[9fans] ^ and $ in structural expressions

[9fans] ^ and $ in structural expressions

[Anthony Martin]

In his structural expression paper, Rob briefly mentions
that ^ and $ match on line boundaries for compatibility
reasons and because of problems with the implementation.

Lately, I've been running into situations where it would
be convenient to have them match what they should, i.e.,
the beginning and end of the selected data.

Rob (and any others that have experimented), could you
please elaborate on the problems that were encountered?

My initiative-to-annoyance ratio is approaching one but
I thought I'd ask here before coding.

Thanks,
  Anthony

Re: [9fans] ^ and $ in structural expressions

[roger peppe]

> Lately, I've been running into situations where it would
> be convenient to have them match what they should, i.e.,
> the beginning and end of the selected data.

i've often wanted something similar, but i'm not sure you can
re-use ^ and $, as for example ,x/;$/d would stop working.

i thought perhaps « and » would do the job, but that's
not considered a good idea.

here are some relevant articles from the archives:

    http://9fans.net/archive/2003/10/241
    http://9fans.net/archive/2003/10/245
    http://9fans.net/archive/2003/10/340
    http://9fans.net/archive/2003/10/341
    http://9fans.net/archive/2003/10/343

basically, it might be useful, but there's no way to fit it into the syntax.

Re: [9fans] ^ and $ in structural expressions

[Hongzheng Wang]

And it might also make regexp too complicated.

On Fri, Feb 29, 2008 at 11:48 PM, roger peppe <rogpeppe@gmail.com> wrote:
> > Lately, I've been running into situations where it would
>  > be convenient to have them match what they should, i.e.,
>  > the beginning and end of the selected data.
>
>  i've often wanted something similar, but i'm not sure you can
>  re-use ^ and $, as for example ,x/;$/d would stop working.
>
>  i thought perhaps « and » would do the job, but that's
>  not considered a good idea.
>
>  here are some relevant articles from the archives:
>
>     http://9fans.net/archive/2003/10/241
>     http://9fans.net/archive/2003/10/245
>     http://9fans.net/archive/2003/10/340
>     http://9fans.net/archive/2003/10/341
>     http://9fans.net/archive/2003/10/343
>
>  basically, it might be useful, but there's no way to fit it into the syntax.
>



-- 
HZ

Re: [9fans] ^ and $ in structural expressions

[erik quanstrom]

i'm not arguing for or against the idea here, just want to point
out that i don't think syntax is the limiting factor.

if « and » are not acceptable, then we can reuse existing characters.
having ^ and $ loose their idempotency would not be a large inconvience.
^^ and $$ could easily do.

adding an expression anchor should not be a major complication.

- erik