Re: [Boost-docs] The beauty of LATEX

Subject: Re: [Boost-docs] The beauty of LATEX
From: Joel de Guzman (joel_at_[hidden])
Date: 2011-10-26 14:07:27

On 10/26/2011 6:12 PM, Daniel James wrote:
> On 26 October 2011 11:02, Joel de Guzman <joel_at_[hidden]> wrote:
>>>
>>> The problem is that to do a decent job of generating documentation, the engine needs to
>>> understand the document structure (for chunking, generating
>>> paragraphs, anchor placement, footnotes etc.), and that can't be known until after. the
>>> templates have been expanded. So you probably need some form of representation, i.e
>>> document's 'DOM' (although I'm not suggesting that you copy XML's DOM), which you can
>>> transform appropriately. As far as I can tell, the language you're proposing isn't
>>> designed to deal with this. I mentioned the 'block' element not to suggest it should be
>>> used here, but to point out it's an attempt at dealing the same problem. You'll need a
>>> much richer representation.
>>
>> Obviously, we disagree here. I believe the wiki can and should be
>> a preprocessor without full knowledge of the full document schema.
>> I think you are overstating the scope of "the problem", which is
>> simply to know what are to be considered blocks and phrases. There
>> are various ways to deal with this. Matias suggested a simple
>> solution. There surely are other solutions.
> 
> Well, I mentioned some other issues as well. But can you explain to me
> something simple? Say, how a list will be transformed from wiki markup
> to the final document. That doesn't touch on anything I mentioned, but
> it will do a better job of illustrating what you're planning. I might
> have misunderstood what your intermediate representation is going to
> be.

Final document? You might want to back up a little bit. This sub-thread
that Matias started is not about the final document nor the backend.
It is about the wiki preprocessor detection of blocks and phrases.

I think the problem is that you tend to bunch up other issues.
Why can't we focus on the OPs question and not get distracted by
other issues? We can't discuss all the issues all at once. For
now, we are concerned with how the wiki preprocessor generates
template code. Anyway, here's an example for a list:

This:

  # One
  # Two
  # Three
      * Three.a
      * Three.b
      * Three.c
  # Four

Might generate this:

  [orderedlist
    [One]
    [Two]
    [Three
      [itemizedlist
        [Three.a]
        [Three.b]
        [Three.c]
      ]
    ]
    [Four]
  ]

Then again, I am not sure how that is related to the wiki
blocks and phrases question at all.

Regards,
-- 
Joel de Guzman
http://www.boostpro.com
http://boost-spirit.com