XML Attributes

[daveejay]

You may well have to declare those characters as entities.

Not sure but there may also be a character set issue in the xml directive at the top of the file.

[Judd]

since I don't know a great deal about XML, how do I do that?

Entities, what?

[daveejay]

Have a look here:

http://www.xml.com/pub/a/2001/01/31/qanda.html

Found from doing a search on Google: entities special characters xml

Are those characters part of the standard ASCII character set?

Dave

[Judd]

So I need to declare them somehow? In a DOCTYPE element?

The reference you posted talks about DTD, I'm using a schema (XSD) how does all that hang together

The sorts of characters I'm using are:

° = ALT+248
± = ALT+177
² = ALT+0178
³ = ALT+0179

[daveejay]

Yes that's right you need to declare them in a DTD (DOCTYPE element).

I suggest you take a few hours out to learn about DTDs, they really are quite simple. Don't think of XML as anything other than self-describing data (in markup form).

I learnt about DTDs a while back - just starting on Schemas. Take a look here:

http://www.xml101.com/dtd/default.asp

And here:

http://lists.w3.org/Archives/Public/...1Jul/0057.html

It would imply you can use a combination of schema and DTD.

I imagine this is the way you will accomplish what you are trying to do.

[Judd]

Thanks for your help so far dave. I haveone more thing I'm not sure about.

It isn't practical for me to prevent the '³' characters from appearing in my XML file. From what I've read, entities allow things to be recognised as '&something;', but that wouldn't be the case in my situation. I want the charcter itself to be recognised.

Am I understanding this correctly?

The alternative is to remove the validation of my documents entirely, which I'm not too keen on...

[daveejay]

Entity declarations are a kind of variable, so you can reference any string (could be a standard footer or disclaimer) with another (much shorter or more convenient) one i.e.

<?xml version="1.0"?>
<!-- Internal DTD Subset-->
<!DOCTYPE page [
<!ENTITY disclaimer "We are not responsible for the content of this page in any way">
]>
<!--DTD end-->
<page>
<content>Blah Blah Blah</content>
<disclaim>&disclaimer;</disclaim>
</page>

So you might implement something like:

<?xml version="1.0"?>
<!-- Internal DTD Subset-->
<!DOCTYPE juddspage [
<!ENTITY degrees "&#248;">
<!ENTITY slash "&#47;">
]>
<!--DTD end-->
<juddspage>
<Unit External="kW&slash;&degrees;C" Internal="kW&slash;&degrees;C" Zero="0." Scale="1." Offset="0.">kW&slash;&degrees;C :: kW&slash;&degrees;C</Unit>
</juddspage>

[Judd]

but I think I haven't explained myself properly.

What you describe is a change to the content of the XML file.
I'm writing large files so it is not really practical to watch every character and replace '°' with '&degree;' etc.

What I need is a way to instruct the parser (through the schema file? DTD?) that if it comes across a '³' or a '°' (for instance) when validating, not to throw a fatal error event. I really don't see why these are considered invalid characters in the first place.

I guess I need to know if this can even be done?

Sorry to be a pain about this!

[daveejay]

Even & and / are invalid characters and have to defined by their ASCII Character codes, although in the case of >, <, & etc. the XML equivalents. &amp; will do fine.

Thinking logically about it, I'm not so sure the parser doesn't like the special characters, could it be objecting to the "/" s before them?

Dave

[Judd]

..our internet pipe went down. I found the 'actual' problem was my encoding statement. IF I specified the encoding as UTF-8 the validation failed, if I used ISO-8859-1 (Latin-1) the validation passed.

However, I had to hack in the XML declaration manually (hardcode it) because using the put_encoding method ddoesn't seem to work (if you don't use it, it defaults to writing the encoding as UTF-16 (even though it doesn't write in that encoding!!!), if you use put_encoding, it doesn't write any encoding statement at all, thus assuming UTF-8 and failing validation....