Sub-optimal query with unnecessary not null

      FROM "Animals" AS "a"
      WHERE ((
          SELECT "t"."Value"
          FROM "text_attributes" AS "t"
          WHERE ("a"."Id" = "t"."AnimalId") AND (("t"."Name" = 'placeOfBirth') AND "t"."Name" IS NOT NULL)
          LIMIT 1) = 'seattle') AND (
          SELECT "t"."Value"
          FROM "text_attributes" AS "t"
          WHERE ("a"."Id" = "t"."AnimalId") AND (("t"."Name" = 'placeOfBirth') AND "t"."Name" IS NOT NULL)
          LIMIT 1) IS NOT NULL

As far as understand the not null checks aren’t required in the above scenario, both in the predicate on testing Name, but also the whole redundant duplicate SQL statement. I understand the issue with database nulls, and if we were doing “not equals” and we would need to consider null values too, but when we are testing equality against a constant this overzealous approach seems inefficient. Is the query builder too generalized, this seems like a low hanging fruit win for performance?

The above is the result of:

   var results = ctx.Animals
                    .Where(a => a.TextAttributes.FirstOrDefault(x => x.Name == "placeOfBirth").Value == "seattle").ToList()

Where Animals has a List<TextAttribute> where TextAttribute has Name/Value string properties.

tl;dr - not null checks aren’t required

[smitpatel]

By mistake when removing type-bug.

[ajcvickers]

@smitpatel Why did you remove customer-reported from this issue?