Martin Honnen <mahotrash@y...> writes:

> LionelAndJen@g... wrote:
. . .
>> ...  UNFORTUNATELY, XSLT
>> translates  doesn&apos;t into doesn't in the output, which then
>> destroys my sql statement.  How do I tell my XSLT to leave well-enough
>> alone ?  ...
. . .
> references in the output the XSLT processor creates then you need to
> look at processor specific extensions to do that. Or you can look into
> XSLT 2.0 and character maps
> http://www.w3.org/TR/xslt20/#character-maps. Note that both ways will
> not simply allow you to preserve entity references in the input,
> instead you will create entity references in the output.

I suppose sql language is not usually envisioned as a translation
target in the xml world.

Nuggets of CDATA in an xml document are not supposed to be decision
points. Nonetheless it's possible but cpu-expensive to look at
individual characters.  (The expense may be reasonable in the OP's
context.)

This is why the exclusion of SDATA character entities was controversial
in early days.

                                    -- Bill