Necessary whitespace?

Also, I'm reminded of Benito van der Zander's post in xquery-talk a while 

In particular, consider the examples:
     12!(12 div.)
     12 div-3
     3!(12 div-.)

At the time, Christian GrĂ¼n said these were all invalid, and Ghislain Fourny 
agreed that they did not parse.

However, I don't think the spec says they're illegal. Note that in the 
putative parse, 'div' is a keyword, so you can't apply the rule that 
requires a symbol separator between an NCName and "-" or ".".

And if you want to apply the "longest possible match" rule, you would have 
to argue that an IntegerLiteral followed by an NCName is "consistent with 
the EBNF", which seems unlikely to fly.

(For what it's worth, the applets reject all of these examples, but that 
doesn't prove anything.)


Received on Tuesday, 22 March 2016 21:25:55 UTC