CancelConnection class

Requests that the server abandons processing of the current command in a blocking manner.

If the cancel request wasn’t successfully dispatched an error message is raised.

Successful dispatch of the cancellation is no guarantee that the request will have any effect, however. If the cancellation is effective, the command being canceled will terminate early and raises an error. If the cancellation fails (say, because the server was already done processing the command), then there will be no visible result at all.

# File lib/pg/cancel_connection.rb, line 39
def cancel
        start
        polling_loop(:poll)
end

Returns the error message most recently generated by an operation on the cancel connection.

Nearly all PG::CancelConnection functions will set a message if they fail. Note that by libpq convention, a nonempty error_message result can consist of multiple lines, and will include a trailing newline.

static VALUE
pg_cancon_error_message(VALUE self)
{
        PGcancelConn *conn = pg_cancon_get_conn(self);
        char *p_err;

        p_err = PQcancelErrorMessage(conn);

        return p_err ? rb_str_new_cstr(p_err) : Qnil;
}

Closes the cancel connection (if it did not finish sending the cancel request yet). Also frees memory used by the PG::CancelConnection object.

static VALUE
pg_cancon_finish(VALUE self)
{
        t_pg_cancon *this = pg_cancon_get_this( self );

        pg_cancon_close_socket_io( self );
        if( this->pg_cancon )
                PQcancelFinish(this->pg_cancon);
        this->pg_cancon = NULL;

        return Qnil;
}

This is to poll libpq so that it can proceed with the cancel connection sequence.

The behavior is the same like PG::Connection#connect_poll .

See also corresponding libpq function

static VALUE
pg_cancon_poll(VALUE self)
{
        PostgresPollingStatusType status;
        PGcancelConn *conn = pg_cancon_get_conn(self);

        pg_cancon_close_socket_io( self );
        status = gvl_PQcancelPoll(conn);

        return INT2FIX((int)status);
}

Resets the PG::CancelConnection so it can be reused for a new cancel connection.

If the PG::CancelConnection is currently used to send a cancel request, then this connection is closed. It will then prepare the PG::CancelConnection object such that it can be used to send a new cancel request.

This can be used to create one PG::CancelConnection for a PG::Connection and reuse it multiple times throughout the lifetime of the original PG::Connection.

static VALUE
pg_cancon_reset(VALUE self)
{
        PGcancelConn *conn = pg_cancon_get_conn(self);

        pg_cancon_close_socket_io( self );
        PQcancelReset(conn);

        return Qnil;
}

Fetch an IO object created from the CancelConnection’s underlying socket. This object can be used per socket_io.wait_readable, socket_io.wait_writable or for IO.select to wait for events while running asynchronous API calls. IO#wait_*able is Fiber.scheduler compatible in contrast to IO.select.

The IO object can change while the connection is established. So be sure not to cache the IO object, but repeat calling conn.socket_io instead.

static VALUE
pg_cancon_socket_io(VALUE self)
{
        t_pg_cancon *this = pg_cancon_get_this( self );

        if ( !RTEST(this->socket_io) ) {
                int sd;
                if( (sd = PQcancelSocket(this->pg_cancon)) < 0){
                        pg_raise_conn_error( rb_eConnectionBad, self, "PQcancelSocket() can't get socket descriptor");
                }
                return pg_wrap_socket_io( sd, self, &this->socket_io, &this->ruby_sd);
        }

        return this->socket_io;
}

Requests that the server abandons processing of the current command in a non-blocking manner.

The behavior is the same like PG::Connection.connect_start .

Use poll to poll the status of the connection.

static VALUE
pg_cancon_start(VALUE self)
{
        PGcancelConn *conn = pg_cancon_get_conn(self);

        pg_cancon_close_socket_io( self );
        if(gvl_PQcancelStart(conn) == 0)
                pg_raise_conn_error( rb_eConnectionBad, self, "PQcancelStart %s", PQcancelErrorMessage(conn));
        return Qnil;
}

Returns the status of the cancel connection.

The status can be one of a number of values. However, only three of these are seen outside of an asynchronous cancel procedure: CONNECTION_ALLOCATED, CONNECTION_OK and CONNECTION_BAD. The initial state of a PG::CancelConnection that’s successfully created is CONNECTION_ALLOCATED. A cancel request that was successfully dispatched has the status CONNECTION_OK. A failed cancel attempt is signaled by status CONNECTION_BAD. An OK status will remain so until finish or reset is called.

See poll with regards to other status codes that might be returned.

Successful dispatch of the cancellation is no guarantee that the request will have any effect, however. If the cancellation is effective, the command being canceled will terminate early and return an error result. If the cancellation fails (say, because the server was already done processing the command), then there will be no visible result at all.

static VALUE
pg_cancon_status(VALUE self)
{
        ConnStatusType status;
        PGcancelConn *conn = pg_cancon_get_conn(self);

        status = PQcancelStatus(conn);

        return INT2NUM(status);
}

Requests that the server abandons processing of the current command in a blocking manner.

This method directly calls PQcancelBlocking of libpq, so that it doesn’t respond to ruby interrupts and doesn’t trigger the Thread.scheduler . It is threrfore recommended to call cancel instead.

static VALUE
pg_cancon_sync_cancel(VALUE self)
{
        PGcancelConn *conn = pg_cancon_get_conn(self);

        pg_cancon_close_socket_io( self );
        if(gvl_PQcancelBlocking(conn) == 0)
                pg_raise_conn_error( rb_eConnectionBad, self, "PQcancelBlocking %s", PQcancelErrorMessage(conn));
        return Qnil;
}