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

Globbing versus Regular Expressions (was: Regular Associations)

• To: xml-dev@l...
• Subject: Globbing versus Regular Expressions (was: Regular Associations)
• From: David Tolpin <dvd@d...>
• Date: Mon, 5 Jan 2004 13:29:41 +0400 (AMT)
• In-reply-to: <830178CE7378FC40BC6F1DDADCFDD1D1D517EB@R...>

> Sounds like you've invented a more verbose version of XPath 1.0. 

Almost so. With a few notes:

- it is not me who invented it, Relax NG is an international standard;

- it is not necessarily more verbose, there are cases when it is significantly
  more concise (or when it does not have an XPath equivalent);

- it does a different thing: XPath is a language for addressing parts of 
  a document, Relax NG, and the way it is used in ARX, the utility we are discussing,
  assesses that a document matches a regular pattern. 

The difference is almost the same as between filename globbing and regular
expressions for character strings. Globs is a subset of regular expressions,
they are not closed and cannot be freely combined. While globs can be shorter in 
a limited number of cases (and it only makes sense when amount of typing is important --
like filename globbing in a shell), their use for anything non-trivial is problematic.
Many Unix shell programmers routinely use sed to manipulate filenames in command's arguments.

  wc -l `ls *.*|grep -v "_(test|check)\.(c|cpp|sc)"`

will count lines in every file with dot in the name in the current directory (globbing), 
except those which name ends with _test or _check and which suffix is (.c, .cpp, or, .sc).

The same is true for XPath. It is good for XSLT (and several other applications), where
its use is constraineed and/or where the length of an expression is crucial (for clarity
or typing speed); however, in many other cases, a regular syntax is preferrable.

Note 1. However, I would find an XSLT-like language  that provides XML regular expressions
(similar to Relax NG) for matching and selection more useful and convenient, if occasionally
less verbose, than XSLT; if string regular expressions (which are peripheral to the area of XSLT)
were added, why XML regular expressions were not?

Note 2. Actually, I mentioned in the original message that my first thought had been to
use XSLT for this purpose (and I meant XPath, of course), and that I had realized later
that Relax NG is more appropriate. However, it was written at the very end of the message.

David Tolpin
http://davidashen.net/

• Follow-Ups:
  • Re: Globbing versus Regular Expressions (was: Regular Associations)
    • From: james anderson <james.anderson@s...>

• References:
  • RE: ANN: Regular Associations
    • From: "Dare Obasanjo" <dareo@m...>

• Prev by Date: Re: Formalism and complexity
• Next by Date: Re: Globbing versus Regular Expressions (was: Regular Associations)
• Previous by thread: RE: ANN: Regular Associations
• Next by thread: Re: Globbing versus Regular Expressions (was: Regular Associations)
• Index(es):
  • Date
  • Thread