2 36.6. Function Overloading #
4 More than one function can be defined with the same SQL name, so long
5 as the arguments they take are different. In other words, function
6 names can be overloaded. Whether or not you use it, this capability
7 entails security precautions when calling functions in databases where
8 some users mistrust other users; see Section 10.3. When a query is
9 executed, the server will determine which function to call from the
10 data types and the number of the provided arguments. Overloading can
11 also be used to simulate functions with a variable number of arguments,
12 up to a finite maximum number.
14 When creating a family of overloaded functions, one should be careful
15 not to create ambiguities. For instance, given the functions:
16 CREATE FUNCTION test(int, real) RETURNS ...
17 CREATE FUNCTION test(smallint, double precision) RETURNS ...
19 it is not immediately clear which function would be called with some
20 trivial input like test(1, 1.5). The currently implemented resolution
21 rules are described in Chapter 10, but it is unwise to design a system
22 that subtly relies on this behavior.
24 A function that takes a single argument of a composite type should
25 generally not have the same name as any attribute (field) of that type.
26 Recall that attribute(table) is considered equivalent to
27 table.attribute. In the case that there is an ambiguity between a
28 function on a composite type and an attribute of the composite type,
29 the attribute will always be used. It is possible to override that
30 choice by schema-qualifying the function name (that is,
31 schema.func(table) ) but it's better to avoid the problem by not
32 choosing conflicting names.
34 Another possible conflict is between variadic and non-variadic
35 functions. For instance, it is possible to create both foo(numeric) and
36 foo(VARIADIC numeric[]). In this case it is unclear which one should be
37 matched to a call providing a single numeric argument, such as
38 foo(10.1). The rule is that the function appearing earlier in the
39 search path is used, or if the two functions are in the same schema,
40 the non-variadic one is preferred.
42 When overloading C-language functions, there is an additional
43 constraint: The C name of each function in the family of overloaded
44 functions must be different from the C names of all other functions,
45 either internal or dynamically loaded. If this rule is violated, the
46 behavior is not portable. You might get a run-time linker error, or one
47 of the functions will get called (usually the internal one). The
48 alternative form of the AS clause for the SQL CREATE FUNCTION command
49 decouples the SQL function name from the function name in the C source
51 CREATE FUNCTION test(int) RETURNS int
52 AS 'filename', 'test_1arg'
54 CREATE FUNCTION test(int, int) RETURNS int
55 AS 'filename', 'test_2arg'
58 The names of the C functions here reflect one of many possible