Re: [O] New tutorial on using Emacs lisp for Org spreadsheet formulas

[Nicolas Goaziou]

Hello,

Achim Gratz <address@hidden> writes:

> Nicolas Goaziou <address@hidden> writes:
>> I don't see what is clumsy with affiliated keywords.
>
> Well, you write the arguments before starting the source block, getting
> everything wardbacks (and I don't think that's easier to parse, but you
> would know better(*)).

I think you're mixing src blocks and Babel. A src block contains,
fontifies and allows to edit source code. Babel arguments (or caption,
or name) are meta-data that applies to the whole src block element, not
to its contents.

> Last but not least I can never remember if any of the "#+" stuff is
> supposed to end with a ":"

Affiliated keywords all do. If one doesn't, it's a bug.

> and whether it wants another ":" in front of the arguments or not.

That's another topic, and I agree that, at one point or another, we'll
have to unify the way parameters are written. IMO, Babel usage is a good
standard.

> (*) Reminds me I'll have to try what happens if I write header
> arguments that are detached from the source block.  Devious, I know,
> but the manual does not really specify anything.

But Org elements does. Affiliated keywords must be attached to their
relative element.

> If you look carefully, that source block above does not contain a
> drawer.  It has a (hypothetical) associated multiline header argument
> that happens to look and feel like a drawer.

Then it's a duck.

>> Though, drawers can contain src-blocks.
>
> Is there any specification that they must fully contain them?

You may want to read comments in org-element.el. I also posted some
explanations when I announced the library.

> Anyway, that was just an idea and if nobody likes it, that's fine with
> me, too.

I just wanted to let you know that the idea of putting "quasi-drawers"
within a src-block is against the model described by the current parser.


Regards,

-- 
Nicolas Goaziou