want a complicated mix of different languages in the same
> function.
>
> Thus I have the following
>

i get your pasted codes to 0 %

kuelseq ()
>  {
>   awk -v ctp="$1" -v rst="$2" -v lseq="$3"  \
>     'BEGIN { fm1="%s%s%s\n" ; fm2="%s\n"
>         n = split(lseq, a, ",")
>         for ( i=1 ; i<=n ; i++ ) laggr[a[i]] }
>
>      { if (NR in laggr) printf(fm1, ctp, $0, rst)
>        else printf(fm2, $0) }'
>  }
>
> I want to call the function above from another bash function
> called "kue", passing data either through redirection or a pipeline.
>
> kue ()
>  {  impl="REDR"
>     if [[ "$impl" == "REDR" ]]; then      # Redirection
>       kuelseq "$ctp" "$rst" "$lseq" < <(printf "%s\n" "$@")
>     elif [[ "$impl" == "PIPL" ]]; then    # Pipeline
>       printf '%s\n' "$@" | kuelseq "$ctp" "$rst" "$lseq"
>     fi
>  }
>
> I am under the impression that I am missing something in the function
> "kuelseq", as I do require a way to pass the data supplied by "kue"
> through "$@" into the function "kuelseq".
>
> How should I handle the input problem to "kuelseq" ?
>
>
>