Support for Repo.delete! with multiple ids (fix #57)

src/Repo.jl

@@ -459,15 +459,32 @@ function delete!(M::Type, nt::NamedTuple; db::Connection=current_connection())
 end
 
 """
-    Repo.delete!(M::Type, pk_range::UnitRange{Int64}; db::Connection=current_connection())
+    Repo.delete!(M::Type, pk_range::UnitRange{Int}; db::Connection=current_connection())
 """
-function delete!(M::Type, pk_range::UnitRange{Int64}; db::Connection=current_connection()) # throw Schema.PrimaryKeyError
+function delete!(M::Type, pk_range::UnitRange{Int}; db::Connection=current_connection()) # throw Schema.PrimaryKeyError
     a = db.adapter
     table = a.from(M)
     key = _field_for_primary_key(db, M)
     execute([a.DELETE a.FROM table a.WHERE key a.BETWEEN pk_range.start a.AND pk_range.stop]; db=db)
 end
 
+"""
+    Repo.delete!(M::Type, pk_tuple::(Tuple{Vararg{Int, N}} where N); db::Connection=current_connection())
+"""
+function delete!(M::Type, pk_tuple::(Tuple{Vararg{Int, N}} where N); db::Connection=current_connection()) # throw Schema.PrimaryKeyError
+    delete!(M, collect(pk_tuple); db)
+end
+
+"""
+    Repo.delete!(M::Type, pk_vector::Vector{Int}; db::Connection=current_connection())
+"""
+function delete!(M::Type, pk_vector::Vector{Int}; db::Connection=current_connection()) # throw Schema.PrimaryKeyError
+    a = db.adapter
+    table = a.from(M)
+    key = _field_for_primary_key(db, M)
+    execute([a.DELETE a.FROM table a.WHERE key a.IN a.Enclosed(pk_vector)]; db=db)
+end
+
 function sql_startswith_insert_update_delete(sql::String)::Bool
     lowercase(first(split(lstrip(sql), ' '))) in ("insert", "update", "delete")
 end

test/adapters/postgresql/repo_test.jl

@@ -83,6 +83,18 @@ df = Repo.query([SELECT * FROM em WHERE em.Name == "Tim"])
 @test size(df) == (1,)
 @test df[1].name == "Tim"
 
+result = Repo.delete!(Employee, [1, 3])
+@test result.num_affected_rows == 2
+
+result = Repo.delete!(Employee, (4, 5))
+@test result.num_affected_rows == 2
+
+result = Repo.delete!(Employee, 6:7)
+@test result.num_affected_rows == 2
+
+df = Repo.query(Employee)
+@test size(df) == (0,)
+
 # coverage up
 Schema.tables[Base.typename(Employee)] = Dict(:table_name => "Employee")
 @test_throws Schema.PrimaryKeyError Repo.get(Employee, 2)

test/adapters/sqlite/repo_test.jl

@@ -89,43 +89,54 @@ df = Repo.query([SELECT * FROM em WHERE em.EmployeeId == 2])
 # using Octo.Adapters.SQLite # CREATE TABLE IF NOT EXISTS
 Repo.query([CREATE TABLE IF NOT EXISTS :temp AS SELECT * FROM :Album])
 
-struct Temp
+struct Album
 end
-Schema.model(Temp, table_name="temp", primary_key="AlbumId")
+Schema.model(Album, table_name="temp", primary_key="AlbumId")
 
-df = Repo.query(Temp)
+df = Repo.query(Album)
 @test size(df) == (347,)
 
 changes = (AlbumId=0, Title="Test Album", ArtistId=0)
-result = Repo.insert!(Temp, changes)
+result = Repo.insert!(Album, changes)
 @test result.num_affected_rows == 1
 
-df = Repo.query(Temp)
+df = Repo.query(Album)
 @test size(df) == (348,)
 
-df = Repo.get(Temp, 6)
+df = Repo.get(Album, 6)
 @test df[1].Title == "Jagged Little Pill"
 
-df = Repo.get(Temp, (Title="Jagged Little Pill",))
+df = Repo.get(Album, (Title="Jagged Little Pill",))
 @test df[1].Title == "Jagged Little Pill"
 
 changes = (AlbumId=6, Title="Texas")
-result = Repo.update!(Temp, changes)
+result = Repo.update!(Album, changes)
 @test result.num_affected_rows == 1
 
-df = Repo.get(Temp, 6)
+df = Repo.get(Album, 6)
 @test df[1].Title == "Texas"
 
-result = Repo.delete!(Temp, changes)
+result = Repo.delete!(Album, changes)
 @test result.num_affected_rows == 1
 
-df = Repo.get(Temp, 6)
-@test size(df) == (0,)
+result = Repo.delete!(Album, [1, 2, 3])
+@test result.num_affected_rows == 3
+
+result = Repo.delete!(Album, (4, 5))
+@test result.num_affected_rows == 2
 
+result = Repo.delete!(Album, 11:15)
+@test result.num_affected_rows == 5
+
+df = Repo.query(Album)
+@test size(df) == (337,)
+
+df = Repo.get(Album, 6)
+@test size(df) == (0,)
 
 ❔  = Octo.PlaceHolder
 
-a = from(Temp)
+a = from(Album)
 df = Repo.query([SELECT * FROM a WHERE a.Title == ❔  AND a.ArtistId == ❔ ], ["The Doors", 140])
 @test size(df) == (1,)