From: Mike Bayer Date: Thu, 25 Jun 2020 15:12:40 +0000 (-0400) Subject: Use index name to determine if an index is for the PK X-Git-Tag: rel_1_4_0b1~259^2 X-Git-Url: http://git.ipfire.org/cgi-bin/gitweb.cgi?a=commitdiff_plain;h=ca56d8dc32f939b2bdb1f590986d4c46d280d186;p=thirdparty%2Fsqlalchemy%2Fsqlalchemy.git Use index name to determine if an index is for the PK Fixed bug in Oracle dialect where indexes that contain the full set of primary key columns would be mistaken as the primary key index itself, which is omitted, even if there were multiples. The check has been refined to compare the name of the primary key constraint against the index name itself, rather than trying to guess based on the columns present in the index. Fixes: #5421 Change-Id: I47c2ccdd0b13977cfd9ef249d4de06371c4fb241 --- diff --git a/doc/build/changelog/unreleased_13/5421.rst b/doc/build/changelog/unreleased_13/5421.rst new file mode 100644 index 0000000000..d0e10dd1dd --- /dev/null +++ b/doc/build/changelog/unreleased_13/5421.rst @@ -0,0 +1,10 @@ +.. change:: + :tags: bug, oracle, reflection + :tickets: 5421 + + Fixed bug in Oracle dialect where indexes that contain the full set of + primary key columns would be mistaken as the primary key index itself, + which is omitted, even if there were multiples. The check has been refined + to compare the name of the primary key constraint against the index name + itself, rather than trying to guess based on the columns present in the + index. \ No newline at end of file diff --git a/lib/sqlalchemy/dialects/oracle/base.py b/lib/sqlalchemy/dialects/oracle/base.py index 5e912a0c2f..34c665fbe5 100644 --- a/lib/sqlalchemy/dialects/oracle/base.py +++ b/lib/sqlalchemy/dialects/oracle/base.py @@ -1898,7 +1898,7 @@ class OracleDialect(default.DefaultDialect): dblink=dblink, info_cache=kw.get("info_cache"), ) - pkeys = pk_constraint["constrained_columns"] + uniqueness = dict(NONUNIQUE=False, UNIQUE=True) enabled = dict(DISABLED=False, ENABLED=True) @@ -1906,9 +1906,22 @@ class OracleDialect(default.DefaultDialect): index = None for rset in rp: + index_name_normalized = self.normalize_name(rset.index_name) + + # skip primary key index. This is refined as of + # [ticket:5421]. Note that ALL_INDEXES.GENERATED will by "Y" + # if the name of this index was generated by Oracle, however + # if a named primary key constraint was created then this flag + # is false. + if ( + pk_constraint + and index_name_normalized == pk_constraint["name"] + ): + continue + if rset.index_name != last_index_name: index = dict( - name=self.normalize_name(rset.index_name), + name=index_name_normalized, column_names=[], dialect_options={}, ) @@ -1930,18 +1943,6 @@ class OracleDialect(default.DefaultDialect): ) last_index_name = rset.index_name - def upper_name_set(names): - return {i.upper() for i in names} - - pk_names = upper_name_set(pkeys) - if pk_names: - - def is_pk_index(index): - # don't include the primary key index - return upper_name_set(index["column_names"]) == pk_names - - indexes = [idx for idx in indexes if not is_pk_index(idx)] - return indexes @reflection.cache diff --git a/test/dialect/oracle/test_reflection.py b/test/dialect/oracle/test_reflection.py index 3d1361adb6..97fb627f9c 100644 --- a/test/dialect/oracle/test_reflection.py +++ b/test/dialect/oracle/test_reflection.py @@ -532,6 +532,83 @@ class RoundTripIndexTest(fixtures.TestBase): __only_on__ = "oracle" __backend__ = True + @testing.provide_metadata + def test_no_pk(self): + metadata = self.metadata + + Table( + "sometable", + metadata, + Column("id_a", Unicode(255)), + Column("id_b", Unicode(255)), + Index("pk_idx_1", "id_a", "id_b", unique=True), + Index("pk_idx_2", "id_b", "id_a", unique=True), + ) + metadata.create_all() + + insp = inspect(testing.db) + eq_( + insp.get_indexes("sometable"), + [ + { + "name": "pk_idx_1", + "column_names": ["id_a", "id_b"], + "dialect_options": {}, + "unique": True, + }, + { + "name": "pk_idx_2", + "column_names": ["id_b", "id_a"], + "dialect_options": {}, + "unique": True, + }, + ], + ) + + @testing.combinations((True,), (False,)) + @testing.provide_metadata + def test_include_indexes_resembling_pk(self, explicit_pk): + metadata = self.metadata + + t = Table( + "sometable", + metadata, + Column("id_a", Unicode(255), primary_key=True), + Column("id_b", Unicode(255), primary_key=True), + Column("group", Unicode(255), primary_key=True), + Column("col", Unicode(255)), + # Oracle won't let you do this unless the indexes have + # the columns in different order + Index("pk_idx_1", "id_b", "id_a", "group", unique=True), + Index("pk_idx_2", "id_b", "group", "id_a", unique=True), + ) + if explicit_pk: + t.append_constraint( + PrimaryKeyConstraint( + "id_a", "id_b", "group", name="some_primary_key" + ) + ) + metadata.create_all() + + insp = inspect(testing.db) + eq_( + insp.get_indexes("sometable"), + [ + { + "name": "pk_idx_1", + "column_names": ["id_b", "id_a", "group"], + "dialect_options": {}, + "unique": True, + }, + { + "name": "pk_idx_2", + "column_names": ["id_b", "group", "id_a"], + "dialect_options": {}, + "unique": True, + }, + ], + ) + @testing.provide_metadata def test_basic(self): metadata = self.metadata @@ -560,8 +637,10 @@ class RoundTripIndexTest(fixtures.TestBase): ) metadata.create_all() + mirror = MetaData(testing.db) mirror.reflect() + metadata.drop_all() mirror.create_all()