Trim leading and trailing whitespace from comment on read

[colindean]

So that the XML and any other representation doesn't have the whitespace.

[colindean]

Looks like trimming at the point of creation could be appropriate. That's Parser.scala:22. It could also be in the definition of the comment extractor at line 43. This is my first time attempting to understand parser combinators, so I'm not quite sure what is right.

[hrj]

Line 43 seems more appropriate, as it sits at a higher level. Line 22 does the tokenisation (lexical phase).