New metadata structure proposal
Christopher Diggins
define dip : ('A 'b ('A -> 'C) -> 'C 'b)
{{
summary pops an item
semantics [$B] $a dip == $B $a
test
usage 1 [2] dip
expected 2 1
result 2 1
test
usage 1 2 [[3] dip] dip
expected 3 1 2
result 3 1 2
history
created
by Christoper
on 7/13/2006
modified
by Christopher
on 7/13/2006
}}
{
swap quote compose apply
}
Frank A. Krueger
it would make some of the structure more evident. That said, I
understand the desire to not have them as it makes the doc feel less
computery and more natural.
Also, what is the difference between test/expected and test/result ?
Also :-), I have never been a big fan of revision entries in code docs
unless it is _completely_ managed by a tool. But I think you argee
with me on that one.
-Frank
> }- Hide quoted text -
>
> - Show quoted text -
Christopher Diggins
Glad you are on the list!
I agree about the colons and in fact I already decided to go that route :-)
Check out: http://cat-language.googlecode.com/svn/trunk/standard.cat
For an example of the current spec. for meta-comments.
Expected and result, was just me being careless. I thought originally
that the IDE would keep track of broken tests in the source, but I
decided to abandon that.
As for revisions, I decided to not to worry about that for now. It
would up to the tools to manage it. The meta-comments are open-ended,
so a tool can choose to use meta-comments for that purpose, though I
probably won't bother with that in my own Cat implementation.
As for your other suggestion regarding comments, I do plan on adding a
link to the group. One idea I have is to post my blog posts (at least
the ones that pertain to Cat) to the mailing list, and simply echo
them on the blog. I'd be curious to hear people's thoughts on that
idea.
- Christopher
Frank A. Krueger
A little slow sometimes!
> I agree about the colons and in fact I already decided to go that route :-)
> Check out:http://cat-language.googlecode.com/svn/trunk/standard.cat
> For an example of the current spec. for meta-comments.
Looking through standard.cat now... I like it! Some moments had to be
spent readjusting my brain to understand function docs coming _after_
the function name, but aside from that they are very readable and
understandable.
> As for your other suggestion regarding comments, I do plan on adding a
> link to the group. One idea I have is to post my blog posts (at least
> the ones that pertain to Cat) to the mailing list, and simply echo
> them on the blog. I'd be curious to hear people's thoughts on that
> idea.
My vote is that it's a good idea.
-Frank