On 7 Feb 2014, at 2:08, Matt Neuburg wrote:
I can't believe this would be expensive
It would add complexity to the parser, its API, and it would affect its run-time.
and it would solve a LOT of Asciidoc grammar problems, where there are many multi-line patterns such as this one.
I have considered allowing \n in regular expressions to match multiple lines. I did this for the firstLineMatch pattern (which can match multiple leading lines, if \n is part of its regexp).
Though it is no simple change and it would affect the run-time, but ideally TM would know when such patterns were involved, and thus the penalty would be insignificant for grammars that do not use such patterns.
The lack of setext header support is definitely a thorn in my eye.