path: root/doc/src/sgml/libpgtcl.sgml

<chapter id="pgtcl">
 <title><application>pgtcl</application> - Tcl Binding Library</title>

 <indexterm zone="pgtcl">
  <primary>libpgtcl</primary>
 </indexterm>

 <indexterm zone="pgtcl">
  <primary>Tcl</primary>
 </indexterm>

 <sect1 id="pgtcl-intro">
  <title>Introduction</title>

  <para>
   <application>pgtcl</application> is a Tcl package for client
   programs to interface with <ProductName>PostgreSQL</ProductName>
   servers.  It makes most of the functionality of
   <application>libpq</application> available to Tcl scripts.
  </para>

  <para>
   This package was originally written by Jolly Chen.
  </para>

  <para>
   <xref linkend="pgtcl-commands-table"> gives an overview over the
   commands available in <application>pgtcl</application>. These
   commands are described further on subsequent pages.
  </para>


<TABLE TOCENTRY="1" id="pgtcl-commands-table">
<TITLE><literal>pgtcl</literal> Commands</TITLE>
<TGROUP COLS="2">
<THEAD>
  <ROW>
    <ENTRY>Command</ENTRY>
    <ENTRY>Description</ENTRY>
  </ROW>
</THEAD>
<TBODY>
  <ROW>
    <ENTRY><function>pg_connect</function></ENTRY>
    <ENTRY>opens a connection to the backend server</ENTRY>
  </ROW>
  <ROW>
    <ENTRY><function>pg_disconnect</function></ENTRY>
    <ENTRY>closes a connection</ENTRY>
  </ROW>
  <ROW>
    <ENTRY><function>pg_conndefaults</function></ENTRY>
    <ENTRY>get connection options and their defaults</ENTRY>
  </ROW>
  <ROW>
    <ENTRY><function>pg_exec</function></ENTRY>
    <ENTRY>send a query to the backend</ENTRY>
  </ROW>
  <ROW>
    <ENTRY><function>pg_result</function></ENTRY>
    <ENTRY>manipulate the results of a query</ENTRY>
  </ROW>
  <ROW>
    <ENTRY><function>pg_select</function></ENTRY>
    <ENTRY>loop over the result of a SELECT statement</ENTRY>
  </ROW>
  <ROW>
    <ENTRY><function>pg_execute</function></ENTRY>
    <ENTRY>send a query and optionally loop over the results</ENTRY>
  </ROW>
  <ROW>
    <ENTRY><function>pg_listen</function></ENTRY>
    <ENTRY>establish a callback for NOTIFY messages</ENTRY>
  </ROW>
  <ROW>
    <ENTRY><function>pg_on_connection_loss</function></ENTRY>
    <ENTRY>establish a callback for unexpected connection loss</ENTRY>
  </ROW>

  <ROW>
    <ENTRY><function>pg_lo_creat</function></ENTRY>
    <ENTRY>create a large object</ENTRY>
  </ROW>
  <ROW>
    <ENTRY><function>pg_lo_open</function></ENTRY>
    <ENTRY>open a large object</ENTRY>
  </ROW>
  <ROW>
    <ENTRY><function>pg_lo_close</function></ENTRY>
    <ENTRY>close a large object</ENTRY>
  </ROW>
  <ROW>
    <ENTRY><function>pg_lo_read</function></ENTRY>
    <ENTRY>read a large object</ENTRY>
  </ROW>
  <ROW>
    <ENTRY><function>pg_lo_write</function></ENTRY>
    <ENTRY>write a large object</ENTRY>
  </ROW>
  <ROW>
    <ENTRY><function>pg_lo_lseek</function></ENTRY>
    <ENTRY>seek to a position in a large object</ENTRY>
  </ROW>
  <ROW>
    <ENTRY><function>pg_lo_tell</function></ENTRY>
    <ENTRY>return the current seek position of a large object</ENTRY>
  </ROW>
  <ROW>
    <ENTRY><function>pg_lo_unlink</function></ENTRY>
    <ENTRY>delete a large object</ENTRY>
  </ROW>
  <ROW>
    <ENTRY><function>pg_lo_import</function></ENTRY>
    <ENTRY>import a Unix file into a large object</ENTRY>
  </ROW>
  <ROW>
    <ENTRY><function>pg_lo_export</function></ENTRY>
    <ENTRY>export a large object into a Unix file</ENTRY>
  </ROW>
</TBODY>
</TGROUP>
</TABLE>

  <para>
   The <function>pg_lo_*</function> routines are interfaces to the
   large object features of <ProductName>PostgreSQL</ProductName>.
   The functions are designed to mimic the analogous file system
   functions in the standard Unix file system interface.  The
   <function>pg_lo_*</function> routines should be used within a
   <command>BEGIN</command>/<command>COMMIT</command> transaction
   block because the file descriptor returned by
   <function>pg_lo_open</function> is only valid for the current
   transaction.  <function>pg_lo_import</function> and
   <function>pg_lo_export</function> <emphasis>must</emphasis> be used
   in a <command>BEGIN</command>/<command>COMMIT</command> transaction
   block.
  </para>

  <para>
   <xref linkend="pgtcl-example"> shows a small example of how to use
   the routines.
  </para>

  <example id="pgtcl-example">
   <title><application>pgtcl</application> Example Program</title>

<programlisting>
# getDBs :
#   get the names of all the databases at a given host and port number
#   with the defaults being the localhost and port 5432
#   return them in alphabetical order
proc getDBs { {host "localhost"} {port "5432"} } {
    # datnames is the list to be result
    set conn [pg_connect template1 -host $host -port $port]
    set res [pg_exec $conn "SELECT datname FROM pg_database ORDER BY datname"]
    set ntups [pg_result $res -numTuples]
    for {set i 0} {$i < $ntups} {incr i} {
	lappend datnames [pg_result $res -getTuple $i]
    }
    pg_result $res -clear
    pg_disconnect $conn
    return $datnames
}
</programlisting>
  </example>
 </sect1>

<Sect1 id="libpgtcl-loading">
<Title>Loading <application>pgtcl</application> into your application</Title>

   <para>
    Before using <application>pgtcl</application> commands, you must load
    <filename>libpgtcl</> into your Tcl application.  This is normally
    done with the Tcl <literal>load</> command.  Here is an example:

<programlisting>
load libpgtcl[info sharedlibextension]
</programlisting>

    The use of <literal>info sharedlibextension</> is recommended in
    preference to hard-wiring <literal>.so</> or  <literal>.sl</> into
    the program.
   </para>

   <para>
    The <literal>load</> command will fail unless the system's dynamic
    loader knows where to look for the <filename>libpgtcl</> shared
    library file.  You may need to work with <command>ldconfig</>, or
    set the environment variable <envar>LD_LIBRARY_PATH</>, or use
    some equivalent facility for your platform to make it work.  Refer
    to the <productname>PostgreSQL</> installation instructions for
    more information.
   </para>

   <para>
    <filename>libpgtcl</> in turn depends on <filename>libpq</>, so the
    dynamic loader must also be able to find the <filename>libpq</> shared
    library.  In practice this is seldom an issue, since both of these
    shared libraries are normally stored in the same directory, but it
    can be a stumbling block in some configurations.
   </para>

   <para>
    If you use a custom executable for your application, you might choose
    to statically bind <filename>libpgtcl</> into the executable and thereby
    avoid the <literal>load</> command and the potential problems of dynamic
    linking.  See the source code for <application>pgtclsh</> for an example.
   </para>

</Sect1>

<Sect1 id="libpgtcl-ref">
<Title><application>pgtcl</application> Command Reference Information</Title>

<REFENTRY ID="PGTCL-PGCONNECT">
<REFMETA>
<REFENTRYTITLE>pg_connect</REFENTRYTITLE>
<REFMISCINFO>PGTCL - Connection Management</REFMISCINFO>
</REFMETA>
<REFNAMEDIV>
<REFNAME>pg_connect
</REFNAME>
<REFPURPOSE>open a connection to the backend server
</REFPURPOSE>
<INDEXTERM ID="IX-PGTCL-PGCONNECT-1"><PRIMARY>pgtcl</PRIMARY><SECONDARY>connecting</SECONDARY></INDEXTERM>
<INDEXTERM ID="IX-PGTCL-PGCONNECT-2"><PRIMARY>pg_connect</PRIMARY></INDEXTERM>
</REFNAMEDIV>
<REFSYNOPSISDIV>
<REFSYNOPSISDIVINFO>
<DATE>1997-12-24</DATE>
</REFSYNOPSISDIVINFO>
<SYNOPSIS>
pg_connect -conninfo <REPLACEABLE CLASS="PARAMETER">connectOptions</REPLACEABLE>
pg_connect <REPLACEABLE CLASS="PARAMETER">dbName</REPLACEABLE> <OPTIONAL>-host <REPLACEABLE CLASS="PARAMETER">hostName</REPLACEABLE></OPTIONAL>
  <OPTIONAL>-port <REPLACEABLE
  CLASS="PARAMETER">portNumber</REPLACEABLE></OPTIONAL> <OPTIONAL>-tty <REPLACEABLE
  CLASS="PARAMETER">pqtty</REPLACEABLE></OPTIONAL>
  <OPTIONAL>-options <REPLACEABLE
  CLASS="PARAMETER">optionalBackendArgs</REPLACEABLE></OPTIONAL>
</SYNOPSIS>

<REFSECT2 ID="R2-PGTCL-PGCONNECT-1">
<REFSECT2INFO>
<DATE>1998-10-07</DATE>
</REFSECT2INFO>
<TITLE>Inputs (new style)
</TITLE>
<VARIABLELIST>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">connectOptions</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>A string of connection options, each written in the form keyword = value.
A list of valid options can be found in <filename>libpq</>'s
<function>PQconnectdb()</> manual entry.
</PARA>
</LISTITEM>
</VARLISTENTRY>
</VARIABLELIST>
</REFSECT2>

<REFSECT2 ID="R2-PGTCL-PGCONNECT-2">
<REFSECT2INFO>
<DATE>1997-12-24</DATE>
</REFSECT2INFO>
<TITLE>Inputs (old style)
</TITLE>
<VARIABLELIST>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">dbName</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>Specifies a valid database name.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
  <OPTIONAL>-host <REPLACEABLE CLASS="PARAMETER">hostName</REPLACEABLE></OPTIONAL>
</TERM>
<LISTITEM>
<PARA>Specifies the domain name of the backend server for <REPLACEABLE CLASS="PARAMETER">dbName</REPLACEABLE>.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
  <OPTIONAL>-port <REPLACEABLE CLASS="PARAMETER">portNumber</REPLACEABLE></OPTIONAL>
</TERM>
<LISTITEM>
<PARA>Specifies the IP port number of the backend server for <REPLACEABLE CLASS="PARAMETER">dbName</REPLACEABLE>.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
  <OPTIONAL>-tty <REPLACEABLE CLASS="PARAMETER">pqtty</REPLACEABLE></OPTIONAL>
</TERM>
<LISTITEM>
<PARA>Specifies file or <acronym>tty</acronym> for optional debug output from backend.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
  <OPTIONAL>-options <REPLACEABLE CLASS="PARAMETER">optionalBackendArgs</REPLACEABLE></OPTIONAL>
</TERM>
<LISTITEM>
<PARA>Specifies options for the backend server for <REPLACEABLE CLASS="PARAMETER">dbName</REPLACEABLE>.
</PARA>
</LISTITEM>
</VARLISTENTRY>
</VARIABLELIST>
</REFSECT2>

<REFSECT2 ID="R2-PGTCL-PGCONNECT-3">
<REFSECT2INFO>
<DATE>1997-12-24</DATE>
</REFSECT2INFO>
<TITLE>Outputs
</TITLE>
<VARIABLELIST>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">dbHandle</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>
If successful, a handle for a database connection is returned.
Handles start with the prefix <literal>pgsql</literal>.
</PARA>
</LISTITEM>
</VARLISTENTRY>
</VARIABLELIST>
</REFSECT2>
</REFSYNOPSISDIV>

<!-- ********************************************************** -->

<REFSECT1 ID="R1-PGTCL-PGCONNECT-1">
<REFSECT1INFO>
<DATE>1997-12-24</DATE>
</REFSECT1INFO>
<TITLE>Description
</TITLE>
<PARA><FUNCTION>pg_connect</FUNCTION> opens a connection to the
<ProductName>PostgreSQL</ProductName> backend.
</Para>

<para>
Two syntaxes are available.  In the older one, each possible option
has a separate option switch in the pg_connect statement.  In the
newer form, a single option string is supplied that can contain
multiple option values.  See <FUNCTION>pg_conndefaults</FUNCTION>
for info about the available options in the newer syntax.
</PARA>
</REFSECT1>
<REFSECT1 ID="R1-PGTCL-PGCONNECT-2">
<TITLE>Usage
</TITLE>
<comment>
 XXX thomas 1997-12-24
</comment>
</REFSECT1>
</REFENTRY>

<REFENTRY ID="PGTCL-PGDISCONNECT">
<REFMETA>
<REFENTRYTITLE>pg_disconnect</REFENTRYTITLE>
<REFMISCINFO>PGTCL - Connection Management</REFMISCINFO>
</REFMETA>
<REFNAMEDIV>
<REFNAME>pg_disconnect
</REFNAME>
<REFPURPOSE>close a connection to the backend server
</REFPURPOSE>
<INDEXTERM ID="IX-PGTCL-PGDISCONNECT-1"><PRIMARY>pgtcl</PRIMARY><SECONDARY>connecting</SECONDARY></INDEXTERM>
<INDEXTERM ID="IX-PGTCL-PGDISCONNECT-2"><PRIMARY>pg_connect</PRIMARY></INDEXTERM>
</REFNAMEDIV>
<REFSYNOPSISDIV>
<REFSYNOPSISDIVINFO>
<DATE>1997-12-24</DATE>
</REFSYNOPSISDIVINFO>
<SYNOPSIS>
pg_disconnect <REPLACEABLE CLASS="PARAMETER">dbHandle</REPLACEABLE>
</SYNOPSIS>

<REFSECT2 ID="R2-PGTCL-PGDISCONNECT-1">
<REFSECT2INFO>
<DATE>1997-12-24</DATE>
</REFSECT2INFO>
<TITLE>Inputs
</TITLE>
<VARIABLELIST>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">dbHandle</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>Specifies a valid database handle.
</PARA>
</LISTITEM>
</VARLISTENTRY>
</VARIABLELIST>
</REFSECT2>

<REFSECT2 ID="R2-PGTCL-PGDISCONNECT-2">
<REFSECT2INFO>
<DATE>1997-12-24</DATE>
</REFSECT2INFO>
<TITLE>Outputs
</TITLE>
<VARIABLELIST>
<VARLISTENTRY>
<TERM>
  None
</TERM>
<LISTITEM>
<PARA>
</PARA>
</LISTITEM>
</VARLISTENTRY>
</VARIABLELIST>
</REFSECT2>
</REFSYNOPSISDIV>

<REFSECT1 ID="R1-PGTCL-PGDISCONNECT-1">
<REFSECT1INFO>
<DATE>1997-12-24</DATE>
</REFSECT1INFO>
<TITLE>Description
</TITLE>
<PARA><FUNCTION>pg_disconnect</FUNCTION> closes a connection to the <ProductName>PostgreSQL</ProductName> backend.
</PARA>
</REFSECT1>

</REFENTRY>




<REFENTRY ID="PGTCL-PGCONNDEFAULTS">
<REFMETA>
<REFENTRYTITLE>pg_conndefaults</REFENTRYTITLE>
<REFMISCINFO>PGTCL - Connection Management</REFMISCINFO>
</REFMETA>
<REFNAMEDIV>
<REFNAME>pg_conndefaults
</REFNAME>
<REFPURPOSE>obtain information about default connection parameters
</REFPURPOSE>
<INDEXTERM ID="IX-PGTCL-PGCONNDEFAULTS-1"><PRIMARY>pgtcl</PRIMARY><SECONDARY>connecting</SECONDARY></INDEXTERM>
<INDEXTERM ID="IX-PGTCL-PGCONNDEFAULTS-2"><PRIMARY>pg_conndefaults</PRIMARY></INDEXTERM>
</REFNAMEDIV>
<REFSYNOPSISDIV>
<REFSYNOPSISDIVINFO>
<DATE>1998-10-07</DATE>
</REFSYNOPSISDIVINFO>
<SYNOPSIS>
pg_conndefaults
</SYNOPSIS>

<REFSECT2 ID="R2-PGTCL-PGCONNDEFAULTS-1">
<REFSECT2INFO>
<DATE>1998-10-07</DATE>
</REFSECT2INFO>
<TITLE>Inputs
</TITLE>
<PARA>
None.
</PARA>
</REFSECT2>

<REFSECT2 ID="R2-PGTCL-PGCONNDEFAULTS-2">
<REFSECT2INFO>
<DATE>1998-10-07</DATE>
</REFSECT2INFO>
<TITLE>Outputs
</TITLE>
<VARIABLELIST>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">option list</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>
The result is a list describing the possible connection options and their
current default values.
Each entry in the list is a sublist of the format:
</Para>
<screen>
{optname label dispchar dispsize value}
</screen>
<Para>
where the <replaceable>optname</> is usable as an option in
<FUNCTION>pg_connect -conninfo</FUNCTION>.
</PARA>
</LISTITEM>
</VARLISTENTRY>
</VARIABLELIST>
</REFSECT2>
</REFSYNOPSISDIV>

<REFSECT1 ID="R1-PGTCL-PGCONNDEFAULTS-1">
<REFSECT1INFO>
<DATE>1998-10-07</DATE>
</REFSECT1INFO>
<TITLE>Description
</TITLE>

<para>
<FUNCTION>pg_conndefaults</FUNCTION> returns info about the connection
options available in <FUNCTION>pg_connect -conninfo</FUNCTION> and the
current default value for each option.
</PARA>
</REFSECT1>
<REFSECT1 ID="R1-PGTCL-PGCONNDEFAULTS-2">
<TITLE>Usage
</TITLE>
<PARA><literal>pg_conndefaults</>
</PARA>
</REFSECT1>
</REFENTRY>

<REFENTRY ID="PGTCL-PGEXEC">
<REFMETA>
<REFENTRYTITLE>pg_exec</REFENTRYTITLE>
<REFMISCINFO>PGTCL - Query Processing</REFMISCINFO>
</REFMETA>
<REFNAMEDIV>
<REFNAME>pg_exec
</REFNAME>
<REFPURPOSE>
send a command string to the server
</REFPURPOSE>
<INDEXTERM ID="IX-PGTCL-PGEXEC-1"><PRIMARY>pgtcl</PRIMARY><SECONDARY>connecting</SECONDARY></INDEXTERM>
<INDEXTERM ID="IX-PGTCL-PGEXEC-2"><PRIMARY>pg_connect</PRIMARY></INDEXTERM>
</REFNAMEDIV>
<REFSYNOPSISDIV>
<REFSYNOPSISDIVINFO>
<DATE>1997-12-24</DATE>
</REFSYNOPSISDIVINFO>
<SYNOPSIS>
pg_exec <REPLACEABLE CLASS="PARAMETER">dbHandle</REPLACEABLE> <REPLACEABLE CLASS="PARAMETER">queryString</REPLACEABLE>
</SYNOPSIS>

<!-- ********************************************************** -->

<REFSECT2 ID="R2-PGTCL-PGEXEC-1">
<REFSECT2INFO>
<DATE>1997-12-24</DATE>
</REFSECT2INFO>
<TITLE>Inputs
</TITLE>
<VARIABLELIST>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">dbHandle</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>Specifies a valid database handle.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">queryString</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>Specifies a valid SQL query.
</PARA>
</LISTITEM>
</VARLISTENTRY>
</VARIABLELIST>
</REFSECT2>

<REFSECT2 ID="R2-PGTCL-PGEXEC-2">
<REFSECT2INFO>
<DATE>1997-12-24</DATE>
</REFSECT2INFO>
<TITLE>Outputs
</TITLE>
<VARIABLELIST>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">resultHandle</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>
A Tcl error will be returned if <application>pgtcl</application> was unable to obtain a backend
response.  Otherwise, a query result object is created and a handle for
it is returned.  This handle can be passed to <FUNCTION>pg_result</FUNCTION>
to obtain the results of the query.
</PARA>
</LISTITEM>
</VARLISTENTRY>
</VARIABLELIST>
</REFSECT2></REFSYNOPSISDIV>

<REFSECT1 ID="R1-PGTCL-PGEXEC-1">
<REFSECT1INFO>
<DATE>1997-12-24</DATE>
</REFSECT1INFO>
<TITLE>Description
</TITLE>
<PARA>
<FUNCTION>pg_exec</FUNCTION> submits a query to the <ProductName>PostgreSQL</ProductName> backend and returns a result.

Query result handles start with the connection handle and add a period
and a result number.
</Para>

<PARA>
Note that lack of a Tcl error is not proof that the query succeeded!
An error message returned by the backend will be processed
as a query result with failure status, not by generating a Tcl error
in pg_exec.
</PARA>
</REFSECT1>
</refentry>

<REFENTRY ID="PGTCL-PGRESULT">
<REFMETA>
<REFENTRYTITLE>pg_result</REFENTRYTITLE>
<REFMISCINFO>PGTCL - Query Processing</REFMISCINFO>
</REFMETA>
<REFNAMEDIV>
<REFNAME>pg_result
</REFNAME>
<REFPURPOSE>
get information about a query result
</REFPURPOSE>
<INDEXTERM ID="IX-PGTCL-PGRESULT-1"><PRIMARY>pgtcl</PRIMARY><SECONDARY>connecting</SECONDARY></INDEXTERM>
<INDEXTERM ID="IX-PGTCL-PGRESULT-2"><PRIMARY>pg_connect</PRIMARY></INDEXTERM>
</REFNAMEDIV>
<REFSYNOPSISDIV>
<REFSYNOPSISDIVINFO>
<DATE>1997-12-24</DATE>
</REFSYNOPSISDIVINFO>
<SYNOPSIS>
pg_result <REPLACEABLE CLASS="PARAMETER">resultHandle</REPLACEABLE> <REPLACEABLE CLASS="PARAMETER">resultOption</REPLACEABLE>
</SYNOPSIS>
<REFSECT2 ID="R2-PGTCL-PGRESULT-1">
<REFSECT2INFO>
<DATE>1997-12-24</DATE>
</REFSECT2INFO>
<TITLE>Inputs
</TITLE>
<VARIABLELIST>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">resultHandle</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>
  The handle for a query result.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">resultOption</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>
Specifies one of several possible options.
</PARA>
</LISTITEM>
</VARLISTENTRY>
</VARIABLELIST>

<REFSECT3>
<TITLE>Options</TITLE>

<VARIABLELIST>
<VARLISTENTRY>
<TERM>
<option>-status</option>
</TERM>
<LISTITEM>
<PARA>
the status of the result.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
<option>-error</option>
</TERM>
<LISTITEM>
<PARA>
the error message, if the status indicates error; otherwise an empty string.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
<option>-conn</option>
</TERM>
<LISTITEM>
<PARA>
the connection that produced the result.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
<option>-oid</option>
</TERM>
<LISTITEM>
<PARA>
if the command was an INSERT, the OID of the 
inserted tuple; otherwise 0.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
<option>-numTuples</option>
</TERM>
<LISTITEM>
<PARA>
the number of tuples returned by the query.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
<option>-numAttrs</option>
</TERM>
<LISTITEM>
<PARA>
the number of attributes in each tuple.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
<option>-assign arrayName</option>
</TERM>
<LISTITEM>
<PARA>
assign the results to an array, using subscripts of the form
<literal>(tupno,attributeName)</literal>.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
<option>-assignbyidx arrayName ?appendstr?</option>
</TERM>
<LISTITEM>
<PARA>
assign the results to an array using the first attribute's value and
the remaining attributes' names as keys.  If <parameter>appendstr</> is given then
it is appended to each key.  In short, all but the first field of each
tuple are stored into the array, using subscripts of the form
<literal>(firstFieldValue,fieldNameAppendStr)</literal>.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
<option>-getTuple tupleNumber</option>
</TERM>
<LISTITEM>
<PARA>
returns the fields of the indicated tuple in a list.  Tuple numbers
start at zero.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
<option>-tupleArray tupleNumber arrayName</option>
</TERM>
<LISTITEM>
<PARA>
stores the fields of the tuple in array <parameter>arrayName</parameter>, indexed by field names.
Tuple numbers start at zero.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
<option>-attributes</option>
</TERM>
<LISTITEM>
<PARA>
returns a list of the names of the tuple attributes.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
<option>-lAttributes</option>
</TERM>
<LISTITEM>
<PARA>
returns a list of sublists, <literal>{name ftype fsize}</literal> for each tuple attribute.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
<option>-clear</option>
</TERM>
<LISTITEM>
<PARA>
clear the result query object.
</PARA>
</LISTITEM>
</VARLISTENTRY>
</VARIABLELIST>
</REFSECT3>
</REFSECT2>

<REFSECT2 ID="R2-PGTCL-PGRESULT-2">
<REFSECT2INFO>
<DATE>1997-12-24</DATE>
</REFSECT2INFO>
<TITLE>Outputs
</TITLE>
<PARA>
The result depends on the selected option, as described above.
</PARA>
</REFSECT2></REFSYNOPSISDIV>

<REFSECT1 ID="R1-PGTCL-PGRESULT-1">
<REFSECT1INFO>
<DATE>1997-12-24</DATE>
</REFSECT1INFO>
<TITLE>Description
</TITLE>
<PARA>
<FUNCTION>pg_result</FUNCTION> returns information about a query result
created by a prior <FUNCTION>pg_exec</FUNCTION>.
</Para>

<para>
You can keep a query result around for as long as you need it, but when
you are done with it, be sure to free it by
executing <FUNCTION>pg_result -clear</FUNCTION>.  Otherwise, you have
a memory leak, and <application>Pgtcl</> will eventually start complaining that you've
created too many query result objects.
</PARA>
</REFSECT1>
</REFENTRY>

<!-- ********************************************************** -->

<REFENTRY ID="PGTCL-PGSELECT">
<REFMETA>
<REFENTRYTITLE>pg_select</REFENTRYTITLE>
<REFMISCINFO>PGTCL - Query Processing</REFMISCINFO>
</REFMETA>
<REFNAMEDIV>
<REFNAME>pg_select
</REFNAME>
<REFPURPOSE>
loop over the result of a SELECT statement
</REFPURPOSE>
<INDEXTERM ID="IX-PGTCL-PGSELECT-1"><PRIMARY>pgtcl</PRIMARY><SECONDARY>connecting</SECONDARY></INDEXTERM>
<INDEXTERM ID="IX-PGTCL-PGSELECT-2"><PRIMARY>pg_connect</PRIMARY></INDEXTERM>
</REFNAMEDIV>
<REFSYNOPSISDIV>
<REFSYNOPSISDIVINFO>
<DATE>1997-12-24</DATE>
</REFSYNOPSISDIVINFO>
<SYNOPSIS>
pg_select <REPLACEABLE CLASS="PARAMETER">dbHandle</REPLACEABLE> <REPLACEABLE CLASS="PARAMETER">queryString</REPLACEABLE> <REPLACEABLE CLASS="PARAMETER">arrayVar</REPLACEABLE> <REPLACEABLE CLASS="PARAMETER">queryProcedure</REPLACEABLE>
</SYNOPSIS>

<REFSECT2 ID="R2-PGTCL-PGSELECT-1">
<REFSECT2INFO>
<DATE>1997-12-24</DATE>
</REFSECT2INFO>
<TITLE>Inputs
</TITLE>
<VARIABLELIST>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">dbHandle</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>Specifies a valid database handle.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">queryString</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>Specifies a valid SQL select query.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">arrayVar</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>Array variable for tuples returned.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">queryProcedure</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>Procedure run on each tuple found.
</PARA>
</LISTITEM>
</VARLISTENTRY>
</VARIABLELIST>

</REFSECT2>

<REFSECT2 ID="R2-PGTCL-PGSELECT-2">
<REFSECT2INFO>
<DATE>1997-12-24</DATE>
</REFSECT2INFO>
<TITLE>Outputs
</TITLE>
<Para>
None.
</Para>
</REFSECT2></REFSYNOPSISDIV>

<REFSECT1 ID="R1-PGTCL-PGSELECT-1">
<REFSECT1INFO>
<DATE>1997-12-24</DATE>
</REFSECT1INFO>
<TITLE>Description
</TITLE>
<PARA>
<FUNCTION>pg_select</FUNCTION> submits a SELECT query to the
<ProductName>PostgreSQL</ProductName> backend, and executes a
given chunk of code for each tuple in the result.
  The <REPLACEABLE CLASS="PARAMETER">queryString</REPLACEABLE>
  must be a SELECT statement.  Anything else returns an error.
  The <REPLACEABLE CLASS="PARAMETER">arrayVar</REPLACEABLE>
  variable is an array name used in the loop.  For each tuple,
  <REPLACEABLE CLASS="PARAMETER">arrayVar</REPLACEABLE> is filled in
  with the tuple field values, using the field names as the array
  indexes.  Then the
  <REPLACEABLE CLASS="PARAMETER">queryProcedure</REPLACEABLE>
  is executed.
</PARA>

<PARA>
 In addition to the field values, the following special entries are
made in the array:

<VARIABLELIST>
<VARLISTENTRY>
<TERM><literal>.headers</></TERM>
<LISTITEM>
<PARA>A list of the column names returned by the SELECT.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM><literal>.numcols</></TERM>
<LISTITEM>
<PARA>The number of columns returned by the SELECT.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM><literal>.tupno</></TERM>
<LISTITEM>
<PARA>The current tuple number, starting at zero and incrementing
for each iteration of the loop body.
</PARA>
</LISTITEM>
</VARLISTENTRY>
</VARIABLELIST>
</PARA>

</REFSECT1>

<REFSECT1 ID="R1-PGTCL-PGSELECT-2">
<TITLE>Usage
</TITLE>
<PARA>
This would work if table <classname>table</> has fields <structfield>control</> and <structfield>name</>
(and, perhaps, other fields):
<ProgramListing>
	pg_select $pgconn "SELECT * FROM table" array {
		puts [format "%5d %s" $array(control) $array(name)]
	}
</ProgramListing>
</PARA>
</REFSECT1>

</REFENTRY>

<!-- ********************************************************** -->

<REFENTRY ID="PGTCL-PGEXECUTE">
<REFMETA>
<REFENTRYTITLE>pg_execute</REFENTRYTITLE>
<REFMISCINFO>PGTCL - Query Processing</REFMISCINFO>
</REFMETA>
<REFNAMEDIV>
<REFNAME>pg_execute
</REFNAME>
<REFPURPOSE>
send a query and optionally loop over the results
</REFPURPOSE>
<INDEXTERM ID="IX-PGTCL-PGEXECUTE-1"><PRIMARY>pgtcl</PRIMARY><SECONDARY>query</SECONDARY></INDEXTERM>
<INDEXTERM ID="IX-PGTCL-PGEXECUTE-2"><PRIMARY>pg_execute</PRIMARY></INDEXTERM>
</REFNAMEDIV>
<REFSYNOPSISDIV>
<REFSYNOPSISDIVINFO>
<DATE>2002-03-06</DATE>
</REFSYNOPSISDIVINFO>
<SYNOPSIS>
pg_execute <OPTIONAL>-array <REPLACEABLE CLASS="PARAMETER">arrayVar</REPLACEABLE></OPTIONAL> <OPTIONAL>-oid <REPLACEABLE CLASS="PARAMETER">oidVar</REPLACEABLE></OPTIONAL> <REPLACEABLE CLASS="PARAMETER">dbHandle</REPLACEABLE> <REPLACEABLE CLASS="PARAMETER">queryString</REPLACEABLE> <OPTIONAL><REPLACEABLE CLASS="PARAMETER">queryProcedure</REPLACEABLE></OPTIONAL>
</SYNOPSIS>

<REFSECT2 ID="R2-PGTCL-PGEXECUTE-1">
<REFSECT2INFO>
<DATE>2002-03-06</DATE>
</REFSECT2INFO>
<TITLE>Inputs
</TITLE>
<VARIABLELIST>
<VARLISTENTRY>
<TERM>
  <OPTIONAL>-array <REPLACEABLE CLASS="PARAMETER">arrayVar</REPLACEABLE></OPTIONAL>
</TERM>
<LISTITEM>
<PARA>Specifies the name of an array variable where result tuples are stored,
indexed by the field names.
This is ignored if <replaceable class=parameter>queryString</> is not a SELECT statement. For SELECT
statements, if this option is not used, result tuples values are stored
in individual variables named according to the field names in the result.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
  <OPTIONAL>-oid <REPLACEABLE CLASS="PARAMETER">oidVar</REPLACEABLE></OPTIONAL>
</TERM>
<LISTITEM>
<PARA>Specifies the name of a variable into which the OID from an INSERT
statement will be stored.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">dbHandle</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>Specifies a valid database handle.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">queryString</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>Specifies a valid SQL query.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
  <OPTIONAL><REPLACEABLE CLASS="PARAMETER">queryProcedure</REPLACEABLE></OPTIONAL>
</TERM>
<LISTITEM>
<PARA>Optional command to execute for each result tuple of a SELECT statement.
</PARA>
</LISTITEM>
</VARLISTENTRY>
</VARIABLELIST>

</REFSECT2>

<REFSECT2 ID="R2-PGTCL-PGEXECUTE-2">
<REFSECT2INFO>
<DATE>2002-03-06</DATE>
</REFSECT2INFO>
<TITLE>Outputs
</TITLE>
<VARIABLELIST>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">ntuples</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>
The number of tuples affected or returned by the query.
</PARA>
</LISTITEM>
</VARLISTENTRY>
</VARIABLELIST>
</REFSECT2></REFSYNOPSISDIV>

<REFSECT1 ID="R1-PGTCL-PGEXECUTE-1">
<REFSECT1INFO>
<DATE>2002-03-06</DATE>
</REFSECT1INFO>
<TITLE>Description
</TITLE>
<PARA>
<FUNCTION>pg_execute</FUNCTION> submits a query to the PostgreSQL backend.
</PARA>
<PARA>
If the query is not a SELECT statement, the query is executed and the
number of tuples affected by the query is returned.  If the query is an
INSERT and a single tuple is inserted, the OID of the inserted tuple is
stored in the <replaceable>oidVar</> variable if the optional <PARAMETER>-oid</PARAMETER>
argument is supplied.
</PARA>
<PARA>
If the query is a SELECT statement, the query is executed. For each tuple
in the result, the tuple field values are stored in the
<PARAMETER>arrayVar</PARAMETER> variable,
if supplied, using the field names as the array indexes, else in variables
named by the field names, and then the optional
<PARAMETER>queryProcedure</PARAMETER> is executed if supplied.
(Omitting the <PARAMETER>queryProcedure</PARAMETER> probably makes sense
only if the query will return a single tuple.)
The number of tuples selected is returned.
</PARA>
<PARA>
The <PARAMETER>queryProcedure</PARAMETER> can use the Tcl
<LITERAL>break</LITERAL>, <LITERAL>continue</LITERAL>, and
<LITERAL>return</LITERAL> commands, with the expected behavior.
Note that if the <PARAMETER>queryProcedure</PARAMETER> executes
<LITERAL>return</LITERAL>, <FUNCTION>pg_execute</FUNCTION> does
not return <PARAMETER>ntuples</PARAMETER>.
</PARA>
<PARA>
<FUNCTION>pg_execute</FUNCTION> is a newer function which provides a
superset of the features of <FUNCTION>pg_select</FUNCTION>, and can
replace <FUNCTION>pg_exec</FUNCTION> in many cases where access to
the result handle is not needed.
</PARA>
<PARA>
For backend-handled errors, <FUNCTION>pg_execute</FUNCTION> will
throw a Tcl error and return two element list. The first element
is an error code such as <LITERAL>PGRES_FATAL_ERROR</LITERAL>, and
the second element is the backend error text. For more serious
errors, such as failure to communicate with the backend,
<FUNCTION>pg_execute</FUNCTION> will throw a Tcl error and return
just the error message text.
</PARA>

</REFSECT1>

<REFSECT1 ID="R1-PGTCL-PGEXECUTE-2">
<TITLE>Usage
</TITLE>
<PARA>
In the following examples, error checking with <LITERAL>catch</LITERAL>
has been omitted for clarity.
</PARA>
<PARA>
Insert a row and save the OID in <varname>result_oid</>:
<ProgramListing>
    pg_execute -oid result_oid $pgconn "insert into mytable values (1)"
</ProgramListing>
</PARA>
<PARA>
Print the item and value fields from each row:
<ProgramListing>
    pg_execute -array d $pgconn "select item, value from mytable" {
       puts "Item=$d(item) Value=$d(value)"
    }
</ProgramListing>
</PARA>
<PARA>
Find the maximum and minimum values and store them in $s(max) and $s(min):
<ProgramListing>
    pg_execute -array s $pgconn "select max(value) as max,\
      min(value) as min from mytable"
</ProgramListing>
</PARA>
<PARA>
Find the maximum and minimum values and store them in $max and $min:
<ProgramListing>
    pg_execute $pgconn "select max(value) as max, min(value) as min from mytable"
</ProgramListing>
</PARA>
</REFSECT1>

</REFENTRY>

<!-- ********************************************************** -->

<REFENTRY ID="PGTCL-PGLISTEN">
<REFMETA>
<REFENTRYTITLE>pg_listen</REFENTRYTITLE>
<REFMISCINFO>PGTCL - Asynchronous Notify</REFMISCINFO>
</REFMETA>
<REFNAMEDIV>
<REFNAME>pg_listen
</REFNAME>
<REFPURPOSE>set or change a callback for asynchronous NOTIFY messages
</REFPURPOSE>
<INDEXTERM ID="IX-PGTCL-PGLISTEN-1"><PRIMARY>pgtcl</PRIMARY><SECONDARY>notify</SECONDARY></INDEXTERM>
<INDEXTERM ID="IX-PGTCL-PGLISTEN-2"><PRIMARY>notify</PRIMARY></INDEXTERM>
</REFNAMEDIV>
<REFSYNOPSISDIV>
<REFSYNOPSISDIVINFO>
<DATE>1998-5-22</DATE>
</REFSYNOPSISDIVINFO>
<SYNOPSIS>
pg_listen <REPLACEABLE CLASS="PARAMETER">dbHandle</REPLACEABLE> <REPLACEABLE CLASS="PARAMETER">notifyName</REPLACEABLE> <REPLACEABLE CLASS="PARAMETER">callbackCommand</REPLACEABLE>
</SYNOPSIS>

<REFSECT2 ID="R2-PGTCL-PGLISTEN-1">
<REFSECT2INFO>
<DATE>1998-5-22</DATE>
</REFSECT2INFO>
<TITLE>Inputs
</TITLE>
<VARIABLELIST>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">dbHandle</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>Specifies a valid database handle.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">notifyName</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>Specifies the notify condition name to start or stop listening to.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">callbackCommand</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>If present, provides the command string to execute
when a matching notification arrives.
</PARA>
</LISTITEM>
</VARLISTENTRY>
</VARIABLELIST>
</REFSECT2>

<REFSECT2 ID="R2-PGTCL-PGLISTEN-2">
<REFSECT2INFO>
<DATE>1998-5-22</DATE>
</REFSECT2INFO>
<TITLE>Outputs
</TITLE>
<VARIABLELIST>
<VARLISTENTRY>
<TERM>
  None
</TERM>
<LISTITEM>
<PARA>
</PARA>
</LISTITEM>
</VARLISTENTRY>
</VARIABLELIST>
</REFSECT2>
</REFSYNOPSISDIV>

<REFSECT1 ID="R1-PGTCL-PGLISTEN-1">
<REFSECT1INFO>
<DATE>1998-5-22</DATE>
</REFSECT1INFO>
<TITLE>Description
</TITLE>
<PARA><FUNCTION>pg_listen</FUNCTION> creates, changes, or cancels a request
to listen for asynchronous NOTIFY messages from the
<ProductName>PostgreSQL</ProductName> backend.  With a <parameter>callbackCommand</>
parameter, the request is established, or the command string of an already
existing request is replaced.  With no <parameter>callbackCommand</> parameter, a prior
request is canceled.
</PARA>

<para>
After a <FUNCTION>pg_listen</FUNCTION> request is established,
the specified command string is executed whenever a NOTIFY message bearing
the given name arrives from the backend.  This occurs when any
<ProductName>PostgreSQL</ProductName> client application issues a NOTIFY command
referencing that name.  (Note that the name can be, but does not have to be,
that of an existing relation in the database.)
The command string is executed from the Tcl idle loop.  That is the normal
idle state of an application written with Tk.  In non-Tk Tcl shells, you can
execute <FUNCTION>update</FUNCTION> or <FUNCTION>vwait</FUNCTION> to cause
the idle loop to be entered.
</Para>

<para>
You should not invoke the SQL statements <command>LISTEN</command> or <command>UNLISTEN</command> directly when
using <FUNCTION>pg_listen</FUNCTION>.  <application>Pgtcl</application> takes care of issuing those
statements for you.  But if you want to send a NOTIFY message yourself,
invoke the SQL NOTIFY statement using <FUNCTION>pg_exec</FUNCTION>.
</PARA>
</REFSECT1>

</REFENTRY>

<!-- ********************************************************** -->

<REFENTRY ID="PGTCL-PGON-CONNECTION-LOSS">
<REFMETA>
<REFENTRYTITLE>pg_on_connection_loss</REFENTRYTITLE>
<REFMISCINFO>PGTCL - Asynchronous Notify</REFMISCINFO>
</REFMETA>
<REFNAMEDIV>
<REFNAME>pg_on_connection_loss
</REFNAME>
<REFPURPOSE>set or change a callback for unexpected connection loss
</REFPURPOSE>
<INDEXTERM
ID="IX-PGTCL-PGON-CONNECTION-LOSS-1"><PRIMARY>pgtcl</PRIMARY><SECONDARY>connection loss</SECONDARY></INDEXTERM>
<INDEXTERM ID="IX-PGTCL-PGON-CONNECTION-LOSS-2"><PRIMARY>connection loss</PRIMARY></INDEXTERM>
</REFNAMEDIV>
<REFSYNOPSISDIV>
<REFSYNOPSISDIVINFO>
<DATE>2002-09-02</DATE>
</REFSYNOPSISDIVINFO>
<SYNOPSIS>
pg_on_connection_loss <REPLACEABLE CLASS="PARAMETER">dbHandle</REPLACEABLE> <REPLACEABLE CLASS="PARAMETER">callbackCommand</REPLACEABLE>
</SYNOPSIS>

<REFSECT2 ID="R2-PGTCL-PGON-CONNECTION-LOSS-1">
<REFSECT2INFO>
<DATE>2002-09-02</DATE>
</REFSECT2INFO>
<TITLE>Inputs
</TITLE>
<VARIABLELIST>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">dbHandle</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>Specifies a valid database handle.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">callbackCommand</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>If present, provides the command string to execute
when connection loss is detected.
</PARA>
</LISTITEM>
</VARLISTENTRY>
</VARIABLELIST>
</REFSECT2>

<REFSECT2 ID="R2-PGTCL-PGON-CONNECTION-LOSS-2">
<REFSECT2INFO>
<DATE>2002-09-02</DATE>
</REFSECT2INFO>
<TITLE>Outputs
</TITLE>
<VARIABLELIST>
<VARLISTENTRY>
<TERM>
  None
</TERM>
<LISTITEM>
<PARA>
</PARA>
</LISTITEM>
</VARLISTENTRY>
</VARIABLELIST>
</REFSECT2>
</REFSYNOPSISDIV>

<REFSECT1 ID="R1-PGTCL-PGON-CONNECTION-LOSS-1">
<REFSECT1INFO>
<DATE>2002-09-02</DATE>
</REFSECT1INFO>
<TITLE>Description
</TITLE>
<PARA><FUNCTION>pg_on_connection_loss</FUNCTION> creates, changes, or cancels
a request to execute a callback command if an unexpected loss of connection
to the database occurs.
With a <parameter>callbackCommand</>
parameter, the request is established, or the command string of an already
existing request is replaced.  With no <parameter>callbackCommand</>
parameter, a prior request is canceled.
</PARA>

<para>
The callback command string is executed from the Tcl idle loop.  That is the
normal idle state of an application written with Tk.  In non-Tk Tcl shells,
you can
execute <FUNCTION>update</FUNCTION> or <FUNCTION>vwait</FUNCTION> to cause
the idle loop to be entered.
</Para>
</REFSECT1>

</REFENTRY>

<!-- ********************************************************** -->

<REFENTRY ID="PGTCL-PGLOCREAT">
<REFMETA>
<REFENTRYTITLE>pg_lo_creat</REFENTRYTITLE>
<REFMISCINFO>PGTCL - Large Objects</REFMISCINFO>
</REFMETA>
<REFNAMEDIV>
<REFNAME>pg_lo_creat
</REFNAME>
<REFPURPOSE>create a large object
</REFPURPOSE>
<INDEXTERM ID="IX-PGTCL-PGLOCREAT-1"><PRIMARY>pgtcl</PRIMARY><SECONDARY>creating</SECONDARY></INDEXTERM>
<INDEXTERM ID="IX-PGTCL-PGLOCREAT-2"><PRIMARY>pg_lo_creat</PRIMARY></INDEXTERM>
</REFNAMEDIV>
<REFSYNOPSISDIV>
<REFSYNOPSISDIVINFO>
<DATE>1997-12-24</DATE>
</REFSYNOPSISDIVINFO>
<SYNOPSIS>
pg_lo_creat <REPLACEABLE CLASS="PARAMETER">conn</REPLACEABLE> <REPLACEABLE CLASS="PARAMETER">mode</REPLACEABLE>
</SYNOPSIS>

<REFSECT2 ID="R2-PGTCL-PGLOCREAT-1">
<REFSECT2INFO>
<DATE>1997-12-24</DATE>
</REFSECT2INFO>
<TITLE>Inputs
</TITLE>
<VARIABLELIST>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">conn</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>Specifies a valid database connection.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">mode</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>Specifies the access mode for the large object</PARA>
</LISTITEM>
</VARLISTENTRY>
</VARIABLELIST>
</REFSECT2>

<REFSECT2 ID="R2-PGTCL-PGLOCREAT-2">
<REFSECT2INFO>
<DATE>1997-12-24</DATE>
</REFSECT2INFO>
<TITLE>Outputs
</TITLE>
<VARIABLELIST>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">objOid</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>
The OID of the large object created.
</PARA>
</LISTITEM>
</VARLISTENTRY>
</VARIABLELIST>
</REFSECT2>
</REFSYNOPSISDIV>

<REFSECT1 ID="R1-PGTCL-PGLOCREAT-1">
<REFSECT1INFO>
<DATE>1997-12-24</DATE>
</REFSECT1INFO>
<TITLE>Description
</TITLE>
<PARA><FUNCTION>pg_lo_creat</FUNCTION> creates an Inversion Large Object.
</PARA>
</REFSECT1>
<REFSECT1 ID="R1-PGTCL-PGLOCREAT-2">
<TITLE>Usage
</TITLE>
<PARA>
mode can be any or'ing together of <literal>INV_READ</> and <literal>INV_WRITE</>.
The <quote>or</quote> operator is <literal>|</literal>.
<ProgramListing>
[pg_lo_creat $conn "INV_READ|INV_WRITE"]
</ProgramListing>
</PARA>
</REFSECT1>
</REFENTRY>

<!-- ********************************************************** -->

<REFENTRY ID="PGTCL-PGLOOPEN">
<REFMETA>
<REFENTRYTITLE>pg_lo_open</REFENTRYTITLE>
<REFMISCINFO>PGTCL - Large Objects</REFMISCINFO>
</REFMETA>
<REFNAMEDIV>
<REFNAME>pg_lo_open
</REFNAME>
<REFPURPOSE>open a large object
</REFPURPOSE>
<INDEXTERM ID="IX-PGTCL-PGLOOPEN-1"><PRIMARY>pgtcl</PRIMARY><SECONDARY>opening</SECONDARY></INDEXTERM>
<INDEXTERM ID="IX-PGTCL-PGLOOPEN-2"><PRIMARY>pg_lo_open</PRIMARY></INDEXTERM>
</REFNAMEDIV>
<REFSYNOPSISDIV>
<REFSYNOPSISDIVINFO>
<DATE>1997-12-24</DATE>
</REFSYNOPSISDIVINFO>
<SYNOPSIS>
pg_lo_open <REPLACEABLE CLASS="PARAMETER">conn</REPLACEABLE> <REPLACEABLE CLASS="PARAMETER">objOid</REPLACEABLE> <REPLACEABLE CLASS="PARAMETER">mode</REPLACEABLE>
</SYNOPSIS>

<REFSECT2 ID="R2-PGTCL-PGLOOPEN-1">
<REFSECT2INFO>
<DATE>1997-12-24</DATE>
</REFSECT2INFO>
<TITLE>Inputs
</TITLE>
<VARIABLELIST>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">conn</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>Specifies a valid database connection.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">objOid</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>Specifies a valid large object OID.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">mode</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>Specifies the access mode for the large object</PARA>
</LISTITEM>
</VARLISTENTRY>
</VARIABLELIST>
</REFSECT2>

<REFSECT2 ID="R2-PGTCL-PGLOOPEN-2">
<REFSECT2INFO>
<DATE>1997-12-24</DATE>
</REFSECT2INFO>
<TITLE>Outputs
</TITLE>
<VARIABLELIST>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">fd</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>
A file descriptor for use in later pg_lo* routines.
</PARA>
</LISTITEM>
</VARLISTENTRY>
</VARIABLELIST>
</REFSECT2>
</REFSYNOPSISDIV>

<REFSECT1 ID="R1-PGTCL-PGLOOPEN-1">
<REFSECT1INFO>
<DATE>1997-12-24</DATE>
</REFSECT1INFO>
<TITLE>Description
</TITLE>
<PARA><FUNCTION>pg_lo_open</FUNCTION> open an Inversion Large Object.
</PARA>
</REFSECT1>
<REFSECT1 ID="R1-PGTCL-PGLOOPEN-2">
<TITLE>Usage
</TITLE>
<PARA>
Mode can be either <literal>r</>, <literal>w</>, or <literal>rw</>.
</PARA>
</REFSECT1>
</REFENTRY>

<!-- ********************************************************** -->

<REFENTRY ID="PGTCL-PGLOCLOSE">
<REFMETA>
<REFENTRYTITLE>pg_lo_close</REFENTRYTITLE>
<REFMISCINFO>PGTCL - Large Objects</REFMISCINFO>
</REFMETA>
<REFNAMEDIV>
<REFNAME>pg_lo_close
</REFNAME>
<REFPURPOSE>close a large object
</REFPURPOSE>
<INDEXTERM ID="IX-PGTCL-PGLOCLOSE-1"><PRIMARY>pgtcl</PRIMARY><SECONDARY>closing</SECONDARY></INDEXTERM>
<INDEXTERM ID="IX-PGTCL-PGLOCLOSE-2"><PRIMARY>pg_lo_close</PRIMARY></INDEXTERM>
</REFNAMEDIV>
<REFSYNOPSISDIV>
<REFSYNOPSISDIVINFO>
<DATE>1997-12-24</DATE>
</REFSYNOPSISDIVINFO>
<SYNOPSIS>
pg_lo_close <REPLACEABLE CLASS="PARAMETER">conn</REPLACEABLE> <REPLACEABLE CLASS="PARAMETER">fd</REPLACEABLE>
</SYNOPSIS>

<REFSECT2 ID="R2-PGTCL-PGLOCLOSE-1">
<REFSECT2INFO>
<DATE>1997-12-24</DATE>
</REFSECT2INFO>
<TITLE>Inputs
</TITLE>
<VARIABLELIST>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">conn</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>Specifies a valid database connection.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">fd</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>
A file descriptor for use in later pg_lo* routines.
</PARA>
</LISTITEM>
</VARLISTENTRY>
</VARIABLELIST>
</REFSECT2>

<REFSECT2 ID="R2-PGTCL-PGLOCLOSE-2">
<REFSECT2INFO>
<DATE>1997-12-24</DATE>
</REFSECT2INFO>
<TITLE>Outputs
</TITLE>
<PARA>None</PARA>
</REFSECT2>
</REFSYNOPSISDIV>

<REFSECT1 ID="R1-PGTCL-PGLOCLOSE-1">
<REFSECT1INFO>
<DATE>1997-12-24</DATE>
</REFSECT1INFO>
<TITLE>Description
</TITLE>
<PARA><FUNCTION>pg_lo_close</FUNCTION> closes an Inversion Large Object.
</PARA>
</REFSECT1>
<REFSECT1 ID="R1-PGTCL-PGLOCLOSE-2">
<TITLE>Usage
</TITLE>
<PARA>
</PARA>
</REFSECT1>
</REFENTRY>

<!-- ********************************************************** -->

<REFENTRY ID="PGTCL-PGLOREAD">
<REFMETA>
<REFENTRYTITLE>pg_lo_read</REFENTRYTITLE>
<REFMISCINFO>PGTCL - Large Objects</REFMISCINFO>
</REFMETA>
<REFNAMEDIV>
<REFNAME>pg_lo_read
</REFNAME>
<REFPURPOSE>read a large object
</REFPURPOSE>
<INDEXTERM ID="IX-PGTCL-PGLOREAD-1"><PRIMARY>pgtcl</PRIMARY><SECONDARY>reading</SECONDARY></INDEXTERM>
<INDEXTERM ID="IX-PGTCL-PGLOREAD-2"><PRIMARY>pg_lo_read</PRIMARY></INDEXTERM>
</REFNAMEDIV>
<REFSYNOPSISDIV>
<REFSYNOPSISDIVINFO>
<DATE>1997-12-24</DATE>
</REFSYNOPSISDIVINFO>
<SYNOPSIS>
pg_lo_read <REPLACEABLE CLASS="PARAMETER">conn</REPLACEABLE> <REPLACEABLE CLASS="PARAMETER">fd</REPLACEABLE> <REPLACEABLE CLASS="PARAMETER">bufVar</REPLACEABLE> <REPLACEABLE CLASS="PARAMETER">len</REPLACEABLE>
</SYNOPSIS>

<REFSECT2 ID="R2-PGTCL-PGLOREAD-1">
<REFSECT2INFO>
<DATE>1997-12-24</DATE>
</REFSECT2INFO>
<TITLE>Inputs
</TITLE>
<VARIABLELIST>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">conn</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>Specifies a valid database connection.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">fd</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>
File descriptor for the large object from pg_lo_open.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">bufVar</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>Specifies a valid buffer variable to contain the large object segment.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">len</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>Specifies the maximum allowable size of the large object segment.</PARA>
</LISTITEM>
</VARLISTENTRY>
</VARIABLELIST>
</REFSECT2>

<REFSECT2 ID="R2-PGTCL-PGLOREAD-2">
<REFSECT2INFO>
<DATE>1997-12-24</DATE>
</REFSECT2INFO>
<TITLE>Outputs
</TITLE>
<PARA>None</PARA>
</REFSECT2>
</REFSYNOPSISDIV>

<REFSECT1 ID="R1-PGTCL-PGLOREAD-1">
<REFSECT1INFO>
<DATE>1997-12-24</DATE>
</REFSECT1INFO>
<TITLE>Description
</TITLE>
<PARA><FUNCTION>pg_lo_read</FUNCTION> reads 
at most <REPLACEABLE CLASS="PARAMETER">len</REPLACEABLE> bytes from a large object into a variable
 named <REPLACEABLE CLASS="PARAMETER">bufVar</REPLACEABLE>.
</PARA>
</REFSECT1>
<REFSECT1 ID="R1-PGTCL-PGLOREAD-2">
<TITLE>Usage
</TITLE>
<PARA>
<REPLACEABLE CLASS="PARAMETER">bufVar</REPLACEABLE> must be a valid variable name.
</PARA>
</REFSECT1>
</REFENTRY>

<!-- ********************************************************** -->

<REFENTRY ID="PGTCL-PGLOWRITE">
<REFMETA>
<REFENTRYTITLE>pg_lo_write</REFENTRYTITLE>
<REFMISCINFO>PGTCL - Large Objects</REFMISCINFO>
</REFMETA>
<REFNAMEDIV>
<REFNAME>pg_lo_write
</REFNAME>
<REFPURPOSE>write a large object
</REFPURPOSE>
<INDEXTERM ID="IX-PGTCL-PGLOWRITE-1"><PRIMARY>pgtcl</PRIMARY><SECONDARY>writing</SECONDARY></INDEXTERM>
<INDEXTERM ID="IX-PGTCL-PGLOWRITE-2"><PRIMARY>pg_lo_write</PRIMARY></INDEXTERM>
</REFNAMEDIV>
<REFSYNOPSISDIV>
<REFSYNOPSISDIVINFO>
<DATE>1997-12-24</DATE>
</REFSYNOPSISDIVINFO>
<SYNOPSIS>
pg_lo_write <REPLACEABLE CLASS="PARAMETER">conn</REPLACEABLE> <REPLACEABLE CLASS="PARAMETER">fd</REPLACEABLE> <REPLACEABLE CLASS="PARAMETER">buf</REPLACEABLE> <REPLACEABLE CLASS="PARAMETER">len</REPLACEABLE>
</SYNOPSIS>

<REFSECT2 ID="R2-PGTCL-PGLOWRITE-1">
<REFSECT2INFO>
<DATE>1997-12-24</DATE>
</REFSECT2INFO>
<TITLE>Inputs
</TITLE>
<VARIABLELIST>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">conn</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>Specifies a valid database connection.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">fd</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>
File descriptor for the large object from pg_lo_open.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">buf</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>Specifies a valid string variable to write to the large object.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">len</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>Specifies the maximum size of the string to write.</PARA>
</LISTITEM>
</VARLISTENTRY>
</VARIABLELIST>
</REFSECT2>

<REFSECT2 ID="R2-PGTCL-PGLOWRITE-2">
<REFSECT2INFO>
<DATE>1997-12-24</DATE>
</REFSECT2INFO>
<TITLE>Outputs
</TITLE>
<PARA>None</PARA>
</REFSECT2>
</REFSYNOPSISDIV>

<REFSECT1 ID="R1-PGTCL-PGLOWRITE-1">
<REFSECT1INFO>
<DATE>1997-12-24</DATE>
</REFSECT1INFO>
<TITLE>Description
</TITLE>
<PARA><FUNCTION>pg_lo_write</FUNCTION> writes 
at most <REPLACEABLE CLASS="PARAMETER">len</REPLACEABLE> bytes to a large object from a variable
 <REPLACEABLE CLASS="PARAMETER">buf</REPLACEABLE>.
</PARA>
</REFSECT1>
<REFSECT1 ID="R1-PGTCL-PGLOWRITE-2">
<TITLE>Usage
</TITLE>
<PARA>
<REPLACEABLE CLASS="PARAMETER">buf</REPLACEABLE> must be 
the actual string to write, not a variable name.
</PARA>
</REFSECT1>
</REFENTRY>

<!-- ********************************************************** -->

<REFENTRY ID="PGTCL-PGLOLSEEK">
<REFMETA>
<REFENTRYTITLE>pg_lo_lseek</REFENTRYTITLE>
<REFMISCINFO>PGTCL - Large Objects</REFMISCINFO>
</REFMETA>
<REFNAMEDIV>
<REFNAME>pg_lo_lseek
</REFNAME>
<REFPURPOSE>seek to a position in a large object
</REFPURPOSE>
<INDEXTERM ID="IX-PGTCL-PGLOLSEEK-1"><PRIMARY>pgtcl</PRIMARY><SECONDARY>positioning</SECONDARY></INDEXTERM>
<INDEXTERM ID="IX-PGTCL-PGLOLSEEK-2"><PRIMARY>pg_lo_lseek</PRIMARY></INDEXTERM>
</REFNAMEDIV>
<REFSYNOPSISDIV>
<REFSYNOPSISDIVINFO>
<DATE>1997-12-24</DATE>
</REFSYNOPSISDIVINFO>
<SYNOPSIS>
pg_lo_lseek <REPLACEABLE CLASS="PARAMETER">conn</REPLACEABLE> <REPLACEABLE CLASS="PARAMETER">fd</REPLACEABLE> <REPLACEABLE CLASS="PARAMETER">offset</REPLACEABLE> <REPLACEABLE CLASS="PARAMETER">whence</REPLACEABLE>
</SYNOPSIS>

<REFSECT2 ID="R2-PGTCL-PGLOLSEEK-1">
<REFSECT2INFO>
<DATE>1997-12-24</DATE>
</REFSECT2INFO>
<TITLE>Inputs
</TITLE>
<VARIABLELIST>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">conn</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>Specifies a valid database connection.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">fd</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>
File descriptor for the large object from pg_lo_open.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">offset</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>Specifies a zero-based offset in bytes.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">whence</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA> whence can be <literal>SEEK_CUR</>, <literal>SEEK_END</>, or <literal>SEEK_SET</> </PARA>
</LISTITEM>
</VARLISTENTRY>
</VARIABLELIST>
</REFSECT2>

<REFSECT2 ID="R2-PGTCL-PGLOLSEEK-2">
<REFSECT2INFO>
<DATE>1997-12-24</DATE>
</REFSECT2INFO>
<TITLE>Outputs
</TITLE>
<PARA>None</PARA>
</REFSECT2>
</REFSYNOPSISDIV>

<REFSECT1 ID="R1-PGTCL-PGLOLSEEK-1">
<REFSECT1INFO>
<DATE>1997-12-24</DATE>
</REFSECT1INFO>
<TITLE>Description
</TITLE>
<PARA><FUNCTION>pg_lo_lseek</FUNCTION> positions 
to <REPLACEABLE CLASS="PARAMETER">offset</REPLACEABLE> bytes from the beginning of the large object.
</PARA>
</REFSECT1>
<REFSECT1 ID="R1-PGTCL-PGLOLSEEK-2">
<TITLE>Usage
</TITLE>
<PARA>
<REPLACEABLE CLASS="PARAMETER">whence</REPLACEABLE> 
can be <literal>SEEK_CUR</literal>, <literal>SEEK_END</>, or <literal>SEEK_SET</literal>.
</PARA>
</REFSECT1>
</REFENTRY>

<!-- ********************************************************** -->

<REFENTRY ID="PGTCL-PGLOTELL">
<REFMETA>
<REFENTRYTITLE>pg_lo_tell</REFENTRYTITLE>
<REFMISCINFO>PGTCL - Large Objects</REFMISCINFO>
</REFMETA>
<REFNAMEDIV>
<REFNAME>pg_lo_tell
</REFNAME>
<REFPURPOSE>return the current seek position of a large object
</REFPURPOSE>
<INDEXTERM ID="IX-PGTCL-PGLOTELL-1"><PRIMARY>pgtcl</PRIMARY><SECONDARY>positioning</SECONDARY></INDEXTERM>
<INDEXTERM ID="IX-PGTCL-PGLOTELL-2"><PRIMARY>pg_lo_tell</PRIMARY></INDEXTERM>
</REFNAMEDIV>
<REFSYNOPSISDIV>
<REFSYNOPSISDIVINFO>
<DATE>1997-12-24</DATE>
</REFSYNOPSISDIVINFO>
<SYNOPSIS>
pg_lo_tell <REPLACEABLE CLASS="PARAMETER">conn</REPLACEABLE> <REPLACEABLE CLASS="PARAMETER">fd</REPLACEABLE>
</SYNOPSIS>

<REFSECT2 ID="R2-PGTCL-PGLOTELL-1">
<REFSECT2INFO>
<DATE>1997-12-24</DATE>
</REFSECT2INFO>
<TITLE>Inputs
</TITLE>
<VARIABLELIST>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">conn</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>Specifies a valid database connection.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">fd</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>
File descriptor for the large object from pg_lo_open.
</PARA>
</LISTITEM>
</VARLISTENTRY>
</VARIABLELIST>
</REFSECT2>

<REFSECT2 ID="R2-PGTCL-PGLOTELL-2">
<REFSECT2INFO>
<DATE>1997-12-24</DATE>
</REFSECT2INFO>
<TITLE>Outputs
</TITLE>
<VARIABLELIST>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">offset</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>A zero-based offset in bytes suitable for input to <Function>pg_lo_lseek</Function>.
</PARA>
</LISTITEM>
</VARLISTENTRY>
</VARIABLELIST>
</REFSECT2>
</REFSYNOPSISDIV>

<REFSECT1 ID="R1-PGTCL-PGLOTELL-1">
<REFSECT1INFO>
<DATE>1997-12-24</DATE>
</REFSECT1INFO>
<TITLE>Description
</TITLE>
<PARA><FUNCTION>pg_lo_tell</FUNCTION> returns the current
to <REPLACEABLE CLASS="PARAMETER">offset</REPLACEABLE> in bytes from the beginning of the large object.
</PARA>
</REFSECT1>
<REFSECT1 ID="R1-PGTCL-PGLOTELL-2">
<TITLE>Usage
</TITLE>
<PARA>
</PARA>
</REFSECT1>
</REFENTRY>

<!-- ********************************************************** -->

<REFENTRY ID="PGTCL-PGLOUNLINK">
<REFMETA>
<REFENTRYTITLE>pg_lo_unlink</REFENTRYTITLE>
<REFMISCINFO>PGTCL - Large Objects</REFMISCINFO>
</REFMETA>
<REFNAMEDIV>
<REFNAME>pg_lo_unlink
</REFNAME>
<REFPURPOSE>delete a large object
</REFPURPOSE>
<INDEXTERM ID="IX-PGTCL-PGLOUNLINK-1"><PRIMARY>pgtcl</PRIMARY><SECONDARY>delete</SECONDARY></INDEXTERM>
<INDEXTERM ID="IX-PGTCL-PGLOUNLINK-2"><PRIMARY>pg_lo_unlink</PRIMARY></INDEXTERM>
</REFNAMEDIV>
<REFSYNOPSISDIV>
<REFSYNOPSISDIVINFO>
<DATE>1997-12-24</DATE>
</REFSYNOPSISDIVINFO>
<SYNOPSIS>
pg_lo_unlink <REPLACEABLE CLASS="PARAMETER">conn</REPLACEABLE> <REPLACEABLE CLASS="PARAMETER">lobjId</REPLACEABLE>
</SYNOPSIS>

<REFSECT2 ID="R2-PGTCL-PGLOUNLINK-1">
<REFSECT2INFO>
<DATE>1997-12-24</DATE>
</REFSECT2INFO>
<TITLE>Inputs
</TITLE>
<VARIABLELIST>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">conn</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>Specifies a valid database connection.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">lobjId</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>
Identifier for a large object.
<comment>
 XXX Is this the same as <parameter>objOid</parameter> in other calls?? - thomas 1998-01-11
</comment>
</PARA>
</LISTITEM>
</VARLISTENTRY>
</VARIABLELIST>
</REFSECT2>

<REFSECT2 ID="R2-PGTCL-PGLOUNLINK-2">
<REFSECT2INFO>
<DATE>1997-12-24</DATE>
</REFSECT2INFO>
<TITLE>Outputs
</TITLE>
<PARA>
None
</PARA>
</REFSECT2>
</REFSYNOPSISDIV>

<REFSECT1 ID="R1-PGTCL-PGLOUNLINK-1">
<REFSECT1INFO>
<DATE>1997-12-24</DATE>
</REFSECT1INFO>
<TITLE>Description
</TITLE>
<PARA><FUNCTION>pg_lo_unlink</FUNCTION> deletes the specified large object.
</PARA>
</REFSECT1>
<REFSECT1 ID="R1-PGTCL-PGLOUNLINK-2">
<TITLE>Usage
</TITLE>
<PARA>
</PARA>
</REFSECT1>
</REFENTRY>

<!-- ********************************************************** -->

<REFENTRY ID="PGTCL-PGLOIMPORT">
<REFMETA>
<REFENTRYTITLE>pg_lo_import</REFENTRYTITLE>
<REFMISCINFO>PGTCL - Large Objects</REFMISCINFO>
</REFMETA>
<REFNAMEDIV>
<REFNAME>pg_lo_import
</REFNAME>
<REFPURPOSE>import a large object from a file
</REFPURPOSE>
<INDEXTERM ID="IX-PGTCL-PGLOIMPORT-1"><PRIMARY>pgtcl</PRIMARY><SECONDARY>import</SECONDARY></INDEXTERM>
<INDEXTERM ID="IX-PGTCL-PGLOIMPORT-2"><PRIMARY>pg_lo_import</PRIMARY></INDEXTERM>
</REFNAMEDIV>
<REFSYNOPSISDIV>
<REFSYNOPSISDIVINFO>
<DATE>1997-12-24</DATE>
</REFSYNOPSISDIVINFO>
<SYNOPSIS>
pg_lo_import <REPLACEABLE CLASS="PARAMETER">conn</REPLACEABLE> <REPLACEABLE CLASS="PARAMETER">filename</REPLACEABLE>
</SYNOPSIS>

<REFSECT2 ID="R2-PGTCL-PGLOIMPORT-1">
<REFSECT2INFO>
<DATE>1997-12-24</DATE>
</REFSECT2INFO>
<TITLE>Inputs
</TITLE>
<VARIABLELIST>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">conn</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>Specifies a valid database connection.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">filename</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>
Unix file name.
</PARA>
</LISTITEM>
</VARLISTENTRY>
</VARIABLELIST>
</REFSECT2>

<REFSECT2 ID="R2-PGTCL-PGLOIMPORT-2">
<REFSECT2INFO>
<DATE>1997-12-24</DATE>
</REFSECT2INFO>
<TITLE>Outputs
</TITLE>
<PARA>
None
<comment>
 XXX Does this return a lobjId? Is that the same as the objOid in other calls? thomas - 1998-01-11
</comment>
</PARA>
</REFSECT2>
</REFSYNOPSISDIV>

<REFSECT1 ID="R1-PGTCL-PGLOIMPORT-1">
<REFSECT1INFO>
<DATE>1997-12-24</DATE>
</REFSECT1INFO>
<TITLE>Description
</TITLE>
<PARA><FUNCTION>pg_lo_import</FUNCTION> reads the specified file and places the contents into a large object.
</PARA>
</REFSECT1>
<REFSECT1 ID="R1-PGTCL-PGLOIMPORT-2">
<TITLE>Usage
</TITLE>
<PARA>
  <Function>pg_lo_import</Function> must be called within a BEGIN/END transaction block.
</PARA>
</REFSECT1>
</REFENTRY>

<!-- ********************************************************** -->

<REFENTRY ID="PGTCL-PGLOEXPORT">
<REFMETA>
<REFENTRYTITLE>pg_lo_export</REFENTRYTITLE>
<REFMISCINFO>PGTCL - Large Objects</REFMISCINFO>
</REFMETA>
<REFNAMEDIV>
<REFNAME>pg_lo_export
</REFNAME>
<REFPURPOSE>export a large object to a file
</REFPURPOSE>
<INDEXTERM ID="IX-PGTCL-PGLOEXPORT-1"><PRIMARY>pgtcl</PRIMARY><SECONDARY>export</SECONDARY></INDEXTERM>
<INDEXTERM ID="IX-PGTCL-PGLOEXPORT-2"><PRIMARY>pg_lo_export</PRIMARY></INDEXTERM>
</REFNAMEDIV>
<REFSYNOPSISDIV>
<REFSYNOPSISDIVINFO>
<DATE>1997-12-24</DATE>
</REFSYNOPSISDIVINFO>
<SYNOPSIS>
pg_lo_export <REPLACEABLE CLASS="PARAMETER">conn</REPLACEABLE> <REPLACEABLE CLASS="PARAMETER">lobjId</REPLACEABLE> <REPLACEABLE CLASS="PARAMETER">filename</REPLACEABLE>
</SYNOPSIS>

<REFSECT2 ID="R2-PGTCL-PGLOEXPORT-1">
<REFSECT2INFO>
<DATE>1997-12-24</DATE>
</REFSECT2INFO>
<TITLE>Inputs
</TITLE>
<VARIABLELIST>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">conn</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>Specifies a valid database connection.
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">lobjId</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>
Large object identifier.
<comment>
 XXX Is this the same as the objOid in other calls?? thomas - 1998-01-11
</comment>
</PARA>
</LISTITEM>
</VARLISTENTRY>
<VARLISTENTRY>
<TERM>
  <REPLACEABLE CLASS="PARAMETER">filename</REPLACEABLE>
</TERM>
<LISTITEM>
<PARA>
Unix file name.
</PARA>
</LISTITEM>
</VARLISTENTRY>
</VARIABLELIST>
</REFSECT2>

<REFSECT2 ID="R2-PGTCL-PGLOEXPORT-2">
<REFSECT2INFO>
<DATE>1997-12-24</DATE>
</REFSECT2INFO>
<TITLE>Outputs
</TITLE>
<PARA>
None
<comment>
 XXX Does this return a lobjId? Is that the same as the objOid in other calls? thomas - 1998-01-11
</comment>
</PARA>
</REFSECT2>
</REFSYNOPSISDIV>

<REFSECT1 ID="R1-PGTCL-PGLOEXPORT-1">
<REFSECT1INFO>
<DATE>1997-12-24</DATE>
</REFSECT1INFO>
<TITLE>Description
</TITLE>
<PARA><FUNCTION>pg_lo_export</FUNCTION> writes the specified large object into a Unix file.
</PARA>
</REFSECT1>
<REFSECT1 ID="R1-PGTCL-PGLOEXPORT-2">
<TITLE>Usage
</TITLE>
<PARA>
  <Function>pg_lo_export</Function> must be called within a BEGIN/END transaction block.
</PARA>
</REFSECT1>
</REFENTRY>

</Sect1>
</Chapter>