More secure handling of U+0000

[nwellnhof]

Following my proposal, this sentence was added to the spec:

For security reasons, a conforming parser must strip or replace the Unicode character U+0000.

Thinking more about it, stripping characters is a security risk in itself. I'd propose to change the spec to read:

For security reasons, the Unicode character NULL (U+0000) must be replaced with the replacement character (U+FFFD).

It might also make sense to put this sentence in a separate section, for example 2.3 Insecure characters.

This matches the behavior of the C library. We could also allow a parser to error out whenever it encounters a NULL character.

This also clarifies the spec. See http://talk.commonmark.org/t/how-exactly-should-parsers-handle-u-0000/993/2