1COPY(7) SQL Commands COPY(7)
2
6 COPY - copy data between a file and a table
7
10 COPY tablename [ ( column [, ...] ) ]
11 FROM { 'filename' | STDIN }
12 [ [ WITH ]
13 [ BINARY ]
14 [ OIDS ]
15 [ DELIMITER [ AS ] 'delimiter' ]
16 [ NULL [ AS ] 'null string' ]
17 [ CSV [ HEADER ]
18 [ QUOTE [ AS ] 'quote' ]
19 [ ESCAPE [ AS ] 'escape' ]
20 [ FORCE NOT NULL column [, ...] ]
21 COPY { tablename [ ( column [, ...] ) ] | ( query ) }
23 TO { 'filename' | STDOUT }
24 [ [ WITH ]
25 [ BINARY ]
26 [ OIDS ]
27 [ DELIMITER [ AS ] 'delimiter' ]
28 [ NULL [ AS ] 'null string' ]
29 [ CSV [ HEADER ]
30 [ QUOTE [ AS ] 'quote' ]
31 [ ESCAPE [ AS ] 'escape' ]
32 [ FORCE QUOTE column [, ...] ]
33
36 COPY moves data between PostgreSQL tables and standard file-system
37 files. COPY TO copies the contents of a table to a file, while COPY
38 FROM copies data from a file to a table (appending the data to whatever
39 is in the table already). COPY TO can also copy the results of a SELECT
40 query.
41 If a list of columns is specified, COPY will only copy the data in the
43 specified columns to or from the file. If there are any columns in the
44 table that are not in the column list, COPY FROM will insert the
45 default values for those columns.
46 COPY with a file name instructs the PostgreSQL server to directly read
48 from or write to a file. The file must be accessible to the server and
49 the name must be specified from the viewpoint of the server. When STDIN
50 or STDOUT is specified, data is transmitted via the connection between
51 the client and the server.
52
54 tablename
55 The name (optionally schema-qualified) of an existing table.
56 column An optional list of columns to be copied. If no column list is
58 specified, all columns of the table will be copied.
59 query A SELECT [select(7)] or VALUES [values(7)] command whose results
61 are to be copied. Note that parentheses are required around the
62 query.
63 filename
65 The absolute path name of the input or output file. Windows
66 users might need to use an E'' string and double backslashes
67 used as path separators.
68 STDIN Specifies that input comes from the client application.
70 STDOUT Specifies that output goes to the client application.
72 BINARY Causes all data to be stored or read in binary format rather
74 than as text. You cannot specify the DELIMITER, NULL, or CSV
75 options in binary mode.
76 OIDS Specifies copying the OID for each row. (An error is raised if
78 OIDS is specified for a table that does not have OIDs, or in the
79 case of copying a query.)
80 delimiter
82 The single ASCII character that separates columns within each
83 row (line) of the file. The default is a tab character in text
84 mode, a comma in CSV mode.
85 null string
87 The string that represents a null value. The default is \N
88 (backslash-N) in text mode, and an unquoted empty string in CSV
89 mode. You might prefer an empty string even in text mode for
90 cases where you don't want to distinguish nulls from empty
91 strings.
92 Note: When using COPY FROM, any data item that matches this
94 string will be stored as a null value, so you should make sure
95 that you use the same string as you used with COPY TO.
96 CSV Selects Comma Separated Value (CSV) mode.
99 HEADER Specifies that the file contains a header line with the names of
101 each column in the file. On output, the first line contains the
102 column names from the table, and on input, the first line is
103 ignored.
104 quote Specifies the ASCII quotation character in CSV mode. The
106 default is double-quote.
107 escape Specifies the ASCII character that should appear before a QUOTE
109 data character value in CSV mode. The default is the QUOTE
110 value (usually double-quote).
111 FORCE QUOTE
113 In CSV COPY TO mode, forces quoting to be used for all non-NULL
114 values in each specified column. NULL output is never quoted.
115 FORCE NOT NULL
117 In CSV COPY FROM mode, process each specified column as though
118 it were quoted and hence not a NULL value. For the default null
119 string in CSV mode (''), this causes missing values to be input
120 as zero-length strings.
121
123 On successful completion, a COPY command returns a command tag of the
124 form
125 COPY count
127 The count is the number of rows copied.
129
131 COPY can only be used with plain tables, not with views. However, you
132 can write COPY (SELECT * FROM viewname) TO ....
133 The BINARY key word causes all data to be stored/read as binary format
135 rather than as text. It is somewhat faster than the normal text mode,
136 but a binary-format file is less portable across machine architectures
137 and PostgreSQL versions. Also, the binary format is very data type
138 specific; for example it will not work to output binary data from a
139 smallint column and read it into an integer column, even though that
140 would work fine in text format.
141 You must have select privilege on the table whose values are read by
143 COPY TO, and insert privilege on the table into which values are
144 inserted by COPY FROM. It is sufficient to have column privileges on
145 the column(s) listed in the command.
146 Files named in a COPY command are read or written directly by the
148 server, not by the client application. Therefore, they must reside on
149 or be accessible to the database server machine, not the client. They
150 must be accessible to and readable or writable by the PostgreSQL user
151 (the user ID the server runs as), not the client. COPY naming a file is
152 only allowed to database superusers, since it allows reading or writing
153 any file that the server has privileges to access.
154 Do not confuse COPY with the psql instruction \copy. \copy invokes COPY
156 FROM STDIN or COPY TO STDOUT, and then fetches/stores the data in a
157 file accessible to the psql client. Thus, file accessibility and access
158 rights depend on the client rather than the server when \copy is used.
159 It is recommended that the file name used in COPY always be specified
161 as an absolute path. This is enforced by the server in the case of COPY
162 TO, but for COPY FROM you do have the option of reading from a file
163 specified by a relative path. The path will be interpreted relative to
164 the working directory of the server process (normally the cluster's
165 data directory), not the client's working directory.
166 COPY FROM will invoke any triggers and check constraints on the desti‐
168 nation table. However, it will not invoke rules.
169 COPY input and output is affected by DateStyle. To ensure portability
171 to other PostgreSQL installations that might use non-default DateStyle
172 settings, DateStyle should be set to ISO before using COPY TO. It is
173 also a good idea to avoid dumping data with IntervalStyle set to
174 sql_standard, because negative interval values might be misinterpreted
175 by a server that has a different setting for IntervalStyle.
176 Input data is interpreted according to the current client encoding, and
178 output data is encoded in the the current client encoding, even if the
179 data does not pass through the client but is read from or written to a
180 file.
181 COPY stops operation at the first error. This should not lead to prob‐
183 lems in the event of a COPY TO, but the target table will already have
184 received earlier rows in a COPY FROM. These rows will not be visible or
185 accessible, but they still occupy disk space. This might amount to a
186 considerable amount of wasted disk space if the failure happened well
187 into a large copy operation. You might wish to invoke VACUUM to recover
188 the wasted space.
189
191 TEXT FORMAT
192 When COPY is used without the BINARY or CSV options, the data read or
193 written is a text file with one line per table row. Columns in a row
194 are separated by the delimiter character. The column values themselves
195 are strings generated by the output function, or acceptable to the
196 input function, of each attribute's data type. The specified null
197 string is used in place of columns that are null. COPY FROM will raise
198 an error if any line of the input file contains more or fewer columns
199 than are expected. If OIDS is specified, the OID is read or written as
200 the first column, preceding the user data columns.
201 End of data can be represented by a single line containing just back‐
203 slash-period (\.). An end-of-data marker is not necessary when reading
204 from a file, since the end of file serves perfectly well; it is needed
205 only when copying data to or from client applications using pre-3.0
206 client protocol.
207 Backslash characters (\) can be used in the COPY data to quote data
209 characters that might otherwise be taken as row or column delimiters.
210 In particular, the following characters must be preceded by a backslash
211 if they appear as part of a column value: backslash itself, newline,
212 carriage return, and the current delimiter character.
213 The specified null string is sent by COPY TO without adding any back‐
215 slashes; conversely, COPY FROM matches the input against the null
216 string before removing backslashes. Therefore, a null string such as \N
217 cannot be confused with the actual data value \N (which would be repre‐
218 sented as \\N).
219 The following special backslash sequences are recognized by COPY FROM:
221 SequenceRepresents\bBackspace (ASCII 8)\fForm feed (ASCII 12)\nNewline
222 (ASCII 10)\rCarriage return (ASCII 13)\tTab (ASCII 9)\vVertical tab
223 (ASCII 11)\digitsBackslash followed by one to three octal digits speci‐
224 fies the character with that numeric code\xdigitsBackslash x followed
225 by one or two hex digits specifies the character with that numeric code
226 Presently, COPY TO will never emit an octal or hex-digits backslash
227 sequence, but it does use the other sequences listed above for those
228 control characters.
229 Any other backslashed character that is not mentioned in the above ta‐
231 ble will be taken to represent itself. However, beware of adding back‐
232 slashes unnecessarily, since that might accidentally produce a string
233 matching the end-of-data marker (\.) or the null string (\N by
234 default). These strings will be recognized before any other backslash
235 processing is done.
236 It is strongly recommended that applications generating COPY data con‐
238 vert data newlines and carriage returns to the \n and \r sequences
239 respectively. At present it is possible to represent a data carriage
240 return by a backslash and carriage return, and to represent a data new‐
241 line by a backslash and newline. However, these representations might
242 not be accepted in future releases. They are also highly vulnerable to
243 corruption if the COPY file is transferred across different machines
244 (for example, from Unix to Windows or vice versa).
245 COPY TO will terminate each row with a Unix-style newline (``\n'').
247 Servers running on Microsoft Windows instead output carriage
248 return/newline (``\r\n''), but only for COPY to a server file; for con‐
249 sistency across platforms, COPY TO STDOUT always sends ``\n'' regard‐
250 less of server platform. COPY FROM can handle lines ending with new‐
251 lines, carriage returns, or carriage return/newlines. To reduce the
252 risk of error due to un-backslashed newlines or carriage returns that
253 were meant as data, COPY FROM will complain if the line endings in the
254 input are not all alike.
255 CSV FORMAT
257 This format is used for importing and exporting the Comma Separated
258 Value (CSV) file format used by many other programs, such as spread‐
259 sheets. Instead of the escaping used by PostgreSQL's standard text
260 mode, it produces and recognizes the common CSV escaping mechanism.
261 The values in each record are separated by the DELIMITER character. If
263 the value contains the delimiter character, the QUOTE character, the
264 NULL string, a carriage return, or line feed character, then the whole
265 value is prefixed and suffixed by the QUOTE character, and any occur‐
266 rence within the value of a QUOTE character or the ESCAPE character is
267 preceded by the escape character. You can also use FORCE QUOTE to
268 force quotes when outputting non-NULL values in specific columns.
269 The CSV format has no standard way to distinguish a NULL value from an
271 empty string. PostgreSQL's COPY handles this by quoting. A NULL is
272 output as the NULL parameter string and is not quoted, while a non-NULL
273 value matching the NULL parameter string is quoted. For example, with
274 the default settings, a NULL is written as an unquoted empty string,
275 while an empty string data value is written with double quotes ("").
276 Reading values follows similar rules. You can use FORCE NOT NULL to
277 prevent NULL input comparisons for specific columns.
278 Because backslash is not a special character in the CSV format, \., the
280 end-of-data marker, could also appear as a data value. To avoid any
281 misinterpretation, a \. data value appearing as a lone entry on a line
282 is automatically quoted on output, and on input, if quoted, is not
283 interpreted as the end-of-data marker. If you are loading a file cre‐
284 ated by another application that has a single unquoted column and might
285 have a value of \., you might need to quote that value in the input
286 file.
287 Note: In CSV mode, all characters are significant. A quoted
289 value surrounded by white space, or any characters other than
290 DELIMITER, will include those characters. This can cause errors
291 if you import data from a system that pads CSV lines with white
292 space out to some fixed width. If such a situation arises you
293 might need to preprocess the CSV file to remove the trailing
294 white space, before importing the data into PostgreSQL.
295 Note: CSV mode will both recognize and produce CSV files with
298 quoted values containing embedded carriage returns and line
299 feeds. Thus the files are not strictly one line per table row
300 like text-mode files.
301 Note: Many programs produce strange and occasionally perverse
304 CSV files, so the file format is more a convention than a stan‐
305 dard. Thus you might encounter some files that cannot be
306 imported using this mechanism, and COPY might produce files that
307 other programs cannot process.
308 BINARY FORMAT
311 The file format used for COPY BINARY changed in PostgreSQL 7.4. The new
312 format consists of a file header, zero or more tuples containing the
313 row data, and a file trailer. Headers and data are now in network byte
314 order.
315 FILE HEADER
317 The file header consists of 15 bytes of fixed fields, followed by a
318 variable-length header extension area. The fixed fields are:
319 Signature
321 11-byte sequence PGCOPY\n\377\r\n\0 — note that the zero byte is
322 a required part of the signature. (The signature is designed to
323 allow easy identification of files that have been munged by a
324 non-8-bit-clean transfer. This signature will be changed by end-
325 of-line-translation filters, dropped zero bytes, dropped high
326 bits, or parity changes.)
327 Flags field
329 32-bit integer bit mask to denote important aspects of the file
330 format. Bits are numbered from 0 (LSB) to 31 (MSB). Note that
331 this field is stored in network byte order (most significant
332 byte first), as are all the integer fields used in the file for‐
333 mat. Bits 16-31 are reserved to denote critical file format
334 issues; a reader should abort if it finds an unexpected bit set
335 in this range. Bits 0-15 are reserved to signal backwards-com‐
336 patible format issues; a reader should simply ignore any unex‐
337 pected bits set in this range. Currently only one flag bit is
338 defined, and the rest must be zero:
339 Bit 16 if 1, OIDs are included in the data; if 0, not
341 Header extension area length
343 32-bit integer, length in bytes of remainder of header, not
344 including self. Currently, this is zero, and the first tuple
345 follows immediately. Future changes to the format might allow
346 additional data to be present in the header. A reader should
347 silently skip over any header extension data it does not know
348 what to do with.
349 The header extension area is envisioned to contain a sequence of self-
351 identifying chunks. The flags field is not intended to tell readers
352 what is in the extension area. Specific design of header extension con‐
353 tents is left for a later release.
354 This design allows for both backwards-compatible header additions (add
356 header extension chunks, or set low-order flag bits) and non-backwards-
357 compatible changes (set high-order flag bits to signal such changes,
358 and add supporting data to the extension area if needed).
359 TUPLES
361 Each tuple begins with a 16-bit integer count of the number of fields
362 in the tuple. (Presently, all tuples in a table will have the same
363 count, but that might not always be true.) Then, repeated for each
364 field in the tuple, there is a 32-bit length word followed by that many
365 bytes of field data. (The length word does not include itself, and can
366 be zero.) As a special case, -1 indicates a NULL field value. No value
367 bytes follow in the NULL case.
368 There is no alignment padding or any other extra data between fields.
370 Presently, all data values in a COPY BINARY file are assumed to be in
372 binary format (format code one). It is anticipated that a future exten‐
373 sion might add a header field that allows per-column format codes to be
374 specified.
375 To determine the appropriate binary format for the actual tuple data
377 you should consult the PostgreSQL source, in particular the *send and
378 *recv functions for each column's data type (typically these functions
379 are found in the src/backend/utils/adt/ directory of the source distri‐
380 bution).
381 If OIDs are included in the file, the OID field immediately follows the
383 field-count word. It is a normal field except that it's not included in
384 the field-count. In particular it has a length word — this will allow
385 handling of 4-byte vs. 8-byte OIDs without too much pain, and will
386 allow OIDs to be shown as null if that ever proves desirable.
387 FILE TRAILER
389 The file trailer consists of a 16-bit integer word containing -1. This
390 is easily distinguished from a tuple's field-count word.
391 A reader should report an error if a field-count word is neither -1 nor
393 the expected number of columns. This provides an extra check against
394 somehow getting out of sync with the data.
395
397 The following example copies a table to the client using the vertical
398 bar (|) as the field delimiter:
399 COPY country TO STDOUT WITH DELIMITER '|';
401 To copy data from a file into the country table:
404 COPY country FROM '/usr1/proj/bray/sql/country_data';
406 To copy into a file just the countries whose names start with 'A':
409 COPY (SELECT * FROM country WHERE country_name LIKE 'A%') TO '/usr1/proj/bray/sql/a_list_countries.copy';
411 Here is a sample of data suitable for copying into a table from STDIN:
414 AF AFGHANISTAN
416 AL ALBANIA
417 DZ ALGERIA
418 ZM ZAMBIA
419 ZW ZIMBABWE
420 Note that the white space on each line is actually a tab character.
422 The following is the same data, output in binary format. The data is
424 shown after filtering through the Unix utility od -c. The table has
425 three columns; the first has type char(2), the second has type text,
426 and the third has type integer. All the rows have a null value in the
427 third column.
428 0000000 P G C O P Y \n 377 \r \n \0 \0 \0 \0 \0 \0
430 0000020 \0 \0 \0 \0 003 \0 \0 \0 002 A F \0 \0 \0 013 A
431 0000040 F G H A N I S T A N 377 377 377 377 \0 003
432 0000060 \0 \0 \0 002 A L \0 \0 \0 007 A L B A N I
433 0000100 A 377 377 377 377 \0 003 \0 \0 \0 002 D Z \0 \0 \0
434 0000120 007 A L G E R I A 377 377 377 377 \0 003 \0 \0
435 0000140 \0 002 Z M \0 \0 \0 006 Z A M B I A 377 377
436 0000160 377 377 \0 003 \0 \0 \0 002 Z W \0 \0 \0 \b Z I
437 0000200 M B A B W E 377 377 377 377 377 377
438
441 There is no COPY statement in the SQL standard.
442 The following syntax was used before PostgreSQL version 7.3 and is
444 still supported:
445 COPY [ BINARY ] tablename [ WITH OIDS ]
447 FROM { 'filename' | STDIN }
448 [ [USING] DELIMITERS 'delimiter' ]
449 [ WITH NULL AS 'null string' ]
450 COPY [ BINARY ] tablename [ WITH OIDS ]
452 TO { 'filename' | STDOUT }
453 [ [USING] DELIMITERS 'delimiter' ]
454 [ WITH NULL AS 'null string' ]
455SQL - Language Statements 2011-09-22 COPY(7)