Re: Metadata.xml files bomb out as a result of the &

From John R. McPherson
DateSat, 2 Feb 2002 17:59:45 +1300
Subject Re: Metadata.xml files bomb out as a result of the &
In-Reply-To (C08FFFF817C9D2118C7C00A0C9ED115501569A2A-EXCHANGE)
On Fri, Feb 01, 2002 at 04:04:25PM -0800, Jared Potter wrote:
> I have noticed that if there is an ampersand(&) in the metadata.xml file, it
> will ignore the whole file. Does anyone know of any hacks to get around
> this, and are there any other characters that I don't know about that do the
> same?

I'm not entirely sure about this, but someone will correct me if
I'm wrong...

The XML parser we use with greenstone will fail to parse if it
encounters "invalid" tokens. I've noticed this if there are non-ascii
characters, but the charset is defined as ascii - a warning message
is printed out during import that the file couldn't be parsed.

I don't know enough about it to know why it doesn't like the ampersand,
but in html these are encoded "safely" by "&", which of course
also has an ampersand in it... I guess you tried this?

Also, where exactly in the xml file is it - is it in the content, or
inside a tag?

> By the way, for those of you who have played with the program at
> I would like to tell you that the search

> What do those who have seen it think? Helpful? Not helpful? I enjoy
> feedback. :)

I had a little look, but there weren't any documents in it yet, or
I couldn't retrieve them. I might have another look during the week....

John McPherson