Re: get() and value()

new topic     » goto parent     » topic index » view thread      » older message » newer message

CChris wrote:
> Since no one has made any comment on this for the last 8 days or perhaps more,
> I'm going to proceed...
> 
> * get() and value() will keep returning a 2 element sequence.

I don't think anyone will disagree with that.

> * get() and value() accept embedded comments in sequence - they crash in
> official
> Eu, which some don't see as a bug;
> * get() and value() ignore leading top level comments. Commments start by "--"
> and end a a \n or \r character, or at end of input;

I don't think anyone will mind if comments in the input data
are accepted, and ignored.
I've actually had occasions where I would have liked to 
add comments to my input data.

> * a new value_from(sequence s,integer starting_point) will allow to scan a
> sequence
> from a starting point between 1 and length(s). This function returns a 4
> element
> sequence: first two as value(), total character read, leading whitespace.

Sounds useful.

> * There's no symmetric get_from(), as it would reduce to a possibly
> inefficient
> sequence of seek() and get(), unless there's some demand for it.

OK.
 
> Additionally, the possibility exists for all of the above to return a new
> error
> code when no value is read and end of input is reached, as opposed to having
> read an unfinished value string. Currently, GET_EOF is returned in both case.
> Adding the extra error code would remove 4-5 lines of code from get.e. Is this
> to be considered?

Sounds OK.
 
> I'll update the whole thing tonight (4pm now).

Go ahead with all of this, 
unless someone seriously disagrees very soon.

Regards,
   Rob Craig
   Rapid Deployment Software
   http://www.RapidEuphoria.com

new topic     » goto parent     » topic index » view thread      » older message » newer message

Search



Quick Links

User menu

Not signed in.

Misc Menu