1CREATE FUNCTION(7) PostgreSQL 9.2.24 Documentation CREATE FUNCTION(7)
2
6 CREATE_FUNCTION - define a new function
7
9 CREATE [ OR REPLACE ] FUNCTION
10 name ( [ [ argmode ] [ argname ] argtype [ { DEFAULT | = } default_expr ] [, ...] ] )
11 [ RETURNS rettype
12 | RETURNS TABLE ( column_name column_type [, ...] ) ]
13 { LANGUAGE lang_name
14 | WINDOW
15 | IMMUTABLE | STABLE | VOLATILE | [ NOT ] LEAKPROOF
16 | CALLED ON NULL INPUT | RETURNS NULL ON NULL INPUT | STRICT
17 | [ EXTERNAL ] SECURITY INVOKER | [ EXTERNAL ] SECURITY DEFINER
18 | COST execution_cost
19 | ROWS result_rows
20 | SET configuration_parameter { TO value | = value | FROM CURRENT }
21 | AS 'definition'
22 | AS 'obj_file', 'link_symbol'
23 } ...
24 [ WITH ( attribute [, ...] ) ]
25
27 CREATE FUNCTION defines a new function. CREATE OR REPLACE FUNCTION
28 will either create a new function, or replace an existing definition.
29 To be able to define a function, the user must have the USAGE privilege
30 on the language.
31 If a schema name is included, then the function is created in the
33 specified schema. Otherwise it is created in the current schema. The
34 name of the new function must not match any existing function with the
35 same input argument types in the same schema. However, functions of
36 different argument types can share a name (this is called overloading).
37 To replace the current definition of an existing function, use CREATE
39 OR REPLACE FUNCTION. It is not possible to change the name or argument
40 types of a function this way (if you tried, you would actually be
41 creating a new, distinct function). Also, CREATE OR REPLACE FUNCTION
42 will not let you change the return type of an existing function. To do
43 that, you must drop and recreate the function. (When using OUT
44 parameters, that means you cannot change the types of any OUT
45 parameters except by dropping the function.)
46 When CREATE OR REPLACE FUNCTION is used to replace an existing
48 function, the ownership and permissions of the function do not change.
49 All other function properties are assigned the values specified or
50 implied in the command. You must own the function to replace it (this
51 includes being a member of the owning role).
52 If you drop and then recreate a function, the new function is not the
54 same entity as the old; you will have to drop existing rules, views,
55 triggers, etc. that refer to the old function. Use CREATE OR REPLACE
56 FUNCTION to change a function definition without breaking objects that
57 refer to the function. Also, ALTER FUNCTION can be used to change most
58 of the auxiliary properties of an existing function.
59 The user that creates the function becomes the owner of the function.
61 To be able to create a function, you must have USAGE privilege on the
63 argument types and the return type.
64
66 name
67 The name (optionally schema-qualified) of the function to create.
68 argmode
70 The mode of an argument: IN, OUT, INOUT, or VARIADIC. If omitted,
71 the default is IN. Only OUT arguments can follow a VARIADIC one.
72 Also, OUT and INOUT arguments cannot be used together with the
73 RETURNS TABLE notation.
74 argname
76 The name of an argument. Some languages (including SQL and
77 PL/pgSQL) let you use the name in the function body. For other
78 languages the name of an input argument is just extra
79 documentation, so far as the function itself is concerned; but you
80 can use input argument names when calling a function to improve
81 readability (see Section 4.3, “Calling Functions”, in the
82 documentation). In any case, the name of an output argument is
83 significant, because it defines the column name in the result row
84 type. (If you omit the name for an output argument, the system will
85 choose a default column name.)
86 argtype
88 The data type(s) of the function's arguments (optionally
89 schema-qualified), if any. The argument types can be base,
90 composite, or domain types, or can reference the type of a table
91 column.
92 Depending on the implementation language it might also be allowed
94 to specify “pseudotypes” such as cstring. Pseudotypes indicate that
95 the actual argument type is either incompletely specified, or
96 outside the set of ordinary SQL data types.
97 The type of a column is referenced by writing
99 table_name.column_name%TYPE. Using this feature can sometimes help
100 make a function independent of changes to the definition of a
101 table.
102 default_expr
104 An expression to be used as default value if the parameter is not
105 specified. The expression has to be coercible to the argument type
106 of the parameter. Only input (including INOUT) parameters can have
107 a default value. All input parameters following a parameter with a
108 default value must have default values as well.
109 rettype
111 The return data type (optionally schema-qualified). The return type
112 can be a base, composite, or domain type, or can reference the type
113 of a table column. Depending on the implementation language it
114 might also be allowed to specify “pseudotypes” such as cstring. If
115 the function is not supposed to return a value, specify void as the
116 return type.
117 When there are OUT or INOUT parameters, the RETURNS clause can be
119 omitted. If present, it must agree with the result type implied by
120 the output parameters: RECORD if there are multiple output
121 parameters, or the same type as the single output parameter.
122 The SETOF modifier indicates that the function will return a set of
124 items, rather than a single item.
125 The type of a column is referenced by writing
127 table_name.column_name%TYPE.
128 column_name
130 The name of an output column in the RETURNS TABLE syntax. This is
131 effectively another way of declaring a named OUT parameter, except
132 that RETURNS TABLE also implies RETURNS SETOF.
133 column_type
135 The data type of an output column in the RETURNS TABLE syntax.
136 lang_name
138 The name of the language that the function is implemented in. Can
139 be SQL, C, internal, or the name of a user-defined procedural
140 language. For backward compatibility, the name can be enclosed by
141 single quotes.
142 WINDOW
144 WINDOW indicates that the function is a window function rather than
145 a plain function. This is currently only useful for functions
146 written in C. The WINDOW attribute cannot be changed when replacing
147 an existing function definition.
148 IMMUTABLE, STABLE, VOLATILE
150 These attributes inform the query optimizer about the behavior of
151 the function. At most one choice can be specified. If none of these
152 appear, VOLATILE is the default assumption.
153 IMMUTABLE indicates that the function cannot modify the database
155 and always returns the same result when given the same argument
156 values; that is, it does not do database lookups or otherwise use
157 information not directly present in its argument list. If this
158 option is given, any call of the function with all-constant
159 arguments can be immediately replaced with the function value.
160 STABLE indicates that the function cannot modify the database, and
162 that within a single table scan it will consistently return the
163 same result for the same argument values, but that its result could
164 change across SQL statements. This is the appropriate selection for
165 functions whose results depend on database lookups, parameter
166 variables (such as the current time zone), etc. (It is
167 inappropriate for AFTER triggers that wish to query rows modified
168 by the current command.) Also note that the current_timestamp
169 family of functions qualify as stable, since their values do not
170 change within a transaction.
171 VOLATILE indicates that the function value can change even within a
173 single table scan, so no optimizations can be made. Relatively few
174 database functions are volatile in this sense; some examples are
175 random(), currval(), timeofday(). But note that any function that
176 has side-effects must be classified volatile, even if its result is
177 quite predictable, to prevent calls from being optimized away; an
178 example is setval().
179 For additional details see Section 35.6, “Function Volatility
181 Categories”, in the documentation.
182 LEAKPROOF
184 LEAKPROOF indicates that the function has no side effects. It
185 reveals no information about its arguments other than by its return
186 value. For example, a function which throws an error message for
187 some argument values but not others, or which includes the argument
188 values in any error message, is not leakproof. The query planner
189 may push leakproof functions (but not others) into views created
190 with the security_barrier option. See CREATE VIEW (CREATE_VIEW(7))
191 and Section 37.4, “Rules and Privileges”, in the documentation.
192 This option can only be set by the superuser.
193 CALLED ON NULL INPUT, RETURNS NULL ON NULL INPUT, STRICT
195 CALLED ON NULL INPUT (the default) indicates that the function will
196 be called normally when some of its arguments are null. It is then
197 the function author's responsibility to check for null values if
198 necessary and respond appropriately.
199 RETURNS NULL ON NULL INPUT or STRICT indicates that the function
201 always returns null whenever any of its arguments are null. If this
202 parameter is specified, the function is not executed when there are
203 null arguments; instead a null result is assumed automatically.
204 [EXTERNAL] SECURITY INVOKER, [EXTERNAL] SECURITY DEFINER
206 SECURITY INVOKER indicates that the function is to be executed with
207 the privileges of the user that calls it. That is the default.
208 SECURITY DEFINER specifies that the function is to be executed with
209 the privileges of the user that created it.
210 The key word EXTERNAL is allowed for SQL conformance, but it is
212 optional since, unlike in SQL, this feature applies to all
213 functions not only external ones.
214 execution_cost
216 A positive number giving the estimated execution cost for the
217 function, in units of cpu_operator_cost. If the function returns a
218 set, this is the cost per returned row. If the cost is not
219 specified, 1 unit is assumed for C-language and internal functions,
220 and 100 units for functions in all other languages. Larger values
221 cause the planner to try to avoid evaluating the function more
222 often than necessary.
223 result_rows
225 A positive number giving the estimated number of rows that the
226 planner should expect the function to return. This is only allowed
227 when the function is declared to return a set. The default
228 assumption is 1000 rows.
229 configuration_parameter, value
231 The SET clause causes the specified configuration parameter to be
232 set to the specified value when the function is entered, and then
233 restored to its prior value when the function exits. SET FROM
234 CURRENT saves the value of the parameter that is current when
235 CREATE FUNCTION is executed as the value to be applied when the
236 function is entered.
237 If a SET clause is attached to a function, then the effects of a
239 SET LOCAL command executed inside the function for the same
240 variable are restricted to the function: the configuration
241 parameter's prior value is still restored at function exit.
242 However, an ordinary SET command (without LOCAL) overrides the SET
243 clause, much as it would do for a previous SET LOCAL command: the
244 effects of such a command will persist after function exit, unless
245 the current transaction is rolled back.
246 See SET(7) and Chapter 18, Server Configuration, in the
248 documentation for more information about allowed parameter names
249 and values.
250 definition
252 A string constant defining the function; the meaning depends on the
253 language. It can be an internal function name, the path to an
254 object file, an SQL command, or text in a procedural language.
255 It is often helpful to use dollar quoting (see Section 4.1.2.4,
257 “Dollar-quoted String Constants”, in the documentation) to write
258 the function definition string, rather than the normal single quote
259 syntax. Without dollar quoting, any single quotes or backslashes in
260 the function definition must be escaped by doubling them.
261 obj_file, link_symbol
263 This form of the AS clause is used for dynamically loadable C
264 language functions when the function name in the C language source
265 code is not the same as the name of the SQL function. The string
266 obj_file is the name of the file containing the dynamically
267 loadable object, and link_symbol is the function's link symbol,
268 that is, the name of the function in the C language source code. If
269 the link symbol is omitted, it is assumed to be the same as the
270 name of the SQL function being defined.
271 When repeated CREATE FUNCTION calls refer to the same object file,
273 the file is only loaded once per session. To unload and reload the
274 file (perhaps during development), start a new session.
275 attribute
277 The historical way to specify optional pieces of information about
278 the function. The following attributes can appear here:
279 isStrict
281 Equivalent to STRICT or RETURNS NULL ON NULL INPUT.
282 isCachable
284 isCachable is an obsolete equivalent of IMMUTABLE; it's still
285 accepted for backwards-compatibility reasons.
286 Attribute names are not case-sensitive.
288 Refer to Section 35.3, “User-defined Functions”, in the documentation
290 for further information on writing functions.
291
293 PostgreSQL allows function overloading; that is, the same name can be
294 used for several different functions so long as they have distinct
295 input argument types. However, the C names of all functions must be
296 different, so you must give overloaded C functions different C names
297 (for example, use the argument types as part of the C names).
298 Two functions are considered the same if they have the same names and
300 input argument types, ignoring any OUT parameters. Thus for example
301 these declarations conflict:
302 CREATE FUNCTION foo(int) ...
304 CREATE FUNCTION foo(int, out text) ...
305 Functions that have different argument type lists will not be
307 considered to conflict at creation time, but if defaults are provided
308 they might conflict in use. For example, consider
309 CREATE FUNCTION foo(int) ...
311 CREATE FUNCTION foo(int, int default 42) ...
312 A call foo(10) will fail due to the ambiguity about which function
314 should be called.
315
317 The full SQL type syntax is allowed for declaring a function's
318 arguments and return value. However, parenthesized type modifiers
319 (e.g., the precision field for type numeric) are discarded by CREATE
320 FUNCTION. Thus for example CREATE FUNCTION foo (varchar(10)) ... is
321 exactly the same as CREATE FUNCTION foo (varchar) ....
322 When replacing an existing function with CREATE OR REPLACE FUNCTION,
324 there are restrictions on changing parameter names. You cannot change
325 the name already assigned to any input parameter (although you can add
326 names to parameters that had none before). If there is more than one
327 output parameter, you cannot change the names of the output parameters,
328 because that would change the column names of the anonymous composite
329 type that describes the function's result. These restrictions are made
330 to ensure that existing calls of the function do not stop working when
331 it is replaced.
332 If a function is declared STRICT with a VARIADIC argument, the
334 strictness check tests that the variadic array as a whole is non-null.
335 The function will still be called if the array has null elements.
336
338 Here are some trivial examples to help you get started. For more
339 information and examples, see Section 35.3, “User-defined Functions”,
340 in the documentation.
341 CREATE FUNCTION add(integer, integer) RETURNS integer
343 AS 'select $1 + $2;'
344 LANGUAGE SQL
345 IMMUTABLE
346 RETURNS NULL ON NULL INPUT;
347 Increment an integer, making use of an argument name, in PL/pgSQL:
349 CREATE OR REPLACE FUNCTION increment(i integer) RETURNS integer AS $$
351 BEGIN
352 RETURN i + 1;
353 END;
354 $$ LANGUAGE plpgsql;
355 Return a record containing multiple output parameters:
357 CREATE FUNCTION dup(in int, out f1 int, out f2 text)
359 AS $$ SELECT $1, CAST($1 AS text) || ' is text' $$
360 LANGUAGE SQL;
361 SELECT * FROM dup(42);
363 You can do the same thing more verbosely with an explicitly named
365 composite type:
366 CREATE TYPE dup_result AS (f1 int, f2 text);
368 CREATE FUNCTION dup(int) RETURNS dup_result
370 AS $$ SELECT $1, CAST($1 AS text) || ' is text' $$
371 LANGUAGE SQL;
372 SELECT * FROM dup(42);
374 Another way to return multiple columns is to use a TABLE function:
376 CREATE FUNCTION dup(int) RETURNS TABLE(f1 int, f2 text)
378 AS $$ SELECT $1, CAST($1 AS text) || ' is text' $$
379 LANGUAGE SQL;
380 SELECT * FROM dup(42);
382 However, a TABLE function is different from the preceding examples,
384 because it actually returns a set of records, not just one record.
385
387 Because a SECURITY DEFINER function is executed with the privileges of
388 the user that created it, care is needed to ensure that the function
389 cannot be misused. For security, search_path should be set to exclude
390 any schemas writable by untrusted users. This prevents malicious users
391 from creating objects that mask objects used by the function.
392 Particularly important in this regard is the temporary-table schema,
393 which is searched first by default, and is normally writable by anyone.
394 A secure arrangement can be had by forcing the temporary schema to be
395 searched last. To do this, write pg_temp as the last entry in
396 search_path. This function illustrates safe usage:
397 CREATE FUNCTION check_password(uname TEXT, pass TEXT)
399 RETURNS BOOLEAN AS $$
400 DECLARE passed BOOLEAN;
401 BEGIN
402 SELECT (pwd = $2) INTO passed
403 FROM pwds
404 WHERE username = $1;
405 RETURN passed;
407 END;
408 $$ LANGUAGE plpgsql
409 SECURITY DEFINER
410 -- Set a secure search_path: trusted schema(s), then 'pg_temp'.
411 SET search_path = admin, pg_temp;
412 Before PostgreSQL version 8.3, the SET option was not available, and so
414 older functions may contain rather complicated logic to save, set, and
415 restore search_path. The SET option is far easier to use for this
416 purpose.
417 Another point to keep in mind is that by default, execute privilege is
419 granted to PUBLIC for newly created functions (see GRANT(7) for more
420 information). Frequently you will wish to restrict use of a security
421 definer function to only some users. To do that, you must revoke the
422 default PUBLIC privileges and then grant execute privilege selectively.
423 To avoid having a window where the new function is accessible to all,
424 create it and set the privileges within a single transaction. For
425 example:
426 BEGIN;
428 CREATE FUNCTION check_password(uname TEXT, pass TEXT) ... SECURITY DEFINER;
429 REVOKE ALL ON FUNCTION check_password(uname TEXT, pass TEXT) FROM PUBLIC;
430 GRANT EXECUTE ON FUNCTION check_password(uname TEXT, pass TEXT) TO admins;
431 COMMIT;
432
434 A CREATE FUNCTION command is defined in SQL:1999 and later. The
435 PostgreSQL version is similar but not fully compatible. The attributes
436 are not portable, neither are the different available languages.
437 For compatibility with some other database systems, argmode can be
439 written either before or after argname. But only the first way is
440 standard-compliant.
441 The SQL standard does not specify parameter defaults. The syntax with
443 the DEFAULT key word is from Oracle, and it is somewhat in the spirit
444 of the standard: SQL/PSM uses it for variable default values. The
445 syntax with = is used in T-SQL and Firebird.
446
448 ALTER FUNCTION (ALTER_FUNCTION(7)), DROP FUNCTION (DROP_FUNCTION(7)),
449 GRANT(7), LOAD(7), REVOKE(7), createlang(1)
450PostgreSQL 9.2.24 2017-11-06 CREATE FUNCTION(7)