[Asterisk-Dev] IAX spec: Text formats and character sets

[Steve Underwood]

Hi,

I raised this with Mark ages ago, when I started putting Chinese into 
IAX2 messages. I thought it should be specified that all text is Unicode 
in UTF-8 form, but he seemed pretty indifferent to specifying anything.

There is no need to have ASCII + UTF-8. ASCII is a subset of UTF-8, so 
they are fully compatible. Its only when you have 8 bit sets, like the 
PC ones, that compatibility is an issue. Just define that all strings in 
IAX2 are UTF-8, and that is the end of it.

Regards,
Steve


Olle E. Johansson wrote:

> Good morning, bonjour, god dag!
>
> The IAX spec has to be more specific in regards to character sets
> used in various messages.
>
> The numbers and CID names are specified as "ASCII encoded text".
> There's not spec of HTML text format. That assumes ISO-8859-1, 8 bit.
>
> * ASCII is 7 bit only. A-Z
> * As far as I know, Caller ID Name is also ISO 8859-1, 8 bit.
> * SIP is UTF8, meaning I can include not only Swedish ÅÄÖ but almost
>   any character set, except the secret magic runes...
>
> We need to set an Asterisk /IAX2 standard for text frames, "numbers" 
> and caller ID names. Asterisk is more and more becoming an 
> international project and we need to work this out before we've come 
> too far in the
> "all text strings are US ASCII" soup and can't clear this up. I've 
> been working quite a long time with network communication and have 
> carried a large number of Swedish keyboards to US programmers that 
> needed to learn that there where more characters in the world than A-Z 
> and even stranger keyboards (as the participants on Astricon Training 
> in Stockholm quickly had to learn :-) )
>
> I would like to see that Asterisk internally uses two sets of names 
> and phone numbers - one alphanumeric UTF 8 and one ASCII. That way we 
> can handle both SIP and IAX2 and know what kind of object we're handling.
> Right now, the SIP channel happily adds UTF8 caller ID names to the 
> CID*name field, or UTF8 extensions/phone numers/user names to the 
> called extension field.
>
> Leif Madsen and I have written a proposal that we would like the 
> community to review, called alphaextensions. In this, we try to define 
> a way to add proper handling of international character sets in 
> extensions (a la SIP) without breaking backwards compatibility with 
> the current dialplans and applications within Asterisk.
>
> This proposal is available at
> http://edvina.net/asterisk/alphanumericextensions.pdf
>
> ...and we would very much like input on that. This became even more 
> important after browsing through the IAX spec (IAX2 spec? :-) ).
>
> Another quick comment: I think we should format the IAX spec not as an 
> Internet Draft, but as an Informational RFC like the early NFS RFCs, 
> since the wording of the first paragraphs tells us that it will never 
> be published as an IETF draft any way...