Re: [Boost-docs] Quickbook document author Unicode a accent R&#xE5 isnot converted.

Subject: Re: [Boost-docs] Quickbook document author Unicode a accent R&#xE5 isnot converted.
From: Daniel James (dnljms_at_[hidden])
Date: 2010-08-25 10:19:09

On 25 August 2010 09:25, John Maddock <boost.regex_at_[hidden]> wrote:
>
> How does it tell when the escape has ended and the subsequent text starts?
> Would a template expansion be better?

It's how unicode is escaped in the rest of quickbook. '\u' is followed
by 4 digits, '\U' by 8. It's a pretty common convention, I think
Python, Java and C# use it. Template expansions are problematic
because they currently can't be used in a lot of places. There's also
problems with using escaped xml, it's specific to the target language
rather than quickbook so it's harder to reason about (e.g. at some
point I want to improve the 'simple' formatting markup, I have no idea
how that can interact with escaped xml), will make it harder to target
different formats and I think it doesn't really work inside
attributes.

Also, square brackets tend to represent the structure of the document
- a lot like tags in html and xml, so characters really should be
represented by something else. It's good to use punctuation in a
consistent manner, so that documents can be scanned over to see how
they're made up. Backslash escapes represent characters, while square
bracket markup represent phrases or blocks.

Daniel