The true power of code templates comes from the ability to evaluate XPath expressions on your selected text.

Let's imagine you are working on a Serbian-German-Latin dictionary that has been transcribed in plain text. Very often, the translations are simple and consist of a German string followed by a comma followed by a Latin string.

For instance:

реч, f. das Wort, verbum.

Of course, you could try to convert this type of regularity with regular expressions as well, but we're learning here about code templates, so sit back and relax.

You've created code templates for <form><orth></orth></form> and for the <gramGrp> and your code currently looks like this:

    <entry>
    <form><orth>реч</orth></form>
    <gramGrp><pos>f.</pos></gramGrp> das Wort, verbum. </entry>

Wouldn't it be nice if you could separate the German translation from the Latin translation in one go and wrap everything in <cit>?

We can, using an oXygen editor variable ${xpath_eval()} and a handy xPath function such as substring-after() and substring-before(). When you select your strings, you want everything before the comma to be marked up as a German translation, and everything after the comma and a space as a Latin translation.

    <cit type="translation">
        <quote xml:lang="de">${xpath_eval(substring-before('${selection}',','))}</quote>
    </cit>
    <cit type="translation">
        <quote xml:lang="la">${xpath_eval(substring-after('${selection}',', '))}</quote>
    </cit>

If you were ever looking for a proof why it's worth learning xPath for anything other than to explore your XML-encoded dictionaries, here it is: