1INSERT(7) PostgreSQL 12.2 Documentation INSERT(7)
2
6 INSERT - create new rows in a table
7
9 [ WITH [ RECURSIVE ] with_query [, ...] ]
10 INSERT INTO table_name [ AS alias ] [ ( column_name [, ...] ) ]
11 [ OVERRIDING { SYSTEM | USER } VALUE ]
12 { DEFAULT VALUES | VALUES ( { expression | DEFAULT } [, ...] ) [, ...] | query }
13 [ ON CONFLICT [ conflict_target ] conflict_action ]
14 [ RETURNING * | output_expression [ [ AS ] output_name ] [, ...] ]
15 where conflict_target can be one of:
17 ( { index_column_name | ( index_expression ) } [ COLLATE collation ] [ opclass ] [, ...] ) [ WHERE index_predicate ]
19 ON CONSTRAINT constraint_name
20 and conflict_action is one of:
22 DO NOTHING
24 DO UPDATE SET { column_name = { expression | DEFAULT } |
25 ( column_name [, ...] ) = [ ROW ] ( { expression | DEFAULT } [, ...] ) |
26 ( column_name [, ...] ) = ( sub-SELECT )
27 } [, ...]
28 [ WHERE condition ]
29
31 INSERT inserts new rows into a table. One can insert one or more rows
32 specified by value expressions, or zero or more rows resulting from a
33 query.
34 The target column names can be listed in any order. If no list of
36 column names is given at all, the default is all the columns of the
37 table in their declared order; or the first N column names, if there
38 are only N columns supplied by the VALUES clause or query. The values
39 supplied by the VALUES clause or query are associated with the explicit
40 or implicit column list left-to-right.
41 Each column not present in the explicit or implicit column list will be
43 filled with a default value, either its declared default value or null
44 if there is none.
45 If the expression for any column is not of the correct data type,
47 automatic type conversion will be attempted.
48 ON CONFLICT can be used to specify an alternative action to raising a
50 unique constraint or exclusion constraint violation error. (See ON
51 CONFLICT Clause below.)
52 The optional RETURNING clause causes INSERT to compute and return
54 value(s) based on each row actually inserted (or updated, if an ON
55 CONFLICT DO UPDATE clause was used). This is primarily useful for
56 obtaining values that were supplied by defaults, such as a serial
57 sequence number. However, any expression using the table's columns is
58 allowed. The syntax of the RETURNING list is identical to that of the
59 output list of SELECT. Only rows that were successfully inserted or
60 updated will be returned. For example, if a row was locked but not
61 updated because an ON CONFLICT DO UPDATE ... WHERE clause condition was
62 not satisfied, the row will not be returned.
63 You must have INSERT privilege on a table in order to insert into it.
65 If ON CONFLICT DO UPDATE is present, UPDATE privilege on the table is
66 also required.
67 If a column list is specified, you only need INSERT privilege on the
69 listed columns. Similarly, when ON CONFLICT DO UPDATE is specified, you
70 only need UPDATE privilege on the column(s) that are listed to be
71 updated. However, ON CONFLICT DO UPDATE also requires SELECT privilege
72 on any column whose values are read in the ON CONFLICT DO UPDATE
73 expressions or condition.
74 Use of the RETURNING clause requires SELECT privilege on all columns
76 mentioned in RETURNING. If you use the query clause to insert rows from
77 a query, you of course need to have SELECT privilege on any table or
78 column used in the query.
79
81 Inserting
82 This section covers parameters that may be used when only inserting new
83 rows. Parameters exclusively used with the ON CONFLICT clause are
84 described separately.
85 with_query
87 The WITH clause allows you to specify one or more subqueries that
88 can be referenced by name in the INSERT query. See Section 7.8 and
89 SELECT(7) for details.
90 It is possible for the query (SELECT statement) to also contain a
92 WITH clause. In such a case both sets of with_query can be
93 referenced within the query, but the second one takes precedence
94 since it is more closely nested.
95 table_name
97 The name (optionally schema-qualified) of an existing table.
98 alias
100 A substitute name for table_name. When an alias is provided, it
101 completely hides the actual name of the table. This is particularly
102 useful when ON CONFLICT DO UPDATE targets a table named excluded,
103 since that will otherwise be taken as the name of the special table
104 representing rows proposed for insertion.
105 column_name
107 The name of a column in the table named by table_name. The column
108 name can be qualified with a subfield name or array subscript, if
109 needed. (Inserting into only some fields of a composite column
110 leaves the other fields null.) When referencing a column with ON
111 CONFLICT DO UPDATE, do not include the table's name in the
112 specification of a target column. For example, INSERT INTO
113 table_name ... ON CONFLICT DO UPDATE SET table_name.col = 1 is
114 invalid (this follows the general behavior for UPDATE).
115 OVERRIDING SYSTEM VALUE
117 Without this clause, it is an error to specify an explicit value
118 (other than DEFAULT) for an identity column defined as GENERATED
119 ALWAYS. This clause overrides that restriction.
120 OVERRIDING USER VALUE
122 If this clause is specified, then any values supplied for identity
123 columns defined as GENERATED BY DEFAULT are ignored and the default
124 sequence-generated values are applied.
125 This clause is useful for example when copying values between
127 tables. Writing INSERT INTO tbl2 OVERRIDING USER VALUE SELECT *
128 FROM tbl1 will copy from tbl1 all columns that are not identity
129 columns in tbl2 while values for the identity columns in tbl2 will
130 be generated by the sequences associated with tbl2.
131 DEFAULT VALUES
133 All columns will be filled with their default values. (An
134 OVERRIDING clause is not permitted in this form.)
135 expression
137 An expression or value to assign to the corresponding column.
138 DEFAULT
140 The corresponding column will be filled with its default value.
141 query
143 A query (SELECT statement) that supplies the rows to be inserted.
144 Refer to the SELECT(7) statement for a description of the syntax.
145 output_expression
147 An expression to be computed and returned by the INSERT command
148 after each row is inserted or updated. The expression can use any
149 column names of the table named by table_name. Write * to return
150 all columns of the inserted or updated row(s).
151 output_name
153 A name to use for a returned column.
154 ON CONFLICT Clause
156 The optional ON CONFLICT clause specifies an alternative action to
157 raising a unique violation or exclusion constraint violation error. For
158 each individual row proposed for insertion, either the insertion
159 proceeds, or, if an arbiter constraint or index specified by
160 conflict_target is violated, the alternative conflict_action is taken.
161 ON CONFLICT DO NOTHING simply avoids inserting a row as its alternative
162 action. ON CONFLICT DO UPDATE updates the existing row that conflicts
163 with the row proposed for insertion as its alternative action.
164 conflict_target can perform unique index inference. When performing
166 inference, it consists of one or more index_column_name columns and/or
167 index_expression expressions, and an optional index_predicate. All
168 table_name unique indexes that, without regard to order, contain
169 exactly the conflict_target-specified columns/expressions are inferred
170 (chosen) as arbiter indexes. If an index_predicate is specified, it
171 must, as a further requirement for inference, satisfy arbiter indexes.
172 Note that this means a non-partial unique index (a unique index without
173 a predicate) will be inferred (and thus used by ON CONFLICT) if such an
174 index satisfying every other criteria is available. If an attempt at
175 inference is unsuccessful, an error is raised.
176 ON CONFLICT DO UPDATE guarantees an atomic INSERT or UPDATE outcome;
178 provided there is no independent error, one of those two outcomes is
179 guaranteed, even under high concurrency. This is also known as UPSERT —
180 “UPDATE or INSERT”.
181 conflict_target
183 Specifies which conflicts ON CONFLICT takes the alternative action
184 on by choosing arbiter indexes. Either performs unique index
185 inference, or names a constraint explicitly. For ON CONFLICT DO
186 NOTHING, it is optional to specify a conflict_target; when omitted,
187 conflicts with all usable constraints (and unique indexes) are
188 handled. For ON CONFLICT DO UPDATE, a conflict_target must be
189 provided.
190 conflict_action
192 conflict_action specifies an alternative ON CONFLICT action. It can
193 be either DO NOTHING, or a DO UPDATE clause specifying the exact
194 details of the UPDATE action to be performed in case of a conflict.
195 The SET and WHERE clauses in ON CONFLICT DO UPDATE have access to
196 the existing row using the table's name (or an alias), and to rows
197 proposed for insertion using the special excluded table. SELECT
198 privilege is required on any column in the target table where
199 corresponding excluded columns are read.
200 Note that the effects of all per-row BEFORE INSERT triggers are
202 reflected in excluded values, since those effects may have
203 contributed to the row being excluded from insertion.
204 index_column_name
206 The name of a table_name column. Used to infer arbiter indexes.
207 Follows CREATE INDEX format. SELECT privilege on index_column_name
208 is required.
209 index_expression
211 Similar to index_column_name, but used to infer expressions on
212 table_name columns appearing within index definitions (not simple
213 columns). Follows CREATE INDEX format. SELECT privilege on any
214 column appearing within index_expression is required.
215 collation
217 When specified, mandates that corresponding index_column_name or
218 index_expression use a particular collation in order to be matched
219 during inference. Typically this is omitted, as collations usually
220 do not affect whether or not a constraint violation occurs. Follows
221 CREATE INDEX format.
222 opclass
224 When specified, mandates that corresponding index_column_name or
225 index_expression use particular operator class in order to be
226 matched during inference. Typically this is omitted, as the
227 equality semantics are often equivalent across a type's operator
228 classes anyway, or because it's sufficient to trust that the
229 defined unique indexes have the pertinent definition of equality.
230 Follows CREATE INDEX format.
231 index_predicate
233 Used to allow inference of partial unique indexes. Any indexes that
234 satisfy the predicate (which need not actually be partial indexes)
235 can be inferred. Follows CREATE INDEX format. SELECT privilege on
236 any column appearing within index_predicate is required.
237 constraint_name
239 Explicitly specifies an arbiter constraint by name, rather than
240 inferring a constraint or index.
241 condition
243 An expression that returns a value of type boolean. Only rows for
244 which this expression returns true will be updated, although all
245 rows will be locked when the ON CONFLICT DO UPDATE action is taken.
246 Note that condition is evaluated last, after a conflict has been
247 identified as a candidate to update.
248 Note that exclusion constraints are not supported as arbiters with ON
250 CONFLICT DO UPDATE. In all cases, only NOT DEFERRABLE constraints and
251 unique indexes are supported as arbiters.
252 INSERT with an ON CONFLICT DO UPDATE clause is a “deterministic”
254 statement. This means that the command will not be allowed to affect
255 any single existing row more than once; a cardinality violation error
256 will be raised when this situation arises. Rows proposed for insertion
257 should not duplicate each other in terms of attributes constrained by
258 an arbiter index or constraint.
259 Note that it is currently not supported for the ON CONFLICT DO UPDATE
261 clause of an INSERT applied to a partitioned table to update the
262 partition key of a conflicting row such that it requires the row be
263 moved to a new partition.
264 Tip
266 It is often preferable to use unique index inference rather than
267 naming a constraint directly using ON CONFLICT ON CONSTRAINT
268 constraint_name. Inference will continue to work correctly when
269 the underlying index is replaced by another more or less equivalent
270 index in an overlapping way, for example when using CREATE UNIQUE
271 INDEX ... CONCURRENTLY before dropping the index being replaced.
272
274 On successful completion, an INSERT command returns a command tag of
275 the form
276 INSERT oid count
278 The count is the number of rows inserted or updated. oid is always 0
280 (it used to be the OID assigned to the inserted row if count was
281 exactly one and the target table was declared WITH OIDS and 0
282 otherwise, but creating a table WITH OIDS is not supported anymore).
283 If the INSERT command contains a RETURNING clause, the result will be
285 similar to that of a SELECT statement containing the columns and values
286 defined in the RETURNING list, computed over the row(s) inserted or
287 updated by the command.
288
290 If the specified table is a partitioned table, each row is routed to
291 the appropriate partition and inserted into it. If the specified table
292 is a partition, an error will occur if one of the input rows violates
293 the partition constraint.
294
296 Insert a single row into table films:
297 INSERT INTO films VALUES
299 ('UA502', 'Bananas', 105, '1971-07-13', 'Comedy', '82 minutes');
300 In this example, the len column is omitted and therefore it will have
302 the default value:
303 INSERT INTO films (code, title, did, date_prod, kind)
305 VALUES ('T_601', 'Yojimbo', 106, '1961-06-16', 'Drama');
306 This example uses the DEFAULT clause for the date columns rather than
308 specifying a value:
309 INSERT INTO films VALUES
311 ('UA502', 'Bananas', 105, DEFAULT, 'Comedy', '82 minutes');
312 INSERT INTO films (code, title, did, date_prod, kind)
313 VALUES ('T_601', 'Yojimbo', 106, DEFAULT, 'Drama');
314 To insert a row consisting entirely of default values:
316 INSERT INTO films DEFAULT VALUES;
318 To insert multiple rows using the multirow VALUES syntax:
320 INSERT INTO films (code, title, did, date_prod, kind) VALUES
322 ('B6717', 'Tampopo', 110, '1985-02-10', 'Comedy'),
323 ('HG120', 'The Dinner Game', 140, DEFAULT, 'Comedy');
324 This example inserts some rows into table films from a table tmp_films
326 with the same column layout as films:
327 INSERT INTO films SELECT * FROM tmp_films WHERE date_prod < '2004-05-07';
329 This example inserts into array columns:
331 -- Create an empty 3x3 gameboard for noughts-and-crosses
333 INSERT INTO tictactoe (game, board[1:3][1:3])
334 VALUES (1, '{{" "," "," "},{" "," "," "},{" "," "," "}}');
335 -- The subscripts in the above example aren't really needed
336 INSERT INTO tictactoe (game, board)
337 VALUES (2, '{{X," "," "},{" ",O," "},{" ",X," "}}');
338 Insert a single row into table distributors, returning the sequence
340 number generated by the DEFAULT clause:
341 INSERT INTO distributors (did, dname) VALUES (DEFAULT, 'XYZ Widgets')
343 RETURNING did;
344 Increment the sales count of the salesperson who manages the account
346 for Acme Corporation, and record the whole updated row along with
347 current time in a log table:
348 WITH upd AS (
350 UPDATE employees SET sales_count = sales_count + 1 WHERE id =
351 (SELECT sales_person FROM accounts WHERE name = 'Acme Corporation')
352 RETURNING *
353 )
354 INSERT INTO employees_log SELECT *, current_timestamp FROM upd;
355 Insert or update new distributors as appropriate. Assumes a unique
357 index has been defined that constrains values appearing in the did
358 column. Note that the special excluded table is used to reference
359 values originally proposed for insertion:
360 INSERT INTO distributors (did, dname)
362 VALUES (5, 'Gizmo Transglobal'), (6, 'Associated Computing, Inc')
363 ON CONFLICT (did) DO UPDATE SET dname = EXCLUDED.dname;
364 Insert a distributor, or do nothing for rows proposed for insertion
366 when an existing, excluded row (a row with a matching constrained
367 column or columns after before row insert triggers fire) exists.
368 Example assumes a unique index has been defined that constrains values
369 appearing in the did column:
370 INSERT INTO distributors (did, dname) VALUES (7, 'Redline GmbH')
372 ON CONFLICT (did) DO NOTHING;
373 Insert or update new distributors as appropriate. Example assumes a
375 unique index has been defined that constrains values appearing in the
376 did column. WHERE clause is used to limit the rows actually updated
377 (any existing row not updated will still be locked, though):
378 -- Don't update existing distributors based in a certain ZIP code
380 INSERT INTO distributors AS d (did, dname) VALUES (8, 'Anvil Distribution')
381 ON CONFLICT (did) DO UPDATE
382 SET dname = EXCLUDED.dname || ' (formerly ' || d.dname || ')'
383 WHERE d.zipcode <> '21201';
384 -- Name a constraint directly in the statement (uses associated
386 -- index to arbitrate taking the DO NOTHING action)
387 INSERT INTO distributors (did, dname) VALUES (9, 'Antwerp Design')
388 ON CONFLICT ON CONSTRAINT distributors_pkey DO NOTHING;
389 Insert new distributor if possible; otherwise DO NOTHING. Example
391 assumes a unique index has been defined that constrains values
392 appearing in the did column on a subset of rows where the is_active
393 Boolean column evaluates to true:
394 -- This statement could infer a partial unique index on "did"
396 -- with a predicate of "WHERE is_active", but it could also
397 -- just use a regular unique constraint on "did"
398 INSERT INTO distributors (did, dname) VALUES (10, 'Conrad International')
399 ON CONFLICT (did) WHERE is_active DO NOTHING;
400
402 INSERT conforms to the SQL standard, except that the RETURNING clause
403 is a PostgreSQL extension, as is the ability to use WITH with INSERT,
404 and the ability to specify an alternative action with ON CONFLICT.
405 Also, the case in which a column name list is omitted, but not all the
406 columns are filled from the VALUES clause or query, is disallowed by
407 the standard.
408 The SQL standard specifies that OVERRIDING SYSTEM VALUE can only be
410 specified if an identity column that is generated always exists.
411 PostgreSQL allows the clause in any case and ignores it if it is not
412 applicable.
413 Possible limitations of the query clause are documented under
415 SELECT(7).
416PostgreSQL 12.2 2020 INSERT(7)