# ticket #3 fix, and #1

**URL:** <https://discourse.citationstyles.org/t/ticket-3-fix-and-1/46>\
**Category:** CSL Development\
**Created:** [March 25, 2005, 4:29pm UTC](https://discourse.citationstyles.org/t/ticket-3-fix-and-1/46 "2005-03-25T16:29:10Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![Bruce\_D\_Arcus](https://avatars.discourse-cdn.com/v4/letter/b/f9ae1b/32.png) [@Bruce\_D\_Arcus](https://discourse.citationstyles.org/u/Bruce_D_Arcus)\
**Post date:** [March 25, 2005, 4:29pm UTC](https://discourse.citationstyles.org/t/ticket-3-fix-and-1/46/1 "2005-03-25T16:29:10Z")

</div>

The ticket 3 fix address most of the problem. However, strickly  
speaking, this:

Some more citations: (Thrift, 1990; Tilly, 2000, Tilly, 2002; Times, 2001).

… should be:

Some more citations: (Thrift, 1990; Tilly, 2000, 2002; Times, 2001).

In other words, the second author name in a group gets left off.

Also, for ticket #1, here’s the current rendering with the test doc:

(C.S., 2001)

It should be:

(Tremblay v. Quebec, 2001)

So the question is:

a) what should be the markup in CSL to indicate this, and…  
b) how to modify the code to use CSL for this sorting

Bruce

---

<div class="post-metadata">

**Author:** ![Bruce\_D\_Arcus](https://avatars.discourse-cdn.com/v4/letter/b/f9ae1b/32.png) [@Bruce\_D\_Arcus](https://discourse.citationstyles.org/u/Bruce_D_Arcus)\
**Post date:** [March 25, 2005, 4:42pm UTC](https://discourse.citationstyles.org/t/ticket-3-fix-and-1/46/2 "2005-03-25T16:42:41Z")

</div>

Bruce D’Arcus wrote:

> So the question is:
> 
> a) what should be the markup in CSL to indicate this, and…

Here’s my thoughts:

The sorting logic should just use the first child of a cs:reftype as the  
first option on sort. The second option ought to be indicated via an  
attribute on that element.

So, for the legal case example in the sample, it might be:

… while for an article, it might be:

\<reftype=“article”\>

The above key would then map to an xpath expression, where the options  
would be defined in the schema.

The trick, though, is how to then handle plain text. For example, a  
book without an author name might have a rule that says it must be  
replaced by “Anonymous” (as in current code).

So how to handle this?

Bruce