Re: [Tools-discuss] boilerplate BCP14 and UTF8

[Michael Richardson <[e-mail]>]

Carsten Bormann <[e-mail]> wrote:
    >> Signed PGP part
    >>
    >> Michael Richardson <[e-mail]> wrote:
    >>> Kramdown gives me this error in some documents:
    >>
    >>> /usr/local/rvm/gems/ruby-2.6.5/gems/kramdown-rfc2629-1.2.12/bin/kramdown-rfc2629:333:in
    >>> `encode': U+00A0 from UTF-8 to US-ASCII (Encoding::UndefinedConversionError)
    >>
    >>> when I include {::boilerplate bcp14}
    >>
    >>> I don't have a BOM in my markdown. I don't think I should.
    >>> It feels like the output ought to be in UTF-8, but isn't somehow.
    >>
    >> I found:
    >> coding: us-ascii

    > How did that get there?

    >> in my markdown, and removed it.  Maybe we can make the error more meaningful?

    > Probably.  Now https://github.com/cabo/kramdown-rfc2629/issues/65

    >> Or maybe the template that is copied in could be US-ASCII only?

    > I’ll need to check how to do the NBSP in “BCP 14” in markdown without
    > leaving ASCII encoding.  But maybe the whole idea of allowing “coding:
    > us-ascii” is not so useful now.

Maybe:

if coding == "us-ascii"
  puts "WARNING: all sorts of things break if you insist on XML output being US-ASCII."
  puts "See https://github.com/cabo/kramdown-rfc2629/issues/65"
end