Re: [xml2rfc] RFC Bibtex format doi numbering incorrect

[Brian E Carpenter <[e-mail]>]

On 29/05/2017 17:49, Julian Reschke wrote:
> On 2017-05-28 23:49, John Levine wrote:
>> In article <[e-mail]> you write:
>>> On 2017-05-28 17:04, John R Levine wrote:
>>>> It's easier to parse stuff out of rfc-index.xml in the same directory.
>>>> Yes, Julian, there is a doi field.
>>>
>>> Which makes something that should be easy hard.
>>
>> I'm sorry, but this is getting ridiculous.
>>
>> Let's say I have a a document with a reference RFC7511 and run it
>> through xml2rfc.  Somehow, it will turn that reference into strings
>> like "Scenic Routing for IPv6" and "April 1, 2015".  Where do those
>> strings come from?
>>
>> Wherever they come from, that's where the DOIs come from because
>> they're all stored in the same place.  Looking up DOIs is not
>> negotiable -- the rules about DOIs are fixed, and not only for the
>> IETF.  Please just fix the code.  It's hard to believe that looking up
>> one more field from the place where you look up a dozen already would
>> be hard.
>> ...
> 
> I'm not looking them up at all. I maintain an XML2RFC processor which is 
> capable of
> 
> a) inserting DOIs when they are missing, and
> b) checking RFC DOIs when they are present.
> 
> Neither should require an access to an external database.

But it must, since the DOIs are arbitrary strings made up by the RFC Editor.
I don't understand your point: wherever you look up other arbitrary
strings, such as the document title, you can look up the DOI string.

I don't see why parsing <doi>10.17487/RFC0001</doi>
is any different from parsing <title>Host Software</title>

Regards
   Brian