Incorrect tag search - illusory duplicates returned

[david-kn]

• Python version: 3.10
• Nautobot version: 1.3.2

Steps to Reproduce

1. Search for an existing tag (in my case substring, e.g. exclude):

• either in Organization-> Tags page or
• when adding tag for any other object (see attached images)

Expected Behavior

Only existing (unique) tag(s) should be returned.

Observed Behavior

There are too many returned tags (not unique) that are duplicate (as for each model would exist individual tag with the same name). The number of returned duplicate tags correspond with a number of attached models to a given tag (in my case 35)

btw. So far, I noticed this only for Tags but not for Statuses which can also be connected with multiple models.

[ubajze]

@david-kn I tried to reproduce the issue but I was not able to do it. I tried to add more tags through the UI.

I am wondering what you get if you run the following in nbshell:

>>> for tag in Tag.objects.filter(name="scan_exclude"):
...  print(tag.pk, tag.content_types.all())

[glennmatthews]

I can confirm this issue. It does not happen in the base Tags list (un-searched), but does appear when a search is issued.

[glennmatthews]

The below diff appears to fix the issue:

diff --git a/nautobot/extras/filters.py b/nautobot/extras/filters.py
index 48a0a4550..338a858a9 100644
--- a/nautobot/extras/filters.py
+++ b/nautobot/extras/filters.py
@@ -977,7 +977,7 @@ class TagFilterSet(NautobotFilterSet):
             return queryset
         return queryset.filter(
             Q(name__icontains=value) | Q(slug__icontains=value) | Q(content_types__model__icontains=value)
-        )
+        ).distinct()
 
 
 #

[david-kn]

Thanks Glen for confirmation :-) To filter and show only distinct values was my inner guess as well.

btw. Isn't such a fix a good candidate also for 1.3.3 release sometimes in a near future as well? :-)
We won't upgrade to 1.3.2 yet because this behavior would be very confusing for many users.