]> git.ipfire.org Git - thirdparty/sqlalchemy/sqlalchemy.git/commitdiff
Pass DDLCompiler IdentifierPreparer to visit_ENUM
authorMike Bayer <mike_mp@zzzcomputing.com>
Mon, 17 Feb 2020 20:21:59 +0000 (15:21 -0500)
committerMike Bayer <mike_mp@zzzcomputing.com>
Mon, 17 Feb 2020 20:35:44 +0000 (15:35 -0500)
Fixed issue where the "schema_translate_map" feature would not work with a
PostgreSQL native enumeration type (i.e. :class:`.Enum`,
:class:`.postgresql.ENUM`) in that while the "CREATE TYPE" statement would
be emitted with the correct schema, the schema would not be rendered in
the CREATE TABLE statement at the point at which the enumeration was
referenced.

Fixes: #5158
Change-Id: I41529785de2e736c70a142c2ae5705060bfed73e

doc/build/changelog/unreleased_13/5158.rst [new file with mode: 0644]
lib/sqlalchemy/dialects/postgresql/base.py
test/dialect/postgresql/test_compiler.py
test/dialect/postgresql/test_types.py

diff --git a/doc/build/changelog/unreleased_13/5158.rst b/doc/build/changelog/unreleased_13/5158.rst
new file mode 100644 (file)
index 0000000..adab86d
--- /dev/null
@@ -0,0 +1,11 @@
+.. change::
+    :tags: bug, postgresql
+    :tickets: 5158
+
+    Fixed issue where the "schema_translate_map" feature would not work with a
+    PostgreSQL native enumeration type (i.e. :class:`.Enum`,
+    :class:`.postgresql.ENUM`) in that while the "CREATE TYPE" statement would
+    be emitted with the correct schema, the schema would not be rendered in
+    the CREATE TABLE statement at the point at which the enumeration was
+    referenced.
+
index ceefc20b01f26254cfd0fca3fe1cd8e48f3b46a2..45911d4c076d55e03cb4e7d10a82085154f82d7a 100644 (file)
@@ -1887,7 +1887,9 @@ class PGDDLCompiler(compiler.DDLCompiler):
                 colspec += " SERIAL"
         else:
             colspec += " " + self.dialect.type_compiler.process(
-                column.type, type_expression=column
+                column.type,
+                type_expression=column,
+                identifier_preparer=self.preparer,
             )
             default = self.get_column_default_string(column)
             if default is not None:
@@ -2149,8 +2151,11 @@ class PGTypeCompiler(compiler.GenericTypeCompiler):
         else:
             return self.visit_ENUM(type_, **kw)
 
-    def visit_ENUM(self, type_, **kw):
-        return self.dialect.identifier_preparer.format_type(type_)
+    def visit_ENUM(self, type_, identifier_preparer=None, **kw):
+        if identifier_preparer is None:
+            identifier_preparer = self.dialect.identifier_preparer
+
+        return identifier_preparer.format_type(type_)
 
     def visit_TIMESTAMP(self, type_, **kw):
         return "TIMESTAMP%s %s" % (
index 4c4c43281e0f822fa22591d0a8d6cc3a28c04c50..aabbc3ac3b0aebbeac79b85e4b1201e1702c410f 100644 (file)
@@ -237,6 +237,22 @@ class CompileTest(fixtures.TestBase, AssertsCompiledSQL):
             schema_translate_map=schema_translate_map,
         )
 
+    def test_create_table_with_schema_type_schema_translate(self):
+        e1 = Enum("x", "y", "z", name="somename")
+        e2 = Enum("x", "y", "z", name="somename", schema="someschema")
+        schema_translate_map = {None: "foo", "someschema": "bar"}
+
+        table = Table(
+            "some_table", MetaData(), Column("q", e1), Column("p", e2)
+        )
+        from sqlalchemy.schema import CreateTable
+
+        self.assert_compile(
+            CreateTable(table),
+            "CREATE TABLE foo.some_table (q foo.somename, p bar.somename)",
+            schema_translate_map=schema_translate_map,
+        )
+
     def test_create_table_with_tablespace(self):
         m = MetaData()
         tbl = Table(
index 6a61fb33be4cbed12c79e923f61032f4eb360bec..be05dec7bcee5178f6da451922c195e716611ce7 100644 (file)
@@ -175,6 +175,58 @@ class EnumTest(fixtures.TestBase, AssertsExecutionResults):
                 [(1, "two"), (2, "three"), (3, "three")],
             )
 
+    @testing.combinations(None, "foo")
+    def test_create_table_schema_translate_map(self, symbol_name):
+        # note we can't use the fixture here because it will not drop
+        # from the correct schema
+        metadata = MetaData()
+
+        t1 = Table(
+            "table",
+            metadata,
+            Column("id", Integer, primary_key=True),
+            Column(
+                "value",
+                Enum(
+                    "one",
+                    "two",
+                    "three",
+                    name="schema_enum",
+                    schema=symbol_name,
+                ),
+            ),
+            schema=symbol_name,
+        )
+        with testing.db.connect() as conn:
+            conn = conn.execution_options(
+                schema_translate_map={symbol_name: testing.config.test_schema}
+            )
+            t1.create(conn)
+            assert "schema_enum" in [
+                e["name"]
+                for e in inspect(conn).get_enums(
+                    schema=testing.config.test_schema
+                )
+            ]
+            t1.create(conn, checkfirst=True)
+
+            conn.execute(t1.insert(), value="two")
+            conn.execute(t1.insert(), value="three")
+            conn.execute(t1.insert(), value="three")
+            eq_(
+                conn.execute(t1.select().order_by(t1.c.id)).fetchall(),
+                [(1, "two"), (2, "three"), (3, "three")],
+            )
+
+            t1.drop(conn)
+            assert "schema_enum" not in [
+                e["name"]
+                for e in inspect(conn).get_enums(
+                    schema=testing.config.test_schema
+                )
+            ]
+            t1.drop(conn, checkfirst=True)
+
     def test_name_required(self):
         metadata = MetaData(testing.db)
         etype = Enum("four", "five", "six", metadata=metadata)