33.7. C-Language Functions

33.7.1. Dynamic Loading

The first time a user-defined function in a particular loadable object file is called in a session, the dynamic loader loads that object file into memory so that the function can be called. The CREATE FUNCTION for a user-defined C function must therefore specify two pieces of information for the function: the name of the loadable object file, and the C name (link symbol) of the specific function to call within that object file. If the C name is not explicitly specified then it is assumed to be the same as the SQL function name.

The following algorithm is used to locate the shared object file based on the name given in the CREATE FUNCTION command:

1. If the name is an absolute path, the given file is loaded.

2. If the name starts with the string $libdir, that part is replaced by the PostgreSQL package library directory name, which is determined at build time.

3. If the name does not contain a directory part, the file is searched for in the path specified by the configuration variable dynamic_library_path.

4. Otherwise (the file was not found in the path, or it contains a non-absolute directory part), the dynamic loader will try to take the name as given, which will most likely fail. (It is unreliable to depend on the current working directory.)

If this sequence does not work, the platform-specific shared library file name extension (often .so) is appended to the given name and this sequence is tried again. If that fails as well, the load will fail.

The user ID the PostgreSQL server runs as must be able to traverse the path to the file you intend to load. Making the file or a higher-level directory not readable and/or not executable by the postgres user is a common mistake.

In any case, the file name that is given in the CREATE FUNCTION command is recorded literally in the system catalogs, so if the file needs to be loaded again the same procedure is applied.

After it is used for the first time, a dynamically loaded object file is retained in memory. Future calls in the same session to the function(s) in that file will only incur the small overhead of a symbol table lookup. If you need to force a reload of an object file, for example after recompiling it, use the LOAD command or begin a fresh session.

It is recommended to locate shared libraries either relative to $libdir or through the dynamic library path. This simplifies version upgrades if the new installation is at a different location. The actual directory that $libdir stands for can be found out with the command pg_config --pkglibdir.

Before PostgreSQL release 7.2, only exact absolute paths to object files could be specified in CREATE FUNCTION. This approach is now deprecated since it makes the function definition unnecessarily unportable. It's best to specify just the shared library name with no path nor extension, and let the search mechanism provide that information instead.

33.7.2. Base Types in C-Language Functions

To know how to write C-language functions, you need to know how PostgreSQL internally represents base data types and how they can be passed to and from functions. Internally, PostgreSQL regards a base type as a "blob of memory". The user-defined functions that you define over a type in turn define the way that PostgreSQL can operate on it. That is, PostgreSQL will only store and retrieve the data from disk and use your user-defined functions to input, process, and output the data.

Base types can have one of three internal formats:

• pass by value, fixed-length

• pass by reference, fixed-length

• pass by reference, variable-length

By-value types can only be 1, 2, or 4 bytes in length (also 8 bytes, if sizeof(Datum) is 8 on your machine). You should be careful to define your types such that they will be the same size (in bytes) on all architectures. For example, the long type is dangerous because it is 4 bytes on some machines and 8 bytes on others, whereas int type is 4 bytes on most Unix machines. A reasonable implementation of the int4 type on Unix machines might be:

/* 4-byte integer, passed by value */
typedef int int4;

On the other hand, fixed-length types of any size may be passed by-reference. For example, here is a sample implementation of a PostgreSQL type:

/* 16-byte structure, passed by reference */
typedef struct
{
    double  x, y;
} Point;

Only pointers to such types can be used when passing them in and out of PostgreSQL functions. To return a value of such a type, allocate the right amount of memory with palloc, fill in the allocated memory, and return a pointer to it. (You can also return an input value that has the same type as the return value directly by returning the pointer to the input value. Never modify the contents of a pass-by-reference input value, however.)

Finally, all variable-length types must also be passed by reference. All variable-length types must begin with a length field of exactly 4 bytes, and all data to be stored within that type must be located in the memory immediately following that length field. The length field contains the total length of the structure, that is, it includes the size of the length field itself.

As an example, we can define the type text as follows:

typedef struct {
    int4 length;
    char data[1];
} text;

Obviously, the data field declared here is not long enough to hold all possible strings. Since it's impossible to declare a variable-size structure in C, we rely on the knowledge that the C compiler won't range-check array subscripts. We just allocate the necessary amount of space and then access the array as if it were declared the right length. (This is a common trick, which you can read about in many textbooks about C.)

When manipulating variable-length types, we must be careful to allocate the correct amount of memory and set the length field correctly. For example, if we wanted to store 40 bytes in a text structure, we might use a code fragment like this:

#include "postgres.h"
...
char buffer[40]; /* our source data */
...
text *destination = (text *) palloc(VARHDRSZ + 40);
destination->length = VARHDRSZ + 40;
memcpy(destination->data, buffer, 40);
...

VARHDRSZ is the same as sizeof(int4), but it's considered good style to use the macro VARHDRSZ to refer to the size of the overhead for a variable-length type.

Table 33-1 specifies which C type corresponds to which SQL type when writing a C-language function that uses a built-in type of PostgreSQL. The "Defined In" column gives the header file that needs to be included to get the type definition. (The actual definition may be in a different file that is included by the listed file. It is recommended that users stick to the defined interface.) Note that you should always include postgres.h first in any source file, because it declares a number of things that you will need anyway.

Now that we've gone over all of the possible structures for base types, we can show some examples of real functions.

33.7.3. Calling Conventions Version 0 for C-Language Functions

We present the "old style" calling convention first --- although this approach is now deprecated, it's easier to get a handle on initially. In the version-0 method, the arguments and result of the C function are just declared in normal C style, but being careful to use the C representation of each SQL data type as shown above.

Here are some examples:

#include "postgres.h"
#include <string.h>

/* by value */
         
int
add_one(int arg)
{
    return arg + 1;
}

/* by reference, fixed length */

float8 *
add_one_float8(float8 *arg)
{
    float8    *result = (float8 *) palloc(sizeof(float8));

    *result = *arg + 1.0;
       
    return result;
}

Point *
makepoint(Point *pointx, Point *pointy)
{
    Point     *new_point = (Point *) palloc(sizeof(Point));

    new_point->x = pointx->x;
    new_point->y = pointy->y;
       
    return new_point;
}

/* by reference, variable length */

text *
copytext(text *t)
{
    /*
     * VARSIZE is the total size of the struct in bytes.
     */
    text *new_t = (text *) palloc(VARSIZE(t));
    VARATT_SIZEP(new_t) = VARSIZE(t);
    /*
     * VARDATA is a pointer to the data region of the struct.
     */
    memcpy((void *) VARDATA(new_t), /* destination */
           (void *) VARDATA(t),     /* source */
           VARSIZE(t)-VARHDRSZ);    /* how many bytes */
    return new_t;
}

text *
concat_text(text *arg1, text *arg2)
{
    int32 new_text_size = VARSIZE(arg1) + VARSIZE(arg2) - VARHDRSZ;
    text *new_text = (text *) palloc(new_text_size);

    VARATT_SIZEP(new_text) = new_text_size;
    memcpy(VARDATA(new_text), VARDATA(arg1), VARSIZE(arg1)-VARHDRSZ);
    memcpy(VARDATA(new_text) + (VARSIZE(arg1)-VARHDRSZ),
           VARDATA(arg2), VARSIZE(arg2)-VARHDRSZ);
    return new_text;
}

Supposing that the above code has been prepared in file funcs.c and compiled into a shared object, we could define the functions to PostgreSQL with commands like this:

CREATE FUNCTION add_one(integer) RETURNS integer
     AS 'DIRECTORY/funcs', 'add_one'
     LANGUAGE C STRICT;

-- note overloading of SQL function name "add_one"
CREATE FUNCTION add_one(double precision) RETURNS double precision
     AS 'DIRECTORY/funcs', 'add_one_float8'
     LANGUAGE C STRICT;

CREATE FUNCTION makepoint(point, point) RETURNS point
     AS 'DIRECTORY/funcs', 'makepoint'
     LANGUAGE C STRICT;
                         
CREATE FUNCTION copytext(text) RETURNS text
     AS 'DIRECTORY/funcs', 'copytext'
     LANGUAGE C STRICT;

CREATE FUNCTION concat_text(text, text) RETURNS text
     AS 'DIRECTORY/funcs', 'concat_text',
     LANGUAGE C STRICT;

Here, DIRECTORY stands for the directory of the shared library file (for instance the PostgreSQL tutorial directory, which contains the code for the examples used in this section). (Better style would be to use just 'funcs' in the AS clause, after having added DIRECTORY to the search path. In any case, we may omit the system-specific extension for a shared library, commonly .so or .sl.)

Notice that we have specified the functions as "strict", meaning that the system should automatically assume a null result if any input value is null. By doing this, we avoid having to check for null inputs in the function code. Without this, we'd have to check for null values explicitly, by checking for a null pointer for each pass-by-reference argument. (For pass-by-value arguments, we don't even have a way to check!)

Although this calling convention is simple to use, it is not very portable; on some architectures there are problems with passing data types that are smaller than int this way. Also, there is no simple way to return a null result, nor to cope with null arguments in any way other than making the function strict. The version-1 convention, presented next, overcomes these objections.

33.7.4. Calling Conventions Version 1 for C-Language Functions

The version-1 calling convention relies on macros to suppress most of the complexity of passing arguments and results. The C declaration of a version-1 function is always

Datum funcname(PG_FUNCTION_ARGS)

In addition, the macro call

PG_FUNCTION_INFO_V1(funcname);

must appear in the same source file. (Conventionally. it's written just before the function itself.) This macro call is not needed for internal-language functions, since PostgreSQL assumes that all internal functions use the version-1 convention. It is, however, required for dynamically-loaded functions.

In a version-1 function, each actual argument is fetched using a PG_GETARG_xxx() macro that corresponds to the argument's data type, and the result is returned using a PG_RETURN_xxx() macro for the return type. PG_GETARG_xxx() takes as its argument the number of the function argument to fetch, where the count starts at 0. PG_RETURN_xxx() takes as its argument the actual value to return.

Here we show the same functions as above, coded in version-1 style:

#include "postgres.h"
#include <string.h>
#include "fmgr.h"

/* by value */

PG_FUNCTION_INFO_V1(add_one);
         
Datum
add_one(PG_FUNCTION_ARGS)
{
    int32   arg = PG_GETARG_INT32(0);

    PG_RETURN_INT32(arg + 1);
}

/* b reference, fixed length */

PG_FUNCTION_INFO_V1(add_one_float8);

Datum
add_one_float8(PG_FUNCTION_ARGS)
{
    /* The macros for FLOAT8 hide its pass-by-reference nature. */
    float8   arg = PG_GETARG_FLOAT8(0);

    PG_RETURN_FLOAT8(arg + 1.0);
}

PG_FUNCTION_INFO_V1(makepoint);

Datum
makepoint(PG_FUNCTION_ARGS)
{
    /* Here, the pass-by-reference nature of Point is not hidden. */
    Point     *pointx = PG_GETARG_POINT_P(0);
    Point     *pointy = PG_GETARG_POINT_P(1);
    Point     *new_point = (Point *) palloc(sizeof(Point));

    new_point->x = pointx->x;
    new_point->y = pointy->y;
       
    PG_RETURN_POINT_P(new_point);
}

/* by reference, variable length */

PG_FUNCTION_INFO_V1(copytext);

Datum
copytext(PG_FUNCTION_ARGS)
{
    text     *t = PG_GETARG_TEXT_P(0);
    /*
     * VARSIZE is the total size of the struct in bytes.
     */
    text     *new_t = (text *) palloc(VARSIZE(t));
    VARATT_SIZEP(new_t) = VARSIZE(t);
    /*
     * VARDATA is a pointer to the data region of the struct.
     */
    memcpy((void *) VARDATA(new_t), /* destination */
           (void *) VARDATA(t),     /* source */
           VARSIZE(t)-VARHDRSZ);    /* how many bytes */
    PG_RETURN_TEXT_P(new_t);
}

PG_FUNCTION_INFO_V1(concat_text);

Datum
concat_text(PG_FUNCTION_ARGS)
{
    text  *arg1 = PG_GETARG_TEXT_P(0);
    text  *arg2 = PG_GETARG_TEXT_P(1);
    int32 new_text_size = VARSIZE(arg1) + VARSIZE(arg2) - VARHDRSZ;
    text *new_text = (text *) palloc(new_text_size);

    VARATT_SIZEP(new_text) = new_text_size;
    memcpy(VARDATA(new_text), VARDATA(arg1), VARSIZE(arg1)-VARHDRSZ);
    memcpy(VARDATA(new_text) + (VARSIZE(arg1)-VARHDRSZ),
           VARDATA(arg2), VARSIZE(arg2)-VARHDRSZ);
    PG_RETURN_TEXT_P(new_text);
}

The CREATE FUNCTION commands are the same as for the version-0 equivalents.

At first glance, the version-1 coding conventions may appear to be just pointless obscurantism. They do, however, offer a number of improvements, because the macros can hide unnecessary detail. An example is that in coding add_one_float8, we no longer need to be aware that float8 is a pass-by-reference type. Another example is that the GETARG macros for variable-length types allow for more efficient fetching of "toasted" (compressed or out-of-line) values.

One big improvement in version-1 functions is better handling of null inputs and results. The macro PG_ARGISNULL(n) allows a function to test whether each input is null. (Of course, doing this is only necessary in functions not declared "strict".) As with the PG_GETARG_xxx() macros, the input arguments are counted beginning at zero. Note that one should refrain from executing PG_GETARG_xxx() until one has verified that the argument isn't null. To return a null result, execute PG_RETURN_NULL(); this works in both strict and nonstrict functions.

Other options provided in the new-style interface are two variants of the PG_GETARG_xxx() macros. The first of these, PG_GETARG_xxx_COPY(), guarantees to return a copy of the specified argument that is safe for writing into. (The normal macros will sometimes return a pointer to a value that is physically stored in a table, which must not be written to. Using the PG_GETARG_xxx_COPY() macros guarantees a writable result.) The second variant consists of the PG_GETARG_xxx_SLICE() macros which take three arguments. The first is the number of the function argument (as above). The second and third are the offset and length of the segment to be returned. Offsets are counted from zero, and a negative length requests that the remainder of the value be returned. These macros provide more efficient access to parts of large values in the case where they have storage type "external". (The storage type of a column can be specified using ALTER TABLE tablename ALTER COLUMN colname SET STORAGE storagetype. storagetype is one of plain, external, extended, or main.)

Finally, the version-1 function call conventions make it possible to return set results (Section 33.7.9) and implement trigger functions (Chapter 35) and procedural-language call handlers (Chapter 47). Version-1 code is also more portable than version-0, because it does not break restrictions on function call protocol in the C standard. For more details see src/backend/utils/fmgr/README in the source distribution.

33.7.5. Writing Code

Before we turn to the more advanced topics, we should discuss some coding rules for PostgreSQL C-language functions. While it may be possible to load functions written in languages other than C into PostgreSQL, this is usually difficult (when it is possible at all) because other languages, such as C++, FORTRAN, or Pascal often do not follow the same calling convention as C. That is, other languages do not pass argument and return values between functions in the same way. For this reason, we will assume that your C-language functions are actually written in C.

The basic rules for writing and building C functions are as follows:

• Use pg_config --includedir-server to find out where the PostgreSQL server header files are installed on your system (or the system that your users will be running on). This option is new with PostgreSQL 7.2. For PostgreSQL 7.1 you should use the option --includedir. (pg_config will exit with a non-zero status if it encounters an unknown option.) For releases prior to 7.1 you will have to guess, but since that was before the current calling conventions were introduced, it is unlikely that you want to support those releases.

• When allocating memory, use the PostgreSQL functions palloc and pfree instead of the corresponding C library functions malloc and free. The memory allocated by palloc will be freed automatically at the end of each transaction, preventing memory leaks.

• Always zero the bytes of your structures using memset. Without this, it's difficult to support hash indexes or hash joins, as you must pick out only the significant bits of your data structure to compute a hash. Even if you initialize all fields of your structure, there may be alignment padding (holes in the structure) that may contain garbage values.

• Most of the internal PostgreSQL types are declared in postgres.h, while the function manager interfaces (PG_FUNCTION_ARGS, etc.) are in fmgr.h, so you will need to include at least these two files. For portability reasons it's best to include postgres.h first, before any other system or user header files. Including postgres.h will also include elog.h and palloc.h for you.

• Symbol names defined within object files must not conflict with each other or with symbols defined in the PostgreSQL server executable. You will have to rename your functions or variables if you get error messages to this effect.

• Compiling and linking your code so that it can be dynamically loaded into PostgreSQL always requires special flags. See Section 33.7.6 for a detailed explanation of how to do it for your particular operating system.

33.7.6. Compiling and Linking Dynamically-Loaded Functions

Before you are able to use your PostgreSQL extension functions written in C, they must be compiled and linked in a special way to produce a file that can be dynamically loaded by the server. To be precise, a shared library needs to be created.

For information beyond what is contained in this section you should read the documentation of your operating system, in particular the manual pages for the C compiler, cc, and the link editor, ld. In addition, the PostgreSQL source code contains several working examples in the contrib directory. If you rely on these examples you will make your modules dependent on the availability of the PostgreSQL source code, however.

Creating shared libraries is generally analogous to linking executables: first the source files are compiled into object files, then the object files are linked together. The object files need to be created as position-independent code (PIC), which conceptually means that they can be placed at an arbitrary location in memory when they are loaded by the executable. (Object files intended for executables are usually not compiled that way.) The command to link a shared library contains special flags to distinguish it from linking an executable. --- At least this is the theory. On some systems the practice is much uglier.

In the following examples we assume that your source code is in a file foo.c and we will create a shared library foo.so. The intermediate object file will be called foo.o unless otherwise noted. A shared library can contain more than one object file, but we only use one here.

The resulting shared library file can then be loaded into PostgreSQL. When specifying the file name to the CREATE FUNCTION command, one must give it the name of the shared library file, not the intermediate object file. Note that the system's standard shared-library extension (usually .so or .sl) can be omitted from the CREATE FUNCTION command, and normally should be omitted for best portability.

Refer back to Section 33.7.1 about where the server expects to find the shared library files.

33.7.7. Composite-Type Arguments in C-Language Functions

Composite types do not have a fixed layout like C structures. Instances of a composite type may contain null fields. In addition, composite types that are part of an inheritance hierarchy may have different fields than other members of the same inheritance hierarchy. Therefore, PostgreSQL provides a function interface for accessing fields of composite types from C.

Suppose we want to write a function to answer the query

SELECT name, c_overpaid(emp, 1500) AS overpaid
    FROM emp
    WHERE name = 'Bill' OR name = 'Sam';

Using call conventions version 0, we can define c_overpaid as:

#include "postgres.h"
#include "executor/executor.h"  /* for GetAttributeByName() */

bool
c_overpaid(TupleTableSlot *t, /* the current row of emp */
           int32 limit)
{
    bool isnull;
    int32 salary;

    salary = DatumGetInt32(GetAttributeByName(t, "salary", &isnull));
    if (isnull)
        return false;
    return salary > limit;
}

In version-1 coding, the above would look like this:

#include "postgres.h"
#include "executor/executor.h"  /* for GetAttributeByName() */

PG_FUNCTION_INFO_V1(c_overpaid);

Datum
c_overpaid(PG_FUNCTION_ARGS)
{
    TupleTableSlot  *t = (TupleTableSlot *) PG_GETARG_POINTER(0);
    int32            limit = PG_GETARG_INT32(1);
    bool isnull;
    int32 salary;

    salary = DatumGetInt32(GetAttributeByName(t, "salary", &isnull));
    if (isnull)
        PG_RETURN_BOOL(false);
    /* Alternatively, we might prefer to do PG_RETURN_NULL() for null salary. */

    PG_RETURN_BOOL(salary > limit);
}

GetAttributeByName is the PostgreSQL system function that returns attributes out of the specified row. It has three arguments: the argument of type TupleTableSlot* passed into the function, the name of the desired attribute, and a return parameter that tells whether the attribute is null. GetAttributeByName returns a Datum value that you can convert to the proper data type by using the appropriate DatumGetXXX() macro.

The following command declares the function c_overpaid in SQL:

CREATE FUNCTION c_overpaid(emp, integer) RETURNS boolean
    AS 'DIRECTORY/funcs', 'c_overpaid'
    LANGUAGE C;

33.7.8. Returning Rows (Composite Types) from C-Language Functions

To return a row or composite-type value from a C-language function, you can use a special API that provides macros and functions to hide most of the complexity of building composite data types. To use this API, the source file must include:

#include "funcapi.h"

The support for returning composite data types (or rows) starts with the AttInMetadata structure. This structure holds arrays of individual attribute information needed to create a row from raw C strings. The information contained in the structure is derived from a TupleDesc structure, but it is stored to avoid redundant computations on each call to a set-returning function (see next section). In the case of a function returning a set, the AttInMetadata structure should be computed once during the first call and saved for reuse in later calls. AttInMetadata also saves a pointer to the original TupleDesc.

typedef struct AttInMetadata
{
    /* full TupleDesc */
    TupleDesc       tupdesc;

    /* array of attribute type input function finfo */
    FmgrInfo       *attinfuncs;

    /* array of attribute type typelem */
    Oid            *attelems;

    /* array of attribute typmod */
    int32    	   *atttypmods;
}	AttInMetadata;

To assist you in populating this structure, several functions and a macro are available. Use

TupleDesc RelationNameGetTupleDesc(const char *relname)

to get a TupleDesc for a named relation, or

TupleDesc TypeGetTupleDesc(Oid typeoid, List *colaliases)

to get a TupleDesc based on a type OID. This can be used to get a TupleDesc for a base or composite type. Then

AttInMetadata *TupleDescGetAttInMetadata(TupleDesc tupdesc)

will return a pointer to an AttInMetadata, initialized based on the given TupleDesc. AttInMetadata can be used in conjunction with C strings to produce a properly formed row value (internally called tuple).

To return a tuple you must create a tuple slot based on the TupleDesc. You can use

TupleTableSlot *TupleDescGetSlot(TupleDesc tupdesc)

to initialize this tuple slot, or obtain one through other (user provided) means. The tuple slot is needed to create a Datum for return by the function. The same slot can (and should) be reused on each call.

After constructing an AttInMetadata structure,

HeapTuple BuildTupleFromCStrings(AttInMetadata *attinmeta, char **values)

can be used to build a HeapTuple given user data in C string form. values is an array of C strings, one for each attribute of the return row. Each C string should be in the form expected by the input function of the attribute data type. In order to return a null value for one of the attributes, the corresponding pointer in the values array should be set to NULL. This function will need to be called again for each row you return.

Building a tuple via TupleDescGetAttInMetadata and BuildTupleFromCStrings is only convenient if your function naturally computes the values to be returned as text strings. If your code naturally computes the values as a set of Datum values, you should instead use the underlying function heap_formtuple to convert the Datum values directly into a tuple. You will still need the TupleDesc and a TupleTableSlot, but not AttInMetadata.

Once you have built a tuple to return from your function, it must be converted into a Datum. Use

TupleGetDatum(TupleTableSlot *slot, HeapTuple tuple)

to get a Datum given a tuple and a slot. This Datum can be returned directly if you intend to return just a single row, or it can be used as the current return value in a set-returning function.

An example appears in the next section.

33.7.9. Returning Sets from C-Language Functions

There is also a special API that provides support for returning sets (multiple rows) from a C-language function. A set-returning function must follow the version-1 calling conventions. Also, source files must include funcapi.h, as above.

A set-returning function (SRF) is called once for each item it returns. The SRF must therefore save enough state to remember what it was doing and return the next item on each call. The structure FuncCallContext is provided to help control this process. Within a function, fcinfo->flinfo->fn_extra is used to hold a pointer to FuncCallContext across calls.

typedef struct
{
    /*
     * Number of times we've been called before
     * 
     * call_cntr is initialized to 0 for you by SRF_FIRSTCALL_INIT(), and
     * incremented for you every time SRF_RETURN_NEXT() is called.
     */
    uint32 call_cntr;

    /*
     * OPTIONAL maximum number of calls
     *
     * max_calls is here for convenience only and setting it is optional.
     * If not set, you must provide alternative means to know when the
     * function is done.
     */
    uint32 max_calls;

    /*
     * OPTIONAL pointer to result slot
     * 
     * slot is for use when returning tuples (i.e., composite data types)
     * and is not needed when returning base data types.
     */
    TupleTableSlot *slot;

    /*
     * OPTIONAL pointer to miscellaneous user-provided context information
     * 
     * user_fctx is for use as a pointer to your own data to retain
     * arbitrary context information between calls of your function.
     */
    void *user_fctx;

    /*
     * OPTIONAL pointer to struct containing attribute type input metadata
     * 
     * attinmeta is for use when returning tuples (i.e., composite data types)
     * and is not needed when returning base data types. It
     * is only needed if you intend to use BuildTupleFromCStrings() to create
     * the return tuple.
     */
    AttInMetadata *attinmeta;

    /*
     * memory context used for structures that must live for multiple calls
     *
     * multi_call_memory_ctx is set by SRF_FIRSTCALL_INIT() for you, and used
     * by SRF_RETURN_DONE() for cleanup. It is the most appropriate memory
     * context for any memory that is to be reused across multiple calls
     * of the SRF.
     */
    MemoryContext multi_call_memory_ctx;
} FuncCallContext;

An SRF uses several functions and macros that automatically manipulate the FuncCallContext structure (and expect to find it via fn_extra). Use

SRF_IS_FIRSTCALL()

to determine if your function is being called for the first or a subsequent time. On the first call (only) use

SRF_FIRSTCALL_INIT()

to initialize the FuncCallContext. On every function call, including the first, use

SRF_PERCALL_SETUP()

to properly set up for using the FuncCallContext and clearing any previously returned data left over from the previous pass.

If your function has data to return, use

SRF_RETURN_NEXT(funcctx, result)

to return it to the caller. (result must be of type Datum, either a single value or a tuple prepared as described above.) Finally, when your function is finished returning data, use

SRF_RETURN_DONE(funcctx)

to clean up and end the SRF.

The memory context that is current when the SRF is called is a transient context that will be cleared between calls. This means that you do not need to call pfree on everything you allocated using palloc; it will go away anyway. However, if you want to allocate any data structures to live across calls, you need to put them somewhere else. The memory context referenced by multi_call_memory_ctx is a suitable location for any data that needs to survive until the SRF is finished running. In most cases, this means that you should switch into multi_call_memory_ctx while doing the first-call setup.

A complete pseudo-code example looks like the following:

Datum
my_set_returning_function(PG_FUNCTION_ARGS)
{
    FuncCallContext  *funcctx;
    Datum             result;
    MemoryContext     oldcontext;
    further declarations as needed

    if (SRF_IS_FIRSTCALL())
    {
        funcctx = SRF_FIRSTCALL_INIT();
        oldcontext = MemoryContextSwitchTo(funcctx->multi_call_memory_ctx);
        /* One-time setup code appears here: */
        user code
        if returning composite
            build TupleDesc, and perhaps AttInMetadata
            obtain slot
            funcctx->slot = slot;
        endif returning composite
        user code
        MemoryContextSwitchTo(oldcontext);
    }

    /* Each-time setup code appears here: */
    user code
    funcctx = SRF_PERCALL_SETUP();
    user code

    /* this is just one way we might test whether we are done: */
    if (funcctx->call_cntr < funcctx->max_calls)
    {
        /* Here we want to return another item: */
        user code
        obtain result Datum
        SRF_RETURN_NEXT(funcctx, result);
    }
    else
    {
        /* Here we are done returning items and just need to clean up: */
        user code
        SRF_RETURN_DONE(funcctx);
    }
}

A complete example of a simple SRF returning a composite type looks like:

PG_FUNCTION_INFO_V1(testpassbyval);

Datum
testpassbyval(PG_FUNCTION_ARGS)
{
    FuncCallContext     *funcctx;
    int                  call_cntr;
    int                  max_calls;
    TupleDesc            tupdesc;
    TupleTableSlot      *slot;
    AttInMetadata       *attinmeta;

     /* stuff done only on the first call of the function */
     if (SRF_IS_FIRSTCALL())
     {
        MemoryContext	oldcontext;

        /* create a function context for cross-call persistence */
        funcctx = SRF_FIRSTCALL_INIT();

        /* switch to memory context appropriate for multiple function calls */
        oldcontext = MemoryContextSwitchTo(funcctx->multi_call_memory_ctx);

        /* total number of tuples to be returned */
        funcctx->max_calls = PG_GETARG_UINT32(0);

        /* Build a tuple description for a __testpassbyval tuple */
        tupdesc = RelationNameGetTupleDesc("__testpassbyval");

        /* allocate a slot for a tuple with this tupdesc */
        slot = TupleDescGetSlot(tupdesc);

        /* assign slot to function context */
        funcctx->slot = slot;

        /*
         * generate attribute metadata needed later to produce tuples from raw
         * C strings
         */
        attinmeta = TupleDescGetAttInMetadata(tupdesc);
        funcctx->attinmeta = attinmeta;

        MemoryContextSwitchTo(oldcontext);
    }

    /* stuff done on every call of the function */
    funcctx = SRF_PERCALL_SETUP();

    call_cntr = funcctx->call_cntr;
    max_calls = funcctx->max_calls;
    slot = funcctx->slot;
    attinmeta = funcctx->attinmeta;
 
    if (call_cntr < max_calls)    /* do when there is more left to send */
    {
        char       **values;
        HeapTuple    tuple;
        Datum        result;

        /*
         * Prepare a values array for storage in our slot.
         * This should be an array of C strings which will
         * be processed later by the type input functions.
         */
        values = (char **) palloc(3 * sizeof(char *));
        values[0] = (char *) palloc(16 * sizeof(char));
        values[1] = (char *) palloc(16 * sizeof(char));
        values[2] = (char *) palloc(16 * sizeof(char));

        snprintf(values[0], 16, "%d", 1 * PG_GETARG_INT32(1));
        snprintf(values[1], 16, "%d", 2 * PG_GETARG_INT32(1));
        snprintf(values[2], 16, "%d", 3 * PG_GETARG_INT32(1));

        /* build a tuple */
        tuple = BuildTupleFromCStrings(attinmeta, values);

        /* make the tuple into a datum */
        result = TupleGetDatum(slot, tuple);

        /* clean up (this is not really necessary) */
        pfree(values[0]);
        pfree(values[1]);
        pfree(values[2]);
        pfree(values);

        SRF_RETURN_NEXT(funcctx, result);
    }
    else    /* do when there is no more left */
    {
        SRF_RETURN_DONE(funcctx);
    }
}

The SQL code to declare this function is:

CREATE TYPE __testpassbyval AS (f1 integer, f2 integer, f3 integer);

CREATE OR REPLACE FUNCTION testpassbyval(integer, integer) RETURNS SETOF __testpassbyval
    AS 'filename', 'testpassbyval'
    LANGUAGE C IMMUTABLE STRICT;

The directory contrib/tablefunc in the source distribution contains more examples of set-returning functions.

33.7.10. Polymorphic Arguments and Return Types

C-language functions may be declared to accept and return the polymorphic types anyelement and anyarray. See Section 33.2.5 for a more detailed explanation of polymorphic functions. When function arguments or return types are defined as polymorphic types, the function author cannot know in advance what data type it will be called with, or need to return. There are two routines provided in fmgr.h to allow a version-1 C function to discover the actual data types of its arguments and the type it is expected to return. The routines are called get_fn_expr_rettype(FmgrInfo *flinfo) and get_fn_expr_argtype(FmgrInfo *flinfo, int argnum). They return the result or argument type OID, or InvalidOid if the information is not available. The structure flinfo is normally accessed as fcinfo->flinfo. The parameter argnum is zero based.

For example, suppose we want to write a function to accept a single element of any type, and return a one-dimensional array of that type:

PG_FUNCTION_INFO_V1(make_array);
Datum
make_array(PG_FUNCTION_ARGS)
{
    ArrayType  *result;
    Oid         element_type = get_fn_expr_argtype(fcinfo->flinfo, 0);
    Datum       element;
    int16       typlen;
    bool        typbyval;
    char        typalign;
    int         ndims;
    int         dims[MAXDIM];
    int         lbs[MAXDIM];

    if (!OidIsValid(element_type))
        elog(ERROR, "could not determine data type of input");

    /* get the provided element */
    element = PG_GETARG_DATUM(0);

    /* we have one dimension */
    ndims = 1;
    /* and one element */
    dims[0] = 1;
    /* and lower bound is 1 */
    lbs[0] = 1;

    /* get required info about the element type */
    get_typlenbyvalalign(element_type, &typlen, &typbyval, &typalign);

    /* now build the array */
    result = construct_md_array(&element, ndims, dims, lbs,
                                element_type, typlen, typbyval, typalign);

    PG_RETURN_ARRAYTYPE_P(result);
}

The following command declares the function make_array in SQL:

CREATE FUNCTION make_array(anyelement) RETURNS anyarray
    AS 'DIRECTORY/funcs', 'make_array'
    LANGUAGE C STRICT;

Note the use of STRICT; this is essential since the code is not bothering to test for a null input.