Re: ASN.1 and XML

[Al Snell <alaric@alaric-snell.com>]

On Sat, 26 May 2001, Thomas B. Passin wrote:

> One of the interesting things about ASN.1 is that it was originally
> developed for humans to communicate with each other about the data
> structures they were developing - that is, not for automated processing.  I
> understand that this origin has lead to quirks in ASN.1 that make it really
> hard to parse in its full glory.

Aye, this is what the work to define the schema language, ASN.1, as ASN.1
types will hopefully fix...

The issue basically boils down to:

1) Using curly braces for everything (well, it could have been <> :-)
2) Using space seperated lists of things whose parsing depends on what
those things have been defined as previously
3) A shortage of restart symbols for error recovery (like semicolons in
most languages; if the parser gets confused, it can report this fact that
and then skip to the next semicolon to carry on from there and see what it
can make of the rest).

There's a chapter in in the Dubuisson book (a free copy of which is lined
to from just about every ASN.1 site).

> 
> Tom P
> 

ABS

-- 
                               Alaric B. Snell
 http://www.alaric-snell.com/  http://RFC.net/  http://www.warhead.org.uk/
   Any sufficiently advanced technology can be emulated in software