Code in ignored

tillheilmann

I have a macro like this to distinguish between types in one of my custom styles...

<macro name="issued">
<choose>
<if type="book chapter" match="any">
...
</if>
<else-if type="article-journal article-magazine" match="any">
...
</else-if>
<else-if type="article-newspaper">
...
</else-if>
<else-if type="manuscript">
...
</else-if>
<else>
...
</else>
</choose>
</macro>

Funny thing is, code in <else></else> (the fallback for all other types) gets ignored. Any ideas?

adamsmith

we'd need to see the full code - there's nothing wrong with the structure as you have it above.

tillheilmann

Okay, I've narrowed it down to the minimum example below. There is something about the syntax of CSL (especially the group tag) that I don't understand because the line

<text variable="publisher-place"/>

(or a similar one) breaks the following <choose></choose> conditional.

<?xml version="1.0" encoding="utf-8"?>
<style xmlns="http://purl.org/net/xbiblio/csl" version="1.0" class="note">
<info>
<title>1</title>
<id>http://www.zotero.org/styles/1</id>;
<updated>2012-02-07T01:46:04+00:00</updated>
</info>
<citation>
<layout>
<names variable="author"/>
<group>
<text variable="publisher-place"/> <!-- comment this line to make it work -->
<choose>
<if variable="issued">
<date variable="issued" form="numeric"/>
</if>
<else>
<text term="no date"/>
</else>
</choose>
</group>
</layout>
</citation>
</style>

adamsmith

you're hitting this:

cs:group implicitly acts as a conditional: cs:group and its child elements are suppressed if a) at least one rendering element in cs:group calls a variable (either directly or via a macro), and b) all variables that are called are empty. This accommodates descriptive cs:text elements

http://citationstyles.org/downloads/specification.html#id56

tillheilmann

Thanks -- and sorry for ignoring the straightforward explanation in the specification.

adamsmith

no worries at all - while I think this makes a lot of sense from an implementation point of view, I think "straightforward" is stretching it ;-), especially when it's "hidden" inside a cs:choose element such as in your case.