Page MenuHomePhabricator

Two sets of duplicate named references created in one article
Closed, InvalidPublic

Description

This is a strange one, as this hasn't happened with any other article that I've noticed.

AWB in the case of William Harding Mayes is creating two sets of duplicate named references, one named "familytreemaker.genealogy.com" and the other named "jacksonpurchasehistory.org",

I have a settings file where you can run it to see the same results. I will attach it shortly.

Event Timeline

Stevietheman raised the priority of this task from to High.
Stevietheman updated the task description. (Show Details)

Here's the settings file. It's pretty basic.

I've found another article where this is happening: Virginia Tech -- AWB is trying to create a duplicate named reference "Virginia Tech Foundation".

Rjwilmsi claimed this task.
Rjwilmsi subscribed.

Following T99291 we made a change to DuplicateUnnamedReferences to only condense refs given name when ref not declared within a template call

Was the settings file I uploaded run? This is an outright bug where a duplicate name is generated for two separate but identical references. I obviously can't save such changes.

I'll need a clearer explanation if you're sure there's a bug here.

On [[William Harding Mayes]] we have two refs given name "familytreemaker.genealogy.com" and also two refs given name "jacksonpurchasehistory.org". As far as I can see there's no issue, the first two for familytreemaker.genealogy.com are identical and given name "familytreemaker.genealogy.com"; the second two for jacksonpurchasehistory.org are unrelated to the familytreemaker.genealogy.com ones, the second two are given name "jacksonpurchasehistory.org". So I see no issue. As the refs are within templates none are condensed due to T99291.

On [[Virginia Tech]] current version no refs with name "Virginia Tech Foundation" are added or changed. So I see no issue.

I understand that we don't want to create two duplicate named refs if one is inside a template, but in the case of the William Harding Mayes article, all the refs in question are inside the same template, and duplicate names are being generated.

First, why are names being generated at all? Secondly, since they are within the same template, why can't this be condensed as expected?