Additional info:

1. Sets are not currently being used, as far as I know.

2. So in order to test this, you'll have to add (on localhost) the set parameter. For instance, in my <templdatedata> I have:

"sets": [

		{
			"label": "SomeIDs",
			"params": ["id", "rfc"]
		}

],

3. And then in the api, on Cite_journal/doc I get back:
   
   "sets": [ { "label": { "en-gb": "SomeIDs" }, "params": [ "id", "rfc" ] } ],

4. But in the api, on Cite_journal, I get back:
   
   "sets": []

Which is an empty array that is appended to what is returned from the api, since what the is retrieved from the database does not contain the sets parameter.

I tested this on my local wiki by adding sets: [] to a local template. I was able to get the full information (with the sets I filled in) in the API response.

If this was tested in production, I wonder if it may have been a case of delay between updating some cache the API is using?

Am I misunderstanding the problem?

Can't reproduce this on latest master.

Edited Unsigned/doc on my local wiki and added a sets section:

"sets": [

		{
			"label": {
				"en": "Default",
				"nl": "Standaard"
			},
			"params": [
				"user",
				"timestamp"
			]
		}

]

• w/api.php?format=jsonfm&action=templatedata&titles=Template:Unsigned/doc
• w/api.php?format=jsonfm&action=templatedata&titles=Template:Unsigned

Both included the new "sets" data.

Thanks for looking into this all.

I just realized this is actually a duplicate bug:

https://bugzilla.wikimedia.org/show_bug.cgi?id=50372

Basically the templatedata property in pp page gets updated when I save changes to cite_web/doc, but not cite_web, so when I made api requests I was getting old td from cite_web and the new changes to cite_web/doc. I can force the change by making a blank edit to the place where td is transcluded i.e. directly in Template:cite_web.

The fact that what I was editing was "sets" at the time was a red herring, so thanks for disabusing me of the notion this was a "sets" issue!

• This bug has been marked as a duplicate of bug 50372 ***