<?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>COPY</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="sql-commit-prepared.html" title="COMMIT PREPARED" /><link rel="next" href="sql-create-access-method.html" title="CREATE ACCESS METHOD" /></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">COPY</th></tr><tr><td width="10%" align="left"><a accesskey="p" href="sql-commit-prepared.html" title="COMMIT PREPARED">Prev</a> </td><td width="10%" align="left"><a accesskey="u" href="sql-commands.html" title="SQL Commands">Up</a></td><th width="60%" align="center">SQL Commands</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="sql-create-access-method.html" title="CREATE ACCESS METHOD">Next</a></td></tr></table><hr></hr></div><div class="refentry" id="SQL-COPY"><div class="titlepage"></div><a id="id-1.9.3.55.1" class="indexterm"></a><div class="refnamediv"><h2><span class="refentrytitle">COPY</span></h2><p>COPY — copy data between a file and a table</p></div><div class="refsynopsisdiv"><h2>Synopsis</h2><pre class="synopsis">
COPY <em class="replaceable"><code>table_name</code></em> [ ( <em class="replaceable"><code>column_name</code></em> [, ...] ) ]
FROM { '<em class="replaceable"><code>filename</code></em>' | PROGRAM '<em class="replaceable"><code>command</code></em>' | STDIN }
[ [ WITH ] ( <em class="replaceable"><code>option</code></em> [, ...] ) ]
COPY { <em class="replaceable"><code>table_name</code></em> [ ( <em class="replaceable"><code>column_name</code></em> [, ...] ) ] | ( <em class="replaceable"><code>query</code></em> ) }
TO { '<em class="replaceable"><code>filename</code></em>' | PROGRAM '<em class="replaceable"><code>command</code></em>' | STDOUT }
[ [ WITH ] ( <em class="replaceable"><code>option</code></em> [, ...] ) ]
<span class="phrase">where <em class="replaceable"><code>option</code></em> can be one of:</span>
FORMAT <em class="replaceable"><code>format_name</code></em>
OIDS [ <em class="replaceable"><code>boolean</code></em> ]
FREEZE [ <em class="replaceable"><code>boolean</code></em> ]
DELIMITER '<em class="replaceable"><code>delimiter_character</code></em>'
NULL '<em class="replaceable"><code>null_string</code></em>'
HEADER [ <em class="replaceable"><code>boolean</code></em> ]
QUOTE '<em class="replaceable"><code>quote_character</code></em>'
ESCAPE '<em class="replaceable"><code>escape_character</code></em>'
FORCE_QUOTE { ( <em class="replaceable"><code>column_name</code></em> [, ...] ) | * }
FORCE_NOT_NULL ( <em class="replaceable"><code>column_name</code></em> [, ...] )
FORCE_NULL ( <em class="replaceable"><code>column_name</code></em> [, ...] )
ENCODING '<em class="replaceable"><code>encoding_name</code></em>'
</pre></div><div class="refsect1" id="id-1.9.3.55.5"><h2>Description</h2><p>
<code class="command">COPY</code> moves data between
<span class="productname">PostgreSQL</span> tables and standard file-system
files. <code class="command">COPY TO</code> copies the contents of a table
<span class="emphasis"><em>to</em></span> a file, while <code class="command">COPY FROM</code> copies
data <span class="emphasis"><em>from</em></span> a file to a table (appending the data to
whatever is in the table already). <code class="command">COPY TO</code>
can also copy the results of a <code class="command">SELECT</code> query.
</p><p>
If a list of columns is specified, <code class="command">COPY</code> will
only copy the data in the specified columns to or from the file.
If there are any columns in the table that are not in the column list,
<code class="command">COPY FROM</code> will insert the default values for
those columns.
</p><p>
<code class="command">COPY</code> with a file name instructs the
<span class="productname">PostgreSQL</span> server to directly read from
or write to a file. The file must be accessible by the
<span class="productname">PostgreSQL</span> user (the user ID the server
runs as) and the name must be specified from the viewpoint of the
server. When <code class="literal">PROGRAM</code> is specified, the server
executes the given command and reads from the standard output of the
program, or writes to the standard input of the program. The command
must be specified from the viewpoint of the server, and be executable
by the <span class="productname">PostgreSQL</span> user. When
<code class="literal">STDIN</code> or <code class="literal">STDOUT</code> is
specified, data is transmitted via the connection between the
client and the server.
</p></div><div class="refsect1" id="id-1.9.3.55.6"><h2>Parameters</h2><div class="variablelist"><dl class="variablelist"><dt><span class="term"><em class="replaceable"><code>table_name</code></em></span></dt><dd><p>
The name (optionally schema-qualified) of an existing table.
</p></dd><dt><span class="term"><em class="replaceable"><code>column_name</code></em></span></dt><dd><p>
An optional list of columns to be copied. If no column list is
specified, all columns of the table will be copied.
</p></dd><dt><span class="term"><em class="replaceable"><code>query</code></em></span></dt><dd><p>
A <a class="xref" href="sql-select.html" title="SELECT"><span class="refentrytitle">SELECT</span></a>, <a class="xref" href="sql-values.html" title="VALUES"><span class="refentrytitle">VALUES</span></a>,
<a class="xref" href="sql-insert.html" title="INSERT"><span class="refentrytitle">INSERT</span></a>, <a class="xref" href="sql-update.html" title="UPDATE"><span class="refentrytitle">UPDATE</span></a> or
<a class="xref" href="sql-delete.html" title="DELETE"><span class="refentrytitle">DELETE</span></a> command whose results are to be
copied. Note that parentheses are required around the query.
</p><p>
For <code class="command">INSERT</code>, <code class="command">UPDATE</code> and
<code class="command">DELETE</code> queries a RETURNING clause must be provided,
and the target relation must not have a conditional rule, nor
an <code class="literal">ALSO</code> rule, nor an <code class="literal">INSTEAD</code> rule
that expands to multiple statements.
</p></dd><dt><span class="term"><em class="replaceable"><code>filename</code></em></span></dt><dd><p>
The path name of the input or output file. An input file name can be
an absolute or relative path, but an output file name must be an absolute
path. Windows users might need to use an <code class="literal">E''</code> string and
double any backslashes used in the path name.
</p></dd><dt><span class="term"><code class="literal">PROGRAM</code></span></dt><dd><p>
A command to execute. In <code class="command">COPY FROM</code>, the input is
read from standard output of the command, and in <code class="command">COPY TO</code>,
the output is written to the standard input of the command.
</p><p>
Note that the command is invoked by the shell, so if you need to pass
any arguments to shell command that come from an untrusted source, you
must be careful to strip or escape any special characters that might
have a special meaning for the shell. For security reasons, it is best
to use a fixed command string, or at least avoid passing any user input
in it.
</p></dd><dt><span class="term"><code class="literal">STDIN</code></span></dt><dd><p>
Specifies that input comes from the client application.
</p></dd><dt><span class="term"><code class="literal">STDOUT</code></span></dt><dd><p>
Specifies that output goes to the client application.
</p></dd><dt><span class="term"><em class="replaceable"><code>boolean</code></em></span></dt><dd><p>
Specifies whether the selected option should be turned on or off.
You can write <code class="literal">TRUE</code>, <code class="literal">ON</code>, or
<code class="literal">1</code> to enable the option, and <code class="literal">FALSE</code>,
<code class="literal">OFF</code>, or <code class="literal">0</code> to disable it. The
<em class="replaceable"><code>boolean</code></em> value can also
be omitted, in which case <code class="literal">TRUE</code> is assumed.
</p></dd><dt><span class="term"><code class="literal">FORMAT</code></span></dt><dd><p>
Selects the data format to be read or written:
<code class="literal">text</code>,
<code class="literal">csv</code> (Comma Separated Values),
or <code class="literal">binary</code>.
The default is <code class="literal">text</code>.
</p></dd><dt><span class="term"><code class="literal">OIDS</code></span></dt><dd><p>
Specifies copying the OID for each row. (An error is raised if
<code class="literal">OIDS</code> is specified for a table that does not
have OIDs, or in the case of copying a <em class="replaceable"><code>query</code></em>.)
</p></dd><dt><span class="term"><code class="literal">FREEZE</code></span></dt><dd><p>
Requests copying the data with rows already frozen, just as they
would be after running the <code class="command">VACUUM FREEZE</code> command.
This is intended as a performance option for initial data loading.
Rows will be frozen only if the table being loaded has been created
or truncated in the current subtransaction, there are no cursors
open and there are no older snapshots held by this transaction. It is
currently not possible to perform a <code class="command">COPY FREEZE</code> on
a partitioned table.
</p><p>
Note that all other sessions will immediately be able to see the data
once it has been successfully loaded. This violates the normal rules
of MVCC visibility and users specifying should be aware of the
potential problems this might cause.
</p></dd><dt><span class="term"><code class="literal">DELIMITER</code></span></dt><dd><p>
Specifies the character that separates columns within each row
(line) of the file. The default is a tab character in text format,
a comma in <code class="literal">CSV</code> format.
This must be a single one-byte character.
This option is not allowed when using <code class="literal">binary</code> format.
</p></dd><dt><span class="term"><code class="literal">NULL</code></span></dt><dd><p>
Specifies the string that represents a null value. The default is
<code class="literal">\N</code> (backslash-N) in text format, and an unquoted empty
string in <code class="literal">CSV</code> format. You might prefer an
empty string even in text format for cases where you don't want to
distinguish nulls from empty strings.
This option is not allowed when using <code class="literal">binary</code> format.
</p><div class="note"><h3 class="title">Note</h3><p>
When using <code class="command">COPY FROM</code>, any data item that matches
this string will be stored as a null value, so you should make
sure that you use the same string as you used with
<code class="command">COPY TO</code>.
</p></div></dd><dt><span class="term"><code class="literal">HEADER</code></span></dt><dd><p>
Specifies that the file contains a header line with the names of each
column in the file. On output, the first line contains the column
names from the table, and on input, the first line is ignored.
This option is allowed only when using <code class="literal">CSV</code> format.
</p></dd><dt><span class="term"><code class="literal">QUOTE</code></span></dt><dd><p>
Specifies the quoting character to be used when a data value is quoted.
The default is double-quote.
This must be a single one-byte character.
This option is allowed only when using <code class="literal">CSV</code> format.
</p></dd><dt><span class="term"><code class="literal">ESCAPE</code></span></dt><dd><p>
Specifies the character that should appear before a
data character that matches the <code class="literal">QUOTE</code> value.
The default is the same as the <code class="literal">QUOTE</code> value (so that
the quoting character is doubled if it appears in the data).
This must be a single one-byte character.
This option is allowed only when using <code class="literal">CSV</code> format.
</p></dd><dt><span class="term"><code class="literal">FORCE_QUOTE</code></span></dt><dd><p>
Forces quoting to be
used for all non-<code class="literal">NULL</code> values in each specified column.
<code class="literal">NULL</code> output is never quoted. If <code class="literal">*</code> is specified,
non-<code class="literal">NULL</code> values will be quoted in all columns.
This option is allowed only in <code class="command">COPY TO</code>, and only when
using <code class="literal">CSV</code> format.
</p></dd><dt><span class="term"><code class="literal">FORCE_NOT_NULL</code></span></dt><dd><p>
Do not match the specified columns' values against the null string.
In the default case where the null string is empty, this means that
empty values will be read as zero-length strings rather than nulls,
even when they are not quoted.
This option is allowed only in <code class="command">COPY FROM</code>, and only when
using <code class="literal">CSV</code> format.
</p></dd><dt><span class="term"><code class="literal">FORCE_NULL</code></span></dt><dd><p>
Match the specified columns' values against the null string, even
if it has been quoted, and if a match is found set the value to
<code class="literal">NULL</code>. In the default case where the null string is empty,
this converts a quoted empty string into NULL.
This option is allowed only in <code class="command">COPY FROM</code>, and only when
using <code class="literal">CSV</code> format.
</p></dd><dt><span class="term"><code class="literal">ENCODING</code></span></dt><dd><p>
Specifies that the file is encoded in the <em class="replaceable"><code>encoding_name</code></em>. If this option is
omitted, the current client encoding is used. See the Notes below
for more details.
</p></dd></dl></div></div><div class="refsect1" id="id-1.9.3.55.7"><h2>Outputs</h2><p>
On successful completion, a <code class="command">COPY</code> command returns a command
tag of the form
</p><pre class="screen">
COPY <em class="replaceable"><code>count</code></em>
</pre><p>
The <em class="replaceable"><code>count</code></em> is the number
of rows copied.
</p><div class="note"><h3 class="title">Note</h3><p>
<span class="application">psql</span> will print this command tag only if the command
was not <code class="literal">COPY ... TO STDOUT</code>, or the
equivalent <span class="application">psql</span> meta-command
<code class="literal">\copy ... to stdout</code>. This is to prevent confusing the
command tag with the data that was just printed.
</p></div></div><div class="refsect1" id="id-1.9.3.55.8"><h2>Notes</h2><p>
<code class="command">COPY TO</code> can only be used with plain tables, not
with views. However, you can write <code class="literal">COPY (SELECT * FROM
<em class="replaceable"><code>viewname</code></em>) TO ...</code>
to copy the current contents of a view.
</p><p>
<code class="command">COPY FROM</code> can be used with plain, foreign, or
partitioned tables or with views that have
<code class="literal">INSTEAD OF INSERT</code> triggers.
</p><p>
<code class="command">COPY</code> only deals with the specific table named;
it does not copy data to or from child tables. Thus for example
<code class="literal">COPY <em class="replaceable"><code>table</code></em> TO</code>
shows the same data as <code class="literal">SELECT * FROM ONLY <em class="replaceable"><code>table</code></em></code>. But <code class="literal">COPY
(SELECT * FROM <em class="replaceable"><code>table</code></em>) TO ...</code>
can be used to dump all of the data in an inheritance hierarchy.
</p><p>
You must have select privilege on the table
whose values are read by <code class="command">COPY TO</code>, and
insert privilege on the table into which values
are inserted by <code class="command">COPY FROM</code>. It is sufficient
to have column privileges on the column(s) listed in the command.
</p><p>
If row-level security is enabled for the table, the relevant
<code class="command">SELECT</code> policies will apply to <code class="literal">COPY
<em class="replaceable"><code>table</code></em> TO</code> statements.
Currently, <code class="command">COPY FROM</code> is not supported for tables
with row-level security. Use equivalent <code class="command">INSERT</code>
statements instead.
</p><p>
Files named in a <code class="command">COPY</code> command are read or written
directly by the server, not by the client application. Therefore,
they must reside on or be accessible to the database server machine,
not the client. They must be accessible to and readable or writable
by the <span class="productname">PostgreSQL</span> user (the user ID the
server runs as), not the client. Similarly,
the command specified with <code class="literal">PROGRAM</code> is executed directly
by the server, not by the client application, must be executable by the
<span class="productname">PostgreSQL</span> user.
<code class="command">COPY</code> naming a file or command is only allowed to
database superusers or users who are granted one of the default roles
<code class="literal">pg_read_server_files</code>,
<code class="literal">pg_write_server_files</code>,
or <code class="literal">pg_execute_server_program</code>, since it allows reading
or writing any file or running a program that the server has privileges to
access.
</p><p>
Do not confuse <code class="command">COPY</code> with the
<span class="application">psql</span> instruction
<code class="command"><a class="link" href="app-psql.html#APP-PSQL-META-COMMANDS-COPY">\copy</a></code>. <code class="command">\copy</code> invokes
<code class="command">COPY FROM STDIN</code> or <code class="command">COPY TO
STDOUT</code>, and then fetches/stores the data in a file
accessible to the <span class="application">psql</span> client. Thus,
file accessibility and access rights depend on the client rather
than the server when <code class="command">\copy</code> is used.
</p><p>
It is recommended that the file name used in <code class="command">COPY</code>
always be specified as an absolute path. This is enforced by the
server in the case of <code class="command">COPY TO</code>, but for
<code class="command">COPY FROM</code> you do have the option of reading from
a file specified by a relative path. The path will be interpreted
relative to the working directory of the server process (normally
the cluster's data directory), not the client's working directory.
</p><p>
Executing a command with <code class="literal">PROGRAM</code> might be restricted
by the operating system's access control mechanisms, such as SELinux.
</p><p>
<code class="command">COPY FROM</code> will invoke any triggers and check
constraints on the destination table. However, it will not invoke rules.
</p><p>
For identity columns, the <code class="command">COPY FROM</code> command will always
write the column values provided in the input data, like
the <code class="command">INSERT</code> option <code class="literal">OVERRIDING SYSTEM
VALUE</code>.
</p><p>
<code class="command">COPY</code> input and output is affected by
<code class="varname">DateStyle</code>. To ensure portability to other
<span class="productname">PostgreSQL</span> installations that might use
non-default <code class="varname">DateStyle</code> settings,
<code class="varname">DateStyle</code> should be set to <code class="literal">ISO</code> before
using <code class="command">COPY TO</code>. It is also a good idea to avoid dumping
data with <code class="varname">IntervalStyle</code> set to
<code class="literal">sql_standard</code>, because negative interval values might be
misinterpreted by a server that has a different setting for
<code class="varname">IntervalStyle</code>.
</p><p>
Input data is interpreted according to <code class="literal">ENCODING</code>
option or the current client encoding, and output data is encoded
in <code class="literal">ENCODING</code> or the current client encoding, even
if the data does not pass through the client but is read from or
written to a file directly by the server.
</p><p>
<code class="command">COPY</code> stops operation at the first error. This
should not lead to problems in the event of a <code class="command">COPY
TO</code>, but the target table will already have received
earlier rows in a <code class="command">COPY FROM</code>. These rows will not
be visible or accessible, but they still occupy disk space. This might
amount to a considerable amount of wasted disk space if the failure
happened well into a large copy operation. You might wish to invoke
<code class="command">VACUUM</code> to recover the wasted space.
</p><p>
<code class="literal">FORCE_NULL</code> and <code class="literal">FORCE_NOT_NULL</code> can be used
simultaneously on the same column. This results in converting quoted
null strings to null values and unquoted null strings to empty strings.
</p></div><div class="refsect1" id="id-1.9.3.55.9"><h2>File Formats</h2><div class="refsect2" id="id-1.9.3.55.9.2"><h3>Text Format</h3><p>
When the <code class="literal">text</code> format is used,
the data read or written is a text file with one line per table row.
Columns in a row are separated by the delimiter character.
The column values themselves are strings generated by the
output function, or acceptable to the input function, of each
attribute's data type. The specified null string is used in
place of columns that are null.
<code class="command">COPY FROM</code> will raise an error if any line of the
input file contains more or fewer columns than are expected.
If <code class="literal">OIDS</code> is specified, the OID is read or written as the first column,
preceding the user data columns.
</p><p>
End of data can be represented by a single line containing just
backslash-period (<code class="literal">\.</code>). An end-of-data marker is
not necessary when reading from a file, since the end of file
serves perfectly well; it is needed only when copying data to or from
client applications using pre-3.0 client protocol.
</p><p>
Backslash characters (<code class="literal">\</code>) can be used in the
<code class="command">COPY</code> data to quote data characters that might
otherwise be taken as row or column delimiters. In particular, the
following characters <span class="emphasis"><em>must</em></span> be preceded by a backslash if
they appear as part of a column value: backslash itself,
newline, carriage return, and the current delimiter character.
</p><p>
The specified null string is sent by <code class="command">COPY TO</code> without
adding any backslashes; conversely, <code class="command">COPY FROM</code> matches
the input against the null string before removing backslashes. Therefore,
a null string such as <code class="literal">\N</code> cannot be confused with
the actual data value <code class="literal">\N</code> (which would be represented
as <code class="literal">\\N</code>).
</p><p>
The following special backslash sequences are recognized by
<code class="command">COPY FROM</code>:
</p><div class="informaltable"><table class="informaltable" border="1"><colgroup><col /><col /></colgroup><thead><tr><th>Sequence</th><th>Represents</th></tr></thead><tbody><tr><td><code class="literal">\b</code></td><td>Backspace (ASCII 8)</td></tr><tr><td><code class="literal">\f</code></td><td>Form feed (ASCII 12)</td></tr><tr><td><code class="literal">\n</code></td><td>Newline (ASCII 10)</td></tr><tr><td><code class="literal">\r</code></td><td>Carriage return (ASCII 13)</td></tr><tr><td><code class="literal">\t</code></td><td>Tab (ASCII 9)</td></tr><tr><td><code class="literal">\v</code></td><td>Vertical tab (ASCII 11)</td></tr><tr><td><code class="literal">\</code><em class="replaceable"><code>digits</code></em></td><td>Backslash followed by one to three octal digits specifies
the character with that numeric code</td></tr><tr><td><code class="literal">\x</code><em class="replaceable"><code>digits</code></em></td><td>Backslash <code class="literal">x</code> followed by one or two hex digits specifies
the character with that numeric code</td></tr></tbody></table></div><p>
Presently, <code class="command">COPY TO</code> will never emit an octal or
hex-digits backslash sequence, but it does use the other sequences
listed above for those control characters.
</p><p>
Any other backslashed character that is not mentioned in the above table
will be taken to represent itself. However, beware of adding backslashes
unnecessarily, since that might accidentally produce a string matching the
end-of-data marker (<code class="literal">\.</code>) or the null string (<code class="literal">\N</code> by
default). These strings will be recognized before any other backslash
processing is done.
</p><p>
It is strongly recommended that applications generating <code class="command">COPY</code> data convert
data newlines and carriage returns to the <code class="literal">\n</code> and
<code class="literal">\r</code> sequences respectively. At present it is
possible to represent a data carriage return by a backslash and carriage
return, and to represent a data newline by a backslash and newline.
However, these representations might not be accepted in future releases.
They are also highly vulnerable to corruption if the <code class="command">COPY</code> file is
transferred across different machines (for example, from Unix to Windows
or vice versa).
</p><p>
<code class="command">COPY TO</code> will terminate each row with a Unix-style
newline (<span class="quote">“<span class="quote"><code class="literal">\n</code></span>”</span>). Servers running on Microsoft Windows instead
output carriage return/newline (<span class="quote">“<span class="quote"><code class="literal">\r\n</code></span>”</span>), but only for
<code class="command">COPY</code> to a server file; for consistency across platforms,
<code class="command">COPY TO STDOUT</code> always sends <span class="quote">“<span class="quote"><code class="literal">\n</code></span>”</span>
regardless of server platform.
<code class="command">COPY FROM</code> can handle lines ending with newlines,
carriage returns, or carriage return/newlines. To reduce the risk of
error due to un-backslashed newlines or carriage returns that were
meant as data, <code class="command">COPY FROM</code> will complain if the line
endings in the input are not all alike.
</p></div><div class="refsect2" id="id-1.9.3.55.9.3"><h3>CSV Format</h3><p>
This format option is used for importing and exporting the Comma
Separated Value (<code class="literal">CSV</code>) file format used by many other
programs, such as spreadsheets. Instead of the escaping rules used by
<span class="productname">PostgreSQL</span>'s standard text format, it
produces and recognizes the common CSV escaping mechanism.
</p><p>
The values in each record are separated by the <code class="literal">DELIMITER</code>
character. If the value contains the delimiter character, the
<code class="literal">QUOTE</code> character, the <code class="literal">NULL</code> string, a carriage
return, or line feed character, then the whole value is prefixed and
suffixed by the <code class="literal">QUOTE</code> character, and any occurrence
within the value of a <code class="literal">QUOTE</code> character or the
<code class="literal">ESCAPE</code> character is preceded by the escape character.
You can also use <code class="literal">FORCE_QUOTE</code> to force quotes when outputting
non-<code class="literal">NULL</code> values in specific columns.
</p><p>
The <code class="literal">CSV</code> format has no standard way to distinguish a
<code class="literal">NULL</code> value from an empty string.
<span class="productname">PostgreSQL</span>'s <code class="command">COPY</code> handles this by quoting.
A <code class="literal">NULL</code> is output as the <code class="literal">NULL</code> parameter string
and is not quoted, while a non-<code class="literal">NULL</code> value matching the
<code class="literal">NULL</code> parameter string is quoted. For example, with the
default settings, a <code class="literal">NULL</code> is written as an unquoted empty
string, while an empty string data value is written with double quotes
(<code class="literal">""</code>). Reading values follows similar rules. You can
use <code class="literal">FORCE_NOT_NULL</code> to prevent <code class="literal">NULL</code> input
comparisons for specific columns. You can also use
<code class="literal">FORCE_NULL</code> to convert quoted null string data values to
<code class="literal">NULL</code>.
</p><p>
Because backslash is not a special character in the <code class="literal">CSV</code>
format, <code class="literal">\.</code>, the end-of-data marker, could also appear
as a data value. To avoid any misinterpretation, a <code class="literal">\.</code>
data value appearing as a lone entry on a line is automatically
quoted on output, and on input, if quoted, is not interpreted as the
end-of-data marker. If you are loading a file created by another
application that has a single unquoted column and might have a
value of <code class="literal">\.</code>, you might need to quote that value in the
input file.
</p><div class="note"><h3 class="title">Note</h3><p>
In <code class="literal">CSV</code> format, all characters are significant. A quoted value
surrounded by white space, or any characters other than
<code class="literal">DELIMITER</code>, will include those characters. This can cause
errors if you import data from a system that pads <code class="literal">CSV</code>
lines with white space out to some fixed width. If such a situation
arises you might need to preprocess the <code class="literal">CSV</code> file to remove
the trailing white space, before importing the data into
<span class="productname">PostgreSQL</span>.
</p></div><div class="note"><h3 class="title">Note</h3><p>
CSV format will both recognize and produce CSV files with quoted
values containing embedded carriage returns and line feeds. Thus
the files are not strictly one line per table row like text-format
files.
</p></div><div class="note"><h3 class="title">Note</h3><p>
Many programs produce strange and occasionally perverse CSV files,
so the file format is more a convention than a standard. Thus you
might encounter some files that cannot be imported using this
mechanism, and <code class="command">COPY</code> might produce files that other
programs cannot process.
</p></div></div><div class="refsect2" id="id-1.9.3.55.9.4"><h3>Binary Format</h3><p>
The <code class="literal">binary</code> format option causes all data to be
stored/read as binary format rather than as text. It is
somewhat faster than the text and <code class="literal">CSV</code> formats,
but a binary-format file is less portable across machine architectures and
<span class="productname">PostgreSQL</span> versions.
Also, the binary format is very data type specific; for example
it will not work to output binary data from a <code class="type">smallint</code> column
and read it into an <code class="type">integer</code> column, even though that would work
fine in text format.
</p><p>
The <code class="literal">binary</code> file format consists
of a file header, zero or more tuples containing the row data, and
a file trailer. Headers and data are in network byte order.
</p><div class="note"><h3 class="title">Note</h3><p>
<span class="productname">PostgreSQL</span> releases before 7.4 used a
different binary file format.
</p></div><div class="refsect3" id="id-1.9.3.55.9.4.5"><h4>File Header</h4><p>
The file header consists of 15 bytes of fixed fields, followed
by a variable-length header extension area. The fixed fields are:
</p><div class="variablelist"><dl class="variablelist"><dt><span class="term">Signature</span></dt><dd><p>
11-byte sequence <code class="literal">PGCOPY\n\377\r\n\0</code> — note that the zero byte
is a required part of the signature. (The signature is designed to allow
easy identification of files that have been munged by a non-8-bit-clean
transfer. This signature will be changed by end-of-line-translation
filters, dropped zero bytes, dropped high bits, or parity changes.)
</p></dd><dt><span class="term">Flags field</span></dt><dd><p>
32-bit integer bit mask to denote important aspects of the file format. Bits
are numbered from 0 (<acronym class="acronym">LSB</acronym>) to 31 (<acronym class="acronym">MSB</acronym>). Note that
this field is stored in network byte order (most significant byte first),
as are all the integer fields used in the file format. Bits
16-31 are reserved to denote critical file format issues; a reader
should abort if it finds an unexpected bit set in this range. Bits 0-15
are reserved to signal backwards-compatible format issues; a reader
should simply ignore any unexpected bits set in this range. Currently
only one flag bit is defined, and the rest must be zero:
</p><div class="variablelist"><dl class="variablelist"><dt><span class="term">Bit 16</span></dt><dd><p>
if 1, OIDs are included in the data; if 0, not
</p></dd></dl></div></dd><dt><span class="term">Header extension area length</span></dt><dd><p>
32-bit integer, length in bytes of remainder of header, not including self.
Currently, this is zero, and the first tuple follows
immediately. Future changes to the format might allow additional data
to be present in the header. A reader should silently skip over any header
extension data it does not know what to do with.
</p></dd></dl></div><p>
</p><p>
The header extension area is envisioned to contain a sequence of
self-identifying chunks. The flags field is not intended to tell readers
what is in the extension area. Specific design of header extension contents
is left for a later release.
</p><p>
This design allows for both backwards-compatible header additions (add
header extension chunks, or set low-order flag bits) and
non-backwards-compatible changes (set high-order flag bits to signal such
changes, and add supporting data to the extension area if needed).
</p></div><div class="refsect3" id="id-1.9.3.55.9.4.6"><h4>Tuples</h4><p>
Each tuple begins with a 16-bit integer count of the number of fields in the
tuple. (Presently, all tuples in a table will have the same count, but that
might not always be true.) Then, repeated for each field in the tuple, there
is a 32-bit length word followed by that many bytes of field data. (The
length word does not include itself, and can be zero.) As a special case,
-1 indicates a NULL field value. No value bytes follow in the NULL case.
</p><p>
There is no alignment padding or any other extra data between fields.
</p><p>
Presently, all data values in a binary-format file are
assumed to be in binary format (format code one). It is anticipated that a
future extension might add a header field that allows per-column format codes
to be specified.
</p><p>
To determine the appropriate binary format for the actual tuple data you
should consult the <span class="productname">PostgreSQL</span> source, in
particular the <code class="function">*send</code> and <code class="function">*recv</code> functions for
each column's data type (typically these functions are found in the
<code class="filename">src/backend/utils/adt/</code> directory of the source
distribution).
</p><p>
If OIDs are included in the file, the OID field immediately follows the
field-count word. It is a normal field except that it's not included
in the field-count. In particular it has a length word — this will allow
handling of 4-byte vs. 8-byte OIDs without too much pain, and will allow
OIDs to be shown as null if that ever proves desirable.
</p></div><div class="refsect3" id="id-1.9.3.55.9.4.7"><h4>File Trailer</h4><p>
The file trailer consists of a 16-bit integer word containing -1. This
is easily distinguished from a tuple's field-count word.
</p><p>
A reader should report an error if a field-count word is neither -1
nor the expected number of columns. This provides an extra
check against somehow getting out of sync with the data.
</p></div></div></div><div class="refsect1" id="id-1.9.3.55.10"><h2>Examples</h2><p>
The following example copies a table to the client
using the vertical bar (<code class="literal">|</code>) as the field delimiter:
</p><pre class="programlisting">
COPY country TO STDOUT (DELIMITER '|');
</pre><p>
</p><p>
To copy data from a file into the <code class="literal">country</code> table:
</p><pre class="programlisting">
COPY country FROM '/usr1/proj/bray/sql/country_data';
</pre><p>
</p><p>
To copy into a file just the countries whose names start with 'A':
</p><pre class="programlisting">
COPY (SELECT * FROM country WHERE country_name LIKE 'A%') TO '/usr1/proj/bray/sql/a_list_countries.copy';
</pre><p>
</p><p>
To copy into a compressed file, you can pipe the output through an external
compression program:
</p><pre class="programlisting">
COPY country TO PROGRAM 'gzip > /usr1/proj/bray/sql/country_data.gz';
</pre><p>
</p><p>
Here is a sample of data suitable for copying into a table from
<code class="literal">STDIN</code>:
</p><pre class="programlisting">
AF AFGHANISTAN
AL ALBANIA
DZ ALGERIA
ZM ZAMBIA
ZW ZIMBABWE
</pre><p>
Note that the white space on each line is actually a tab character.
</p><p>
The following is the same data, output in binary format.
The data is shown after filtering through the
Unix utility <code class="command">od -c</code>. The table has three columns;
the first has type <code class="type">char(2)</code>, the second has type <code class="type">text</code>,
and the third has type <code class="type">integer</code>. All the rows have a null value
in the third column.
</p><pre class="programlisting">
0000000 P G C O P Y \n 377 \r \n \0 \0 \0 \0 \0 \0
0000020 \0 \0 \0 \0 003 \0 \0 \0 002 A F \0 \0 \0 013 A
0000040 F G H A N I S T A N 377 377 377 377 \0 003
0000060 \0 \0 \0 002 A L \0 \0 \0 007 A L B A N I
0000100 A 377 377 377 377 \0 003 \0 \0 \0 002 D Z \0 \0 \0
0000120 007 A L G E R I A 377 377 377 377 \0 003 \0 \0
0000140 \0 002 Z M \0 \0 \0 006 Z A M B I A 377 377
0000160 377 377 \0 003 \0 \0 \0 002 Z W \0 \0 \0 \b Z I
0000200 M B A B W E 377 377 377 377 377 377
</pre></div><div class="refsect1" id="id-1.9.3.55.11"><h2>Compatibility</h2><p>
There is no <code class="command">COPY</code> statement in the SQL standard.
</p><p>
The following syntax was used before <span class="productname">PostgreSQL</span>
version 9.0 and is still supported:
</p><pre class="synopsis">
COPY <em class="replaceable"><code>table_name</code></em> [ ( <em class="replaceable"><code>column_name</code></em> [, ...] ) ]
FROM { '<em class="replaceable"><code>filename</code></em>' | STDIN }
[ [ WITH ]
[ BINARY ]
[ OIDS ]
[ DELIMITER [ AS ] '<em class="replaceable"><code>delimiter</code></em>' ]
[ NULL [ AS ] '<em class="replaceable"><code>null string</code></em>' ]
[ CSV [ HEADER ]
[ QUOTE [ AS ] '<em class="replaceable"><code>quote</code></em>' ]
[ ESCAPE [ AS ] '<em class="replaceable"><code>escape</code></em>' ]
[ FORCE NOT NULL <em class="replaceable"><code>column_name</code></em> [, ...] ] ] ]
COPY { <em class="replaceable"><code>table_name</code></em> [ ( <em class="replaceable"><code>column_name</code></em> [, ...] ) ] | ( <em class="replaceable"><code>query</code></em> ) }
TO { '<em class="replaceable"><code>filename</code></em>' | STDOUT }
[ [ WITH ]
[ BINARY ]
[ OIDS ]
[ DELIMITER [ AS ] '<em class="replaceable"><code>delimiter</code></em>' ]
[ NULL [ AS ] '<em class="replaceable"><code>null string</code></em>' ]
[ CSV [ HEADER ]
[ QUOTE [ AS ] '<em class="replaceable"><code>quote</code></em>' ]
[ ESCAPE [ AS ] '<em class="replaceable"><code>escape</code></em>' ]
[ FORCE QUOTE { <em class="replaceable"><code>column_name</code></em> [, ...] | * } ] ] ]
</pre><p>
Note that in this syntax, <code class="literal">BINARY</code> and <code class="literal">CSV</code> are
treated as independent keywords, not as arguments of a <code class="literal">FORMAT</code>
option.
</p><p>
The following syntax was used before <span class="productname">PostgreSQL</span>
version 7.3 and is still supported:
</p><pre class="synopsis">
COPY [ BINARY ] <em class="replaceable"><code>table_name</code></em> [ WITH OIDS ]
FROM { '<em class="replaceable"><code>filename</code></em>' | STDIN }
[ [USING] DELIMITERS '<em class="replaceable"><code>delimiter</code></em>' ]
[ WITH NULL AS '<em class="replaceable"><code>null string</code></em>' ]
COPY [ BINARY ] <em class="replaceable"><code>table_name</code></em> [ WITH OIDS ]
TO { '<em class="replaceable"><code>filename</code></em>' | STDOUT }
[ [USING] DELIMITERS '<em class="replaceable"><code>delimiter</code></em>' ]
[ WITH NULL AS '<em class="replaceable"><code>null string</code></em>' ]
</pre></div></div><div class="navfooter"><hr /><table width="100%" summary="Navigation footer"><tr><td width="40%" align="left"><a accesskey="p" href="sql-commit-prepared.html">Prev</a> </td><td width="20%" align="center"><a accesskey="u" href="sql-commands.html">Up</a></td><td width="40%" align="right"> <a accesskey="n" href="sql-create-access-method.html">Next</a></td></tr><tr><td width="40%" align="left" valign="top">COMMIT PREPARED </td><td width="20%" align="center"><a accesskey="h" href="index.html">Home</a></td><td width="40%" align="right" valign="top"> CREATE ACCESS METHOD</td></tr></table></div></body></html>
