[wikka-community] line breaks mid 'markup'

Dave Pawson dave.pawson
Thu Sep 10 10:56:15 GMT 2009

2009/9/10 Dario Taraborelli <dartar at wikkawiki.org>:

> there are many ways in which you can force a markup parser (not just
> Wikka's) to do things it was not designed to, that doesn't mean we encourage
> our users to take these cases as the rule.
> Sorry if it sounds disappointing.

Not disappointing, no. Just something that could be addressed,
I guess in this case through documentation.

>> But it does contradict the newlines within links regex? If they are
>> allowed
>> in one place, why not in another would be the logic I guess.
> let me stress it again. Literal newlines in the links regex are not
> officially supported.

Nor discouraged through documentation?

> The table markup parser is pretty robust, but it currently doesn't support
> multiline elements (such as lists) within table cells. This is why using a
> literal \n within a table cell (as well as within a link) is not supported.

Again undocumented  Dario? I guess this is my concern.
I went to the documentation to find out what was allowed.

>>> In the next release (using trunk as a codebase, after merging back
>>> 1.2-specific changes), the regular expressions used by the formatter will
>>> be
>>> stored in a library to simplify their maintenance and documentation. This
>>> should facilitate expressing formatting rules in other ways (hopefully
>>> including BNF).

Great. Look forward to it, Another oddity I found. Right at the top

"Note: Anything between 2 sets of double-quotes is not formatted."

Surely this is an alternative to monospaced text? #mono#
Nothing is said about it being block or inline level though.

>> Summary: ?\n is allowed in a link, but nowhere else within an inline?
> No, \n is *not* officially supported within a forced link. I was not behind
> the decision to losen the link regex, I would be happy to enforce a stricter
> markup.
> I agree that leaving undocumented formatter options lurking around may be an
> annoying source of confusion.

I'd call it a weakness awaiting to be exploited, as per my earlier browser
Like many users, if it's not documented I make my own assumptions, quite
likely different from others, and use them.
   I guess the annoyance comes when the devs make a decision to tighten
things up (parser or documentation) and they make a different decision
to the user? The user then has to change all his|her markup?

>> AFAIK that makes it the one exception?
>> Everything else about the syntax is line based.
>> ?Or are there more exceptions?
> the table row syntax is line based. Tables (as multi-row elements) are by
> definition multilines.
> literal linebreaks are allowed within text formatters (such as **strong**
> and //emphasis//), <<floats<<, ?%%code blocks%%

Another assumption I made. I took these as being inlines, hence
** line
 broken  **
I said was wrong and reported as an error.

Not very clear Dario, is it.


Dave Pawson
Docbook FAQ.

More information about the community mailing list