Re: [xml2rfc-dev] When is @ascii required?

[Carsten Bormann <[e-mail]>]

On Oct 27, 2019, at 19:40, Henrik Levkowetz <[e-mail]> wrote:
> 
> Hi Carsten,
> 
> On 2019-10-27 18:24, Carsten Bormann wrote:
>> I’m not quite sure I understand when I have to provide the ascii=“” attribute.
>> 
>> I can have authors that have umlauts without further ado.
>> If I have organizations with umlauts, I need to put in ascii=“”, or I get:
>> 
>> t2trg-sworn.xml(20): Error: Found non-ascii content without matching ascii attribute in <organization>: Universit?t Bremen TZI
> 
> This is inconsistent.

Thanks.  So I’ll continue those experiments with the next xml2rfc version.

(I think that the general direction to allow Latin input but require manual ASCII transliteration for non-Latin is right — most English-speaking people should be able to approximately render Latin outside ßðÐþÞŋƣƦ and maybe İıĸ; there are transliteration libraries for other scripts as well but employing them is probably not a very stable approach.)

>> (Note also that the error message has some mojibake.)
> 
> Is this from the web service or on the command line?

Command line.  I notice that I had a Python 2.7 install running; with 3.7 (I know) I get:

t2trg-sworn.xml(20): Error: Found non-ascii content without matching ascii attribute in <organization>: b'Universit?t Bremen TZI’

which is interesting.

Grüße, Carsten