postgresql/doc/src/sgml/libpq.sgml

 <chapter id="libpq-chapter">
  <title id="libpq">libpq</title>

  <para>
   <filename>libpq</filename> is the <acronym>C</acronym>
   application programmer's interface to
   <productname>PostgreSQL</productname>.  <filename>libpq</filename> is a set
   of library routines that allow client programs to pass queries to the
   <productname>Postgres</productname> backend server and to receive the
   results of these queries.  <filename>libpq</filename> is also the
   underlying engine for several other <productname>PostgreSQL</productname>
   application interfaces, including <filename>libpq++</filename> (C++),
   <filename>libpgtcl</filename> (Tcl), <productname>Perl</productname>, and
   <filename>ecpg</filename>.  So some aspects of libpq's behavior will be
   important to you if you use one of those packages.
  </para>

  <para>
   Three short programs are included at the end of this section to show how
   to write programs that use <filename>libpq</filename>.  There are several
   complete examples of <filename>libpq</filename> applications in the
   following directories:

   <programlisting>
../src/test/regress
../src/test/examples
../src/bin/psql
   </programlisting>
  </para>

  <para>
   Frontend programs which use <filename>libpq</filename> must include the
   header file <filename>libpq-fe.h</filename> and must link with the
   <filename>libpq</filename> library.
  </para>

  <sect1>
   <title>Database Connection Functions</title>

   <para>
    The following routines deal with making a connection to
    a <productname>Postgres</productname> backend server.  The application
    program can have several backend connections open at one time.
    (One reason to do that is to access more than one database.)
    Each connection is represented by a PGconn object which is obtained
    from PQconnectdb() or PQsetdbLogin().  Note that these functions
    will always return a non-null object pointer, unless perhaps
    there is too little memory even to allocate the PGconn object.
    The  PQstatus function should be called
    to check whether  a  connection  was  successfully made
    before queries are sent via the connection object.

    <itemizedlist>
     <listitem>
      <para>
       <function>PQconnectdb</function> 
       Makes a new connection to the database server.
       <synopsis>
PGconn *PQconnectdb(const char *conninfo)
       </synopsis>

   This routine opens a new database connection using the parameters taken
   from the string <literal>conninfo</literal>.  Unlike PQsetdbLogin() below,
   the parameter set can be extended without changing the function signature,
   so use either of this routine or the non-blocking analogues PQconnectStart
   / PQconnectPoll is prefered for application programming.  The passed string
   can be empty to use all default parameters, or it can contain one or more
   parameter settings separated by whitespace.
   </para>

   <para>
   Each parameter setting is in the form <literal>keyword = value</literal>.
   (To write a null value or a value containing
   spaces, surround it with single quotes, e.g.,
   <literal>keyword = 'a value'</literal>.
   Single quotes within the value must be written as <literal>\'</literal>.
   Spaces around the equal sign are optional.)  The currently recognized
   parameter keywords are:

   <variablelist>
    <varlistentry>
     <term><literal>host</literal></term>
     <listitem>
     <para>
      Name of host to connect to. If a non-zero-length string is
	   specified, TCP/IP
      communication is used.  Using this parameter causes a hostname look-up.
      See hostaddr.
     </para>
     </listitem>
    </varlistentry>

    <varlistentry>
     <term><literal>hostaddr</literal></term>
     <listitem>
     <para>
      IP address of host to connect to. This should be in standard
      numbers-and-dots form, as used by the BSD functions inet_aton et al. If
      a non-zero-length string is specified, TCP/IP communication is used.
     </para>
     <para>
      Using hostaddr instead of host allows the application to avoid a host
      name look-up, which may be important in applications with time
      constraints. However, Kerberos authentication requires the host
      name. The following therefore applies. If host is specified without
      hostaddr, a hostname look-up is forced. If hostaddr is specified without
      host, the value for hostaddr gives the remote address; if Kerberos is
      used, this causes a reverse name query. If both host and hostaddr are
      specified, the value for hostaddr gives the remote address; the value
      for host is ignored, unless Kerberos is used, in which case that value
      is used for Kerberos authentication. Note that authentication is likely
      to fail if libpq is passed a host name which is not the name of the
      machine at hostaddr.
     </para>
     <para>
      Without both a host name and host address, libpq will connect using a
      local Unix domain socket.
     </para>
     </listitem>
    </varlistentry>

    <varlistentry>
     <term><literal>port</literal></term>
     <listitem>
     <para>
      Port number to connect to at the server host,
      or socket filename extension for Unix-domain connections.
     </para>
     </listitem>
    </varlistentry>

    <varlistentry>
     <term><literal>dbname</literal></term>
     <listitem>
     <para>
      The database name.
     </para>
     </listitem>
    </varlistentry>

    <varlistentry>
     <term><literal>user</literal></term> 
     <listitem>
     <para>
      User name to connect as.
     </para>
     </listitem>
    </varlistentry>

    <varlistentry>
     <term><literal>password</literal></term>
     <listitem>
     <para>
      Password to be used if the server demands password authentication.
     </para>
     </listitem>
    </varlistentry>

    <varlistentry>
     <term><literal>options</literal></term>
     <listitem>
      <para>
       Trace/debug options to be sent to the server.
      </para>
     </listitem>
    </varlistentry>

    <varlistentry>
     <term><literal>tty</literal></term>
     <listitem>
     <para>
      A file or tty for optional debug output from the backend.
     </para>
     </listitem>
    </varlistentry>
   </variablelist>

   If  any  parameter is unspecified, then the corresponding
   environment variable (see "Environment Variables" section)
   is checked. If the  environment  variable is not set either,
   then hardwired defaults are used.
   The return value is a pointer to an abstract struct
   representing the connection to the backend.
   </para>
  </listitem>

  <listitem>
   <para>
   <function>PQsetdbLogin</function>
       Makes a new connection to the database server.
<synopsis>
PGconn *PQsetdbLogin(const char *pghost,
                     const char *pgport,
                     const char *pgoptions,
                     const char *pgtty,
                     const char *dbName,
                     const char *login,
                     const char *pwd)
</synopsis>

   This is the predecessor of <function>PQconnectdb</function> with a fixed number
   of parameters but the same functionality.   
   </para>
  </listitem>

  <listitem>
   <para>
   <function>PQsetdb</function> Makes a new connection to the database server.
<synopsis>
PGconn *PQsetdb(char *pghost,
                char *pgport,
                char *pgoptions,
                char *pgtty,
                char *dbName)
</synopsis>
   This is a macro that calls <function>PQsetdbLogin()</function> with null pointers
   for the login and pwd parameters.  It is provided primarily
   for backward compatibility with old programs.
   </para>
  </listitem>

 <listitem>
  <para>
   <function>PQconnectStart</function>
   <function>PQconnectPoll</function>
   Make a connection to the database server in a non-blocking manner.
<synopsis>
PGconn *PQconnectStart(const char *conninfo)
</synopsis>
<synopsis>
PostgresPollingStatusType *PQconnectPoll(PQconn *conn)
</synopsis>
   These two routines are used to open a connection to a database server such
   that your application's thread of execution is not blocked on remote I/O
   whilst doing so.
  </para>
  <para>
   The database connection is made using the parameters taken from the string
   <literal>conninfo</literal>, passed to PQconnectStart. This string is in
   the same format as described above for PQconnectdb.
  </para>
  <para>
   Neither PQconnectStart nor PQconnectPoll will block, as long as a number of
   restrictions are met:
   <itemizedlist>
    <listitem>
     <para>
      The hostaddr and host parameters are used appropriately to ensure that
      name and reverse name queries are not made. See the documentation of
      these parameters under PQconnectdb above for details.
     </para>
    </listitem>

    <listitem>
     <para>
      If you call PQtrace, ensure that the stream object into which you trace
      will not block.
     </para>
    </listitem>

    <listitem>
     <para>
      You ensure for yourself that the socket is in the appropriate state
      before calling PQconnectPoll, as described below.
     </para>
    </listitem>
   </itemizedlist>
  </para>

  <para>
   To begin, call <literal>conn=PQconnectStart("&lt;connection_info_string&gt;")</literal>.
   If conn is NULL, then libpq has been unable to allocate a new PGconn
   structure. Otherwise, a valid PGconn pointer is returned (though not yet
   representing a valid connection to the database). On return from
   PQconnectStart, call status=PQstatus(conn). If status equals
   CONNECTION_BAD, PQconnectStart has failed.
  </para>
  <para>
   If PQconnectStart succeeds, the next stage is to poll libpq so that it may
   proceed with the connection sequence.  Loop thus: Consider a connection
   'inactive' by default. If PQconnectPoll last returned PGRES_POLLING_ACTIVE,
   consider it 'active' instead. If PQconnectPoll(conn) last returned
   PGRES_POLLING_READING, perform a select for reading on PQsocket(conn). If
   it last returned PGRES_POLLING_WRITING, perform a select for writing on
   PQsocket(conn). If you have yet to call PQconnectPoll, i.e. after the call
   to PQconnectStart, behave as if it last returned PGRES_POLLING_WRITING.  If
   the select shows that the socket is ready, consider it 'active'. If it has
   been decided that this connection is 'active', call PQconnectPoll(conn)
   again. If this call returns PGRES_POLLING_FAILED, the connection procedure
   has failed.  If this call returns PGRES_POLLING_OK, the connection has been
   successfully made.
  </para>
  <para>
    Note that the use of select() to ensure that the socket is ready is merely
    a (likely) example; those with other facilities available, such as a
    poll() call, may of course use that instead.
  </para>
  <para>
    At any time during connection, the status of the connection may be
    checked, by calling PQstatus. If this is CONNECTION_BAD, then the
    connection procedure has failed; if this is CONNECTION_OK, then the
    connection is ready.  Either of these states should be equally detectable
    from the return value of PQconnectPoll, as above. Other states may be
    shown during (and only during) an asynchronous connection procedure. These
    indicate the current stage of the connection procedure, and may be useful
    to provide feedback to the user for example. These statuses may include:
    <itemizedlist>
     <listitem>
      <para>
      CONNECTION_STARTED: Waiting for connection to be made.
      </para>
     </listitem>
     <listitem>
      <para>
      CONNECTION_MADE: Connection OK; waiting to send.
      </para>
     </listitem>
     <listitem>
      <para>
      CONNECTION_AWAITING_RESPONSE: Waiting for a response from the postmaster.
      </para>
     </listitem>
     <listitem>
      <para>
      CONNECTION_AUTH_OK: Received authentication; waiting for backend startup.
      </para>
     </listitem>
     <listitem>
      <para>
      CONNECTION_SETENV: Negotiating environment.
      </para>
     </listitem>
    </itemizedlist>

    Note that, although these constants will remain (in order to maintain
    compatibility) an application should never rely upon these appearing in a
    particular order, or at all, or on the status always being one of these
    documented values. An application may do something like this:
<programlisting>
    switch(PQstatus(conn))
    {
        case CONNECTION_STARTED:
            feedback = "Connecting...";
	    break;

        case CONNECTION_MADE:
            feedback = "Connected to server...";
            break;
.
.
.
        default:
	    feedback = "Connecting...";
    }
</programlisting>
  </para>
  <para>
   Note that if PQconnectStart returns a non-NULL pointer, you must call
   PQfinish when you are finished with it, in order to dispose of
   the structure and any associated memory blocks. This must be done even if a
   call to PQconnectStart or PQconnectPoll failed.
  </para>
  <para>
   PQconnectPoll will currently block if libpq is compiled with USE_SSL
   defined. This restriction may be removed in the future.
  </para>
  <para>
   PQconnectPoll will currently block under Windows, unless libpq is compiled
   with WIN32_NON_BLOCKING_CONNECTIONS defined. This code has not yet been
   tested under Windows, and so it is currently off by default. This may be
   changed in the future.
  </para>
  <para>
   These functions leave the socket in a non-blocking state as if 
   <function>PQsetnonblocking</function> had been called.
  </para>
 </listitem>

  <listitem>
   <para>
   <function>PQconndefaults</function> Returns the default connection options.
<synopsis>
PQconninfoOption *PQconndefaults(void)

struct PQconninfoOption
{
    char   *keyword;   /* The keyword of the option */
    char   *envvar;    /* Fallback environment variable name */
    char   *compiled;  /* Fallback compiled in default value */
    char   *val;       /* Option's current value, or NULL */
    char   *label;     /* Label for field in connect dialog */
    char   *dispchar;  /* Character to display for this field
                          in a connect dialog. Values are:
                          ""        Display entered value as is
                          "*"       Password field - hide value
                          "D"       Debug option - don't show by default */
    int     dispsize;  /* Field size in characters for dialog */
}
</synopsis>
   Returns a connection options array.  This may
   be used to determine all possible PQconnectdb options and their
   current default values.  The return value points to an array of
   PQconninfoOption structs, which ends with an entry having a NULL
   keyword pointer.  Note that the default values ("val" fields)
   will depend on environment variables and other context.
   Callers must treat the connection options data as read-only.
   </para>
   <para>
    After processing the options array, free it by passing it to
    PQconninfoFree().  If this is not done, a small amount of memory
    is leaked for each call to PQconndefaults().
   </para>
   <para>
    In PostgreSQL versions before 7.0, PQconndefaults() returned a pointer
    to a static array, rather than a dynamically allocated array.  That
    wasn't thread-safe, so the behavior has been changed.
   </para>
  </listitem>

  <listitem>
   <para>
   <function>PQfinish</function>
   Close  the  connection to the backend.  Also frees
   memory used by the PGconn object.
<synopsis>
void PQfinish(PGconn *conn)
</synopsis>
   Note that even if the backend connection attempt fails (as
   indicated by PQstatus), the application should call PQfinish
   to free the memory used by the PGconn object.
   The PGconn pointer should not be used after PQfinish has been called.
   </para>
  </listitem>

  <listitem>
   <para>
   <function>PQreset</function>
   Reset the communication  port  with  the  backend.
<synopsis>
void PQreset(PGconn *conn)
</synopsis>
   This function will close the connection
   to the backend and attempt to  reestablish  a  new
   connection to the same postmaster, using all the same
   parameters previously used.  This may be useful for
   error recovery if a working connection is lost.
   </para>
  </listitem>

  <listitem>
   <para>
   <function>PQresetStart</function>
   <function>PQresetPoll</function>
   Reset the communication  port  with  the  backend, in a non-blocking manner.
<synopsis>
int PQresetStart(PGconn *conn);
</synopsis>
<synopsis>
PostgresPollingStatusType PQresetPoll(PGconn *conn);
</synopsis>
    These functions will close the connection to the backend and attempt to
    reestablish a new connection to the same postmaster, using all the same
    parameters previously used. This may be useful for error recovery if a
    working connection is lost. They differ from PQreset (above) in that they
    act in a non-blocking manner. These functions suffer from the same
    restrictions as PQconnectStart and PQconnectPoll.
   </para>
   <para>
    Call PQresetStart. If it returns 0, the reset has failed. If it returns 1,
    poll the reset using PQresetPoll in exactly the same way as you would
    create the connection using PQconnectPoll.
   </para>
  </listitem>

 </itemizedlist>
</para>

<para>
libpq application programmers should be careful to
maintain the PGconn abstraction.  Use the accessor functions below to get
at the contents of PGconn.  Avoid directly referencing the fields of the
PGconn structure because they are subject to change in the future.
(Beginning in <productname>PostgreSQL</productname> release 6.4, the
definition of struct PGconn is not even provided in <filename>libpq-fe.h</filename>.
If you have old code that accesses PGconn fields directly, you can keep using it
by including <filename>libpq-int.h</filename> too, but you are encouraged to fix the code
soon.)
<itemizedlist>
<listitem>
<para>
<function>PQdb</function>  
         Returns the database name of the connection.
<synopsis>
char *PQdb(const PGconn *conn)
</synopsis>
PQdb and the next several functions return the values established
at connection.  These values are fixed for the life of the PGconn
object.
</para>
</listitem>

<listitem>
<para>
<function>PQuser</function>
         Returns the user name of the connection.
<synopsis>
char *PQuser(const PGconn *conn)
</synopsis>
</para>
</listitem>

<listitem>
<para>
<function>PQpass</function>
         Returns the password of the connection.
<synopsis>
char *PQpass(const PGconn *conn)
</synopsis>
</para>
</listitem>

<listitem>
<para>
<function>PQhost</function>
         Returns the server host name of the connection.
<synopsis>
char *PQhost(const PGconn *conn)
</synopsis>
</para>
</listitem>

<listitem>
<para>
<function>PQport</function>
         Returns the port of the connection.
<synopsis>
char *PQport(const PGconn *conn)
</synopsis>
</para>
</listitem>

<listitem>
<para>
<function>PQtty</function>
         Returns the debug tty of the connection.
<synopsis>
char *PQtty(const PGconn *conn)
</synopsis>
</para>
</listitem>

<listitem>
<para>
<function>PQoptions</function>
       Returns the backend options used in  the  connection.
<synopsis>
char *PQoptions(const PGconn *conn)
</synopsis>
</para>
</listitem>

<listitem>
<para>
<function>PQstatus</function>
         Returns the status of the connection. 
<synopsis>
ConnStatusType PQstatus(const PGconn *conn)
</synopsis>
</para>

      <para>
       The status can be one of a number of values.
       However, only two of these are
       seen outside of an asynchronous connection procedure -
       <literal>CONNECTION_OK</literal> or
       <literal>CONNECTION_BAD</literal>. A good
       connection to the database has the status CONNECTION_OK.
       A failed connection
       attempt is signaled by status
       <literal>CONNECTION_BAD</literal>.
       Ordinarily, an OK status will remain so until
       <function>PQfinish</function>, but a
       communications failure might result in the status changing to
       <literal>CONNECTION_BAD</literal> prematurely.
       In that case the application
       could try to recover by calling <function>PQreset</function>.
      </para>

      <para>
       See the entry for PQconnectStart and PQconnectPoll with regards
       to other status codes
       that might be seen.
      </para>
     </listitem>

     <listitem>
      <para>
       <function>PQerrorMessage</function>
       Returns the error message most recently generated by
       an operation on the connection.
       <synopsis>
char *PQerrorMessage(const PGconn* conn);
       </synopsis>
      </para>

      <para>
       Nearly all libpq functions will set
       <function>PQerrorMessage</function> if they fail.
       Note that by libpq convention, a non-empty
       <function>PQerrorMessage</function> will
       include a trailing newline.
      </para>
     </listitem>

     <listitem>
      <para>
       <function>PQbackendPID</function>
       Returns the process <acronym>ID</acronym> of the backend server 
      handling this connection.
       <synopsis>
int PQbackendPID(const PGconn *conn);
       </synopsis>
       The backend <acronym>PID</acronym> is useful for debugging
       purposes and for comparison
       to NOTIFY messages (which include the <acronym>PID</acronym> of
       the notifying backend).
       Note that the <acronym>PID</acronym> belongs to a process
       executing on the database
       server host, not the local host!
      </para>
     </listitem>
    </itemizedlist>
   </para>
  </sect1>

<sect1>
<title>Query Execution Functions</title>

<para>
Once a connection to a database server has been successfully
established, the functions described here are used to perform
SQL queries and commands.
<itemizedlist>
<listitem>
<para>
<function>PQexec</function>
          Submit a query to <productname>Postgres</productname>
          and wait for the result.
<synopsis>
PGresult *PQexec(PGconn *conn,
                 const char *query);
</synopsis>
          Returns  a  PGresult pointer or possibly a NULL pointer.
          A non-NULL pointer will generally be returned except in
          out-of-memory conditions or serious errors such as inability
          to send the query to the backend.
          If a NULL is returned, it
	  should be treated like a PGRES_FATAL_ERROR result.  Use
	  PQerrorMessage to get more information about the error.
</para>
</listitem>
</itemizedlist>
</para>

<para>
The <function>PGresult</function> structure encapsulates the query result
returned by the backend.
<filename>libpq</filename> application programmers should be careful to
maintain the PGresult abstraction.  Use the accessor functions below to get
at the contents of PGresult.  Avoid directly referencing the fields of the
PGresult structure because they are subject to change in the future.
(Beginning in <productname>Postgres</productname> release 6.4, the
definition of struct PGresult is not even provided in libpq-fe.h.  If you
have old code that accesses PGresult fields directly, you can keep using it
by including libpq-int.h too, but you are encouraged to fix the code
soon.)

<itemizedlist>
<listitem>
<para>
<function>PQresultStatus</function>
          Returns the result status of the query.
<synopsis>
ExecStatusType PQresultStatus(const PGresult *res)
</synopsis>
PQresultStatus can return one of the following values:
<itemizedlist>
 <listitem>
  <para><literal>PGRES_EMPTY_QUERY</literal> -- The string sent to the backend was empty.</para>
 </listitem>
 <listitem>
  <para><literal>PGRES_COMMAND_OK</literal> -- Successful completion of a command returning no data</para>
 </listitem>
 <listitem>
  <para><literal>PGRES_TUPLES_OK</literal> -- The query successfully executed</para>
 </listitem>
 <listitem>
  <para><literal>PGRES_COPY_OUT</literal> -- Copy Out (from server) data transfer started</para>
 </listitem>
 <listitem>
  <para><literal>PGRES_COPY_IN</literal> -- Copy In (to server) data transfer started</para>
 </listitem>
 <listitem>
  <para><literal>PGRES_BAD_RESPONSE</literal> -- The server's response was not understood</para>
 </listitem>
 <listitem>
  <para><literal>PGRES_NONFATAL_ERROR</literal></para>
 </listitem>
 <listitem>
  <para><literal>PGRES_FATAL_ERROR</literal></para>
 </listitem>
</itemizedlist>

If  the result status is <literal>PGRES_TUPLES_OK</literal>, then the
routines described below can be  used  to  retrieve  the
tuples returned by the query.  Note that a SELECT that
happens to retrieve zero tuples still shows <literal>PGRES_TUPLES_OK</literal>.
<literal>PGRES_COMMAND_OK</literal> is for commands that can never return tuples
(INSERT, UPDATE, etc.). A response of <literal>PGRES_EMPTY_QUERY</literal> often
exposes a bug in the client software.
</para>
</listitem>

<listitem>
<para>
<function>PQresStatus</function>
	Converts the enumerated type returned by PQresultStatus into
	a string constant describing the status code.
<synopsis>
char *PQresStatus(ExecStatusType status);
</synopsis>
</para>
</listitem>

<listitem>
<para>
<function>PQresultErrorMessage</function>
returns the error message associated with the query, or an empty string
if there was no error.
<synopsis>
char *PQresultErrorMessage(const PGresult *res);
</synopsis>
Immediately following a <function>PQexec</function> or <function>PQgetResult</function>
call, <function>PQerrorMessage</function> (on the connection) will return the same
string as <function>PQresultErrorMessage</function> (on the result).  However, a
PGresult will retain its error message
until destroyed, whereas the connection's error message will change when
subsequent operations are done.  Use <function>PQresultErrorMessage</function> when you want to
know the status associated with a particular PGresult; use <function>PQerrorMessage</function>
when you want to know the status from the latest operation on the connection.
</para>
</listitem>

<listitem>
<para>
<function>PQntuples</function>
          Returns the number of tuples (instances)
          in the query result.
<synopsis>
int PQntuples(const PGresult *res);
</synopsis>
</para>
</listitem>

<listitem>
<para>
<function>PQnfields</function>
          Returns   the   number    of    fields
          (attributes) in each tuple of the query result.
<synopsis>
int PQnfields(const PGresult *res);
</synopsis>
</para>
</listitem>

<listitem>
<para>
<function>PQbinaryTuples</function>
          Returns 1 if the PGresult contains binary tuple data,
	  0 if it contains ASCII data.
<synopsis>
int PQbinaryTuples(const PGresult *res);
</synopsis>
Currently, binary tuple data can only be returned by a query that
extracts data from a <acronym>BINARY</acronym> cursor.
</para>
</listitem>

<listitem>
<para>
<function>PQfname</function>
 Returns the field (attribute) name associated with the given field  index.
 Field  indices start at 0.
<synopsis>
char *PQfname(const PGresult *res,
                    int field_index);
</synopsis>
</para>
</listitem>

<listitem>
<para>
<function>PQfnumber</function>
            Returns  the  field  (attribute)  index
          associated with the given field name.
<synopsis>
int PQfnumber(const PGresult *res,
              const char *field_name);
</synopsis>
</para>

<para>
        -1 is returned if the given name does not match any field.
</para>
</listitem>

<listitem>
<para>
<function>PQftype</function>
            Returns the field type associated with the
          given  field  index.  The  integer  returned is an
          internal coding of the type.  Field indices  start
          at 0.
<synopsis>
Oid PQftype(const PGresult *res,
            int field_num);
</synopsis>
You can query the system table <literal>pg_type</literal> to obtain
the name and properties of the various datatypes. The <acronym>OID</acronym>s
of the built-in datatypes are defined in <filename>src/include/catalog/pg_type.h</filename>
in the source tree.
</para>
</listitem>

<listitem>
<para>
<function>PQfsize</function>
          Returns  the  size  in bytes of the field
          associated with the given field index.
          Field indices start at 0.
<synopsis>
int PQfsize(const PGresult *res,
            int field_index);
</synopsis>
	PQfsize returns the space allocated for this field in a database
	tuple, in other words the size of the server's binary representation
	of the data type.  -1 is returned if the field is variable size.
</para>
</listitem>

<listitem>
<para>
<function>PQfmod</function>
          Returns  the type-specific modification data of the field
          associated with the given field index.
          Field indices start at 0.
<synopsis>
int PQfmod(const PGresult *res,
           int field_index);
</synopsis>
</para>
</listitem>

<listitem>
<para>
<function>PQgetvalue</function>
            Returns a single field  (attribute)  value of one tuple
	    of a PGresult.
	    Tuple and field indices start at 0.
<synopsis>
char* PQgetvalue(const PGresult *res,
                 int tup_num,
                 int field_num);
</synopsis>
For most queries, the value returned by <function>PQgetvalue</function>
is a null-terminated <acronym>ASCII</acronym> string  representation
of the attribute value.  But if <function>PQbinaryTuples()</function> is 1,
the  value  returned  by <function>PQgetvalue</function>  is  the  binary
representation of the
type in the internal format of the backend server
(but not including the size word, if the field is variable-length).
It  is then the programmer's responsibility to cast and
convert the data to the correct C type.  The pointer
returned  by  <function>PQgetvalue</function> points to storage that is
part of the PGresult structure.  One should not modify it,
and one must explicitly 
copy the value into other storage if it is to
be used past the lifetime of the  PGresult  structure itself.
</para>
</listitem>

<listitem>
<para>
<function>PQgetlength</function>
          Returns   the   length  of  a  field (attribute) in bytes.
          Tuple and field indices start at 0.
<synopsis>
int PQgetlength(const PGresult *res,
                int tup_num,
                int field_num);
</synopsis>
This is the actual data length for the particular data value, that is the
size of the object pointed to by PQgetvalue.  Note that for ASCII-represented
values, this size has little to do with the binary size reported by PQfsize.
</para>
</listitem>

<listitem>
<para>
<function>PQgetisnull</function>
           Tests a field for a NULL entry.
           Tuple and field indices start at 0.
<synopsis>
int PQgetisnull(const PGresult *res,
                int tup_num,
                int field_num);
</synopsis>
            This function returns  1 if the field contains a NULL, 0 if
            it contains a non-null value.  (Note that PQgetvalue
            will return an empty string, not a null pointer, for a NULL
            field.)
</para>
</listitem>

<listitem>
<para>
<function>PQcmdStatus</function>
          Returns the command status string from the SQL command that
	  generated the PGresult.
<synopsis>
char * PQcmdStatus(const PGresult *res);
</synopsis>
</para>
</listitem>

<listitem>
<para>
<function>PQcmdTuples</function>
	  Returns the number of rows affected by the SQL command.
<synopsis>
char * PQcmdTuples(const PGresult *res);
</synopsis>
          If the <acronym>SQL</acronym> command that generated the
	  PGresult was INSERT, UPDATE or DELETE, this returns a
	  string containing the number of rows affected.  If the
          command was anything else, it returns the empty string.
</para>
</listitem>

<listitem>
<para>
<function>PQoidValue</function>
          Returns the object id of  the  tuple
          inserted,  if  the <acronym>SQL</acronym> command was an INSERT.
          Otherwise, returns <literal>InvalidOid</literal>.
<synopsis>
Oid PQoidValue(const PGresult *res);
</synopsis>
          The type <type>Oid</type> and the constant <literal>InvalidOid</literal>
          will be defined if you include the <application>libpq</application>
          header file. They will both be some integer type.
</para>
</listitem>

<listitem>
<para>
<function>PQoidStatus</function>
          Returns a string with the object id of  the  tuple
          inserted,  if  the <acronym>SQL</acronym> command was an INSERT.
          Otherwise, returns an empty string.
<synopsis>
char * PQoidStatus(const PGresult *res);
</synopsis>
This function is deprecated in favor of <function>PQoidValue</function>
and is not thread-safe.
</para>
</listitem>

<listitem>
<para>
<function>PQprint</function>
          Prints out all the  tuples  and,  optionally,  the
          attribute  names  to  the specified output stream.
<synopsis>
void PQprint(FILE* fout,      /* output stream */
             const PGresult *res,
             const PQprintOpt *po);

struct {
    pqbool  header;      /* print output field headings and row count */
    pqbool  align;       /* fill align the fields */
    pqbool  standard;    /* old brain dead format */
    pqbool  html3;       /* output html tables */
    pqbool  expanded;    /* expand tables */
    pqbool  pager;       /* use pager for output if needed */
    char    *fieldSep;   /* field separator */
    char    *tableOpt;   /* insert to HTML &lt;table ...&gt; */
    char    *caption;    /* HTML &lt;caption&gt; */
    char    **fieldName; /* null terminated array of replacement field names */
} PQprintOpt;
</synopsis>
This function was formerly used by <application>psql</application>
to print query results, but this is no longer the case and this
function is no longer actively supported.
</para>
</listitem>

<listitem>
<para>
<function>PQclear</function>
          Frees  the  storage  associated with the PGresult.
          Every query result should be freed via PQclear  when
          it  is  no  longer needed.
<synopsis>
void PQclear(PQresult *res);
</synopsis>
          You can keep a PGresult object around for as long as you
          need it; it does not go away when you issue a new query,
          nor even if you close the connection.  To get rid of it,
          you must call <function>PQclear</function>.  Failure to do this will
          result in memory leaks in  the  frontend  application.
</para>
</listitem>

<listitem>
<para>
<function>PQmakeEmptyPGresult</function>
          Constructs an empty PGresult object with the given status.
<synopsis>
PGresult* PQmakeEmptyPGresult(PGconn *conn, ExecStatusType status);
</synopsis>
This is libpq's internal routine to allocate and initialize an empty
PGresult object.  It is exported because some applications find it
useful to generate result objects (particularly objects with error
status) themselves.  If conn is not NULL and status indicates an error,
the connection's current errorMessage is copied into the PGresult.
Note that PQclear should eventually be called on the object, just
as with a PGresult returned by libpq itself.
</para>
</listitem>

</itemizedlist>
</para>
</sect1>

<sect1>
<title>Asynchronous Query Processing</title>

<para>
The <function>PQexec</function> function is adequate for submitting queries in
simple synchronous
applications.  It has a couple of major deficiencies however:

<itemizedlist>
<listitem>
<para>
<function>PQexec</function> waits for the query to be completed.  The application may have other
work to do (such as maintaining a user interface), in which case it won't
want to block waiting for the response.
</para>
</listitem>
<listitem>
<para>
Since control is buried inside <function>PQexec</function>, it is hard for the frontend
to decide it would like to try to cancel the ongoing query.  (It can be
done from a signal handler, but not otherwise.)
</para>
</listitem>
<listitem>
<para>
<function>PQexec</function> can return only one PGresult structure.  If the submitted query
string contains multiple <acronym>SQL</acronym> commands, all but the last PGresult are
discarded by <function>PQexec</function>.
</para>
</listitem>
</itemizedlist>
</para>

<para>
Applications that do not like these limitations can instead use the
underlying functions that <function>PQexec</function> is built from:
<function>PQsendQuery</function> and <function>PQgetResult</function>.
</para>
<para>
Older programs that used this functionality as well as 
<function>PQputline</function> and <function>PQputnbytes</function>
could block waiting to send data to the backend, to
address that issue, the function <function>PQsetnonblocking</function>
was added.
</para>
<para>
Old applications can neglect to use <function>PQsetnonblocking</function>
and get the older potentially blocking behavior.  Newer programs can use 
<function>PQsetnonblocking</function> to achieve a completely non-blocking
connection to the backend.

<itemizedlist>
 <listitem>
   <para>
    <function>PQsetnonblocking</function> Sets the state of the connection
    to non-blocking.
<synopsis>
int PQsetnonblocking(PGconn *conn)
</synopsis>
    this function will ensure that calls to 
    <function>PQputline</function>, <function>PQputnbytes</function>,
    <function>PQsendQuery</function> and <function>PQendcopy</function>
    will not block but instead return an error if they need to be called
    again.
   </para>
   <para>
    When a database connection has been set to non-blocking mode and
    <function>PQexec</function> is called, it will temporarily set the state
    of the connection to blocking until the <function>PQexec</function> 
    completes. 
   </para>
   <para>
    More of libpq is expected to be made safe for 
    <function>PQsetnonblocking</function> functionality in the near future.
  </para>
 </listitem>

<listitem>
<para>
<function>PQisnonblocking</function>
       Returns the blocking status of the database connection.
<synopsis>
int PQisnonblocking(const PGconn *conn)
</synopsis>
       Returns TRUE if the connection is set to non-blocking mode,
       FALSE if blocking.
</para>
</listitem>

<listitem>
<para>
<function>PQsendQuery</function>
          Submit a query to <productname>Postgres</productname> without
	  waiting for the result(s).  TRUE is returned if the query was
	  successfully dispatched, FALSE if not (in which case, use
	  PQerrorMessage to get more information about the failure).
<synopsis>
int PQsendQuery(PGconn *conn,
                const char *query);
</synopsis>
	  After successfully calling <function>PQsendQuery</function>, call
          <function>PQgetResult</function> one or more
	  times to obtain the query results.  <function>PQsendQuery</function> may not be called
	  again (on the same connection) until <function>PQgetResult</function> has returned NULL,
	  indicating that the query is done.
</para>
</listitem>

<listitem>
<para>
<function>PQgetResult</function>
          Wait for the next result from a prior <function>PQsendQuery</function>,
	  and return it.  NULL is returned when the query is complete
	  and there will be no more results.
<synopsis>
PGresult *PQgetResult(PGconn *conn);
</synopsis>
	  <function>PQgetResult</function> must be called repeatedly until it returns NULL,
	  indicating that the query is done.  (If called when no query is
	  active, <function>PQgetResult</function> will just return NULL at once.)
	  Each non-null result from <function>PQgetResult</function> should be processed using
	  the same PGresult accessor functions previously described.
	  Don't forget to free each result object with <function>PQclear</function> when done with it.
	  Note that <function>PQgetResult</function> will block only if a query is active and the
	  necessary response data has not yet been read by <function>PQconsumeInput</function>.
</para>
</listitem>

</itemizedlist>
</para>

<para>
Using <function>PQsendQuery</function> and <function>PQgetResult</function>
solves one of <function>PQexec</function>'s problems:
If a query string contains multiple <acronym>SQL</acronym> commands, the results of those
commands can be obtained individually.  (This allows a simple form of
overlapped processing, by the way: the frontend can be handling the
results of one query while the backend is still working on later
queries in the same query string.)  However, calling <function>PQgetResult</function> will
still cause the frontend to block until the backend completes the
next <acronym>SQL</acronym> command.  This can be avoided by proper use of three more
functions:

<itemizedlist>
<listitem>
<para>
<function>PQconsumeInput</function>
	  If input is available from the backend, consume it.
<synopsis>
int PQconsumeInput(PGconn *conn);
</synopsis>
<function>PQconsumeInput</function> normally returns 1 indicating "no error",
but returns 0 if there was some kind of trouble (in which case
<function>PQerrorMessage</function> is set).  Note that the result does not say
whether any input data was actually collected. After calling
<function>PQconsumeInput</function>, the application may check
<function>PQisBusy</function> and/or <function>PQnotifies</function> to see if
their state has changed.
</para>
<para>
<function>PQconsumeInput</function> may be called even if the application is not
prepared to deal with a result or notification just yet.  The
routine will read available data and save it in a buffer, thereby
causing a <function>select</function>(2) read-ready indication to go away.  The
application can thus use <function>PQconsumeInput</function> to clear the
<function>select</function> condition immediately, and then examine the results at leisure.
</para>
</listitem>

<listitem>
<para>
<function>PQisBusy</function>
Returns 1 if a query is busy, that is, <function>PQgetResult</function> would block
waiting for input.  A 0 return indicates that <function>PQgetResult</function> can
be called with assurance of not blocking.
<synopsis>
int PQisBusy(PGconn *conn);
</synopsis>
<function>PQisBusy</function> will not itself attempt to read data from the backend;
therefore <function>PQconsumeInput</function> must be invoked first, or the busy
state will never end.
</para>
</listitem>

<listitem>
<para>
<function>PQflush</function> Attempt to flush any data queued to the backend,
returns 0 if successful (or if the send queue is empty) or EOF if it failed for
some reason.
<synopsis>
int PQflush(PGconn *conn);
</synopsis>
<function>PQflush</function> needs to be called on a non-blocking connection 
before calling <function>select</function> to determine if a responce has
arrived.  If 0 is returned it ensures that there is no data queued to the 
backend that has not actually been sent.  Only applications that have used
<function>PQsetnonblocking</function> have a need for this.
</para>
</listitem>

<listitem>
<para>
<function>PQsocket</function>
	  Obtain the file descriptor number for the backend connection socket.
	  A valid descriptor will be &gt;= 0; a result of -1 indicates that
	  no backend connection is currently open.
<synopsis>
int PQsocket(const PGconn *conn);
</synopsis>
<function>PQsocket</function> should be used to obtain the backend socket descriptor
in preparation for executing <function>select</function>(2).  This allows an
application using a blocking connection to wait for either backend responses or
other conditions.
If the result of <function>select</function>(2) indicates that data can be read from
the backend socket, then <function>PQconsumeInput</function> should be called to read the
data; after which, <function>PQisBusy</function>, <function>PQgetResult</function>,
and/or <function>PQnotifies</function> can be used to process the response.
</para>
<para>
Non-blocking connections (that have used <function>PQsetnonblocking</function>)
should not use <function>select</function> until <function>PQflush</function>
has returned 0 indicating that there is no buffered data waiting to be sent
to the backend.
</para>
</listitem>

</itemizedlist>
</para>

<para>
A typical frontend using these functions will have a main loop that uses
<function>select</function>(2) to wait for all the conditions that it must
respond to.  One of the conditions will be input available from the backend,
which in <function>select</function>'s terms is readable data on the file
descriptor identified by <function>PQsocket</function>.
When the main loop detects input ready, it should call
<function>PQconsumeInput</function> to read the input.  It can then call
<function>PQisBusy</function>, followed by <function>PQgetResult</function>
if <function>PQisBusy</function> returns false (0).  It can also call
<function>PQnotifies</function> to detect NOTIFY messages (see "Asynchronous
Notification", below).
</para>

<para>
A frontend that uses <function>PQsendQuery</function>/<function>PQgetResult</function>
can also attempt to cancel a query that is still being processed by the backend.
</para>

<para>
<itemizedlist>
<listitem>
<para>
<function>PQrequestCancel</function>
	  Request that <productname>Postgres</productname> abandon
	  processing of the current query.
<synopsis>
int PQrequestCancel(PGconn *conn);
</synopsis>
The return value is 1 if the cancel request was successfully
dispatched, 0 if not.  (If not, <function>PQerrorMessage</function> tells why not.)
Successful dispatch is no guarantee that the request will have any
effect, however.  Regardless of the return value of <function>PQrequestCancel</function>,
the application must continue with the normal result-reading
sequence using <function>PQgetResult</function>.  If the cancellation
is effective, the current query will terminate early and return
an error result.  If the cancellation fails (say, because the
backend was already done processing the query), then there will
be no visible result at all.
</para>
</listitem>
</itemizedlist>
</para>

<para>
Note that if the current query is part of a transaction, cancellation
will abort the whole transaction.
</para>

<para>
<function>PQrequestCancel</function> can safely be invoked from a signal handler.
So, it is also possible to use it in conjunction with plain
<function>PQexec</function>, if the decision to cancel can be made in a signal
handler.  For example, <application>psql</application> invokes
<function>PQrequestCancel</function> from a SIGINT signal handler, thus allowing
interactive cancellation of queries that it issues through <function>PQexec</function>.
Note that <function>PQrequestCancel</function> will have no effect if the connection
is not currently open or the backend is not currently processing a query.
</para>

</sect1>

<sect1>
<title>Fast Path</title>

<para>
<productname>PostgreSQL</productname> provides a fast path interface to send
function calls to the backend.  This is a trapdoor into system internals and
can be a potential security hole.  Most users will not need this feature.

<itemizedlist>
<listitem>
<para>
<function>PQfn</function>
	Request execution of a backend function via the fast path interface.
<synopsis>
PGresult* PQfn(PGconn* conn,
               int fnid,
               int *result_buf,
               int *result_len,
               int result_is_int,
               const PQArgBlock *args,
               int nargs);
</synopsis>
     The fnid argument is the object identifier of the function to be
     executed.
     result_buf is the buffer in which
     to place the return value.  The caller must  have  allocated
     sufficient space to store the return value (there is no check!).
     The actual result length will be returned in the integer pointed
     to  by  result_len.   If a 4-byte integer result is expected, set
     result_is_int to 1; otherwise set it to 0.  (Setting result_is_int to 1
     tells libpq to byte-swap the value if necessary, so that it is
     delivered as a proper int value for the client machine.  When
     result_is_int is 0, the byte string sent by the backend is returned
     unmodified.)
     args and nargs specify the arguments to be passed to the function.
<synopsis>
typedef struct {
    int len;
    int isint;
    union {
        int *ptr;
        int integer;
    } u;
} PQArgBlock;
</synopsis>
     <function>PQfn</function> always returns a valid PGresult*. The resultStatus
     should be checked before the result is used.   The
     caller is responsible for  freeing  the  PGresult  with
     <function>PQclear</function> when it is no longer needed.
</para>
</listitem>
</itemizedlist>
</para>

</sect1>

<sect1>
<title>Asynchronous Notification</title>

<para>
<productname>PostgreSQL</productname> supports asynchronous notification via the
LISTEN and NOTIFY commands.  A backend registers its interest in a particular
notification condition with the LISTEN command (and can stop listening
with the UNLISTEN command).  All backends listening on a
particular condition will be notified asynchronously when a NOTIFY of that
condition name is executed by any backend.  No additional information is
passed from the notifier to the listener.  Thus, typically, any actual data
that needs to be communicated is transferred through a database relation.
Commonly the condition name is the same as the associated relation, but it is
not necessary for there to be any associated relation.
</para>

<para>
<filename>libpq</filename> applications submit LISTEN and UNLISTEN
commands as ordinary SQL queries.  Subsequently, arrival of NOTIFY
messages can be detected by calling PQnotifies().

<itemizedlist>
<listitem>
<para>
<function>PQnotifies</function>
          Returns  the next notification from a list of unhandled
          notification messages received from the backend.  Returns NULL if
          there are no pending notifications.  Once a notification is
	  returned from PQnotifies, it is considered handled and will be
	  removed from the list of notifications.
<synopsis>
PGnotify* PQnotifies(PGconn *conn);

typedef struct pgNotify {
    char relname[NAMEDATALEN];       /* name of relation
                                      * containing data */
    int  be_pid;                     /* process id of backend */
} PGnotify;
</synopsis>
After processing a PGnotify object returned by <function>PQnotifies</function>,
be sure to free it with <function>free()</function> to avoid a memory leak.
</para>
<note>
<para>
 In <productname>PostgreSQL</productname> 6.4 and later,
 the <literal>be_pid</literal> is the notifying backend's,
 whereas in earlier versions it was always your own backend's <acronym>PID</acronym>.
</para>
</note>
</listitem>
</itemizedlist>
</para>

<para>
The  second  sample program gives an example of the use
of asynchronous notification.
</para>

<para>
<function>PQnotifies()</function> does not actually read backend data; it just
returns messages previously absorbed by another <application>libpq</application>
function.  In prior releases of <application>libpq</application>, the only way
to ensure timely receipt of NOTIFY messages was to constantly submit queries,
even empty ones, and then check <function>PQnotifies()</function> after each
<function>PQexec()</function>.  While this still works, it is
deprecated as a waste of processing power.
</para>
<para>
A better way to check for NOTIFY
messages when you have no useful queries to make is to call
<function>PQconsumeInput()</function>, then check
<function>PQnotifies()</function>.
You can use <function>select</function>(2) to wait for backend data to
arrive, thereby using no <acronym>CPU</acronym> power unless there is something
to do.  (See <function>PQsocket()</function> to obtain the file descriptor
number to use with <function>select</function>.)
Note that this will work OK whether you submit queries with
<function>PQsendQuery</function>/<function>PQgetResult</function> or simply
use <function>PQexec</function>.  You should, however, remember to
check <function>PQnotifies()</function> after each
<function>PQgetResult</function> or <function>PQexec</function>, to see
if any notifications came in during the processing of the query.
</para>

</sect1>

<sect1>
<title>Functions Associated with the COPY Command</title>

<para>
 The COPY command in <productname>PostgreSQL</productname> has options to  read  from
 or  write  to  the  network  connection  used by <filename>libpq</filename>.
 Therefore, functions are necessary to access this  network
 connection directly so applications may take advantage of this capability.
</para>

<para>
 These functions should be executed only after obtaining a <literal>PGRES_COPY_OUT</literal>
 or <literal>PGRES_COPY_IN</literal> result object from <function>PQexec</function>
 or <function>PQgetResult</function>.
</para>

<para>
<itemizedlist>
<listitem>
<para>
<function>PQgetline</function>
          Reads  a  newline-terminated  line  of  characters
          (transmitted  by the backend server) into a buffer
          string of size length.
<synopsis>
int PQgetline(PGconn *conn,
              char *string,
              int length)
</synopsis>
Like <function>fgets</function>(3),  this  routine copies up to length-1 characters
into string. It is like <function>gets</function>(3), however, in that it converts
the terminating newline into a null character.
<function>PQgetline</function> returns <literal>EOF</literal> at EOF, 0 if the
entire line has been read, and 1 if the buffer is full but the
terminating newline has not yet been read.
</para>
<para>
Notice that the application must check to see if a
new line consists of  the  two characters  "\.",
which  indicates  that the backend server has finished sending
the results  of  the  copy  command.
If  the  application might
receive lines that are more than length-1  characters  long,
care is needed to be sure one recognizes the "\." line correctly
(and does not, for example, mistake the end of a long data line
for a terminator line).
The code in
<filename>
src/bin/psql/copy.c
</filename>
contains example routines that correctly handle the  copy protocol.
</para>
</listitem>

<listitem>
<para>
<function>PQgetlineAsync</function>
          Reads  a  newline-terminated  line  of  characters
          (transmitted  by the backend server) into a buffer
          without blocking.
<synopsis>
int PQgetlineAsync(PGconn *conn,
                   char *buffer,
                   int bufsize)
</synopsis>
This routine is similar to <function>PQgetline</function>, but it can be used
by applications
that must read COPY data asynchronously, that is without blocking.
Having issued the COPY command and gotten a <literal>PGRES_COPY_OUT</literal>
response, the
application should call <function>PQconsumeInput</function> and
<function>PQgetlineAsync</function> until the
end-of-data signal is detected.  Unlike <function>PQgetline</function>, this routine takes
responsibility for detecting end-of-data.
On each call, <function>PQgetlineAsync</function> will return data if a complete newline-
terminated data line is available in libpq's input buffer, or if the
incoming data line is too long to fit in the buffer offered by the caller.
Otherwise, no data is returned until the rest of the line arrives.
</para>
<para>
The routine returns -1 if the end-of-copy-data marker has been recognized,
or 0 if no data is available, or a positive number giving the number of
bytes of data returned.  If -1 is returned, the caller must next call
<function>PQendcopy</function>, and then return to normal processing.
The data returned will not extend beyond a newline character.  If possible
a whole line will be returned at one time.  But if the buffer offered by
the caller is too small to hold a line sent by the backend, then a partial
data line will be returned.  This can be detected by testing whether the
last returned byte is <quote><literal>\n</literal></quote> or not.
The returned string is not null-terminated.  (If you want to add a
terminating null, be sure to pass a bufsize one smaller than the room
actually available.)
</para>
</listitem>

<listitem>
<para>
<function>PQputline</function>
Sends  a  null-terminated  string  to  the backend server.
Returns 0 if OK, <literal>EOF</literal> if unable to send the string.
<synopsis>
int PQputline(PGconn *conn,
              const char *string);
</synopsis>
Note the application must explicitly  send  the  two
characters  <quote><literal>\.</literal></quote> on a final line  to indicate to
the backend that it has finished sending its data.
</para>
</listitem>

<listitem>
<para>
<function>PQputnbytes</function>
Sends  a  non-null-terminated  string  to  the backend server.
Returns 0 if OK, EOF if unable to send the string.
<synopsis>
int PQputnbytes(PGconn *conn,
                const char *buffer,
                int nbytes);
</synopsis>
This is exactly like <function>PQputline</function>, except that the data buffer need
not be null-terminated since the number of bytes to send is
specified directly.
</para>
</listitem>

<listitem>
<para>
<function>PQendcopy</function>
 Syncs with the backend.  This function waits until
 the  backend  has  finished  the  copy.  It should
 either be issued when the  last  string  has  been
 sent  to  the  backend using <function>PQputline</function> or when the
 last string has been  received  from  the  backend
 using <function>PGgetline</function>.  It must be issued or the backend
 may get <quote>out of sync</quote>  with  the  frontend.   Upon
 return from this function, the backend is ready to
 receive the next query.
 The return value is 0  on  successful  completion,
 nonzero otherwise.
<synopsis>
int PQendcopy(PGconn *conn);
</synopsis>
</para>

<para>
As an example:

<programlisting>
PQexec(conn, "create table foo (a int4, b char(16), d float8)");
PQexec(conn, "copy foo from stdin");
PQputline(conn, "3\thello world\t4.5\n");
PQputline(conn,"4\tgoodbye world\t7.11\n");
...
PQputline(conn,"\\.\n");
PQendcopy(conn);
</programlisting>
</para>
</listitem>
</itemizedlist>
</para>

<para>
When using <function>PQgetResult</function>, the application should respond to
a <literal>PGRES_COPY_OUT</literal> result by executing <function>PQgetline</function>
repeatedly, followed by <function>PQendcopy</function> after the terminator line is seen.
It should then return to the <function>PQgetResult</function> loop until
<function>PQgetResult</function> returns NULL. Similarly a <literal>PGRES_COPY_IN</literal>
result is processed by a series of <function>PQputline</function> calls followed by
<function>PQendcopy</function>, then return to the <function>PQgetResult</function> loop.
This arrangement will ensure that
a copy in or copy out command embedded in a series of <acronym>SQL</acronym> commands
will be executed correctly.
</para>
<para>
Older applications are likely to submit a copy in or copy out
via <function>PQexec</function> and assume that the transaction is done after
<function>PQendcopy</function>.
This will work correctly only if the copy in/out is the only
<acronym>SQL</acronym> command in the query string.
</para>

</sect1>

<sect1>
<title><application>libpq</application> Tracing Functions</title>

<para>
<itemizedlist>
<listitem>
<para>
<function>PQtrace</function>
          Enable  tracing of the frontend/backend communication to a debugging file stream.
<synopsis>
void PQtrace(PGconn *conn
             FILE *debug_port)
</synopsis>
</para>
</listitem>

<listitem>
<para>
<function>PQuntrace</function>
          Disable tracing started by PQtrace
<synopsis>
void PQuntrace(PGconn *conn)
</synopsis>
</para>
</listitem>
</itemizedlist>
</para>

</sect1>

<sect1>
<title>
<application>libpq</application> Control Functions</title>

<para>
<itemizedlist>
<listitem>
<para>
<function>PQsetNoticeProcessor</function>
Control reporting of notice and warning messages generated by libpq.
<synopsis>
typedef void (*PQnoticeProcessor) (void *arg, const char *message);

PQnoticeProcessor
PQsetNoticeProcessor(PGconn *conn,
                     PQnoticeProcessor proc,
                     void *arg);
</synopsis>
</para>
</listitem>
</itemizedlist>
</para>

<para>
By default, <application>libpq</application> prints <quote>notice</quote>
messages from the backend on <filename>stderr</filename>,
as well as a few error messages that it generates by itself.
This behavior can be overridden by supplying a callback function that
does something else with the messages.  The callback function is passed
the text of the error message (which includes a trailing newline), plus
a void pointer that is the same one passed to
<function>PQsetNoticeProcessor</function>.
(This pointer can be used to access application-specific state if needed.)
The default notice processor is simply
<programlisting>
static void
defaultNoticeProcessor(void * arg, const char * message)
{
    fprintf(stderr, "%s", message);
}
</programlisting>
To use a special notice processor, call
<function>PQsetNoticeProcessor</function> just after
creation of a new PGconn object.
</para>

<para>
The return value is the pointer to the previous notice processor.
If you supply a callback function pointer of NULL, no action is taken,
but the current pointer is returned.
</para>

<para>
Once you have set a notice processor, you should expect that that function
could be called as long as either the PGconn object or PGresult objects
made from it exist.  At creation of a PGresult, the PGconn's current
notice processor pointer is copied into the PGresult for possible use by
routines like <function>PQgetvalue</function>.
</para>

</sect1>

<sect1 id="libpq-envars">
<title>Environment Variables</title>

<para>
The following environment variables can be used to select default
connection parameter values, which will be used by <function>PQconnectdb</function> or
<function>PQsetdbLogin</function> if no value is directly specified by the calling code.
These are useful to avoid hard-coding database names into simple
application programs.

<itemizedlist>
<listitem>
<para>
<envar>PGHOST</envar> sets the default server name.
If a non-zero-length string is specified, TCP/IP communication is used.
Without a host name, libpq will connect using a local Unix domain socket.
</para>
</listitem>
<listitem>
<para>
<envar>PGPORT</envar>  sets the default port or local Unix domain socket
file extension for communicating with the <productname>PostgreSQL</productname>
backend.
</para>
</listitem>
<listitem>
<para>
<envar>PGDATABASE</envar>  sets the default 
<productname>PostgreSQL</productname> database name.
</para>
</listitem>
<listitem>
<para>
<envar>PGUSER</envar>
sets the username used to connect to the database and for authentication.
</para>
</listitem>
<listitem>
<para>
<envar>PGPASSWORD</envar>
sets the password used if the backend demands password authentication.
</para>
</listitem>
<listitem>
<para>
<envar>PGREALM</envar> sets the Kerberos realm to  use  with  
<productname>PostgreSQL</productname>, if  it is different from the local realm.
If <envar>PGREALM</envar> is set, <productname>PostgreSQL</productname> 
applications  will  attempt authentication  with  servers for this realm and use
separate ticket files to avoid conflicts with  local
ticket  files.   This  environment  variable is only
used if Kerberos authentication is selected by the backend.
</para>
</listitem>
<listitem>
<para>
<envar>PGOPTIONS</envar> sets additional runtime  options  for  
the <productname>PostgreSQL</productname> backend.
</para>
</listitem>
<listitem>
<para>
<envar>PGTTY</envar> sets the file or tty on which  debugging  
messages from the backend server are displayed.
</para>
</listitem>
</itemizedlist>
</para>

<para>
The following environment variables can be used to specify user-level default
behavior for every Postgres session:

<itemizedlist>
<listitem>
<para>
<envar>PGDATESTYLE</envar>
sets the default style of date/time representation.
</para>
</listitem>
<listitem>
<para>
<envar>PGTZ</envar>
sets the default time zone.
</para>
</listitem>
<listitem>
<para>
<envar>PGCLIENTENCODING</envar>
sets the default client encoding (if MULTIBYTE support was selected
when configuring Postgres).
</para>
</listitem>
</itemizedlist>
</para>

<para>
The following environment variables can be used to specify default internal
behavior for every Postgres session:

<itemizedlist>
<listitem>
<para>
<envar>PGGEQO</envar>
sets the default mode for the genetic optimizer.
</para>
</listitem>
</itemizedlist>
</para>

<para>
Refer to the <command>SET</command> <acronym>SQL</acronym> command
for information on correct values for these environment variables.
</para>

</sect1>

<sect1 id="libpq-threading">
<title>Threading Behavior</title>

<para>
<filename>libpq</filename> is thread-safe as of
<productname>PostgreSQL</productname> 7.0, so long as no two threads
attempt to manipulate the same PGconn object at the same time.  In particular,
you can't issue concurrent queries from different threads through the same
connection object.  (If you need to run concurrent queries, start up multiple
connections.)
</para>

<para>
PGresult objects are read-only after creation, and so can be passed around
freely between threads.
</para>

<para>
The deprecated functions <function>PQoidStatus</function> and
<function>fe_setauthsvc</function> are not thread-safe and should not be
used in multi-thread programs.  <function>PQoidStatus</function> can be
replaced by <function>PQoidValue</function>.  There is no good reason to
call <function>fe_setauthsvc</function> at all.
</para>

</sect1>

<sect1>
<title>Sample Programs</title>

<sect2>
<title>Sample Program 1</title>

<para>
<programlisting>
/*
 * testlibpq.c Test the C version of Libpq, the Postgres frontend
 * library.
 *
 *
 */
#include &lt;stdio.h&gt;
#include "libpq-fe.h"

void
exit_nicely(PGconn *conn)
{
    PQfinish(conn);
    exit(1);
}

main()
{
    char       *pghost,
               *pgport,
               *pgoptions,
               *pgtty;
    char       *dbName;
    int         nFields;
    int         i,
                j;

    /* FILE *debug; */

    PGconn     *conn;
    PGresult   *res;

    /*
     * begin, by setting the parameters for a backend connection if the
     * parameters are null, then the system will try to use reasonable
     * defaults by looking up environment variables or, failing that,
     * using hardwired constants
     */
    pghost = NULL;              /* host name of the backend server */
    pgport = NULL;              /* port of the backend server */
    pgoptions = NULL;           /* special options to start up the backend
                                 * server */
    pgtty = NULL;               /* debugging tty for the backend server */
    dbName = "template1";

    /* make a connection to the database */
    conn = PQsetdb(pghost, pgport, pgoptions, pgtty, dbName);

    /*
     * check to see that the backend connection was successfully made
     */
    if (PQstatus(conn) == CONNECTION_BAD)
    {
        fprintf(stderr, "Connection to database '%s' failed.\n", dbName);
        fprintf(stderr, "%s", PQerrorMessage(conn));
        exit_nicely(conn);
    }

    /* debug = fopen("/tmp/trace.out","w"); */
    /* PQtrace(conn, debug);  */

    /* start a transaction block */
    res = PQexec(conn, "BEGIN");
    if (!res || PQresultStatus(res) != PGRES_COMMAND_OK)
    {
        fprintf(stderr, "BEGIN command failed\n");
        PQclear(res);
        exit_nicely(conn);
    }

    /*
     * should PQclear PGresult whenever it is no longer needed to avoid
     * memory leaks
     */
    PQclear(res);

    /*
     * fetch instances from the pg_database, the system catalog of
     * databases
     */
    res = PQexec(conn, "DECLARE mycursor CURSOR FOR select * from pg_database");
    if (!res || PQresultStatus(res) != PGRES_COMMAND_OK)
    {
        fprintf(stderr, "DECLARE CURSOR command failed\n");
        PQclear(res);
        exit_nicely(conn);
    }
    PQclear(res);
    res = PQexec(conn, "FETCH ALL in mycursor");
    if (!res || PQresultStatus(res) != PGRES_TUPLES_OK)
    {
        fprintf(stderr, "FETCH ALL command didn't return tuples properly\n");
        PQclear(res);
        exit_nicely(conn);
    }

    /* first, print out the attribute names */
    nFields = PQnfields(res);
    for (i = 0; i &lt; nFields; i++)
        printf("%-15s", PQfname(res, i));
    printf("\n\n");

    /* next, print out the instances */
    for (i = 0; i &lt; PQntuples(res); i++)
    {
        for (j = 0; j &lt; nFields; j++)
            printf("%-15s", PQgetvalue(res, i, j));
        printf("\n");
    }
    PQclear(res);

    /* close the cursor */
    res = PQexec(conn, "CLOSE mycursor");
    PQclear(res);

    /* commit the transaction */
    res = PQexec(conn, "COMMIT");
    PQclear(res);

    /* close the connection to the database and cleanup */
    PQfinish(conn);

    /* fclose(debug); */
}
</programlisting>
</para>
</sect2>

<sect2>
<title>Sample Program 2</title>

<para>
<programlisting>
/*
 * testlibpq2.c
 *  Test of the asynchronous notification interface
 *
 * Start this program, then from psql in another window do
 *   NOTIFY TBL2;
 *
 * Or, if you want to get fancy, try this:
 * Populate a database with the following:
 *
 *   CREATE TABLE TBL1 (i int4);
 *
 *   CREATE TABLE TBL2 (i int4);
 *
 *   CREATE RULE r1 AS ON INSERT TO TBL1 DO
 *     (INSERT INTO TBL2 values (new.i); NOTIFY TBL2);
 *
 * and do
 *
 *   INSERT INTO TBL1 values (10);
 *
 */
#include &lt;stdio.h&gt;
#include "libpq-fe.h"

void
exit_nicely(PGconn *conn)
{
    PQfinish(conn);
    exit(1);
}

main()
{
    char       *pghost,
               *pgport,
               *pgoptions,
               *pgtty;
    char       *dbName;
    int         nFields;
    int         i,
                j;

    PGconn     *conn;
    PGresult   *res;
    PGnotify   *notify;

    /*
     * begin, by setting the parameters for a backend connection if the
     * parameters are null, then the system will try to use reasonable
     * defaults by looking up environment variables or, failing that,
     * using hardwired constants
     */
    pghost = NULL;              /* host name of the backend server */
    pgport = NULL;              /* port of the backend server */
    pgoptions = NULL;           /* special options to start up the backend
                                 * server */
    pgtty = NULL;               /* debugging tty for the backend server */
    dbName = getenv("USER");    /* change this to the name of your test
                                 * database */

    /* make a connection to the database */
    conn = PQsetdb(pghost, pgport, pgoptions, pgtty, dbName);

    /*
     * check to see that the backend connection was successfully made
     */
    if (PQstatus(conn) == CONNECTION_BAD)
    {
        fprintf(stderr, "Connection to database '%s' failed.\n", dbName);
        fprintf(stderr, "%s", PQerrorMessage(conn));
        exit_nicely(conn);
    }

    res = PQexec(conn, "LISTEN TBL2");
    if (!res || PQresultStatus(res) != PGRES_COMMAND_OK)
    {
        fprintf(stderr, "LISTEN command failed\n");
        PQclear(res);
        exit_nicely(conn);
    }

    /*
     * should PQclear PGresult whenever it is no longer needed to avoid
     * memory leaks
     */
    PQclear(res);

    while (1)
    {

        /*
         * wait a little bit between checks; waiting with select()
         * would be more efficient.
         */
        sleep(1);
        /* collect any asynchronous backend messages */
        PQconsumeInput(conn);
        /* check for asynchronous notify messages */
        while ((notify = PQnotifies(conn)) != NULL)
        {
            fprintf(stderr,
                 "ASYNC NOTIFY of '%s' from backend pid '%d' received\n",
                    notify-&gt;relname, notify-&gt;be_pid);
            free(notify);
        }
    }

    /* close the connection to the database and cleanup */
    PQfinish(conn);

}
</programlisting>
</para>
</sect2>

<sect2>
<title>Sample Program 3</title>

<para>
<programlisting>
/*
 * testlibpq3.c Test the C version of Libpq, the Postgres frontend
 * library. tests the binary cursor interface
 *
 *
 *
 * populate a database by doing the following:
 *
 * CREATE TABLE test1 (i int4, d float4, p polygon);
 *
 * INSERT INTO test1 values (1, 3.567, '(3.0, 4.0, 1.0,
 * 2.0)'::polygon);
 *
 * INSERT INTO test1 values (2, 89.05, '(4.0, 3.0, 2.0,
 * 1.0)'::polygon);
 *
 * the expected output is:
 *
 * tuple 0: got i = (4 bytes) 1, d = (4 bytes) 3.567000, p = (4
 * bytes) 2 points   boundbox = (hi=3.000000/4.000000, lo =
 * 1.000000,2.000000) tuple 1: got i = (4 bytes) 2, d = (4 bytes)
 * 89.050003, p = (4 bytes) 2 points   boundbox =
 * (hi=4.000000/3.000000, lo = 2.000000,1.000000)
 *
 *
 */
#include &lt;stdio.h&gt;
#include "libpq-fe.h"
#include "utils/geo-decls.h"    /* for the POLYGON type */

void
exit_nicely(PGconn *conn)
{
    PQfinish(conn);
    exit(1);
}

main()
{
    char       *pghost,
               *pgport,
               *pgoptions,
               *pgtty;
    char       *dbName;
    int         nFields;
    int         i,
                j;
    int         i_fnum,
                d_fnum,
                p_fnum;
    PGconn     *conn;
    PGresult   *res;

    /*
     * begin, by setting the parameters for a backend connection if the
     * parameters are null, then the system will try to use reasonable
     * defaults by looking up environment variables or, failing that,
     * using hardwired constants
     */
    pghost = NULL;              /* host name of the backend server */
    pgport = NULL;              /* port of the backend server */
    pgoptions = NULL;           /* special options to start up the backend
                                 * server */
    pgtty = NULL;               /* debugging tty for the backend server */

    dbName = getenv("USER");    /* change this to the name of your test
                                 * database */

    /* make a connection to the database */
    conn = PQsetdb(pghost, pgport, pgoptions, pgtty, dbName);

    /*
     * check to see that the backend connection was successfully made
     */
    if (PQstatus(conn) == CONNECTION_BAD)
    {
        fprintf(stderr, "Connection to database '%s' failed.\n", dbName);
        fprintf(stderr, "%s", PQerrorMessage(conn));
        exit_nicely(conn);
    }

    /* start a transaction block */
    res = PQexec(conn, "BEGIN");
    if (!res || PQresultStatus(res) != PGRES_COMMAND_OK)
    {
        fprintf(stderr, "BEGIN command failed\n");
        PQclear(res);
        exit_nicely(conn);
    }

    /*
     * should PQclear PGresult whenever it is no longer needed to avoid
     * memory leaks
     */
    PQclear(res);

    /*
     * fetch instances from the pg_database, the system catalog of
     * databases
     */
    res = PQexec(conn, "DECLARE mycursor BINARY CURSOR FOR select * from test1");
    if (!res || PQresultStatus(res) != PGRES_COMMAND_OK)
    {
        fprintf(stderr, "DECLARE CURSOR command failed\n");
        PQclear(res);
        exit_nicely(conn);
    }
    PQclear(res);

    res = PQexec(conn, "FETCH ALL in mycursor");
    if (!res || PQresultStatus(res) != PGRES_TUPLES_OK)
    {
        fprintf(stderr, "FETCH ALL command didn't return tuples properly\n");
        PQclear(res);
        exit_nicely(conn);
    }

    i_fnum = PQfnumber(res, "i");
    d_fnum = PQfnumber(res, "d");
    p_fnum = PQfnumber(res, "p");

    for (i = 0; i &lt; 3; i++)
    {
        printf("type[%d] = %d, size[%d] = %d\n",
               i, PQftype(res, i),
               i, PQfsize(res, i));
    }
    for (i = 0; i &lt; PQntuples(res); i++)
    {
        int        *ival;
        float      *dval;
        int         plen;
        POLYGON    *pval;

        /* we hard-wire this to the 3 fields we know about */
        ival = (int *) PQgetvalue(res, i, i_fnum);
        dval = (float *) PQgetvalue(res, i, d_fnum);
        plen = PQgetlength(res, i, p_fnum);

        /*
         * plen doesn't include the length field so need to
         * increment by VARHDSZ
         */
        pval = (POLYGON *) malloc(plen + VARHDRSZ);
        pval-&gt;size = plen;
        memmove((char *) &amp;pval-&gt;npts, PQgetvalue(res, i, p_fnum), plen);
        printf("tuple %d: got\n", i);
        printf(" i = (%d bytes) %d,\n",
               PQgetlength(res, i, i_fnum), *ival);
        printf(" d = (%d bytes) %f,\n",
               PQgetlength(res, i, d_fnum), *dval);
        printf(" p = (%d bytes) %d points \tboundbox = (hi=%f/%f, lo = %f,%f)\n",
               PQgetlength(res, i, d_fnum),
               pval-&gt;npts,
               pval-&gt;boundbox.xh,
               pval-&gt;boundbox.yh,
               pval-&gt;boundbox.xl,
               pval-&gt;boundbox.yl);
    }
    PQclear(res);

    /* close the cursor */
    res = PQexec(conn, "CLOSE mycursor");
    PQclear(res);

    /* commit the transaction */
    res = PQexec(conn, "COMMIT");
    PQclear(res);

    /* close the connection to the database and cleanup */
    PQfinish(conn);

}
</programlisting>
</para>

</sect2>
</sect1>
</chapter>

<!-- Keep this comment at the end of the file
Local variables:
mode:sgml
sgml-omittag:nil
sgml-shorttag:t
sgml-minimize-attributes:nil
sgml-always-quote-attributes:t
sgml-indent-step:1
sgml-indent-data:t
sgml-parent-document:nil
sgml-default-dtd-file:"./reference.ced"
sgml-exposed-tags:nil
sgml-local-catalogs:("/usr/lib/sgml/CATALOG")
sgml-local-ecat-files:nil
End:
-->