emacs-orgmode
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Re: [O] Why does a column's name need to start with a letter? Change


From: frozenlock
Subject: Re: Re: [O] Why does a column's name need to start with a letter? Changed source with no perceivable
Date: Sat, 05 Mar 2011 17:01:25 +0000

Yes, it's very well explained, thank you.
I'll try to encode my names by removing any "-" and by adding a dummy letter at the beginning.

Is there another way of doing simple database in org-mode? By putting a component's name in a cell, I would like if the other columns could fill them-self with the component's characteristics.

(My apologies for the double e-mail, the first one wasn't sent to the mailing list)


On , Carsten Dominik <address@hidden> wrote:
> Hi Christian,
>
>
>
> On 5.3.2011, at 04:35, Frozenlock wrote:
>
>
>
> > Greetings,
>
> >
>
> > First and foremost, I must say I'm a new Org-mode user... as well as a new emacs user... and have only limited experience with lisp.
>
> >
>
> > I'm using a table in org-mode as a database, from which I retrieve information as needed from other tables.
>
> > This database stores multiples components with names starting by numbers and with the character "-" in them. For example: "10K-AN-D8".
>
> > I want to be able to refer to its column with $10K-AN-D8. Yet, this is impossible at the moment.
>
> >
>
> > I've looked in org-table.el and found, at line 2038:
>
> >
>
> > (if (string-match "^[a-zA-Z][a-zA-Z0-9]*$" name)
>
> >
>
> > which, once replaced by this:
>
> >
>
> > (if (string-match "^[-_a-zA-Z0-9]*$" name)
>
> >
>
> > Provides me with the ability to name (and refer) columns in any way I want.
>
> >
>
> > However, I'm reluctant to use this feature; surely there was a reason for this limitation.
>
> > Could someone point it out for me?
>
>
>
> This is limitation is present to disambiguate formulas.
>
> In you example $10K-AN-D8, D8 is already a field reference,
>
> so Org does not know if you mean $10K-AN - @8$4.
>
> Furthermore, you could easily arrive at variables like
>
> $10-AN-D8, and then what should the $10 mean?
>
>
>
> Of cause one could disambiguate heuristically by checking
>
> which names have been defined.  In fact, name replacement
>
> happens first, and this is why your patch appears to
>
> work.  But the side effect would be that introducing
>
> new names could change the interpretation of an
>
> existing equation.  All this is unstable and
>
> unpredictable.
>
>
>
> Hope this makes it clear.
>
>
>
> - Carsten
reply via email to

[Prev in Thread] Current Thread [Next in Thread]