sql: qualify table name for COMMENT ON INDEX, COLUMN event logs

[the-ericwang35]

Fixes #57740.

Release note (bug fix): qualify table name for COMMENT ON INDEX, COLUMN.

Previously, event logs were not capturing the qualified table names
for COMMENT ON INDEX and COMMENT ON COLUMN commands.
This PR changes the event logs to use the qualified table name.
Tests were also added for these changes.

[cockroach-teamcity]

This change is 

[ajwerner]

Reviewed 3 of 3 files at r1.
Reviewable status: complete! 1 of 0 LGTMs obtained

---

pkg/sql/logictest/testdata/logic_test/event_log, line 941 at r1 (raw file):

WHERE "eventType" = 'comment_on_index'
----
1  {"Comment": "This is an index.", "EventType": "comment_on_index", "IndexName": "b_index", "Statement": "COMMENT ON INDEX \"\".\"\".b_index IS 'This is an index.'", "TableName": "defaultdb.public.a", "User": "root"}

this ""."".b_index is suspect. I think I understand why it's happening but it's somewhat unfortunate. I think we want to update the way that tree.TableIndexName formats itself around here:

cockroach/pkg/sql/sem/tree/table_name.go

Lines 191 to 195 in c08088d

	// The table is not specified. The schema/catalog can still be specified.
	if n.Table.ExplicitSchema || ctx.alwaysFormatTablePrefix() {
	ctx.FormatNode(&n.Table.ObjectNamePrefix)
	ctx.WriteByte('.')
	}

.

The issue here is that the TableIndexName doesn't end up being an annotated node so we don't end up knowing how to resolve this appropriately. In the short term, I think I'm in favor of omitting the ""."". prefix when formatting this statement if the prefix is empty regardless of the flags. In the longer term I'd like us to annotate the node to properly resolve the prefix. Also, the logic to not include a prefix in that node except when we don't know the object name is particularly bizarre.

I'm thinking that for now this is what it is and is not touched by this code but I filed #58496 to follow up.

[the-ericwang35]

bors r=ajwerner

[craig]

Build succeeded:

• GitHub CI (Cockroach)