Help with disambiguation

Alan_Bes

Hi,
I'm needing help to set up the disambiguation correctly.
I'm using a Harvard variation that requires the following rule:
Only when family name is identical AND in the same year should initials be used.

So for example:
(1)
John Jones 2000 and Gene Jones 2000 needs to read as (Jones, J 2000) and (Jones, G 2000) - No problem with disambiguate-add-givenname="true"

(2)
However when the years are different such as John Jones 2000 and Gene Jones 2001 they need to read as (Jones 2000) and (Jones 2001)
But if I use the option above I get (Jones, J 2000) and (Jones, G 2001)

How do I get both conditions to work?

Thanks in advance

fbennett

There is a CSL option givenname-diambiguation-rule, which is valid on the cs:citation element. If it is set to "by-cite", you should get the result you're after. It is documented here:

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

Alan_Bes

Thanks, but I'm still not getting the result
I'll put the whole line here in case I'm missing out on something

If I use the following with your suggestion of "by-cite":
<citation et-al-min="3" et-al-use-first="1" et-al-subsequent-min="3" et-al-subsequent-use-first="1" disambiguate-add-year-suffix="true" disambiguate-add-names="true" disambiguate-add-givenname="true" givenname-disambiguation-rule="by-cite" collapse="year">

I incorrectly get
(Jones 2000) and (Jones 2000)

If I use givenname-disambiguation-rule="primary-name":
I correctly get
(Jones, J 2000) and (Jones, G 2000)
but when years are different I incorrectly get
(Jones, J 2001) and (Jones, G 2000)
That needs to be (Jones 2001)and (Jones 2000)

fbennett

Thanks for posting the config line. I'll do a test here and post back.

fbennett

It works fine in a test file here, using the options listed above.

The first thing is to check that you have switched to another style, then back to this one before testing. (Changes aren't picked up automatically -- but "by-cite" is the default, so this should have worked for you in the first place ...)

The second thing is to validate the style:

http://simonster.github.io/csl-validator.js/

(Zotero pre-validates styles, on installation, but if you made changes in the CSL Editor, they might not be re-validated -- I'm not sure. In any case, though, syntax errors can certainly cause formatting glitches, so it's worth double-checking.)

Alan_Bes

Thanks so much.

Part of the problem I see is that when I use the default test pane I get different results than when I use the editor at http://editor.citationstyles.org/codeEditor/

Which one should I be using?

Alan_Bes

Okay, narrowed the problem down to first and subsequent citations - shows correctly for subsequent citations, not first citations

Alan_Bes

I'll start a new thread with the first and subsequent problem with author names

fbennett

Post a copy of the style you are using to http://gist.github.com and I'll take a look.

Alan_Bes

Thanks I will appreciate that.
https://gist.github.com/anonymous/ed22c6bd0ae2f455fd18/download#file-gistfile1-txt

Alan_Bes

Here's what I'm getting in Word

References in database
J Jones 2000
G Jones 2000
A Smith 2000
B Smith 2001

1st instance
(Jones 2000) (Jones 2000) (Smith 2000) (Smith 2001)
Subsequent instances
(Jones, G 2000) (Jones, J 2000) (Smith 2000) (Smith 2001)

I'm needing the subsequent instances to be displayed as the first as well.

Thanks again for your help.

fbennett

Neither of those links works, unfortunately. The first gives an empty page, and the second shows a page of JavaScript code.

Here are the steps:

(1) Open Zotero Preferences -> Advanced, and open the CSL Editor pane.
(2) Select your style in the pane.
(3) Click in the upper box, where the CSL code is shown. Right-click for the context menu, and choose "Select all".
(4) Copy the selected text to the clipboard with Ctrl-C (or Cmd-C on the Mac, I think).
(5) In a browser, visit http://gist.github.com, and paste the clipboard text of the style into the gist.
(6) Save the page as a "Public Gist".
(7) Post the URL from the browser address bar back here.

Alan_Bes

I edited it and it is now corrected.

fbennett

Aha. The revised link works. Thanks -- I'll take a look.

Alan_Bes

Thanks

fbennett

Good news, in a way. I can reproduce the failure.

The bug appears when et-al-subsequent-min is set to a lower value than et-al-min. If both are set to "4", the bug goes away. When et-al-subsequent-min is reduced to "3", the bug appears again (the first references lose the author initials).

As a short-term solution, you can set et-al-subsequent-min to "4" in the style. Meanwhile, I'll dig into this today, and figure out why it's doing this to us.

Alan_Bes

Thanks for looking into this.
Would be great to get a permanent fix as my thesis reference rules require that the subsequent be set to 3.

fbennett

That setting should definitely work, and it will be fixed soon. Disambiguation is very accurate these days, bug reports are getting scarce, and it is that much more important to get this one squashed. More soon!

fbennett

I think I've got it. You can try out the amended processor by installing the processor patch plugin. Its only effect is to swap in the latest CSL processor version. You should be able to remove it at the next Zotero update.

Here are the results it's delivering now, with a pair of single-author works and a pair of three-author works, separated by two separate works authored by the same "Other" author:
..[0] (Aalto, A 2000)
..[1] (Aalto, E 2000)
..[2] (Other 2000a)
..[3] (Aalto, A 2000)
..[4] (Aalto, E 2000)
..[5] (Bluthers, C, Conxton, Dragnet 2000)
..[6] (Bluthers, Z, Conxton, Dragnet 2000)
..[7] (Other 2000b)
..[8] (Bluthers, C et al. 2000)
>>[9] (Bluthers, Z et al. 2000)