Re: [Json] Call for Consensus: Proposed Text for "8.1 Character Encoding"

[Peter Cordell <[e-mail]>]

On 14/03/2017 10:46, Carsten Bormann wrote:
>>    00 00        UTF-32BE
>>    00 xx        UTF-16BE
>>    xx 00 00 00  UTF-32LE
>>    xx 00 00 xx  UTF-16LE
>>    xx 00 xx     UTF-16LE
>>    xx xx        UTF-8
>>    xx EOF       UTF-8 (For Carsten's recent comment)
>
> Wow, people are still posting incorrect match tables :-)

Well, it was posted to show the rat hole that specifying detection is, 
and illustrate why we might not want to go there.

> I think that Julian (with my amendment maybe) nailed it, no need to make ever more wronger proposals here.

I think Julian's table missed the case of a string starting with a 
character like U+0100.  If that did turn out to be the route chosen, it 
seemed prudent to point out the omission earlier rather than later. 
(I'll resist an update with xx 00 EOF UTF-16LE.)

 > (Again, in the age of unit testing, all this is much easier to get
 > right for a coder than for a spec writer.)

Only if you don't miss cases to test!

Pete Cordell
Codalogic Ltd
Rules for Describing JSON Content, http://json-content-rules.org