can't produce '&' in output

[David Schwartz]

I need to produce '&' in the urls I'm producing to support
parameterization. I get the following results with one processor
running on my Mac:

<xsl:text disable-output-escaping="yes">http:...&....</xsl:text>
produces http:...&amp;....
<xsl:text disable-output-escaping="yes">http:...&amp;....</xsl:text>
produces http:...&amp;....

Not using the disable-output-escaping attribute produces the same
results as does not using the xsl:text tag. Just using '&' produces an
error as the processor is expecting a ';' to close the entity.


However, just using '&' produces a '&' in the output just fine using
XALAN on WinXP.

Why the difference? Which is 'right'?

TIA,
David

 

[David Schwartz]

However, just using '&' produces a '&' in the output just fine using
XALAN on WinXP.

My apologies. I actually have to use &amp; (or &#032, I'd imagine)
with XALAN as you'd expect. So, I guess I should just conclude that
the processor running on the Mac is non-compliant?

 

[Martin Honnen]

I need to produce '&' in the urls I'm producing to support
parameterization. I get the following results with one processor
running on my Mac:

<xsl:text disable-output-escaping="yes">http:...&....</xsl:text>
produces http:...&amp;....
<xsl:text disable-output-escaping="yes">http:...&amp;....</xsl:text>
produces http:...&amp;....

Not using the disable-output-escaping attribute produces the same
results as does not using the xsl:text tag. Just using '&' produces an
error as the processor is expecting a ';' to close the entity.


However, just using '&' produces a '&' in the output just fine using
XALAN on WinXP.

Why the difference? Which is 'right'?

disable-output-escaping is an optional serialization feature an XSLT
processor does not need to support at all or that is only supported for
certain settings in which the XSLT processor controls serialization.

As for enforcing '&' in URLs, if you create HTML with your XSLT
transformation then any HTML user agent will be able to deal with e.g.
<a href="http://example.com/?arg1=foo&amp;arg2=bar">link</a>
just fine, there is no need to make efforts to avoid the escaping.

 

[David Schwartz]

As for enforcing '&' in URLs, if you create HTML with your XSLT

transformation then any HTML user agent will be able to deal with e.g.
<a href="http://example.com/?arg1=foo&arg2=bar">link</a>
just fine, there is no need to make efforts to avoid the escaping.

Right but the problem is that the following is what's getting
generated:
<a href="http://example.com/?arg1=foo&amp;arg2=bar">

 

[Martin Honnen]

Right but the problem is that the following is what's getting
generated:
<a href="http://example.com/?arg1=foo&amp;arg2=bar">

That is what I had in my post, somehow the &amp; (& a m p got
unescaped. Anyway, &amp; is correct HTML so it is not a problem for
validators and browsers:
<URL:http://validator.w3.org/check?uri=h...ct+automatically)&doctype=Inline&group=0&ss=1>

 

[Joseph Kesselman]

Right but the problem is that the following is what's getting
generated:
<a href="http://example.com/?arg1=foo&amp;arg2=bar">

That's correct XML/XHTML. You could try specifying HTML output, which in
some processors *may* suppress escaping for attribute values which are
known to represent URIs.

 

[Richard Tobin]

Right but the problem is that the following is what's getting
generated:
<a href="http://example.com/?arg1=foo&amp;arg2=bar">

[/QUOTE]

That's correct XML/XHTML.

Furthermore, you can't suppress output escaping in an attribute value:

It is an error for output escaping to be disabled for a text node
that is used for something other than a text node in the result
tree. Thus, it is an error to disable output escaping for an
xsl:value-of or xsl:text element that is used to generate the
string-value of a comment, processing instruction or attribute node;
[...]
an XSLT processor may signal the error; if it does not signal the
error, it must recover by ignoring the disable-output-escaping
attribute.

See http://www.w3.org/TR/xslt#disable-output-escaping

-- Richard

 

[David Schwartz]

That is what I had in my post, somehow the &amp; (& a m p got
unescaped. Anyway, &amp; is correct HTML so it is not a problem for
validators and browsers:
<URL:http://validator.w3.org/check?uri=http://home.arcor.de/martin...>

FWIW, what I posted used the named entity. Whatever is processing
these posts converted it to '&'.