The option should not be needed when the API is working correctly – it's supposed to be able to detect existing signatures, and not add an unnecessary one. The question is why it's not accepting the signature in your examples. The code for that is here, unfortunately it's a bit complex under the hood: https://codesearch.wmcloud.org/search/?q=isSingleCommentSignedBy

(The check for signatures uses the same algorithm as all the other code in DiscussionTools, e.g. for inserting replies in the right place, so if there was such an option, that would allow adding messages using the discussiontoolsedit API that can't be correctly replied to using the discussiontoolsedit API, and I think that would be bad.)

I think in my examples, we are subst-ing two self-signing templates, which would explain why DiscussionTools can't detect the signatures.

If you're disinclined to have me write a patch to add noautosignature to the discussiontoolsedit API, perhaps I can get consensus to just remove the signatures from the templates. Let me work on that.

I think in my examples, we are subst-ing two self-signing templates, which would explain why DiscussionTools can't detect the signatures.

Even so, it should work, I wrote this feature specifically to support cases like this – see the tasks tagged on rEDTO0ecc8a4c051d: Improve detecting already signed comments, specifically T291421 mentions substitution and T282983 signatures inside wrapper templates.

Your edits have a HTML comment, a line break, and a category at the end; which is a bit weird, but even that should still be supported.

I tested by posting {{subst:Wikipedia:Teahouse/AfC Invitation|sign=~~~~}} through the normal interface, and I didn't get an extra signature: https://en.wikipedia.org/w/index.php?title=User_talk:Matma_Rex&diff=prev&oldid=1220608587

It could be helpful if you could share more details about the API requests you're sending.

Sounds like DiscussionTools can detect the wikitext of subst'd templates? That's pretty cool.

OK, I fiddled around with this and the steps to reproduce is something like...

• make sure the target page doesn't exist yet
• run this API query:

action: discussiontoolsedit
format: json
paction: addtopic
page: User talk:NovemBot
sectiontitle: 
wikitext: {{Talk header}}

== Your submission at [[Wikipedia:Articles for creation|Articles for creation]]: [[Draft:Sandbox|Sandbox]] ({{subst:CURRENTMONTHNAME}} {{subst:CURRENTDAY}}) ==
{{subst:Afc decline|full=Draft:Sandbox|cv=no|reason=v|details=|reason2=bio|details2=|comment=|sig=yes}}

{{subst:Wikipedia:Teahouse/AFC invitation|sign=~~~~}}
summary: Notification: Your [[Draft:Sandbox|Articles for Creation submission]] has been declined ([[WP:AFCH|AFCH]])
autosubscribe: yes
token: [redacted for privacy]

I guess that {{Talk header}} is throwing it off

edit: Confirmed. It doesn't double signature when I remove the {{Talk header}}. I may take {{Talk header}} out of the gadget code.

You’re basically adding two comments in two topics, one of which is unsigned (the talk header). The PHPDoc in rEDTO includes/CommentUtils.php:729-735 (at 1130ab2491fa8bbfcb8113fbc0a35cd1725a9356) starts with Assuming that the thread item set contains exactly one comment (or multiple comments with identical signatures […]) – this assumption doesn’t hold. Maybe it would work if you signed the talk header, but of course signing a talk header doesn’t make any sense.

Change #1024512 had a related patch set uploaded (by Bartosz Dziewoński; author: Bartosz Dziewoński):

[mediawiki/extensions/DiscussionTools@master] Fix signature check not to look at the content of previous section

https://gerrit.wikimedia.org/r/1024512

Awesome. Thanks for writing a patch for my unusual use case. Feel free to add reviewers to the patch if needed.

Change #1024512 merged by jenkins-bot:

[mediawiki/extensions/DiscussionTools@master] Fix signature check not to look at the content of previous section

https://gerrit.wikimedia.org/r/1024512