[XML-DEV Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message]

Re: Many different syntaxes in XML - is that good languagedesi

• From: Pete Cordell <pete++xmldev@codalogic.com>
• To: Norman Gray <norman.gray@glasgow.ac.uk>,Roger L Costello <costello@m...>
• Date: Mon, 7 Mar 2022 15:52:46 +0000

On 07/03/2022 14:20, Norman Gray wrote:

As Peter mentioned, those aren't different 'syntaxes'.  There's just one syntax being described here -- the document syntax -- which has a relatively small number of lexical elements.  Thus 'the syntax' is the set of rules about how all of those lexical elements may be combined.  And as Shaun mentions, XML is remarkable for being an embedded syntax (embedded within the flow of document text, that is), which has only two characters which shift to 'markup' (TeX, which is doing the same thing in a way, has 16 different categories of character).

A possibly interesting tangent: there are fewer lexical elements here than might be obvious, depending on your point of view.  One of these is '<!', which is what SGML described as 'markup declaration open', meaning a bit of markup that isn't an element, but is instead somewhat 'meta'.  The following 'markup declaration' could be 'DOCTYPE' or '[CDATA[' (and a couple more).  Entertainingly, it could also be empty: '<!>', though that empty markup declaration could contain commented text: the comment characters in SGML were '--' as open and close characters, which could appear anywhere; thus you could, I recall have

• Follow-Ups:
  • Re: Many different syntaxes in XML - is that good languagedesign?
    • From: Peter Flynn <peter@silmaril.ie>
  • Re: Many different syntaxes in XML - is that good language design?
    • From: Norman Gray <norman.gray@glasgow.ac.uk>

• References:
  • Many different syntaxes in XML - is that good language design?
    • From: Roger L Costello <costello@mitre.org>
  • Re: Many different syntaxes in XML - is that good language design?
    • From: Norman Gray <norman.gray@glasgow.ac.uk>