1COPY(7) PostgreSQL 11.3 Documentation COPY(7)
2
6 COPY - copy data between a file and a table
7
9 COPY table_name [ ( column_name [, ...] ) ]
10 FROM { 'filename' | PROGRAM 'command' | STDIN }
11 [ [ WITH ] ( option [, ...] ) ]
12 COPY { table_name [ ( column_name [, ...] ) ] | ( query ) }
14 TO { 'filename' | PROGRAM 'command' | STDOUT }
15 [ [ WITH ] ( option [, ...] ) ]
16 where option can be one of:
18 FORMAT format_name
20 OIDS [ boolean ]
21 FREEZE [ boolean ]
22 DELIMITER 'delimiter_character'
23 NULL 'null_string'
24 HEADER [ boolean ]
25 QUOTE 'quote_character'
26 ESCAPE 'escape_character'
27 FORCE_QUOTE { ( column_name [, ...] ) | * }
28 FORCE_NOT_NULL ( column_name [, ...] )
29 FORCE_NULL ( column_name [, ...] )
30 ENCODING 'encoding_name'
31
33 COPY moves data between PostgreSQL tables and standard file-system
34 files. COPY TO copies the contents of a table to a file, while COPY
35 FROM copies data from a file to a table (appending the data to whatever
36 is in the table already). COPY TO can also copy the results of a
37 SELECT query.
38 If a list of columns is specified, COPY will only copy the data in the
40 specified columns to or from the file. If there are any columns in the
41 table that are not in the column list, COPY FROM will insert the
42 default values for those columns.
43 COPY with a file name instructs the PostgreSQL server to directly read
45 from or write to a file. The file must be accessible by the PostgreSQL
46 user (the user ID the server runs as) and the name must be specified
47 from the viewpoint of the server. When PROGRAM is specified, the server
48 executes the given command and reads from the standard output of the
49 program, or writes to the standard input of the program. The command
50 must be specified from the viewpoint of the server, and be executable
51 by the PostgreSQL user. When STDIN or STDOUT is specified, data is
52 transmitted via the connection between the client and the server.
53
55 table_name
56 The name (optionally schema-qualified) of an existing table.
57 column_name
59 An optional list of columns to be copied. If no column list is
60 specified, all columns of the table will be copied.
61 query
63 A SELECT(7), VALUES(7), INSERT(7), UPDATE(7) or DELETE(7) command
64 whose results are to be copied. Note that parentheses are required
65 around the query.
66 For INSERT, UPDATE and DELETE queries a RETURNING clause must be
68 provided, and the target relation must not have a conditional rule,
69 nor an ALSO rule, nor an INSTEAD rule that expands to multiple
70 statements.
71 filename
73 The path name of the input or output file. An input file name can
74 be an absolute or relative path, but an output file name must be an
75 absolute path. Windows users might need to use an E'' string and
76 double any backslashes used in the path name.
77 PROGRAM
79 A command to execute. In COPY FROM, the input is read from standard
80 output of the command, and in COPY TO, the output is written to the
81 standard input of the command.
82 Note that the command is invoked by the shell, so if you need to
84 pass any arguments to shell command that come from an untrusted
85 source, you must be careful to strip or escape any special
86 characters that might have a special meaning for the shell. For
87 security reasons, it is best to use a fixed command string, or at
88 least avoid passing any user input in it.
89 STDIN
91 Specifies that input comes from the client application.
92 STDOUT
94 Specifies that output goes to the client application.
95 boolean
97 Specifies whether the selected option should be turned on or off.
98 You can write TRUE, ON, or 1 to enable the option, and FALSE, OFF,
99 or 0 to disable it. The boolean value can also be omitted, in which
100 case TRUE is assumed.
101 FORMAT
103 Selects the data format to be read or written: text, csv (Comma
104 Separated Values), or binary. The default is text.
105 OIDS
107 Specifies copying the OID for each row. (An error is raised if OIDS
108 is specified for a table that does not have OIDs, or in the case of
109 copying a query.)
110 FREEZE
112 Requests copying the data with rows already frozen, just as they
113 would be after running the VACUUM FREEZE command. This is intended
114 as a performance option for initial data loading. Rows will be
115 frozen only if the table being loaded has been created or truncated
116 in the current subtransaction, there are no cursors open and there
117 are no older snapshots held by this transaction. It is currently
118 not possible to perform a COPY FREEZE on a partitioned table.
119 Note that all other sessions will immediately be able to see the
121 data once it has been successfully loaded. This violates the normal
122 rules of MVCC visibility and users specifying should be aware of
123 the potential problems this might cause.
124 DELIMITER
126 Specifies the character that separates columns within each row
127 (line) of the file. The default is a tab character in text format,
128 a comma in CSV format. This must be a single one-byte character.
129 This option is not allowed when using binary format.
130 NULL
132 Specifies the string that represents a null value. The default is
133 \N (backslash-N) in text format, and an unquoted empty string in
134 CSV format. You might prefer an empty string even in text format
135 for cases where you don't want to distinguish nulls from empty
136 strings. This option is not allowed when using binary format.
137 Note
139 When using COPY FROM, any data item that matches this string
140 will be stored as a null value, so you should make sure that
141 you use the same string as you used with COPY TO.
142 HEADER
144 Specifies that the file contains a header line with the names of
145 each column in the file. On output, the first line contains the
146 column names from the table, and on input, the first line is
147 ignored. This option is allowed only when using CSV format.
148 QUOTE
150 Specifies the quoting character to be used when a data value is
151 quoted. The default is double-quote. This must be a single one-byte
152 character. This option is allowed only when using CSV format.
153 ESCAPE
155 Specifies the character that should appear before a data character
156 that matches the QUOTE value. The default is the same as the QUOTE
157 value (so that the quoting character is doubled if it appears in
158 the data). This must be a single one-byte character. This option is
159 allowed only when using CSV format.
160 FORCE_QUOTE
162 Forces quoting to be used for all non-NULL values in each specified
163 column. NULL output is never quoted. If * is specified, non-NULL
164 values will be quoted in all columns. This option is allowed only
165 in COPY TO, and only when using CSV format.
166 FORCE_NOT_NULL
168 Do not match the specified columns' values against the null string.
169 In the default case where the null string is empty, this means that
170 empty values will be read as zero-length strings rather than nulls,
171 even when they are not quoted. This option is allowed only in COPY
172 FROM, and only when using CSV format.
173 FORCE_NULL
175 Match the specified columns' values against the null string, even
176 if it has been quoted, and if a match is found set the value to
177 NULL. In the default case where the null string is empty, this
178 converts a quoted empty string into NULL. This option is allowed
179 only in COPY FROM, and only when using CSV format.
180 ENCODING
182 Specifies that the file is encoded in the encoding_name. If this
183 option is omitted, the current client encoding is used. See the
184 Notes below for more details.
185
187 On successful completion, a COPY command returns a command tag of the
188 form
189 COPY count
191 The count is the number of rows copied.
193 Note
195 psql will print this command tag only if the command was not COPY
196 ... TO STDOUT, or the equivalent psql meta-command \copy ... to
197 stdout. This is to prevent confusing the command tag with the data
198 that was just printed.
199
201 COPY TO can only be used with plain tables, not with views. However,
202 you can write COPY (SELECT * FROM viewname) TO ... to copy the current
203 contents of a view.
204 COPY FROM can be used with plain, foreign, or partitioned tables or
206 with views that have INSTEAD OF INSERT triggers.
207 COPY only deals with the specific table named; it does not copy data to
209 or from child tables. Thus for example COPY table TO shows the same
210 data as SELECT * FROM ONLY table. But COPY (SELECT * FROM table) TO ...
211 can be used to dump all of the data in an inheritance hierarchy.
212 You must have select privilege on the table whose values are read by
214 COPY TO, and insert privilege on the table into which values are
215 inserted by COPY FROM. It is sufficient to have column privileges on
216 the column(s) listed in the command.
217 If row-level security is enabled for the table, the relevant SELECT
219 policies will apply to COPY table TO statements. Currently, COPY FROM
220 is not supported for tables with row-level security. Use equivalent
221 INSERT statements instead.
222 Files named in a COPY command are read or written directly by the
224 server, not by the client application. Therefore, they must reside on
225 or be accessible to the database server machine, not the client. They
226 must be accessible to and readable or writable by the PostgreSQL user
227 (the user ID the server runs as), not the client. Similarly, the
228 command specified with PROGRAM is executed directly by the server, not
229 by the client application, must be executable by the PostgreSQL user.
230 COPY naming a file or command is only allowed to database superusers or
231 users who are granted one of the default roles pg_read_server_files,
232 pg_write_server_files, or pg_execute_server_program, since it allows
233 reading or writing any file or running a program that the server has
234 privileges to access.
235 Do not confuse COPY with the psql instruction \copy. \copy invokes
237 COPY FROM STDIN or COPY TO STDOUT, and then fetches/stores the data in
238 a file accessible to the psql client. Thus, file accessibility and
239 access rights depend on the client rather than the server when \copy is
240 used.
241 It is recommended that the file name used in COPY always be specified
243 as an absolute path. This is enforced by the server in the case of COPY
244 TO, but for COPY FROM you do have the option of reading from a file
245 specified by a relative path. The path will be interpreted relative to
246 the working directory of the server process (normally the cluster's
247 data directory), not the client's working directory.
248 Executing a command with PROGRAM might be restricted by the operating
250 system's access control mechanisms, such as SELinux.
251 COPY FROM will invoke any triggers and check constraints on the
253 destination table. However, it will not invoke rules.
254 For identity columns, the COPY FROM command will always write the
256 column values provided in the input data, like the INSERT option
257 OVERRIDING SYSTEM VALUE.
258 COPY input and output is affected by DateStyle. To ensure portability
260 to other PostgreSQL installations that might use non-default DateStyle
261 settings, DateStyle should be set to ISO before using COPY TO. It is
262 also a good idea to avoid dumping data with IntervalStyle set to
263 sql_standard, because negative interval values might be misinterpreted
264 by a server that has a different setting for IntervalStyle.
265 Input data is interpreted according to ENCODING option or the current
267 client encoding, and output data is encoded in ENCODING or the current
268 client encoding, even if the data does not pass through the client but
269 is read from or written to a file directly by the server.
270 COPY stops operation at the first error. This should not lead to
272 problems in the event of a COPY TO, but the target table will already
273 have received earlier rows in a COPY FROM. These rows will not be
274 visible or accessible, but they still occupy disk space. This might
275 amount to a considerable amount of wasted disk space if the failure
276 happened well into a large copy operation. You might wish to invoke
277 VACUUM to recover the wasted space.
278 FORCE_NULL and FORCE_NOT_NULL can be used simultaneously on the same
280 column. This results in converting quoted null strings to null values
281 and unquoted null strings to empty strings.
282
284 Text Format
285 When the text format is used, the data read or written is a text file
286 with one line per table row. Columns in a row are separated by the
287 delimiter character. The column values themselves are strings generated
288 by the output function, or acceptable to the input function, of each
289 attribute's data type. The specified null string is used in place of
290 columns that are null. COPY FROM will raise an error if any line of
291 the input file contains more or fewer columns than are expected. If
292 OIDS is specified, the OID is read or written as the first column,
293 preceding the user data columns.
294 End of data can be represented by a single line containing just
296 backslash-period (\.). An end-of-data marker is not necessary when
297 reading from a file, since the end of file serves perfectly well; it is
298 needed only when copying data to or from client applications using
299 pre-3.0 client protocol.
300 Backslash characters (\) can be used in the COPY data to quote data
302 characters that might otherwise be taken as row or column delimiters.
303 In particular, the following characters must be preceded by a backslash
304 if they appear as part of a column value: backslash itself, newline,
305 carriage return, and the current delimiter character.
306 The specified null string is sent by COPY TO without adding any
308 backslashes; conversely, COPY FROM matches the input against the null
309 string before removing backslashes. Therefore, a null string such as \N
310 cannot be confused with the actual data value \N (which would be
311 represented as \\N).
312 The following special backslash sequences are recognized by COPY FROM:
314 ┌─────────┬────────────────────────────┐
316 │Sequence │ Represents │
317 ├─────────┼────────────────────────────┤
318 │\b │ Backspace (ASCII 8) │
319 ├─────────┼────────────────────────────┤
320 │\f │ Form feed (ASCII 12) │
321 ├─────────┼────────────────────────────┤
322 │\n │ Newline (ASCII 10) │
323 ├─────────┼────────────────────────────┤
324 │\r │ Carriage return (ASCII 13) │
325 ├─────────┼────────────────────────────┤
326 │\t │ Tab (ASCII 9) │
327 ├─────────┼────────────────────────────┤
328 │\v │ Vertical tab (ASCII 11) │
329 ├─────────┼────────────────────────────┤
330 │\digits │ Backslash followed by one │
331 │ │ to three octal digits │
332 │ │ specifies │
333 │ │ the character with │
334 │ │ that numeric code │
335 ├─────────┼────────────────────────────┤
336 │\xdigits │ Backslash x followed by │
337 │ │ one or two hex digits │
338 │ │ specifies │
339 │ │ the character with │
340 │ │ that numeric code │
341 └─────────┴────────────────────────────┘
342 Presently, COPY TO will never emit an octal or hex-digits backslash
343 sequence, but it does use the other sequences listed above for those
344 control characters.
345 Any other backslashed character that is not mentioned in the above
347 table will be taken to represent itself. However, beware of adding
348 backslashes unnecessarily, since that might accidentally produce a
349 string matching the end-of-data marker (\.) or the null string (\N by
350 default). These strings will be recognized before any other backslash
351 processing is done.
352 It is strongly recommended that applications generating COPY data
354 convert data newlines and carriage returns to the \n and \r sequences
355 respectively. At present it is possible to represent a data carriage
356 return by a backslash and carriage return, and to represent a data
357 newline by a backslash and newline. However, these representations
358 might not be accepted in future releases. They are also highly
359 vulnerable to corruption if the COPY file is transferred across
360 different machines (for example, from Unix to Windows or vice versa).
361 COPY TO will terminate each row with a Unix-style newline (“\n”).
363 Servers running on Microsoft Windows instead output carriage
364 return/newline (“\r\n”), but only for COPY to a server file; for
365 consistency across platforms, COPY TO STDOUT always sends “\n”
366 regardless of server platform. COPY FROM can handle lines ending with
367 newlines, carriage returns, or carriage return/newlines. To reduce the
368 risk of error due to un-backslashed newlines or carriage returns that
369 were meant as data, COPY FROM will complain if the line endings in the
370 input are not all alike.
371 CSV Format
373 This format option is used for importing and exporting the Comma
374 Separated Value (CSV) file format used by many other programs, such as
375 spreadsheets. Instead of the escaping rules used by PostgreSQL's
376 standard text format, it produces and recognizes the common CSV
377 escaping mechanism.
378 The values in each record are separated by the DELIMITER character. If
380 the value contains the delimiter character, the QUOTE character, the
381 NULL string, a carriage return, or line feed character, then the whole
382 value is prefixed and suffixed by the QUOTE character, and any
383 occurrence within the value of a QUOTE character or the ESCAPE
384 character is preceded by the escape character. You can also use
385 FORCE_QUOTE to force quotes when outputting non-NULL values in specific
386 columns.
387 The CSV format has no standard way to distinguish a NULL value from an
389 empty string. PostgreSQL's COPY handles this by quoting. A NULL is
390 output as the NULL parameter string and is not quoted, while a non-NULL
391 value matching the NULL parameter string is quoted. For example, with
392 the default settings, a NULL is written as an unquoted empty string,
393 while an empty string data value is written with double quotes ("").
394 Reading values follows similar rules. You can use FORCE_NOT_NULL to
395 prevent NULL input comparisons for specific columns. You can also use
396 FORCE_NULL to convert quoted null string data values to NULL.
397 Because backslash is not a special character in the CSV format, \., the
399 end-of-data marker, could also appear as a data value. To avoid any
400 misinterpretation, a \. data value appearing as a lone entry on a line
401 is automatically quoted on output, and on input, if quoted, is not
402 interpreted as the end-of-data marker. If you are loading a file
403 created by another application that has a single unquoted column and
404 might have a value of \., you might need to quote that value in the
405 input file.
406 Note
408 In CSV format, all characters are significant. A quoted value
409 surrounded by white space, or any characters other than DELIMITER,
410 will include those characters. This can cause errors if you import
411 data from a system that pads CSV lines with white space out to some
412 fixed width. If such a situation arises you might need to
413 preprocess the CSV file to remove the trailing white space, before
414 importing the data into PostgreSQL.
415 Note
417 CSV format will both recognize and produce CSV files with quoted
418 values containing embedded carriage returns and line feeds. Thus
419 the files are not strictly one line per table row like text-format
420 files.
421 Note
423 Many programs produce strange and occasionally perverse CSV files,
424 so the file format is more a convention than a standard. Thus you
425 might encounter some files that cannot be imported using this
426 mechanism, and COPY might produce files that other programs cannot
427 process.
428 Binary Format
430 The binary format option causes all data to be stored/read as binary
431 format rather than as text. It is somewhat faster than the text and CSV
432 formats, but a binary-format file is less portable across machine
433 architectures and PostgreSQL versions. Also, the binary format is very
434 data type specific; for example it will not work to output binary data
435 from a smallint column and read it into an integer column, even though
436 that would work fine in text format.
437 The binary file format consists of a file header, zero or more tuples
439 containing the row data, and a file trailer. Headers and data are in
440 network byte order.
441 Note
443 PostgreSQL releases before 7.4 used a different binary file format.
444 File Header
446 The file header consists of 15 bytes of fixed fields, followed by a
447 variable-length header extension area. The fixed fields are:
448 Signature
450 11-byte sequence PGCOPY\n\377\r\n\0 — note that the zero byte
451 is a required part of the signature. (The signature is designed
452 to allow easy identification of files that have been munged by
453 a non-8-bit-clean transfer. This signature will be changed by
454 end-of-line-translation filters, dropped zero bytes, dropped
455 high bits, or parity changes.)
456 Flags field
458 32-bit integer bit mask to denote important aspects of the file
459 format. Bits are numbered from 0 (LSB) to 31 (MSB). Note that
460 this field is stored in network byte order (most significant
461 byte first), as are all the integer fields used in the file
462 format. Bits 16-31 are reserved to denote critical file format
463 issues; a reader should abort if it finds an unexpected bit set
464 in this range. Bits 0-15 are reserved to signal
465 backwards-compatible format issues; a reader should simply
466 ignore any unexpected bits set in this range. Currently only
467 one flag bit is defined, and the rest must be zero:
468 Bit 16
470 if 1, OIDs are included in the data; if 0, not
471 Header extension area length
473 32-bit integer, length in bytes of remainder of header, not
474 including self. Currently, this is zero, and the first tuple
475 follows immediately. Future changes to the format might allow
476 additional data to be present in the header. A reader should
477 silently skip over any header extension data it does not know
478 what to do with.
479 The header extension area is envisioned to contain a sequence of
481 self-identifying chunks. The flags field is not intended to tell
482 readers what is in the extension area. Specific design of header
483 extension contents is left for a later release.
484 This design allows for both backwards-compatible header additions
486 (add header extension chunks, or set low-order flag bits) and
487 non-backwards-compatible changes (set high-order flag bits to
488 signal such changes, and add supporting data to the extension area
489 if needed).
490 Tuples
492 Each tuple begins with a 16-bit integer count of the number of
493 fields in the tuple. (Presently, all tuples in a table will have
494 the same count, but that might not always be true.) Then, repeated
495 for each field in the tuple, there is a 32-bit length word followed
496 by that many bytes of field data. (The length word does not include
497 itself, and can be zero.) As a special case, -1 indicates a NULL
498 field value. No value bytes follow in the NULL case.
499 There is no alignment padding or any other extra data between
501 fields.
502 Presently, all data values in a binary-format file are assumed to
504 be in binary format (format code one). It is anticipated that a
505 future extension might add a header field that allows per-column
506 format codes to be specified.
507 To determine the appropriate binary format for the actual tuple
509 data you should consult the PostgreSQL source, in particular the
510 *send and *recv functions for each column's data type (typically
511 these functions are found in the src/backend/utils/adt/ directory
512 of the source distribution).
513 If OIDs are included in the file, the OID field immediately follows
515 the field-count word. It is a normal field except that it's not
516 included in the field-count. In particular it has a length word —
517 this will allow handling of 4-byte vs. 8-byte OIDs without too much
518 pain, and will allow OIDs to be shown as null if that ever proves
519 desirable.
520 File Trailer
522 The file trailer consists of a 16-bit integer word containing -1.
523 This is easily distinguished from a tuple's field-count word.
524 A reader should report an error if a field-count word is neither -1
526 nor the expected number of columns. This provides an extra check
527 against somehow getting out of sync with the data.
528
530 The following example copies a table to the client using the vertical
531 bar (|) as the field delimiter:
532 COPY country TO STDOUT (DELIMITER '|');
534 To copy data from a file into the country table:
536 COPY country FROM '/usr1/proj/bray/sql/country_data';
538 To copy into a file just the countries whose names start with 'A':
540 COPY (SELECT * FROM country WHERE country_name LIKE 'A%') TO '/usr1/proj/bray/sql/a_list_countries.copy';
542 To copy into a compressed file, you can pipe the output through an
544 external compression program:
545 COPY country TO PROGRAM 'gzip > /usr1/proj/bray/sql/country_data.gz';
547 Here is a sample of data suitable for copying into a table from STDIN:
549 AF AFGHANISTAN
551 AL ALBANIA
552 DZ ALGERIA
553 ZM ZAMBIA
554 ZW ZIMBABWE
555 Note that the white space on each line is actually a tab character.
557 The following is the same data, output in binary format. The data is
559 shown after filtering through the Unix utility od -c. The table has
560 three columns; the first has type char(2), the second has type text,
561 and the third has type integer. All the rows have a null value in the
562 third column.
563 0000000 P G C O P Y \n 377 \r \n \0 \0 \0 \0 \0 \0
565 0000020 \0 \0 \0 \0 003 \0 \0 \0 002 A F \0 \0 \0 013 A
566 0000040 F G H A N I S T A N 377 377 377 377 \0 003
567 0000060 \0 \0 \0 002 A L \0 \0 \0 007 A L B A N I
568 0000100 A 377 377 377 377 \0 003 \0 \0 \0 002 D Z \0 \0 \0
569 0000120 007 A L G E R I A 377 377 377 377 \0 003 \0 \0
570 0000140 \0 002 Z M \0 \0 \0 006 Z A M B I A 377 377
571 0000160 377 377 \0 003 \0 \0 \0 002 Z W \0 \0 \0 \b Z I
572 0000200 M B A B W E 377 377 377 377 377 377
573
575 There is no COPY statement in the SQL standard.
576 The following syntax was used before PostgreSQL version 9.0 and is
578 still supported:
579 COPY table_name [ ( column_name [, ...] ) ]
581 FROM { 'filename' | STDIN }
582 [ [ WITH ]
583 [ BINARY ]
584 [ OIDS ]
585 [ DELIMITER [ AS ] 'delimiter' ]
586 [ NULL [ AS ] 'null string' ]
587 [ CSV [ HEADER ]
588 [ QUOTE [ AS ] 'quote' ]
589 [ ESCAPE [ AS ] 'escape' ]
590 [ FORCE NOT NULL column_name [, ...] ] ] ]
591 COPY { table_name [ ( column_name [, ...] ) ] | ( query ) }
593 TO { 'filename' | STDOUT }
594 [ [ WITH ]
595 [ BINARY ]
596 [ OIDS ]
597 [ DELIMITER [ AS ] 'delimiter' ]
598 [ NULL [ AS ] 'null string' ]
599 [ CSV [ HEADER ]
600 [ QUOTE [ AS ] 'quote' ]
601 [ ESCAPE [ AS ] 'escape' ]
602 [ FORCE QUOTE { column_name [, ...] | * } ] ] ]
603 Note that in this syntax, BINARY and CSV are treated as independent
605 keywords, not as arguments of a FORMAT option.
606 The following syntax was used before PostgreSQL version 7.3 and is
608 still supported:
609 COPY [ BINARY ] table_name [ WITH OIDS ]
611 FROM { 'filename' | STDIN }
612 [ [USING] DELIMITERS 'delimiter' ]
613 [ WITH NULL AS 'null string' ]
614 COPY [ BINARY ] table_name [ WITH OIDS ]
616 TO { 'filename' | STDOUT }
617 [ [USING] DELIMITERS 'delimiter' ]
618 [ WITH NULL AS 'null string' ]
619PostgreSQL 11.3 2019 COPY(7)