Custom SQL Constructs and Compilation Extension

Provides an API for creation of custom ClauseElements and compilers.

Synopsis

Usage involves the creation of one or more ClauseElement subclasses and one or more callables defining its compilation:

  1. from sqlalchemy.ext.compiler import compiles
  2. from sqlalchemy.sql.expression import ColumnClause
  4. class MyColumn(ColumnClause):
  5.     inherit_cache = True
  7. @compiles(MyColumn)
  8. def compile_mycolumn(element, compiler, **kw):
  9.     return "[%s]" % element.name

Above, MyColumn extends ColumnClause, the base expression element for named column objects. The compiles decorator registers itself with the MyColumn class so that it is invoked when the object is compiled to a string:

  1. from sqlalchemy import select
  3. s = select(MyColumn('x'), MyColumn('y'))
  4. print(str(s))

Produces:

  1. SELECT [x], [y]

Dialect-specific compilation rules

Compilers can also be made dialect-specific. The appropriate compiler will be invoked for the dialect in use:

  1. from sqlalchemy.schema import DDLElement
  3. class AlterColumn(DDLElement):
  4.     inherit_cache = False
  6.     def __init__(self, column, cmd):
  7.         self.column = column
  8.         self.cmd = cmd
  10. @compiles(AlterColumn)
  11. def visit_alter_column(element, compiler, **kw):
  12.     return "ALTER COLUMN %s ..." % element.column.name
  14. @compiles(AlterColumn, 'postgresql')
  15. def visit_alter_column(element, compiler, **kw):
  16.     return "ALTER TABLE %s ALTER COLUMN %s ..." % (element.table.name,
  17.                                                    element.column.name)

The second visit_alter_table will be invoked when any postgresql dialect is used.

Compiling sub-elements of a custom expression construct

The compiler argument is the Compiled object in use. This object can be inspected for any information about the in-progress compilation, including compiler.dialect, compiler.statement etc. The SQLCompiler and DDLCompiler both include a process() method which can be used for compilation of embedded attributes:

  1. from sqlalchemy.sql.expression import Executable, ClauseElement
  3. class InsertFromSelect(Executable, ClauseElement):
  4.     inherit_cache = False
  6.     def __init__(self, table, select):
  7.         self.table = table
  8.         self.select = select
  10. @compiles(InsertFromSelect)
  11. def visit_insert_from_select(element, compiler, **kw):
  12.     return "INSERT INTO %s (%s)" % (
  13.         compiler.process(element.table, asfrom=True, **kw),
  14.         compiler.process(element.select, **kw)
  15.     )
  17. insert = InsertFromSelect(t1, select(t1).where(t1.c.x>5))
  18. print(insert)

Produces:

  1. "INSERT INTO mytable (SELECT mytable.x, mytable.y, mytable.z
  2.                       FROM mytable WHERE mytable.x > :x_1)"

Note

The above InsertFromSelect construct is only an example, this actual functionality is already available using the Insert.from_select() method.

Cross Compiling between SQL and DDL compilers

SQL and DDL constructs are each compiled using different base compilers - SQLCompiler and DDLCompiler. A common need is to access the compilation rules of SQL expressions from within a DDL expression. The DDLCompiler includes an accessor sql_compiler for this reason, such as below where we generate a CHECK constraint that embeds a SQL expression:

  1. @compiles(MyConstraint)
  2. def compile_my_constraint(constraint, ddlcompiler, **kw):
  3.     kw['literal_binds'] = True
  4.     return "CONSTRAINT %s CHECK (%s)" % (
  5.         constraint.name,
  6.         ddlcompiler.sql_compiler.process(
  7.             constraint.expression, **kw)
  8.     )

Above, we add an additional flag to the process step as called by SQLCompiler.process(), which is the literal_binds flag. This indicates that any SQL expression which refers to a BindParameter object or other “literal” object such as those which refer to strings or integers should be rendered in-place, rather than being referred to as a bound parameter; when emitting DDL, bound parameters are typically not supported.

Changing the default compilation of existing constructs

The compiler extension applies just as well to the existing constructs. When overriding the compilation of a built in SQL construct, the @compiles decorator is invoked upon the appropriate class (be sure to use the class, i.e. Insert or Select, instead of the creation function such as insert() or select()).

Within the new compilation function, to get at the “original” compilation routine, use the appropriate visit_XXX method - this because compiler.process() will call upon the overriding routine and cause an endless loop. Such as, to add “prefix” to all insert statements:

  1. from sqlalchemy.sql.expression import Insert
  3. @compiles(Insert)
  4. def prefix_inserts(insert, compiler, **kw):
  5.     return compiler.visit_insert(insert.prefix_with("some prefix"), **kw)

The above compiler will prefix all INSERT statements with “some prefix” when compiled.

Changing Compilation of Types

compiler works for types, too, such as below where we implement the MS-SQL specific ‘max’ keyword for String/VARCHAR:

  1. @compiles(String, 'mssql')
  2. @compiles(VARCHAR, 'mssql')
  3. def compile_varchar(element, compiler, **kw):
  4.     if element.length == 'max':
  5.         return "VARCHAR('max')"
  6.     else:
  7.         return compiler.visit_VARCHAR(element, **kw)
  9. foo = Table('foo', metadata,
  10.     Column('data', VARCHAR('max'))
  11. )

Subclassing Guidelines

A big part of using the compiler extension is subclassing SQLAlchemy expression constructs. To make this easier, the expression and schema packages feature a set of “bases” intended for common tasks. A synopsis is as follows:

  • ClauseElement - This is the root expression class. Any SQL expression can be derived from this base, and is probably the best choice for longer constructs such as specialized INSERT statements.

  • ColumnElement - The root of all “column-like” elements. Anything that you’d place in the “columns” clause of a SELECT statement (as well as order by and group by) can derive from this - the object will automatically have Python “comparison” behavior.

    ColumnElement classes want to have a type member which is expression’s return type. This can be established at the instance level in the constructor, or at the class level if its generally constant:

    1. class timestamp(ColumnElement):
    2.     type = TIMESTAMP()
    3.     inherit_cache = True

  • FunctionElement - This is a hybrid of a ColumnElement and a “from clause” like object, and represents a SQL function or stored procedure type of call. Since most databases support statements along the line of “SELECT FROM <some function>” FunctionElement adds in the ability to be used in the FROM clause of a select() construct:

    1. from sqlalchemy.sql.expression import FunctionElement
    3. class coalesce(FunctionElement):
    4.     name = 'coalesce'
    5.     inherit_cache = True
    7. @compiles(coalesce)
    8. def compile(element, compiler, **kw):
    9.     return "coalesce(%s)" % compiler.process(element.clauses, **kw)
    11. @compiles(coalesce, 'oracle')
    12. def compile(element, compiler, **kw):
    13.     if len(element.clauses) > 2:
    14.         raise TypeError("coalesce only supports two arguments on Oracle")
    15.     return "nvl(%s)" % compiler.process(element.clauses, **kw)

  • ExecutableDDLElement - The root of all DDL expressions, like CREATE TABLE, ALTER TABLE, etc. Compilation of ExecutableDDLElement subclasses is issued by a DDLCompiler instead of a SQLCompiler. ExecutableDDLElement can also be used as an event hook in conjunction with event hooks like DDLEvents.before_create() and DDLEvents.after_create(), allowing the construct to be invoked automatically during CREATE TABLE and DROP TABLE sequences.

    See also

    Customizing DDL - contains examples of associating DDL objects (which are themselves ExecutableDDLElement instances) with DDLEvents event hooks.

  • Executable - This is a mixin which should be used with any expression class that represents a “standalone” SQL statement that can be passed directly to an execute() method. It is already implicit within DDLElement and FunctionElement.

Most of the above constructs also respond to SQL statement caching. A subclassed construct will want to define the caching behavior for the object, which usually means setting the flag inherit_cache to the value of False or True. See the next section Enabling Caching Support for Custom Constructs for background.

Enabling Caching Support for Custom Constructs

SQLAlchemy as of version 1.4 includes a SQL compilation caching facility which will allow equivalent SQL constructs to cache their stringified form, along with other structural information used to fetch results from the statement.

For reasons discussed at Object will not produce a cache key, Performance Implications, the implementation of this caching system takes a conservative approach towards including custom SQL constructs and/or subclasses within the caching system. This includes that any user-defined SQL constructs, including all the examples for this extension, will not participate in caching by default unless they positively assert that they are able to do so. The HasCacheKey.inherit_cache attribute when set to True at the class level of a specific subclass will indicate that instances of this class may be safely cached, using the cache key generation scheme of the immediate superclass. This applies for example to the “synopsis” example indicated previously:

  1. class MyColumn(ColumnClause):
  2.     inherit_cache = True
  4. @compiles(MyColumn)
  5. def compile_mycolumn(element, compiler, **kw):
  6.     return "[%s]" % element.name

Above, the MyColumn class does not include any new state that affects its SQL compilation; the cache key of MyColumn instances will make use of that of the ColumnClause superclass, meaning it will take into account the class of the object (MyColumn), the string name and datatype of the object:

  1. >>> MyColumn("some_name", String())._generate_cache_key()
  2. CacheKey(
  3.     key=('0', <class '__main__.MyColumn'>,
  4.     'name', 'some_name',
  5.     'type', (<class 'sqlalchemy.sql.sqltypes.String'>,
  6.              ('length', None), ('collation', None))
  7. ), bindparams=[])

For objects that are likely to be used liberally as components within many larger statements, such as Column subclasses and custom SQL datatypes, it’s important that caching be enabled as much as possible, as this may otherwise negatively affect performance.

An example of an object that does contain state which affects its SQL compilation is the one illustrated at Compiling sub-elements of a custom expression construct; this is an “INSERT FROM SELECT” construct that combines together a Table as well as a Select construct, each of which independently affect the SQL string generation of the construct. For this class, the example illustrates that it simply does not participate in caching:

  1. class InsertFromSelect(Executable, ClauseElement):
  2.     inherit_cache = False
  4.     def __init__(self, table, select):
  5.         self.table = table
  6.         self.select = select
  8. @compiles(InsertFromSelect)
  9. def visit_insert_from_select(element, compiler, **kw):
  10.     return "INSERT INTO %s (%s)" % (
  11.         compiler.process(element.table, asfrom=True, **kw),
  12.         compiler.process(element.select, **kw)
  13.     )

While it is also possible that the above InsertFromSelect could be made to produce a cache key that is composed of that of the Table and Select components together, the API for this is not at the moment fully public. However, for an “INSERT FROM SELECT” construct, which is only used by itself for specific operations, caching is not as critical as in the previous example.

For objects that are used in relative isolation and are generally standalone, such as custom DML constructs like an “INSERT FROM SELECT”, caching is generally less critical as the lack of caching for such a construct will have only localized implications for that specific operation.

Further Examples

“UTC timestamp” function

A function that works like “CURRENT_TIMESTAMP” except applies the appropriate conversions so that the time is in UTC time. Timestamps are best stored in relational databases as UTC, without time zones. UTC so that your database doesn’t think time has gone backwards in the hour when daylight savings ends, without timezones because timezones are like character encodings - they’re best applied only at the endpoints of an application (i.e. convert to UTC upon user input, re-apply desired timezone upon display).

For PostgreSQL and Microsoft SQL Server:

  1. from sqlalchemy.sql import expression
  2. from sqlalchemy.ext.compiler import compiles
  3. from sqlalchemy.types import DateTime
  5. class utcnow(expression.FunctionElement):
  6.     type = DateTime()
  7.     inherit_cache = True
  9. @compiles(utcnow, 'postgresql')
  10. def pg_utcnow(element, compiler, **kw):
  11.     return "TIMEZONE('utc', CURRENT_TIMESTAMP)"
  13. @compiles(utcnow, 'mssql')
  14. def ms_utcnow(element, compiler, **kw):
  15.     return "GETUTCDATE()"

Example usage:

  1. from sqlalchemy import (
  2.             Table, Column, Integer, String, DateTime, MetaData
  3.         )
  4. metadata = MetaData()
  5. event = Table("event", metadata,
  6.     Column("id", Integer, primary_key=True),
  7.     Column("description", String(50), nullable=False),
  8.     Column("timestamp", DateTime, server_default=utcnow())
  9. )

“GREATEST” function

The “GREATEST” function is given any number of arguments and returns the one that is of the highest value - its equivalent to Python’s max function. A SQL standard version versus a CASE based version which only accommodates two arguments:

  1. from sqlalchemy.sql import expression, case
  2. from sqlalchemy.ext.compiler import compiles
  3. from sqlalchemy.types import Numeric
  5. class greatest(expression.FunctionElement):
  6.     type = Numeric()
  7.     name = 'greatest'
  8.     inherit_cache = True
  10. @compiles(greatest)
  11. def default_greatest(element, compiler, **kw):
  12.     return compiler.visit_function(element)
  14. @compiles(greatest, 'sqlite')
  15. @compiles(greatest, 'mssql')
  16. @compiles(greatest, 'oracle')
  17. def case_greatest(element, compiler, **kw):
  18.     arg1, arg2 = list(element.clauses)
  19.     return compiler.process(case([(arg1 > arg2, arg1)], else_=arg2), **kw)

Example usage:

  1. Session.query(Account).\
  2.         filter(
  3.             greatest(
  4.                 Account.checking_balance,
  5.                 Account.savings_balance) > 10000
  6.         )

“false” expression

Render a “false” constant expression, rendering as “0” on platforms that don’t have a “false” constant:

  1. from sqlalchemy.sql import expression
  2. from sqlalchemy.ext.compiler import compiles
  4. class sql_false(expression.ColumnElement):
  5.     inherit_cache = True
  7. @compiles(sql_false)
  8. def default_false(element, compiler, **kw):
  9.     return "false"
  11. @compiles(sql_false, 'mssql')
  12. @compiles(sql_false, 'mysql')
  13. @compiles(sql_false, 'oracle')
  14. def int_false(element, compiler, **kw):
  15.     return "0"

Example usage:

  1. from sqlalchemy import select, union_all
  3. exp = union_all(
  4.     select(users.c.name, sql_false().label("enrolled")),
  5.     select(customers.c.name, customers.c.enrolled)
  6. )
Object NameDescription

compiles(class, *specs)

Register a function as a compiler for a given ClauseElement type.

deregister(class)

Remove all custom compilers associated with a given ClauseElement type.

function sqlalchemy.ext.compiler.compiles(class__, *specs_)

Register a function as a compiler for a given ClauseElement type.

function sqlalchemy.ext.compiler.deregister(class_)

Remove all custom compilers associated with a given ClauseElement type.