Fix composition of markup lists containing markup command list calls

Fix composition of markup lists containing markup command list calls

Composed markup lists in the parser apply normal markup commands to
markup lists, like
\with-color #red { a b c }

If those markup lists contained calls to markup list commands (like
\column-lines), the resulting behavior was undefined, at worst
producing segfaults (cf issue 2393).  This patch series cleans up the
inconsistencies.  It contains the following commits:


Allow unbraced forms of composed markup lists

Since composing markup lists may now work at run time, there is no
particular need for the associated syntactic restrictions.

Allow markup lists to be composed at run-time

Using a markup command on a markup list generated by a markup list
command was not previously possible.  Now things like

  \with-color #red { \column-lines { x y z } }

work as well.

Move map-markup-command-list into parser internals


Simplify interpret-markup-list


Use fold in map-markup-command-list


Text_interface::is_markup should reject markup-list-command calls

[lemzwerg, 2013-03-24 05:55:00]

LGTM.

[dak, 2013-03-24 12:08:01]

Combine successive composed markup list commands

[dak, 2013-03-24 14:33:22]

I just changed the summary of the commit messages since one commit in the middle
referred to a syntax that was only provided by a commit later in the sequence. 
Just a technicality.

[janek, 2013-03-24 21:50:52]

judging mostly from description, LGTM.

[Ian Hulin (gmail), 2013-03-25 11:37:59]

LGTM, apart from comment on the doc-string.

Cheers, Ian

https://codereview.appspot.com/7799048/diff/3001/scm/define-markup-commands.scm
File scm/define-markup-commands.scm (right):

https://codereview.appspot.com/7799048/diff/3001/scm/define-markup-commands.s...
scm/define-markup-commands.scm:4286: (let ((key (make-symbol "key")))
I don't understand the last sentence of the doc-string at first reading.  Is it
something like:
"Although the means of getting such results differ, the results themselves
should be comparable"

[dak, 2013-03-25 21:07:45]

Address Ian's comment.  Hopefully.

[Ian Hulin (gmail), 2013-03-27 15:00:17]

On 2013/03/25 21:07:45, dak wrote:
> Address Ian's comment.  Hopefully.

LGTM

Cheers, Ian

[dak, 2013-03-27 21:08:40]

Some comments, make robust about throw not actually happening