Insert, Updates, Deletes

    Top level “INSERT”, “UPDATE”, “DELETE” constructors.

    function sqlalchemy.sql.expression.delete(table: _DMLTableArgument) →

    Construct Delete object.

    E.g.:

    Similar functionality is available via the method on Table.

    • Parameters:

      table – The table to delete rows from.

    See also

    - in the SQLAlchemy Unified Tutorial

    function sqlalchemy.sql.expression.insert(table: _DMLTableArgument) →

    Construct an Insert object.

    E.g.:

    3. stmt = (
    4.     insert(user_table).
    5.     values(name='username', fullname='Full Username')
    6. )

    Similar functionality is available via the method on Table.

    See also

    - in the SQLAlchemy Unified Tutorial

    • Parameters:

      • table – which is the subject of the insert.

      • values – collection of values to be inserted; see Insert.values() for a description of allowed formats here. Can be omitted entirely; a construct will also dynamically render the VALUES clause at execution time based on the parameters passed to Connection.execute().

      • inline – if True, no attempt will be made to retrieve the SQL-generated default values to be provided within the statement; in particular, this allows SQL expressions to be rendered ‘inline’ within the statement without the need to pre-execute them beforehand; for backends that support “returning”, this turns off the “implicit returning” feature for the statement.

    If both and compile-time bind parameters are present, the compile-time bind parameters override the information specified within insert.values on a per-key basis.

    The keys within can be either Column objects or their string identifiers. Each key may reference one of:

    • a literal data value (i.e. string, number, etc.);

    • a Column object;

    • a SELECT statement.

    If a SELECT statement is specified which references this INSERT statement’s table, the statement will be correlated against the INSERT statement.

    See also

    - in the SQLAlchemy Unified Tutorial

    function sqlalchemy.sql.expression.update(table: _DMLTableArgument) →

    Construct an Update object.

    E.g.:

    1. from sqlalchemy import update
    3. stmt = (
    4.     update(user_table).
    5.     where(user_table.c.id == 5).
    6.     values(name='user #5')
    7. )

    Similar functionality is available via the method on Table.

    • Parameters:

      table – A object representing the database table to be updated.

    See also

    Using UPDATE and DELETE Statements - in the

    DML Class Documentation Constructors

    Class documentation for the constructors listed at .

    Object NameDescription

    Delete

    Represent a DELETE construct.

    Represent an INSERT construct.

    Update

    Represent an Update construct.

    Form the base for INSERT, UPDATE, and DELETE statements.

    ValuesBase

    Supplies support for to INSERT and UPDATE constructs.

    class sqlalchemy.sql.expression.Delete

    Represent a DELETE construct.

    The Delete object is created using the function.

    Members

    where(),

    Class signature

    class sqlalchemy.sql.expression.Delete (sqlalchemy.sql.expression.DMLWhereBase, )

    • method sqlalchemy.sql.expression.Delete.where(*whereclause: _ColumnExpressionArgument[bool]) → SelfDMLWhereBase

      inherited from the DMLWhereBase.where() method of DMLWhereBase

      Return a new construct with the given expression(s) added to its WHERE clause, joined to the existing clause via AND, if any.

      Both and Delete.where() support multiple-table forms, including database-specific UPDATE...FROM as well as DELETE..USING. For backends that don’t have multiple-table support, a backend agnostic approach to using multiple tables is to make use of correlated subqueries. See the linked tutorial sections below for examples.

      See also

      UPDATE..FROM

    • method sqlalchemy.sql.expression.Delete.returning(*cols: _ColumnsClauseArgument[Any], **_UpdateBase\_kw: Any_) →

      inherited from the UpdateBase.returning() method of

      Add a RETURNING or equivalent clause to this statement.

      e.g.:

      1. >>> stmt = (
      2. ...     table.update()
      3. ...     .where(table.c.data == "value")
      4. ...     .values(status="X")
      5. ...     .returning(table.c.server_flag, table.c.updated_timestamp)
      6. ... )
      7. >>> print(stmt)
      8. UPDATE some_table SET status=:status
      9. WHERE some_table.data = :data_1
      10. RETURNING some_table.server_flag, some_table.updated_timestamp

      The method may be invoked multiple times to add new entries to the list of expressions to be returned.

      New in version 1.4.0b2: The method may be invoked multiple times to add new entries to the list of expressions to be returned.

      The given collection of column expressions should be derived from the table that is the target of the INSERT, UPDATE, or DELETE. While objects are typical, the elements can also be expressions:

      1. >>> stmt = table.insert().returning(
      2. ...     (table.c.first_name + " " + table.c.last_name).label("fullname")
      3. ... )
      4. >>> print(stmt)
      5. INSERT INTO some_table (first_name, last_name)
      6. VALUES (:first_name, :last_name)
      7. RETURNING some_table.first_name || :first_name_1 || some_table.last_name AS fullname

      Upon compilation, a RETURNING clause, or database equivalent, will be rendered within the statement. For INSERT and UPDATE, the values are the newly inserted/updated values. For DELETE, the values are those of the rows which were deleted.

      Upon execution, the values of the columns to be returned are made available via the result set and can be iterated using CursorResult.fetchone() and similar. For DBAPIs which do not natively support returning values (i.e. cx_oracle), SQLAlchemy will approximate this behavior at the result level so that a reasonable amount of behavioral neutrality is provided.

      Note that not all databases/DBAPIs support RETURNING. For those backends with no support, an exception is raised upon compilation and/or execution. For those who do support it, the functionality across backends varies greatly, including restrictions on executemany() and other statements which return multiple rows. Please read the documentation notes for the database in use in order to determine the availability of RETURNING.

      See also

      - an alternative method tailored towards efficient fetching of server-side defaults and triggers for single-row INSERTs or UPDATEs.

      INSERT…RETURNING - in the

    class sqlalchemy.sql.expression.Insert

    Represent an INSERT construct.

    The Insert object is created using the function.

    Members

    values(), , from_select(), , select

    Class signature

    class (sqlalchemy.sql.expression.ValuesBase)

    • method values(*args: Union[Dict[_DMLColumnArgument, Any], Sequence[Any]], **kwargs: Any) → SelfValuesBase

      inherited from the method of ValuesBase

      Specify a fixed VALUES clause for an INSERT statement, or the SET clause for an UPDATE.

      Note that the and Update constructs support per-execution time formatting of the VALUES and/or SET clauses, based on the arguments passed to . However, the ValuesBase.values() method can be used to “fix” a particular set of parameters into the statement.

      Multiple calls to will produce a new construct, each one with the parameter list modified to include the new parameters sent. In the typical case of a single dictionary of parameters, the newly passed keys will replace the same keys in the previous construct. In the case of a list-based “multiple values” construct, each new list of values is extended onto the existing list of values.

      • Parameters:

        • **kwargs –

          key value pairs representing the string key of a Column mapped to the value to be rendered into the VALUES or SET clause:

          1. users.insert().values(name="some name")
          3. users.update().where(users.c.id==5).values(name="some name")

        • *args –

          As an alternative to passing key/value parameters, a dictionary, tuple, or list of dictionaries or tuples can be passed as a single positional argument in order to form the VALUES or SET clause of the statement. The forms that are accepted vary based on whether this is an or an Update construct.

          For either an or Update construct, a single dictionary can be passed, which works the same as that of the kwargs form:

          1. users.insert().values({"name": "some name"})
          3. users.update().values({"name": "some new name"})

          Also for either form but more typically for the construct, a tuple that contains an entry for every column in the table is also accepted:

          1. users.insert().values((5, "some name"))

          The Insert construct also supports being passed a list of dictionaries or full-table-tuples, which on the server will render the less common SQL syntax of “multiple values” - this syntax is supported on backends such as SQLite, PostgreSQL, MySQL, but not necessarily others:

          1. users.insert().values([
          2.                     {"name": "some name"},
          3.                     {"name": "some other name"},
          4.                     {"name": "yet another name"},
          5.                 ])

          The above form would render a multiple VALUES statement similar to:

          1. INSERT INTO users (name) VALUES
          2.                 (:name_1),
          3.                 (:name_2),
          4.                 (:name_3)

    • method returning(*cols: _ColumnsClauseArgument[Any], **_UpdateBase\_kw: Any_) → UpdateBase

      inherited from the method of UpdateBase

      Add a or equivalent clause to this statement.

      e.g.:

      1. >>> stmt = (
      2. ...     table.update()
      3. ...     .where(table.c.data == "value")
      4. ...     .values(status="X")
      5. ...     .returning(table.c.server_flag, table.c.updated_timestamp)
      6. ... )
      8. WHERE some_table.data = :data_1
      9. RETURNING some_table.server_flag, some_table.updated_timestamp

      The method may be invoked multiple times to add new entries to the list of expressions to be returned.

      New in version 1.4.0b2: The method may be invoked multiple times to add new entries to the list of expressions to be returned.

      The given collection of column expressions should be derived from the table that is the target of the INSERT, UPDATE, or DELETE. While Column objects are typical, the elements can also be expressions:

      Upon compilation, a RETURNING clause, or database equivalent, will be rendered within the statement. For INSERT and UPDATE, the values are the newly inserted/updated values. For DELETE, the values are those of the rows which were deleted.

      Upon execution, the values of the columns to be returned are made available via the result set and can be iterated using and similar. For DBAPIs which do not natively support returning values (i.e. cx_oracle), SQLAlchemy will approximate this behavior at the result level so that a reasonable amount of behavioral neutrality is provided.

      Note that not all databases/DBAPIs support RETURNING. For those backends with no support, an exception is raised upon compilation and/or execution. For those who do support it, the functionality across backends varies greatly, including restrictions on executemany() and other statements which return multiple rows. Please read the documentation notes for the database in use in order to determine the availability of RETURNING.

      See also

      UpdateBase.return_defaults() - an alternative method tailored towards efficient fetching of server-side defaults and triggers for single-row INSERTs or UPDATEs.

      - in the SQLAlchemy Unified Tutorial

    • method from_select(names: List[str], select: Selectable, include_defaults: bool = True) → SelfInsert

      Return a new construct which represents an INSERT...FROM SELECT statement.

      e.g.:

      1. sel = select(table1.c.a, table1.c.b).where(table1.c.c > 5)
      2. ins = table2.insert().from_select(['a', 'b'], sel)

      • Parameters:

        • names – a sequence of string column names or Column objects representing the target columns.

        • select – a construct, FromClause or other construct which resolves into a , such as an ORM Query object, etc. The order of columns returned from this FROM clause should correspond to the order of columns sent as the names parameter; while this is not checked before passing along to the database, the database would normally raise an exception if these column lists don’t correspond.

        • include_defaults –

          if True, non-server default values and SQL expressions as specified on objects (as documented in Column INSERT/UPDATE Defaults) not otherwise specified in the list of names will be rendered into the INSERT and SELECT statements, so that these values are also included in the data to be inserted.

          Note

          A Python-side default that uses a Python callable function will only be invoked once for the whole statement, and not per row.

          New in version 1.0.0: - now renders Python-side and SQL expression column defaults into the SELECT statement for columns otherwise not included in the list of column names.

    1. Changed in version 1.0.0: an INSERT that uses FROM SELECT implies that the [insert.inline](#sqlalchemy.sql.expression.insert.params.inline "sqlalchemy.sql.expression.insert") flag is set to True, indicating that the statement will not attempt to fetch the “last inserted primary key” or other defaults. The statement deals with an arbitrary number of rows, so the [CursorResult.inserted\_primary\_key]($3743e3464fa80ce7.md#sqlalchemy.engine.CursorResult.inserted_primary_key "sqlalchemy.engine.CursorResult.inserted_primary_key") accessor does not apply.

    • method sqlalchemy.sql.expression.Insert.inline() → SelfInsert

      Make this construct “inline” .

      When set, no attempt will be made to retrieve the SQL-generated default values to be provided within the statement; in particular, this allows SQL expressions to be rendered ‘inline’ within the statement without the need to pre-execute them beforehand; for backends that support “returning”, this turns off the “implicit returning” feature for the statement.

      Changed in version 1.4: the Insert.inline parameter is now superseded by the method.

    • attribute sqlalchemy.sql.expression.Insert.select: Optional[[Any]] = None

      SELECT statement for INSERT .. FROM SELECT

    class sqlalchemy.sql.expression.Update

    Represent an Update construct.

    The Update object is created using the function.

    Members

    returning(), , values(), , ordered_values()

    Class signature

    class (sqlalchemy.sql.expression.DMLWhereBase, sqlalchemy.sql.expression.ValuesBase)

    • method returning(*cols: _ColumnsClauseArgument[Any], **_UpdateBase\_kw: Any_) → UpdateBase

      inherited from the method of UpdateBase

      Add a or equivalent clause to this statement.

      e.g.:

      1. >>> stmt = (
      2. ...     table.update()
      3. ...     .where(table.c.data == "value")
      4. ...     .values(status="X")
      5. ...     .returning(table.c.server_flag, table.c.updated_timestamp)
      6. ... )
      7. >>> print(stmt)
      8. UPDATE some_table SET status=:status
      9. WHERE some_table.data = :data_1
      10. RETURNING some_table.server_flag, some_table.updated_timestamp

      The method may be invoked multiple times to add new entries to the list of expressions to be returned.

      New in version 1.4.0b2: The method may be invoked multiple times to add new entries to the list of expressions to be returned.

      The given collection of column expressions should be derived from the table that is the target of the INSERT, UPDATE, or DELETE. While Column objects are typical, the elements can also be expressions:

      1. >>> stmt = table.insert().returning(
      2. ...     (table.c.first_name + " " + table.c.last_name).label("fullname")
      3. ... )
      4. >>> print(stmt)
      5. INSERT INTO some_table (first_name, last_name)
      6. VALUES (:first_name, :last_name)
      7. RETURNING some_table.first_name || :first_name_1 || some_table.last_name AS fullname

      Upon compilation, a RETURNING clause, or database equivalent, will be rendered within the statement. For INSERT and UPDATE, the values are the newly inserted/updated values. For DELETE, the values are those of the rows which were deleted.

      Upon execution, the values of the columns to be returned are made available via the result set and can be iterated using and similar. For DBAPIs which do not natively support returning values (i.e. cx_oracle), SQLAlchemy will approximate this behavior at the result level so that a reasonable amount of behavioral neutrality is provided.

      Note that not all databases/DBAPIs support RETURNING. For those backends with no support, an exception is raised upon compilation and/or execution. For those who do support it, the functionality across backends varies greatly, including restrictions on executemany() and other statements which return multiple rows. Please read the documentation notes for the database in use in order to determine the availability of RETURNING.

      See also

      UpdateBase.return_defaults() - an alternative method tailored towards efficient fetching of server-side defaults and triggers for single-row INSERTs or UPDATEs.

      - in the SQLAlchemy Unified Tutorial

    • method where(*whereclause: _ColumnExpressionArgument[bool]) → SelfDMLWhereBase

      inherited from the DMLWhereBase.where() method of DMLWhereBase

      Return a new construct with the given expression(s) added to its WHERE clause, joined to the existing clause via AND, if any.

      Both Update.where() and support multiple-table forms, including database-specific UPDATE...FROM as well as DELETE..USING. For backends that don’t have multiple-table support, a backend agnostic approach to using multiple tables is to make use of correlated subqueries. See the linked tutorial sections below for examples.

      See also

      Correlated Updates

      Multiple Table Deletes

    • method values(*args: Union[Dict[_DMLColumnArgument, Any], Sequence[Any]], **kwargs: Any) → SelfValuesBase

      inherited from the method of ValuesBase

      Specify a fixed VALUES clause for an INSERT statement, or the SET clause for an UPDATE.

      Note that the and Update constructs support per-execution time formatting of the VALUES and/or SET clauses, based on the arguments passed to . However, the ValuesBase.values() method can be used to “fix” a particular set of parameters into the statement.

      Multiple calls to will produce a new construct, each one with the parameter list modified to include the new parameters sent. In the typical case of a single dictionary of parameters, the newly passed keys will replace the same keys in the previous construct. In the case of a list-based “multiple values” construct, each new list of values is extended onto the existing list of values.

      • Parameters:

        • **kwargs –

          key value pairs representing the string key of a Column mapped to the value to be rendered into the VALUES or SET clause:

          1. users.insert().values(name="some name")
          3. users.update().where(users.c.id==5).values(name="some name")

        • *args –

          As an alternative to passing key/value parameters, a dictionary, tuple, or list of dictionaries or tuples can be passed as a single positional argument in order to form the VALUES or SET clause of the statement. The forms that are accepted vary based on whether this is an or an Update construct.

          For either an or Update construct, a single dictionary can be passed, which works the same as that of the kwargs form:

          1. users.insert().values({"name": "some name"})
          3. users.update().values({"name": "some new name"})

          Also for either form but more typically for the construct, a tuple that contains an entry for every column in the table is also accepted:

          1. users.insert().values((5, "some name"))

          The Insert construct also supports being passed a list of dictionaries or full-table-tuples, which on the server will render the less common SQL syntax of “multiple values” - this syntax is supported on backends such as SQLite, PostgreSQL, MySQL, but not necessarily others:

          1. users.insert().values([
          2.                     {"name": "some name"},
          3.                     {"name": "some other name"},
          4.                     {"name": "yet another name"},
          5.                 ])

          The above form would render a multiple VALUES statement similar to:

          1. INSERT INTO users (name) VALUES
          2.                 (:name_1),
          3.                 (:name_2),
          4.                 (:name_3)

          It is essential to note that passing multiple values is NOT the same as using traditional executemany() form. The above syntax is a special syntax not typically used. To emit an INSERT statement against multiple rows, the normal method is to pass a multiple values list to the method, which is supported by all database backends and is generally more efficient for a very large number of parameters.

    • method sqlalchemy.sql.expression.Update.inline() → SelfUpdate

      Make this construct “inline” .

      When set, SQL defaults present on Column objects via the default keyword will be compiled ‘inline’ into the statement and not pre-executed. This means that their values will not be available in the dictionary returned from .

      Changed in version 1.4: the update.inline parameter is now superseded by the method.

    • method sqlalchemy.sql.expression.Update.ordered_values(*args: [_DMLColumnArgument, Any]) → SelfUpdate

      Specify the VALUES clause of this UPDATE statement with an explicit parameter ordering that will be maintained in the SET clause of the resulting UPDATE statement.

      E.g.:

      1. stmt = table.update().ordered_values(
      2.     ("name", "ed"), ("ident": "foo")
      3. )

      See also

      Parameter Ordered Updates - full example of the method.

      Changed in version 1.4: The Update.ordered_values() method supersedes the parameter, which will be removed in SQLAlchemy 2.0.

    class sqlalchemy.sql.expression.UpdateBase

    Form the base for INSERT, UPDATE, and DELETE statements.

    Members

    entity_description, , params(), , returning(), , with_dialect_options(),

    class sqlalchemy.sql.expression.UpdateBase (sqlalchemy.sql.roles.DMLRole, , sqlalchemy.sql.expression.HasCompileState, sqlalchemy.sql.base.DialectKWArgs, , sqlalchemy.sql.expression.Generative, sqlalchemy.sql.expression.ExecutableReturnsRows, sqlalchemy.sql.expression.ClauseElement)

    • attribute entity_description

      Return a plugin-enabled description of the table and/or entity which this DML construct is operating against.

      This attribute is generally useful when using the ORM, as an extended structure which includes information about mapped entities is returned. The section contains more background.

      For a Core statement, the structure returned by this accessor is derived from the UpdateBase.table attribute, and refers to the Table being inserted, updated, or deleted:

      New in version 1.4.33.

      See also

      Select.column_descriptions - entity information for a construct

      Inspecting entities and columns from ORM-enabled SELECT and DML statements - ORM background

    • attribute exported_columns

      Return the RETURNING columns as a column collection for this statement.

      New in version 1.4.

    • method sqlalchemy.sql.expression.UpdateBase.params(*arg: Any, **kw: Any) → NoReturn

      Set the parameters for the statement.

      This method raises NotImplementedError on the base class, and is overridden by to provide the SET/VALUES clause of UPDATE and INSERT.

    • method sqlalchemy.sql.expression.UpdateBase.return_defaults(*cols: _DMLColumnArgument, supplemental_cols: Optional[Iterable[_DMLColumnArgument]] = None) → SelfUpdateBase

      Make use of a clause for the purpose of fetching server-side expressions and defaults, for supporting backends only.

      Deep Alchemy

      The UpdateBase.return_defaults() method is used by the ORM for its internal work in fetching newly generated primary key and server default values, in particular to provide the underyling implementation of the ORM feature as well as to allow RETURNING support with bulk ORM inserts. Its behavior is fairly idiosyncratic and is not really intended for general use. End users should stick with using UpdateBase.returning() in order to add RETURNING clauses to their INSERT, UPDATE and DELETE statements.

      Normally, a single row INSERT statement will automatically populate the attribute when executed, which stores the primary key of the row that was just inserted in the form of a Row object with column names as named tuple keys (and the view fully populated as well). The dialect in use chooses the strategy to use in order to populate this data; if it was generated using server-side defaults and / or SQL expressions, dialect-specific approaches such as cursor.lastrowid or RETURNING are typically used to acquire the new primary key value.

      However, when the statement is modified by calling UpdateBase.return_defaults() before executing the statement, additional behaviors take place only for backends that support RETURNING and for objects that maintain the Table.implicit_returning parameter at its default value of True. In these cases, when the is returned from the statement’s execution, not only will CursorResult.inserted_primary_key be populated as always, the attribute will also be populated with a Row named-tuple representing the full range of server generated values from that single row, including values for any columns that specify or which make use of Column.default using a SQL expression.

      When invoking INSERT statements with multiple rows using , the UpdateBase.return_defaults() modifier will have the effect of the and CursorResult.returned_defaults_rows attributes being fully populated with lists of objects representing newly inserted primary key values as well as newly inserted server generated values for each row inserted. The CursorResult.inserted_primary_key and attributes will also continue to be populated with the first row of these two collections.

      If the backend does not support RETURNING or the Table in use has disabled , then no RETURNING clause is added and no additional data is fetched, however the INSERT, UPDATE or DELETE statement proceeds normally.

      E.g.:

      1. stmt = table.insert().values(data='newdata').return_defaults()
      5. server_created_at = result.returned_defaults['created_at']

      When used against an UPDATE statement UpdateBase.return_defaults() instead looks for columns that include or Column.server_onupdate parameters assigned, when constructing the columns that will be included in the RETURNING clause by default if explicit columns were not specified. When used against a DELETE statement, no columns are included in RETURNING by default, they instead must be specified explicitly as there are no columns that normally change values when a DELETE statement proceeds.

      New in version 2.0: is supported for DELETE statements also and has been moved from ValuesBase to .

      The UpdateBase.return_defaults() method is mutually exclusive against the method and errors will be raised during the SQL compilation process if both are used at the same time on one statement. The RETURNING clause of the INSERT, UPDATE or DELETE statement is therefore controlled by only one of these methods at a time.

      The UpdateBase.return_defaults() method differs from in these ways:

      1. UpdateBase.return_defaults() method causes the collection to be populated with the first row from the RETURNING result. This attribute is not populated when using UpdateBase.returning().

      2. is compatible with existing logic used to fetch auto-generated primary key values that are then populated into the CursorResult.inserted_primary_key attribute. By contrast, using will have the effect of the CursorResult.inserted_primary_key attribute being left unpopulated.

      3. can be called against any backend. Backends that don’t support RETURNING will skip the usage of the feature, rather than raising an exception. The return value of CursorResult.returned_defaults will be None for backends that don’t support RETURNING or for which the target sets Table.implicit_returning to False.

      4. An INSERT statement invoked with executemany() is supported if the backend database driver supports the feature which is now supported by most SQLAlchemy-included backends. When executemany is used, the CursorResult.returned_defaults_rows and accessors will return the inserted defaults and primary keys.

        New in version 1.4: Added CursorResult.returned_defaults_rows and accessors. In version 2.0, the underlying implementation which fetches and populates the data for these attributes was generalized to be supported by most backends, whereas in 1.4 they were only supported by the psycopg2 driver.

    1. See also
    3. [UpdateBase.returning()](#sqlalchemy.sql.expression.UpdateBase.returning "sqlalchemy.sql.expression.UpdateBase.returning")
    5. [CursorResult.returned\_defaults]($3743e3464fa80ce7.md#sqlalchemy.engine.CursorResult.returned_defaults "sqlalchemy.engine.CursorResult.returned_defaults")
    7. [CursorResult.returned\_defaults\_rows]($3743e3464fa80ce7.md#sqlalchemy.engine.CursorResult.returned_defaults_rows "sqlalchemy.engine.CursorResult.returned_defaults_rows")
    9. [CursorResult.inserted\_primary\_key]($3743e3464fa80ce7.md#sqlalchemy.engine.CursorResult.inserted_primary_key "sqlalchemy.engine.CursorResult.inserted_primary_key")
    11. [CursorResult.inserted\_primary\_key\_rows]($3743e3464fa80ce7.md#sqlalchemy.engine.CursorResult.inserted_primary_key_rows "sqlalchemy.engine.CursorResult.inserted_primary_key_rows")

    • method sqlalchemy.sql.expression.UpdateBase.returning(*cols: _ColumnsClauseArgument[Any], **_UpdateBase\_kw: Any_) →

      Add a RETURNING or equivalent clause to this statement.

      e.g.:

      1. >>> stmt = (
      2. ...     table.update()
      3. ...     .where(table.c.data == "value")
      4. ...     .values(status="X")
      5. ...     .returning(table.c.server_flag, table.c.updated_timestamp)
      6. ... )
      7. >>> print(stmt)
      8. UPDATE some_table SET status=:status
      9. WHERE some_table.data = :data_1
      10. RETURNING some_table.server_flag, some_table.updated_timestamp

      The method may be invoked multiple times to add new entries to the list of expressions to be returned.

      New in version 1.4.0b2: The method may be invoked multiple times to add new entries to the list of expressions to be returned.

      The given collection of column expressions should be derived from the table that is the target of the INSERT, UPDATE, or DELETE. While objects are typical, the elements can also be expressions:

      1. >>> stmt = table.insert().returning(
      2. ...     (table.c.first_name + " " + table.c.last_name).label("fullname")
      3. ... )
      4. >>> print(stmt)
      5. INSERT INTO some_table (first_name, last_name)
      6. VALUES (:first_name, :last_name)
      7. RETURNING some_table.first_name || :first_name_1 || some_table.last_name AS fullname

      Upon compilation, a RETURNING clause, or database equivalent, will be rendered within the statement. For INSERT and UPDATE, the values are the newly inserted/updated values. For DELETE, the values are those of the rows which were deleted.

      Upon execution, the values of the columns to be returned are made available via the result set and can be iterated using CursorResult.fetchone() and similar. For DBAPIs which do not natively support returning values (i.e. cx_oracle), SQLAlchemy will approximate this behavior at the result level so that a reasonable amount of behavioral neutrality is provided.

      Note that not all databases/DBAPIs support RETURNING. For those backends with no support, an exception is raised upon compilation and/or execution. For those who do support it, the functionality across backends varies greatly, including restrictions on executemany() and other statements which return multiple rows. Please read the documentation notes for the database in use in order to determine the availability of RETURNING.

      See also

      - an alternative method tailored towards efficient fetching of server-side defaults and triggers for single-row INSERTs or UPDATEs.

      INSERT…RETURNING - in the

    • attribute sqlalchemy.sql.expression.UpdateBase.returning_column_descriptions

      Return a description of the columns which this DML construct is RETURNING against, in other words the expressions established as part of UpdateBase.returning().

      This attribute is generally useful when using the ORM, as an extended structure which includes information about mapped entities is returned. The section contains more background.

      For a Core statement, the structure returned by this accessor is derived from the same objects that are returned by the UpdateBase.exported_columns accessor:

      1. >>> stmt = insert(user_table).returning(user_table.c.id, user_table.c.name)
      2. >>> stmt.entity_description
      3. [
      4.     {
      5.         "name": "id",
      6.         "type": Integer,
      7.         "expr": Column("id", Integer(), table=<user>, ...)
      8.     },
      9.     {
      10.         "name": "name",
      11.         "type": String(),
      12.         "expr": Column("name", String(), table=<user>, ...)
      13.     },
      14. ]

      New in version 1.4.33.

      See also

      Select.column_descriptions - entity information for a construct

      Inspecting entities and columns from ORM-enabled SELECT and DML statements - ORM background

    • method with_dialect_options(**opt: Any) → SelfUpdateBase

      Add dialect options to this INSERT/UPDATE/DELETE object.

      e.g.:

      1. upd = table.update().dialect_options(mysql_limit=10)

    • method sqlalchemy.sql.expression.UpdateBase.with_hint(text: str, selectable: Optional[_DMLTableArgument] = None, dialect_name: str = ‘*‘) → SelfUpdateBase

      Add a table hint for a single table to this INSERT/UPDATE/DELETE statement.

      Note

      currently applies only to Microsoft SQL Server. For MySQL INSERT/UPDATE/DELETE hints, use UpdateBase.prefix_with().

      The text of the hint is rendered in the appropriate location for the database backend in use, relative to the Table that is the subject of this statement, or optionally to that of the given passed as the selectable argument.

      The dialect_name option will limit the rendering of a particular hint to a particular backend. Such as, to add a hint that only takes effect for SQL Server:

      1. mytable.insert().with_hint("WITH (PAGLOCK)", dialect_name="mssql")

      • Parameters:

        • text – Text of the hint.

        • selectable – optional Table that specifies an element of the FROM clause within an UPDATE or DELETE to be the subject of the hint - applies only to certain backends.

        • dialect_name – defaults to *, if specified as the name of a particular dialect, will apply these hints only when that dialect is in use.

    class sqlalchemy.sql.expression.ValuesBase

    Supplies support for to INSERT and UPDATE constructs.

    Members

    select,

    Class signature

    class sqlalchemy.sql.expression.ValuesBase ()

    • attribute sqlalchemy.sql.expression.ValuesBase.select: Optional[[Any]] = None

      SELECT statement for INSERT .. FROM SELECT

    • method sqlalchemy.sql.expression.ValuesBase.values(*args: Union[Dict[_DMLColumnArgument, Any], [Any]], **kwargs: Any) → SelfValuesBase

      Specify a fixed VALUES clause for an INSERT statement, or the SET clause for an UPDATE.

      Note that the Insert and constructs support per-execution time formatting of the VALUES and/or SET clauses, based on the arguments passed to Connection.execute(). However, the method can be used to “fix” a particular set of parameters into the statement.

      Multiple calls to ValuesBase.values() will produce a new construct, each one with the parameter list modified to include the new parameters sent. In the typical case of a single dictionary of parameters, the newly passed keys will replace the same keys in the previous construct. In the case of a list-based “multiple values” construct, each new list of values is extended onto the existing list of values.

      • Parameters:

        • **kwargs –

          key value pairs representing the string key of a mapped to the value to be rendered into the VALUES or SET clause:

          1. users.insert().values(name="some name")
          3. users.update().where(users.c.id==5).values(name="some name")

        • *args –

          As an alternative to passing key/value parameters, a dictionary, tuple, or list of dictionaries or tuples can be passed as a single positional argument in order to form the VALUES or SET clause of the statement. The forms that are accepted vary based on whether this is an Insert or an construct.

          For either an Insert or construct, a single dictionary can be passed, which works the same as that of the kwargs form:

          1. users.insert().values({"name": "some name"})
          3. users.update().values({"name": "some new name"})

          Also for either form but more typically for the Insert construct, a tuple that contains an entry for every column in the table is also accepted:

          1. users.insert().values((5, "some name"))

          The construct also supports being passed a list of dictionaries or full-table-tuples, which on the server will render the less common SQL syntax of “multiple values” - this syntax is supported on backends such as SQLite, PostgreSQL, MySQL, but not necessarily others:

          The above form would render a multiple VALUES statement similar to:

          1. INSERT INTO users (name) VALUES
          2.                 (:name_1),
          3.                 (:name_2),

          It is essential to note that passing multiple values is NOT the same as using traditional executemany() form. The above syntax is a special syntax not typically used. To emit an INSERT statement against multiple rows, the normal method is to pass a multiple values list to the Connection.execute() method, which is supported by all database backends and is generally more efficient for a very large number of parameters.