Re: How to substitute in a string for the first n matches?

[Wolfgang Laun]

I was just beginning to write a message to carp about this m x n (not
*really* quadratic) and the lack of an early stop if there aren't n
matching substrings. So:

function repn( string, n, regexp, rep,   res ){
    while( n-- > 0 && match(string, regexp) ){
       res = res substr(string, 1, RSTART-1) rep
       string = substr(string, RSTART+RLENGTH)
    }
    return res string
}

Cheers
Wolfgang

On Mon, 29 Mar 2021 at 16:24, Peng Yu <pengyu.ut@gmail.com> wrote:

> This can be slow as the time complexity should be linear wrt to the
> length of the string (suppose the regex to match is randomly spersed
> in the input string), but this implementation is quadratic.
>
> Should gensub() allows a more flexible "h" parameter so that it can
> accept a range like 1-3 (besides just numbers and "g"/"G") which means
> replacing the 1st, 2nd, and 3rd matches?
>
> On Mon, Mar 29, 2021 at 1:33 AM <arnold@skeeve.com> wrote:
> >
> > This is the right idea. For production use it should be
> > wrapped in a nice function.
> >
> > Arnold
> >
> > david kerns <david.t.kerns@gmail.com> wrote:
> >
> > > don't hate me!
> > >
> > > awk -v n=4 'BEGIN{x="rrrrrrrrrr";for(i=0;i<n;i++){sub("r","x",x)}print
> x}'
> > >
> > > On Fri, Mar 26, 2021 at 1:57 PM Peng Yu <pengyu.ut@gmail.com> wrote:
> > >
> > > > Hi,
> > > >
> > > > man gawk says.
> > > >
> > > > """
> > > > gensub(r, s, h [, t])
> > > >
> > > > h is a number indicating which match of r to replace.
> > > > """
> > > >
> > > > Is there a way to perform string substitution for the first n
> matches in
> > > > awk?
> > > >
> > > > --
> > > > Regards,
> > > > Peng
> > > >
> > > >
>
>
>
> --
> Regards,
> Peng
>
>

-- 
Wolfgang Laun