[
Lists Home |
Date Index |
Thread Index
]
- From: Peter Murray-Rust <peter@ursus.demon.co.uk>
- To: XML-Dev ML <xml-dev@XML.ORG>
- Date: Wed, 17 May 2000 08:26:06 +0100
At 10:50 AM 5/16/00 -0700, Pierre, Sebastian wrote:
>Hi all!
>
>Still asking some silly questions...sorry!
No need to apologize on XML-DEV. Most of us started by asking questions and
we are happy to (try to) answer them. I have learnt at least two new things
from this discussion.
>
> - When defining parameter entities, the parser (IE5...I know;) complains
about non valid characters such as '#' or even '('.
> For example I have this declaration :
> ( link | see | anchor | em | quote | citation | copyright | date |
name | #PCDATA )*
> for an element content is invalid because of the # in #PCDATA.
>
> - I don't know if I can imbricate more than one level of parenthesis in
element content declaration, ie:
> ( (first-name, last-name)+, address )*
> [This example is silly, but the important fact is imbricated parenthesis]
>
I find it a useful strategy to have 2 or more parsers set up to answer
questions like this. There are many high quality parsers as OpenSource or
for free. [Some parsers such as AElfred - which I use a lot - are
deliberately lightweight and give a few false positives and false
negatives.] I recently spent some hours trying to understand whitespace in
parameter entities - I was convinced there was a parser bug until I ran it
through 2 independent ones.
[When it comes to using this strategy with other applications, such as
XSLT, make sure that you know what versions of the spec, the tool, etc. you
are using.]
P.
***************************************************************************
This is xml-dev, the mailing list for XML developers.
To unsubscribe, mailto:majordomo@xml.org&BODY=unsubscribe%20xml-dev
List archives are available at http://xml.org/archives/xml-dev/
***************************************************************************
|