<?xml version="1.0" encoding="UTF-8" standalone="no"?>
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"><html xmlns="http://www.w3.org/1999/xhtml"><head><meta http-equiv="Content-Type" content="text/html; charset=UTF-8" /><title>34.4. Asynchronous Command Processing</title><link rel="stylesheet" type="text/css" href="stylesheet.css" /><link rev="made" href="pgsql-docs@lists.postgresql.org" /><meta name="generator" content="DocBook XSL Stylesheets Vsnapshot" /><link rel="prev" href="libpq-exec.html" title="34.3. Command Execution Functions" /><link rel="next" href="libpq-single-row-mode.html" title="34.5. Retrieving Query Results Row-By-Row" /></head><body><div xmlns="http://www.w3.org/TR/xhtml1/transitional" class="navheader"><table width="100%" summary="Navigation header"><tr><th colspan="5" align="center">34.4. Asynchronous Command Processing</th></tr><tr><td width="10%" align="left"><a accesskey="p" href="libpq-exec.html" title="34.3. Command Execution Functions">Prev</a> </td><td width="10%" align="left"><a accesskey="u" href="libpq.html" title="Chapter 34. libpq - C Library">Up</a></td><th width="60%" align="center">Chapter 34. <span xmlns="http://www.w3.org/1999/xhtml" class="application">libpq</span> - C Library</th><td width="10%" align="right"><a accesskey="h" href="index.html" title="PostgreSQL 11.5 Documentation">Home</a></td><td width="10%" align="right"> <a accesskey="n" href="libpq-single-row-mode.html" title="34.5. Retrieving Query Results Row-By-Row">Next</a></td></tr></table><hr></hr></div><div class="sect1" id="LIBPQ-ASYNC"><div class="titlepage"><div><div><h2 class="title" style="clear: both">34.4. Asynchronous Command Processing</h2></div></div></div><a id="id-1.7.3.11.2" class="indexterm"></a><p>
The <code class="function">PQexec</code> function is adequate for submitting
commands in normal, synchronous applications. It has a few
deficiencies, however, that can be of importance to some users:
</p><div class="itemizedlist"><ul class="itemizedlist" style="list-style-type: disc; "><li class="listitem"><p>
<code class="function">PQexec</code> waits for the command to be completed.
The application might have other work to do (such as maintaining a
user interface), in which case it won't want to block waiting for
the response.
</p></li><li class="listitem"><p>
Since the execution of the client application is suspended while it
waits for the result, it is hard for the application to decide that
it would like to try to cancel the ongoing command. (It can be done
from a signal handler, but not otherwise.)
</p></li><li class="listitem"><p>
<code class="function">PQexec</code> can return only one
<code class="structname">PGresult</code> structure. If the submitted command
string contains multiple <acronym class="acronym">SQL</acronym> commands, all but
the last <code class="structname">PGresult</code> are discarded by
<code class="function">PQexec</code>.
</p></li><li class="listitem"><p>
<code class="function">PQexec</code> always collects the command's entire result,
buffering it in a single <code class="structname">PGresult</code>. While
this simplifies error-handling logic for the application, it can be
impractical for results containing many rows.
</p></li></ul></div><p>
</p><p>
Applications that do not like these limitations can instead use the
underlying functions that <code class="function">PQexec</code> is built from:
<code class="function">PQsendQuery</code> and <code class="function">PQgetResult</code>.
There are also
<code class="function">PQsendQueryParams</code>,
<code class="function">PQsendPrepare</code>,
<code class="function">PQsendQueryPrepared</code>,
<code class="function">PQsendDescribePrepared</code>, and
<code class="function">PQsendDescribePortal</code>,
which can be used with <code class="function">PQgetResult</code> to duplicate
the functionality of
<code class="function">PQexecParams</code>,
<code class="function">PQprepare</code>,
<code class="function">PQexecPrepared</code>,
<code class="function">PQdescribePrepared</code>, and
<code class="function">PQdescribePortal</code>
respectively.
</p><div class="variablelist"><dl class="variablelist"><dt id="LIBPQ-PQSENDQUERY"><span class="term">
<code class="function">PQsendQuery</code>
<a id="id-1.7.3.11.4.15.1.1.2" class="indexterm"></a>
</span></dt><dd><p>
Submits a command to the server without waiting for the result(s).
1 is returned if the command was successfully dispatched and 0 if
not (in which case, use <code class="function">PQerrorMessage</code> to get more
information about the failure).
</p><pre class="synopsis">
int PQsendQuery(PGconn *conn, const char *command);
</pre><p>
After successfully calling <code class="function">PQsendQuery</code>, call
<code class="function">PQgetResult</code> one or more times to obtain the
results. <code class="function">PQsendQuery</code> cannot be called again
(on the same connection) until <code class="function">PQgetResult</code>
has returned a null pointer, indicating that the command is done.
</p></dd><dt id="LIBPQ-PQSENDQUERYPARAMS"><span class="term">
<code class="function">PQsendQueryParams</code>
<a id="id-1.7.3.11.4.15.2.1.2" class="indexterm"></a>
</span></dt><dd><p>
Submits a command and separate parameters to the server without
waiting for the result(s).
</p><pre class="synopsis">
int PQsendQueryParams(PGconn *conn,
const char *command,
int nParams,
const Oid *paramTypes,
const char * const *paramValues,
const int *paramLengths,
const int *paramFormats,
int resultFormat);
</pre><p>
This is equivalent to <code class="function">PQsendQuery</code> except that
query parameters can be specified separately from the query string.
The function's parameters are handled identically to
<code class="function">PQexecParams</code>. Like
<code class="function">PQexecParams</code>, it will not work on 2.0-protocol
connections, and it allows only one command in the query string.
</p></dd><dt id="LIBPQ-PQSENDPREPARE"><span class="term">
<code class="function">PQsendPrepare</code>
<a id="id-1.7.3.11.4.15.3.1.2" class="indexterm"></a>
</span></dt><dd><p>
Sends a request to create a prepared statement with the given
parameters, without waiting for completion.
</p><pre class="synopsis">
int PQsendPrepare(PGconn *conn,
const char *stmtName,
const char *query,
int nParams,
const Oid *paramTypes);
</pre><p>
This is an asynchronous version of <code class="function">PQprepare</code>: it
returns 1 if it was able to dispatch the request, and 0 if not.
After a successful call, call <code class="function">PQgetResult</code> to
determine whether the server successfully created the prepared
statement. The function's parameters are handled identically to
<code class="function">PQprepare</code>. Like
<code class="function">PQprepare</code>, it will not work on 2.0-protocol
connections.
</p></dd><dt id="LIBPQ-PQSENDQUERYPREPARED"><span class="term">
<code class="function">PQsendQueryPrepared</code>
<a id="id-1.7.3.11.4.15.4.1.2" class="indexterm"></a>
</span></dt><dd><p>
Sends a request to execute a prepared statement with given
parameters, without waiting for the result(s).
</p><pre class="synopsis">
int PQsendQueryPrepared(PGconn *conn,
const char *stmtName,
int nParams,
const char * const *paramValues,
const int *paramLengths,
const int *paramFormats,
int resultFormat);
</pre><p>
This is similar to <code class="function">PQsendQueryParams</code>, but
the command to be executed is specified by naming a
previously-prepared statement, instead of giving a query string.
The function's parameters are handled identically to
<code class="function">PQexecPrepared</code>. Like
<code class="function">PQexecPrepared</code>, it will not work on
2.0-protocol connections.
</p></dd><dt id="LIBPQ-PQSENDDESCRIBEPREPARED"><span class="term">
<code class="function">PQsendDescribePrepared</code>
<a id="id-1.7.3.11.4.15.5.1.2" class="indexterm"></a>
</span></dt><dd><p>
Submits a request to obtain information about the specified
prepared statement, without waiting for completion.
</p><pre class="synopsis">
int PQsendDescribePrepared(PGconn *conn, const char *stmtName);
</pre><p>
This is an asynchronous version of <code class="function">PQdescribePrepared</code>:
it returns 1 if it was able to dispatch the request, and 0 if not.
After a successful call, call <code class="function">PQgetResult</code> to
obtain the results. The function's parameters are handled
identically to <code class="function">PQdescribePrepared</code>. Like
<code class="function">PQdescribePrepared</code>, it will not work on
2.0-protocol connections.
</p></dd><dt id="LIBPQ-PQSENDDESCRIBEPORTAL"><span class="term">
<code class="function">PQsendDescribePortal</code>
<a id="id-1.7.3.11.4.15.6.1.2" class="indexterm"></a>
</span></dt><dd><p>
Submits a request to obtain information about the specified
portal, without waiting for completion.
</p><pre class="synopsis">
int PQsendDescribePortal(PGconn *conn, const char *portalName);
</pre><p>
This is an asynchronous version of <code class="function">PQdescribePortal</code>:
it returns 1 if it was able to dispatch the request, and 0 if not.
After a successful call, call <code class="function">PQgetResult</code> to
obtain the results. The function's parameters are handled
identically to <code class="function">PQdescribePortal</code>. Like
<code class="function">PQdescribePortal</code>, it will not work on
2.0-protocol connections.
</p></dd><dt id="LIBPQ-PQGETRESULT"><span class="term">
<code class="function">PQgetResult</code>
<a id="id-1.7.3.11.4.15.7.1.2" class="indexterm"></a>
</span></dt><dd><p>
Waits for the next result from a prior
<code class="function">PQsendQuery</code>,
<code class="function">PQsendQueryParams</code>,
<code class="function">PQsendPrepare</code>,
<code class="function">PQsendQueryPrepared</code>,
<code class="function">PQsendDescribePrepared</code>, or
<code class="function">PQsendDescribePortal</code>
call, and returns it.
A null pointer is returned when the command is complete and there
will be no more results.
</p><pre class="synopsis">
PGresult *PQgetResult(PGconn *conn);
</pre><p>
</p><p>
<code class="function">PQgetResult</code> must be called repeatedly until
it returns a null pointer, indicating that the command is done.
(If called when no command is active,
<code class="function">PQgetResult</code> will just return a null pointer
at once.) Each non-null result from
<code class="function">PQgetResult</code> should be processed using the
same <code class="structname">PGresult</code> accessor functions previously
described. Don't forget to free each result object with
<code class="function">PQclear</code> when done with it. Note that
<code class="function">PQgetResult</code> will block only if a command is
active and the necessary response data has not yet been read by
<code class="function">PQconsumeInput</code>.
</p><div class="note"><h3 class="title">Note</h3><p>
Even when <code class="function">PQresultStatus</code> indicates a fatal
error, <code class="function">PQgetResult</code> should be called until it
returns a null pointer, to allow <span class="application">libpq</span> to
process the error information completely.
</p></div></dd></dl></div><p>
</p><p>
Using <code class="function">PQsendQuery</code> and
<code class="function">PQgetResult</code> solves one of
<code class="function">PQexec</code>'s problems: If a command string contains
multiple <acronym class="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 client can be handling the results of one
command while the server is still working on later queries in the same
command string.)
</p><p>
Another frequently-desired feature that can be obtained with
<code class="function">PQsendQuery</code> and <code class="function">PQgetResult</code>
is retrieving large query results a row at a time. This is discussed
in <a class="xref" href="libpq-single-row-mode.html" title="34.5. Retrieving Query Results Row-By-Row">Section 34.5</a>.
</p><p>
By itself, calling <code class="function">PQgetResult</code>
will still cause the client to block until the server completes the
next <acronym class="acronym">SQL</acronym> command. This can be avoided by proper
use of two more functions:
</p><div class="variablelist"><dl class="variablelist"><dt id="LIBPQ-PQCONSUMEINPUT"><span class="term">
<code class="function">PQconsumeInput</code>
<a id="id-1.7.3.11.7.3.1.1.2" class="indexterm"></a>
</span></dt><dd><p>
If input is available from the server, consume it.
</p><pre class="synopsis">
int PQconsumeInput(PGconn *conn);
</pre><p>
</p><p>
<code class="function">PQconsumeInput</code> normally returns 1 indicating
<span class="quote">“<span class="quote">no error</span>”</span>, but returns 0 if there was some kind of
trouble (in which case <code class="function">PQerrorMessage</code> can be
consulted). Note that the result does not say whether any input
data was actually collected. After calling
<code class="function">PQconsumeInput</code>, the application can check
<code class="function">PQisBusy</code> and/or
<code class="function">PQnotifies</code> to see if their state has changed.
</p><p>
<code class="function">PQconsumeInput</code> can be called even if the
application is not prepared to deal with a result or notification
just yet. The function will read available data and save it in
a buffer, thereby causing a <code class="function">select()</code>
read-ready indication to go away. The application can thus use
<code class="function">PQconsumeInput</code> to clear the
<code class="function">select()</code> condition immediately, and then
examine the results at leisure.
</p></dd><dt id="LIBPQ-PQISBUSY"><span class="term">
<code class="function">PQisBusy</code>
<a id="id-1.7.3.11.7.3.2.1.2" class="indexterm"></a>
</span></dt><dd><p>
Returns 1 if a command is busy, that is,
<code class="function">PQgetResult</code> would block waiting for input.
A 0 return indicates that <code class="function">PQgetResult</code> can be
called with assurance of not blocking.
</p><pre class="synopsis">
int PQisBusy(PGconn *conn);
</pre><p>
</p><p>
<code class="function">PQisBusy</code> will not itself attempt to read data
from the server; therefore <code class="function">PQconsumeInput</code>
must be invoked first, or the busy state will never end.
</p></dd></dl></div><p>
</p><p>
A typical application using these functions will have a main loop that
uses <code class="function">select()</code> or <code class="function">poll()</code> to wait for
all the conditions that it must respond to. One of the conditions
will be input available from the server, which in terms of
<code class="function">select()</code> means readable data on the file
descriptor identified by <code class="function">PQsocket</code>. When the main
loop detects input ready, it should call
<code class="function">PQconsumeInput</code> to read the input. It can then
call <code class="function">PQisBusy</code>, followed by
<code class="function">PQgetResult</code> if <code class="function">PQisBusy</code>
returns false (0). It can also call <code class="function">PQnotifies</code>
to detect <code class="command">NOTIFY</code> messages (see <a class="xref" href="libpq-notify.html" title="34.8. Asynchronous Notification">Section 34.8</a>).
</p><p>
A client that uses
<code class="function">PQsendQuery</code>/<code class="function">PQgetResult</code>
can also attempt to cancel a command that is still being processed
by the server; see <a class="xref" href="libpq-cancel.html" title="34.6. Canceling Queries in Progress">Section 34.6</a>. But regardless of
the return value of <code class="function">PQcancel</code>, the application
must continue with the normal result-reading sequence using
<code class="function">PQgetResult</code>. A successful cancellation will
simply cause the command to terminate sooner than it would have
otherwise.
</p><p>
By using the functions described above, it is possible to avoid
blocking while waiting for input from the database server. However,
it is still possible that the application will block waiting to send
output to the server. This is relatively uncommon but can happen if
very long SQL commands or data values are sent. (It is much more
probable if the application sends data via <code class="command">COPY IN</code>,
however.) To prevent this possibility and achieve completely
nonblocking database operation, the following additional functions
can be used.
</p><div class="variablelist"><dl class="variablelist"><dt id="LIBPQ-PQSETNONBLOCKING"><span class="term">
<code class="function">PQsetnonblocking</code>
<a id="id-1.7.3.11.10.2.1.1.2" class="indexterm"></a>
</span></dt><dd><p>
Sets the nonblocking status of the connection.
</p><pre class="synopsis">
int PQsetnonblocking(PGconn *conn, int arg);
</pre><p>
</p><p>
Sets the state of the connection to nonblocking if
<em class="parameter"><code>arg</code></em> is 1, or blocking if
<em class="parameter"><code>arg</code></em> is 0. Returns 0 if OK, -1 if error.
</p><p>
In the nonblocking state, calls to
<code class="function">PQsendQuery</code>, <code class="function">PQputline</code>,
<code class="function">PQputnbytes</code>, <code class="function">PQputCopyData</code>,
and <code class="function">PQendcopy</code> will not block but instead return
an error if they need to be called again.
</p><p>
Note that <code class="function">PQexec</code> does not honor nonblocking
mode; if it is called, it will act in blocking fashion anyway.
</p></dd><dt id="LIBPQ-PQISNONBLOCKING"><span class="term">
<code class="function">PQisnonblocking</code>
<a id="id-1.7.3.11.10.2.2.1.2" class="indexterm"></a>
</span></dt><dd><p>
Returns the blocking status of the database connection.
</p><pre class="synopsis">
int PQisnonblocking(const PGconn *conn);
</pre><p>
</p><p>
Returns 1 if the connection is set to nonblocking mode and 0 if
blocking.
</p></dd><dt id="LIBPQ-PQFLUSH"><span class="term">
<code class="function">PQflush</code>
<a id="id-1.7.3.11.10.2.3.1.2" class="indexterm"></a>
</span></dt><dd><p>
Attempts to flush any queued output data to the server. Returns
0 if successful (or if the send queue is empty), -1 if it failed
for some reason, or 1 if it was unable to send all the data in
the send queue yet (this case can only occur if the connection
is nonblocking).
</p><pre class="synopsis">
int PQflush(PGconn *conn);
</pre><p>
</p></dd></dl></div><p>
</p><p>
After sending any command or data on a nonblocking connection, call
<code class="function">PQflush</code>. If it returns 1, wait for the socket
to become read- or write-ready. If it becomes write-ready, call
<code class="function">PQflush</code> again. If it becomes read-ready, call
<code class="function">PQconsumeInput</code>, then call
<code class="function">PQflush</code> again. Repeat until
<code class="function">PQflush</code> returns 0. (It is necessary to check for
read-ready and drain the input with <code class="function">PQconsumeInput</code>,
because the server can block trying to send us data, e.g. NOTICE
messages, and won't read our data until we read its.) Once
<code class="function">PQflush</code> returns 0, wait for the socket to be
read-ready and then read the response as described above.
</p></div><div class="navfooter"><hr /><table width="100%" summary="Navigation footer"><tr><td width="40%" align="left"><a accesskey="p" href="libpq-exec.html">Prev</a> </td><td width="20%" align="center"><a accesskey="u" href="libpq.html">Up</a></td><td width="40%" align="right"> <a accesskey="n" href="libpq-single-row-mode.html">Next</a></td></tr><tr><td width="40%" align="left" valign="top">34.3. Command Execution Functions </td><td width="20%" align="center"><a accesskey="h" href="index.html">Home</a></td><td width="40%" align="right" valign="top"> 34.5. Retrieving Query Results Row-By-Row</td></tr></table></div></body></html>
