Working on civil and common law citation styles, I have run into an
ambiguity in the specification. Before relying on the current behavior
of citeproc-js, I would like to request clarification.
Concerning the variable attribute to cs:if and cs:if-else, the
specification currently says this:***
variable
Tests whether the given variables (Appendix I - Variables) contain
non-empty values.
http://citationstyles.org/downloads/specification.html#choose
For this test, citeproc-js currently tests only for a non-nil value in
the named variable. For the title and container-title variables, a
short form can exist (delivered on the item as shortTitle and
journalAbbreviation respectively), even though the named variable is
nil. In that case, the cs:if or cs:else-if test will return false,
but an attempt to render the variable in the short form will produce
output.
While this seems odd, it provides an intuitively straightforward way
of distinguishing common law and civil law case cites. The former
always have a case name. The latter do not (ever), but many cases do
have a short name that is used as part of the cite in the initial
reference. The two types of case are formatted very differently (even
within a single legal style like the Bluebook), and the current
behavior of the test for title allows the two to be distinguished.
Would it be acceptable to confirm this in the specification with
language like the following:
variable
Tests whether the long form of the given variables (Appendix I -
Variables) contain non-empty values.
If this is agreeable to everyone, I think we’re set to offer basic
support for both common law and civil law materials. If not, we’ll
need to explore some other means of distinguishing common law and
civil law cites – either a new item type (which seems wrong, since
both categories of cite refer to court judgements), or a special field
to hold the nickname of civil law cases (which would further clutter
up the UI and could take quite awhile to bring about).
Working on civil and common law citation styles, I have run into an
ambiguity in the specification. Before relying on the current behavior
of citeproc-js, I would like to request clarification.
Concerning the variable attribute to cs:if and cs:if-else, the
specification currently says this:
variable
Tests whether the given variables (Appendix I - Variables) contain
non-empty values.
Redirecting…
For this test, citeproc-js currently tests only for a non-nil value in
the named variable. For the title and container-title variables, a
short form can exist (delivered on the item as shortTitle and
journalAbbreviation respectively), even though the named variable is
nil. In that case, the cs:if or cs:else-if test will return false,
but an attempt to render the variable in the short form will produce
output.
While this seems odd, it provides an intuitively straightforward way
of distinguishing common law and civil law case cites. The former
always have a case name. The latter do not (ever), but many cases do
have a short name that is used as part of the cite in the initial
reference. The two types of case are formatted very differently (even
within a single legal style like the Bluebook), and the current
behavior of the test for title allows the two to be distinguished.
Would it be acceptable to confirm this in the specification with
language like the following:
variable
Tests whether the long form of the given variables (Appendix I -
Variables) contain non-empty values.
Maybe “the default (long) form …”?
Bruce
I hadn’t realized before that you can’t test for the short form of
variables, but it seems that this hasn’t been a big problem in the past
(with a quick search I could only find this thread:
http://forums.zotero.org/discussion/10313/conditional-test-for-shorttitle/).
I’m okay with the spec change.
Rintze
Working on civil and common law citation styles, I have run into an
ambiguity in the specification. Before relying on the current behavior
of citeproc-js, I would like to request clarification.
Concerning the variable attribute to cs:if and cs:if-else, the
specification currently says this:
variable
Tests whether the given variables (Appendix I - Variables) contain
non-empty values.
Redirecting…
For this test, citeproc-js currently tests only for a non-nil value in
the named variable. For the title and container-title variables, a
short form can exist (delivered on the item as shortTitle and
journalAbbreviation respectively), even though the named variable is
nil. In that case, the cs:if or cs:else-if test will return false,
but an attempt to render the variable in the short form will produce
output.
While this seems odd, it provides an intuitively straightforward way
of distinguishing common law and civil law case cites. The former
always have a case name. The latter do not (ever), but many cases do
have a short name that is used as part of the cite in the initial
reference. The two types of case are formatted very differently (even
within a single legal style like the Bluebook), and the current
behavior of the test for title allows the two to be distinguished.
Would it be acceptable to confirm this in the specification with
language like the following:
variable
Tests whether the long form of the given variables (Appendix I -
Variables) contain non-empty values.
Maybe “the default (long) form …”?
That rushing sound that you hear …
I hadn’t realized before that you can’t test for the short form of
variables, but it seems that this hasn’t been a big problem in the past
(with a quick search I could only find this thread:
Conditional test for Short-title - Zotero Forums).
I’m okay with the spec change.
… is a huge sigh of relief coming from this end. 
I’ll apply and push the amended language suggested by Bruce later today.
Frank
Committed (at last):
RintzeOn Fri, Jul 23, 2010 at 9:12 PM, Frank Bennett <@Frank_Bennett>wrote: