2 $Header: /cvsroot/pgsql/doc/src/sgml/syntax.sgml,v 1.86 2003/11/04 09:55:39 petere Exp $
5 <chapter id="sql-syntax">
6 <title>SQL Syntax</title>
8 <indexterm zone="sql-syntax">
9 <primary>syntax</primary>
10 <secondary>SQL</secondary>
14 This chapter describes the syntax of SQL. It forms the foundation
15 for understanding the following chapters which will go into detail
16 about how the SQL commands are applied to define and modify data.
20 We also advise users who are already familiar with SQL to read this
21 chapter carefully because there are several rules and concepts that
22 are implemented inconsistently among SQL databases or that are
23 specific to <productname>PostgreSQL</productname>.
26 <sect1 id="sql-syntax-lexical">
27 <title>Lexical Structure</title>
30 <primary>token</primary>
34 SQL input consists of a sequence of
35 <firstterm>commands</firstterm>. A command is composed of a
36 sequence of <firstterm>tokens</firstterm>, terminated by a
37 semicolon (<quote>;</quote>). The end of the input stream also
38 terminates a command. Which tokens are valid depends on the syntax
39 of the particular command.
43 A token can be a <firstterm>key word</firstterm>, an
44 <firstterm>identifier</firstterm>, a <firstterm>quoted
45 identifier</firstterm>, a <firstterm>literal</firstterm> (or
46 constant), or a special character symbol. Tokens are normally
47 separated by whitespace (space, tab, newline), but need not be if
48 there is no ambiguity (which is generally only the case if a
49 special character is adjacent to some other token type).
53 Additionally, <firstterm>comments</firstterm> can occur in SQL
54 input. They are not tokens, they are effectively equivalent to
59 For example, the following is (syntactically) valid SQL input:
61 SELECT * FROM MY_TABLE;
62 UPDATE MY_TABLE SET A = 5;
63 INSERT INTO MY_TABLE VALUES (3, 'hi there');
65 This is a sequence of three commands, one per line (although this
66 is not required; more than one command can be on a line, and
67 commands can usefully be split across lines).
71 The SQL syntax is not very consistent regarding what tokens
72 identify commands and which are operands or parameters. The first
73 few tokens are generally the command name, so in the above example
74 we would usually speak of a <quote>SELECT</quote>, an
75 <quote>UPDATE</quote>, and an <quote>INSERT</quote> command. But
76 for instance the <command>UPDATE</command> command always requires
77 a <token>SET</token> token to appear in a certain position, and
78 this particular variation of <command>INSERT</command> also
79 requires a <token>VALUES</token> in order to be complete. The
80 precise syntax rules for each command are described in <xref linkend="reference">.
83 <sect2 id="sql-syntax-identifiers">
84 <title>Identifiers and Key Words</title>
86 <indexterm zone="sql-syntax-identifiers">
87 <primary>identifier</primary>
88 <secondary>syntax of</secondary>
91 <indexterm zone="sql-syntax-identifiers">
92 <primary>name</primary>
93 <secondary>syntax of</secondary>
96 <indexterm zone="sql-syntax-identifiers">
97 <primary>key word</primary>
98 <secondary>syntax of</secondary>
102 Tokens such as <token>SELECT</token>, <token>UPDATE</token>, or
103 <token>VALUES</token> in the example above are examples of
104 <firstterm>key words</firstterm>, that is, words that have a fixed
105 meaning in the SQL language. The tokens <token>MY_TABLE</token>
106 and <token>A</token> are examples of
107 <firstterm>identifiers</firstterm>. They identify names of
108 tables, columns, or other database objects, depending on the
109 command they are used in. Therefore they are sometimes simply
110 called <quote>names</quote>. Key words and identifiers have the
111 same lexical structure, meaning that one cannot know whether a
112 token is an identifier or a key word without knowing the language.
113 A complete list of key words can be found in <xref
114 linkend="sql-keywords-appendix">.
118 SQL identifiers and key words must begin with a letter
119 (<literal>a</literal>-<literal>z</literal>, but also letters with
120 diacritical marks and non-Latin letters) or an underscore
121 (<literal>_</literal>). Subsequent characters in an identifier or
122 key word can be letters, underscores, digits
123 (<literal>0</literal>-<literal>9</literal>), or dollar signs
124 (<literal>$</>). Note that dollar signs are not allowed in identifiers
125 according to the letter of the SQL standard, so their use may render
126 applications less portable.
127 The SQL standard will not define a key word that contains
128 digits or starts or ends with an underscore, so identifiers of this
129 form are safe against possible conflict with future extensions of the
134 <indexterm><primary>identifier</primary><secondary>length</secondary></indexterm>
135 The system uses no more than <symbol>NAMEDATALEN</symbol>-1
136 characters of an identifier; longer names can be written in
137 commands, but they will be truncated. By default,
138 <symbol>NAMEDATALEN</symbol> is 64 so the maximum identifier
139 length is 63. If this limit is problematic, it can be raised by
140 changing the <symbol>NAMEDATALEN</symbol> constant in
141 <filename>src/include/postgres_ext.h</filename>.
146 <primary>case sensitivity</primary>
147 <secondary>of SQL commands</secondary>
149 Identifier and key word names are case insensitive. Therefore
151 UPDATE MY_TABLE SET A = 5;
153 can equivalently be written as
155 uPDaTE my_TabLE SeT a = 5;
157 A convention often used is to write key words in upper
158 case and names in lower case, e.g.,
160 UPDATE my_table SET a = 5;
166 <primary>quotation marks</primary>
167 <secondary>and identifiers</secondary>
169 There is a second kind of identifier: the <firstterm>delimited
170 identifier</firstterm> or <firstterm>quoted
171 identifier</firstterm>. It is formed by enclosing an arbitrary
172 sequence of characters in double-quotes
173 (<literal>"</literal>). <!-- " font-lock mania --> A delimited
174 identifier is always an identifier, never a key word. So
175 <literal>"select"</literal> could be used to refer to a column or
176 table named <quote>select</quote>, whereas an unquoted
177 <literal>select</literal> would be taken as a key word and
178 would therefore provoke a parse error when used where a table or
179 column name is expected. The example can be written with quoted
180 identifiers like this:
182 UPDATE "my_table" SET "a" = 5;
187 Quoted identifiers can contain any character other than a double
188 quote itself. To include a double quote, write two double quotes.
189 This allows constructing table or column names that would
190 otherwise not be possible, such as ones containing spaces or
191 ampersands. The length limitation still applies.
195 Quoting an identifier also makes it case-sensitive, whereas
196 unquoted names are always folded to lower case. For example, the
197 identifiers <literal>FOO</literal>, <literal>foo</literal>, and
198 <literal>"foo"</literal> are considered the same by
199 <productname>PostgreSQL</productname>, but
200 <literal>"Foo"</literal> and <literal>"FOO"</literal> are
201 different from these three and each other. (The folding of
202 unquoted names to lower case in <productname>PostgreSQL</> is
203 incompatible with the SQL standard, which says that unquoted names
204 should be folded to upper case. Thus, <literal>foo</literal>
205 should be equivalent to <literal>"FOO"</literal> not
206 <literal>"foo"</literal> according to the standard. If you want
207 to write portable applications you are advised to always quote a
208 particular name or never quote it.
213 <sect2 id="sql-syntax-constants">
214 <title>Constants</title>
216 <indexterm zone="sql-syntax-constants">
217 <primary>constant</primary>
221 There are three kinds of <firstterm>implicitly-typed
222 constants</firstterm> in <productname>PostgreSQL</productname>:
223 strings, bit strings, and numbers.
224 Constants can also be specified with explicit types, which can
225 enable more accurate representation and more efficient handling by
226 the system. The implicit constants are described below; explicit
227 constants are discussed afterwards.
230 <sect3 id="sql-syntax-strings">
231 <title>String Constants</title>
233 <indexterm zone="sql-syntax-strings">
234 <primary>character string</primary>
235 <secondary>constant</secondary>
240 <primary>quotation marks</primary>
241 <secondary>escaping</secondary>
243 A string constant in SQL is an arbitrary sequence of characters
244 bounded by single quotes (<quote>'</quote>), e.g., <literal>'This
245 is a string'</literal>. SQL allows single quotes to be embedded
246 in strings by typing two adjacent single quotes (e.g.,
247 <literal>'Dianne''s horse'</literal>). In
248 <productname>PostgreSQL</productname> single quotes may
249 alternatively be escaped with a backslash (<quote>\</quote>,
250 e.g., <literal>'Dianne\'s horse'</literal>).
254 C-style backslash escapes are also available:
255 <literal>\b</literal> is a backspace, <literal>\f</literal> is a
256 form feed, <literal>\n</literal> is a newline,
257 <literal>\r</literal> is a carriage return, <literal>\t</literal>
258 is a tab, and <literal>\<replaceable>xxx</replaceable></literal>,
259 where <replaceable>xxx</replaceable> is an octal number, is a
260 byte with the corresponding code. (It is your responsibility
261 that the byte sequences you create are valid characters in the
262 server character set encoding.) Any other character following a
263 backslash is taken literally. Thus, to include a backslash in a
264 string constant, type two backslashes.
268 The character with the code zero cannot be in a string constant.
272 Two string constants that are only separated by whitespace
273 <emphasis>with at least one newline</emphasis> are concatenated
274 and effectively treated as if the string had been written in one
275 constant. For example:
288 is not valid syntax. (This slightly bizarre behavior is specified
289 by <acronym>SQL</acronym>; <productname>PostgreSQL</productname> is
290 following the standard.)
294 <sect3 id="sql-syntax-bit-strings">
295 <title>Bit-String Constants</title>
297 <indexterm zone="sql-syntax-bit-strings">
298 <primary>bit string</primary>
299 <secondary>constant</secondary>
303 Bit-string constants look like string constants with a
304 <literal>B</literal> (upper or lower case) immediately before the
305 opening quote (no intervening whitespace), e.g.,
306 <literal>B'1001'</literal>. The only characters allowed within
307 bit-string constants are <literal>0</literal> and
308 <literal>1</literal>.
312 Alternatively, bit-string constants can be specified in hexadecimal
313 notation, using a leading <literal>X</literal> (upper or lower case),
314 e.g., <literal>X'1FF'</literal>. This notation is equivalent to
315 a bit-string constant with four binary digits for each hexadecimal digit.
319 Both forms of bit-string constant can be continued
320 across lines in the same way as regular string constants.
325 <title>Numeric Constants</title>
328 <primary>number</primary>
329 <secondary>constant</secondary>
333 Numeric constants are accepted in these general forms:
335 <replaceable>digits</replaceable>
336 <replaceable>digits</replaceable>.<optional><replaceable>digits</replaceable></optional><optional>e<optional>+-</optional><replaceable>digits</replaceable></optional>
337 <optional><replaceable>digits</replaceable></optional>.<replaceable>digits</replaceable><optional>e<optional>+-</optional><replaceable>digits</replaceable></optional>
338 <replaceable>digits</replaceable>e<optional>+-</optional><replaceable>digits</replaceable>
340 where <replaceable>digits</replaceable> is one or more decimal
341 digits (0 through 9). At least one digit must be before or after the
342 decimal point, if one is used. At least one digit must follow the
343 exponent marker (<literal>e</literal>), if one is present.
344 There may not be any spaces or other characters embedded in the
345 constant. Note that any leading plus or minus sign is not actually
346 considered part of the constant; it is an operator applied to the
351 These are some examples of valid numeric constants:
363 <indexterm><primary>integer</primary></indexterm>
364 <indexterm><primary>bigint</primary></indexterm>
365 <indexterm><primary>numeric</primary></indexterm>
366 A numeric constant that contains neither a decimal point nor an
367 exponent is initially presumed to be type <type>integer</> if its
368 value fits in type <type>integer</> (32 bits); otherwise it is
369 presumed to be type <type>bigint</> if its
370 value fits in type <type>bigint</> (64 bits); otherwise it is
371 taken to be type <type>numeric</>. Constants that contain decimal
372 points and/or exponents are always initially presumed to be type
377 The initially assigned data type of a numeric constant is just a
378 starting point for the type resolution algorithms. In most cases
379 the constant will be automatically coerced to the most
380 appropriate type depending on context. When necessary, you can
381 force a numeric value to be interpreted as a specific data type
382 by casting it.<indexterm><primary>type cast</primary></indexterm>
383 For example, you can force a numeric value to be treated as type
384 <type>real</> (<type>float4</>) by writing
387 REAL '1.23' -- string style
388 1.23::REAL -- PostgreSQL (historical) style
393 <sect3 id="sql-syntax-constants-generic">
394 <title>Constants of Other Types</title>
397 <primary>data type</primary>
398 <secondary>constant</secondary>
402 A constant of an <emphasis>arbitrary</emphasis> type can be
403 entered using any one of the following notations:
405 <replaceable>type</replaceable> '<replaceable>string</replaceable>'
406 '<replaceable>string</replaceable>'::<replaceable>type</replaceable>
407 CAST ( '<replaceable>string</replaceable>' AS <replaceable>type</replaceable> )
409 The string's text is passed to the input conversion
410 routine for the type called <replaceable>type</replaceable>. The
411 result is a constant of the indicated type. The explicit type
412 cast may be omitted if there is no ambiguity as to the type the
413 constant must be (for example, when it is passed as an argument
414 to a non-overloaded function), in which case it is automatically
419 It is also possible to specify a type coercion using a function-like
422 <replaceable>typename</replaceable> ( '<replaceable>string</replaceable>' )
424 but not all type names may be used in this way; see <xref
425 linkend="sql-syntax-type-casts"> for details.
429 The <literal>::</literal>, <literal>CAST()</literal>, and
430 function-call syntaxes can also be used to specify run-time type
431 conversions of arbitrary expressions, as discussed in <xref
432 linkend="sql-syntax-type-casts">. But the form
433 <literal><replaceable>type</replaceable> '<replaceable>string</replaceable>'</literal>
434 can only be used to specify the type of a literal constant.
435 Another restriction on
436 <literal><replaceable>type</replaceable> '<replaceable>string</replaceable>'</literal>
437 is that it does not work for array types; use <literal>::</literal>
438 or <literal>CAST()</literal> to specify the type of an array constant.
443 <sect2 id="sql-syntax-operators">
444 <title>Operators</title>
446 <indexterm zone="sql-syntax-operators">
447 <primary>operator</primary>
448 <secondary>syntax</secondary>
452 An operator is a sequence of up to <symbol>NAMEDATALEN</symbol>-1
453 (63 by default) characters from the following list:
455 + - * / < > = ~ ! @ # % ^ & | ` ?
458 There are a few restrictions on operator names, however:
462 <literal>--</literal> and <literal>/*</literal> cannot appear
463 anywhere in an operator name, since they will be taken as the
470 A multiple-character operator name cannot end in <literal>+</> or <literal>-</>,
471 unless the name also contains at least one of these characters:
473 ~ ! @ # % ^ & | ` ?
475 For example, <literal>@-</literal> is an allowed operator name,
476 but <literal>*-</literal> is not. This restriction allows
477 <productname>PostgreSQL</productname> to parse SQL-compliant
478 queries without requiring spaces between tokens.
485 When working with non-SQL-standard operator names, you will usually
486 need to separate adjacent operators with spaces to avoid ambiguity.
487 For example, if you have defined a left unary operator named <literal>@</literal>,
488 you cannot write <literal>X*@Y</literal>; you must write
489 <literal>X* @Y</literal> to ensure that
490 <productname>PostgreSQL</productname> reads it as two operator names
496 <title>Special Characters</title>
499 Some characters that are not alphanumeric have a special meaning
500 that is different from being an operator. Details on the usage can
501 be found at the location where the respective syntax element is
502 described. This section only exists to advise the existence and
503 summarize the purposes of these characters.
508 A dollar sign (<literal>$</literal>) followed by digits is used
509 to represent a positional parameter in the body of a function
510 definition or a prepared statement. In other contexts the
511 dollar sign may be part of an identifier.
517 Parentheses (<literal>()</literal>) have their usual meaning to
518 group expressions and enforce precedence. In some cases
519 parentheses are required as part of the fixed syntax of a
520 particular SQL command.
526 Brackets (<literal>[]</literal>) are used to select the elements
527 of an array. See <xref linkend="arrays"> for more information
534 Commas (<literal>,</literal>) are used in some syntactical
535 constructs to separate the elements of a list.
541 The semicolon (<literal>;</literal>) terminates an SQL command.
542 It cannot appear anywhere within a command, except within a
543 string constant or quoted identifier.
549 The colon (<literal>:</literal>) is used to select
550 <quote>slices</quote> from arrays. (See <xref
551 linkend="arrays">.) In certain SQL dialects (such as Embedded
552 SQL), the colon is used to prefix variable names.
558 The asterisk (<literal>*</literal>) has a special meaning when
559 used in the <command>SELECT</command> command or with the
560 <function>COUNT</function> aggregate function.
566 The period (<literal>.</literal>) is used in numeric
567 constants, and to separate schema, table, and column names.
575 <sect2 id="sql-syntax-comments">
576 <title>Comments</title>
578 <indexterm zone="sql-syntax-comments">
579 <primary>comment</primary>
580 <secondary sortas="SQL">in SQL</secondary>
584 A comment is an arbitrary sequence of characters beginning with
585 double dashes and extending to the end of the line, e.g.:
587 -- This is a standard SQL comment
592 Alternatively, C-style block comments can be used:
595 * with nesting: /* nested block comment */
598 where the comment begins with <literal>/*</literal> and extends to
599 the matching occurrence of <literal>*/</literal>. These block
600 comments nest, as specified in the SQL standard but unlike C, so that one can
601 comment out larger blocks of code that may contain existing block
606 A comment is removed from the input stream before further syntax
607 analysis and is effectively replaced by whitespace.
611 <sect2 id="sql-precedence">
612 <title>Lexical Precedence</title>
614 <indexterm zone="sql-precedence">
615 <primary>operator</primary>
616 <secondary>precedence</secondary>
620 <xref linkend="sql-precedence-table"> shows the precedence and
621 associativity of the operators in <productname>PostgreSQL</>.
622 Most operators have the same precedence and are left-associative.
623 The precedence and associativity of the operators is hard-wired
624 into the parser. This may lead to non-intuitive behavior; for
625 example the Boolean operators <literal><</> and
626 <literal>></> have a different precedence than the Boolean
627 operators <literal><=</> and <literal>>=</>. Also, you will
628 sometimes need to add parentheses when using combinations of
629 binary and unary operators. For instance
637 because the parser has no idea -- until it is too late -- that
638 <token>!</token> is defined as a postfix operator, not an infix one.
639 To get the desired behavior in this case, you must write
643 This is the price one pays for extensibility.
646 <table id="sql-precedence-table">
647 <title>Operator Precedence (decreasing)</title>
652 <entry>Operator/Element</entry>
653 <entry>Associativity</entry>
654 <entry>Description</entry>
660 <entry><token>.</token></entry>
662 <entry>table/column name separator</entry>
666 <entry><token>::</token></entry>
668 <entry><productname>PostgreSQL</productname>-style typecast</entry>
672 <entry><token>[</token> <token>]</token></entry>
674 <entry>array element selection</entry>
678 <entry><token>-</token></entry>
680 <entry>unary minus</entry>
684 <entry><token>^</token></entry>
686 <entry>exponentiation</entry>
690 <entry><token>*</token> <token>/</token> <token>%</token></entry>
692 <entry>multiplication, division, modulo</entry>
696 <entry><token>+</token> <token>-</token></entry>
698 <entry>addition, subtraction</entry>
702 <entry><token>IS</token></entry>
704 <entry><literal>IS TRUE</>, <literal>IS FALSE</>, <literal>IS UNKNOWN</>, <literal>IS NULL</></entry>
708 <entry><token>ISNULL</token></entry>
710 <entry>test for null</entry>
714 <entry><token>NOTNULL</token></entry>
716 <entry>test for not null</entry>
720 <entry>(any other)</entry>
722 <entry>all other native and user-defined operators</entry>
726 <entry><token>IN</token></entry>
728 <entry>set membership</entry>
732 <entry><token>BETWEEN</token></entry>
734 <entry>containment</entry>
738 <entry><token>OVERLAPS</token></entry>
740 <entry>time interval overlap</entry>
744 <entry><token>LIKE</token> <token>ILIKE</token> <token>SIMILAR</token></entry>
746 <entry>string pattern matching</entry>
750 <entry><token><</token> <token>></token></entry>
752 <entry>less than, greater than</entry>
756 <entry><token>=</token></entry>
758 <entry>equality, assignment</entry>
762 <entry><token>NOT</token></entry>
764 <entry>logical negation</entry>
768 <entry><token>AND</token></entry>
770 <entry>logical conjunction</entry>
774 <entry><token>OR</token></entry>
776 <entry>logical disjunction</entry>
783 Note that the operator precedence rules also apply to user-defined
784 operators that have the same names as the built-in operators
785 mentioned above. For example, if you define a
786 <quote>+</quote> operator for some custom data type it will have
787 the same precedence as the built-in <quote>+</quote> operator, no
788 matter what yours does.
792 When a schema-qualified operator name is used in the
793 <literal>OPERATOR</> syntax, as for example in
795 SELECT 3 OPERATOR(pg_catalog.+) 4;
797 the <literal>OPERATOR</> construct is taken to have the default precedence
798 shown in <xref linkend="sql-precedence-table"> for <quote>any other</> operator. This is true no matter
799 which specific operator name appears inside <literal>OPERATOR()</>.
804 <sect1 id="sql-expressions">
805 <title>Value Expressions</title>
807 <indexterm zone="sql-expressions">
808 <primary>expression</primary>
809 <secondary>syntax</secondary>
812 <indexterm zone="sql-expressions">
813 <primary>value expression</primary>
817 <primary>scalar</primary>
818 <see>expression</see>
822 Value expressions are used in a variety of contexts, such
823 as in the target list of the <command>SELECT</command> command, as
824 new column values in <command>INSERT</command> or
825 <command>UPDATE</command>, or in search conditions in a number of
826 commands. The result of a value expression is sometimes called a
827 <firstterm>scalar</firstterm>, to distinguish it from the result of
828 a table expression (which is a table). Value expressions are
829 therefore also called <firstterm>scalar expressions</firstterm> (or
830 even simply <firstterm>expressions</firstterm>). The expression
831 syntax allows the calculation of values from primitive parts using
832 arithmetic, logical, set, and other operations.
836 A value expression is one of the following:
841 A constant or literal value; see <xref linkend="sql-syntax-constants">.
853 A positional parameter reference, in the body of a function definition
854 or prepared statement.
860 An operator invocation.
872 An aggregate expression.
890 An array constructor.
896 Another value expression in parentheses, useful to group
897 subexpressions and override
898 precedence.<indexterm><primary>parenthesis</></>
905 In addition to this list, there are a number of constructs that can
906 be classified as an expression but do not follow any general syntax
907 rules. These generally have the semantics of a function or
908 operator and are explained in the appropriate location in <xref
909 linkend="functions">. An example is the <literal>IS NULL</literal>
914 We have already discussed constants in <xref
915 linkend="sql-syntax-constants">. The following sections discuss
916 the remaining options.
920 <title>Column References</title>
923 <primary>column reference</primary>
927 A column can be referenced in the form
929 <replaceable>correlation</replaceable>.<replaceable>columnname</replaceable>
933 <replaceable>correlation</replaceable>.<replaceable>columnname</replaceable>[<replaceable>subscript</replaceable>]
935 (Here, the brackets <literal>[ ]</literal> are meant to appear literally.)
939 <replaceable>correlation</replaceable> is the name of a
940 table (possibly qualified), or an alias for a table defined by means of a
941 <literal>FROM</literal> clause, or
942 the key words <literal>NEW</literal> or <literal>OLD</literal>.
943 (<literal>NEW</literal> and <literal>OLD</literal> can only appear in rewrite rules,
944 while other correlation names can be used in any SQL statement.)
945 The correlation name and separating dot may be omitted if the column name
946 is unique across all the tables being used in the current query. (See also <xref linkend="queries">.)
950 If <replaceable>column</replaceable> is of an array type, then the
951 optional <replaceable>subscript</replaceable> selects a specific
952 element or elements in the array. If no subscript is provided, then the
953 whole array is selected. (See <xref linkend="arrays"> for more about
959 <title>Positional Parameters</title>
962 <primary>parameter</primary>
963 <secondary>syntax</secondary>
971 A positional parameter reference is used to indicate a value
972 that is supplied externally to an SQL statement. Parameters are
973 used in SQL function definitions and in prepared queries. Some
974 client libraries also support specifying data values separately
975 from the SQL command string, in which case parameters are used to
976 refer to the out-of-line data values.
977 The form of a parameter reference is:
979 $<replaceable>number</replaceable>
984 For example, consider the definition of a function,
985 <function>dept</function>, as
988 CREATE FUNCTION dept(text) RETURNS dept
989 AS 'SELECT * FROM dept WHERE name = $1'
993 Here the <literal>$1</literal> will be replaced by the first
994 function argument when the function is invoked.
999 <title>Operator Invocations</title>
1002 <primary>operator</primary>
1003 <secondary>invocation</secondary>
1007 There are three possible syntaxes for an operator invocation:
1009 <member><replaceable>expression</replaceable> <replaceable>operator</replaceable> <replaceable>expression</replaceable> (binary infix operator)</member>
1010 <member><replaceable>operator</replaceable> <replaceable>expression</replaceable> (unary prefix operator)</member>
1011 <member><replaceable>expression</replaceable> <replaceable>operator</replaceable> (unary postfix operator)</member>
1013 where the <replaceable>operator</replaceable> token follows the syntax
1014 rules of <xref linkend="sql-syntax-operators">, or is one of the
1015 key words <token>AND</token>, <token>OR</token>, and
1016 <token>NOT</token>, or is a qualified operator name
1018 <literal>OPERATOR(</><replaceable>schema</><literal>.</><replaceable>operatorname</><literal>)</>
1020 Which particular operators exist and whether
1021 they are unary or binary depends on what operators have been
1022 defined by the system or the user. <xref linkend="functions">
1023 describes the built-in operators.
1028 <title>Function Calls</title>
1031 <primary>function</primary>
1032 <secondary>invocation</secondary>
1036 The syntax for a function call is the name of a function
1037 (possibly qualified with a schema name), followed by its argument list
1038 enclosed in parentheses:
1041 <replaceable>function</replaceable> (<optional><replaceable>expression</replaceable> <optional>, <replaceable>expression</replaceable> ... </optional></optional> )
1046 For example, the following computes the square root of 2:
1053 The list of built-in functions is in <xref linkend="functions">.
1054 Other functions may be added by the user.
1058 <sect2 id="syntax-aggregates">
1059 <title>Aggregate Expressions</title>
1061 <indexterm zone="syntax-aggregates">
1062 <primary>aggregate function</primary>
1063 <secondary>invocation</secondary>
1067 An <firstterm>aggregate expression</firstterm> represents the
1068 application of an aggregate function across the rows selected by a
1069 query. An aggregate function reduces multiple inputs to a single
1070 output value, such as the sum or average of the inputs. The
1071 syntax of an aggregate expression is one of the following:
1074 <replaceable>aggregate_name</replaceable> (<replaceable>expression</replaceable>)
1075 <replaceable>aggregate_name</replaceable> (ALL <replaceable>expression</replaceable>)
1076 <replaceable>aggregate_name</replaceable> (DISTINCT <replaceable>expression</replaceable>)
1077 <replaceable>aggregate_name</replaceable> ( * )
1080 where <replaceable>aggregate_name</replaceable> is a previously
1081 defined aggregate (possibly a qualified name), and
1082 <replaceable>expression</replaceable> is
1083 any value expression that does not itself contain an aggregate
1088 The first form of aggregate expression invokes the aggregate
1089 across all input rows for which the given expression yields a
1090 non-null value. (Actually, it is up to the aggregate function
1091 whether to ignore null values or not --- but all the standard ones do.)
1092 The second form is the same as the first, since
1093 <literal>ALL</literal> is the default. The third form invokes the
1094 aggregate for all distinct non-null values of the expression found
1095 in the input rows. The last form invokes the aggregate once for
1096 each input row regardless of null or non-null values; since no
1097 particular input value is specified, it is generally only useful
1098 for the <function>count()</function> aggregate function.
1102 For example, <literal>count(*)</literal> yields the total number
1103 of input rows; <literal>count(f1)</literal> yields the number of
1104 input rows in which <literal>f1</literal> is non-null;
1105 <literal>count(distinct f1)</literal> yields the number of
1106 distinct non-null values of <literal>f1</literal>.
1110 The predefined aggregate functions are described in <xref
1111 linkend="functions-aggregate">. Other aggregate functions may be added
1116 An aggregate expression may only appear in the result list or
1117 <literal>HAVING</> clause of a <command>SELECT</> command.
1118 It is forbidden in other clauses, such as <literal>WHERE</>,
1119 because those clauses are logically evaluated before the results
1120 of aggregates are formed.
1124 When an aggregate expression appears in a subquery (see
1125 <xref linkend="sql-syntax-scalar-subqueries"> and
1126 <xref linkend="functions-subquery">), the aggregate is normally
1127 evaluated over the rows of the subquery. But an exception occurs
1128 if the aggregate's argument contains only outer-level variables:
1129 the aggregate then belongs to the nearest such outer level, and is
1130 evaluated over the rows of that query. The aggregate expression
1131 as a whole is then an outer reference for the subquery it appears in,
1132 and acts as a constant over any one evaluation of that subquery.
1133 The restriction about
1134 appearing only in the result list or <literal>HAVING</> clause
1135 applies with respect to the query level that the aggregate belongs to.
1139 <sect2 id="sql-syntax-type-casts">
1140 <title>Type Casts</title>
1143 <primary>data type</primary>
1144 <secondary>type cast</secondary>
1148 <primary>type cast</primary>
1152 A type cast specifies a conversion from one data type to another.
1153 <productname>PostgreSQL</productname> accepts two equivalent syntaxes
1156 CAST ( <replaceable>expression</replaceable> AS <replaceable>type</replaceable> )
1157 <replaceable>expression</replaceable>::<replaceable>type</replaceable>
1159 The <literal>CAST</> syntax conforms to SQL; the syntax with
1160 <literal>::</literal> is historical <productname>PostgreSQL</productname>
1165 When a cast is applied to a value expression of a known type, it
1166 represents a run-time type conversion. The cast will succeed only
1167 if a suitable type conversion function is available. Notice that this
1168 is subtly different from the use of casts with constants, as shown in
1169 <xref linkend="sql-syntax-constants-generic">. A cast applied to an
1170 unadorned string literal represents the initial assignment of a type
1171 to a literal constant value, and so it will succeed for any type
1172 (if the contents of the string literal are acceptable input syntax for the
1177 An explicit type cast may usually be omitted if there is no ambiguity as
1178 to the type that a value expression must produce (for example, when it is
1179 assigned to a table column); the system will automatically apply a
1180 type cast in such cases. However, automatic casting is only done for
1181 casts that are marked <quote>OK to apply implicitly</>
1182 in the system catalogs. Other casts must be invoked with
1183 explicit casting syntax. This restriction is intended to prevent
1184 surprising conversions from being applied silently.
1188 It is also possible to specify a type cast using a function-like
1191 <replaceable>typename</replaceable> ( <replaceable>expression</replaceable> )
1193 However, this only works for types whose names are also valid as
1194 function names. For example, <literal>double precision</literal>
1195 can't be used this way, but the equivalent <literal>float8</literal>
1196 can. Also, the names <literal>interval</>, <literal>time</>, and
1197 <literal>timestamp</> can only be used in this fashion if they are
1198 double-quoted, because of syntactic conflicts. Therefore, the use of
1199 the function-like cast syntax leads to inconsistencies and should
1200 probably be avoided in new applications.
1202 (The function-like syntax is in fact just a function call. When
1203 one of the two standard cast syntaxes is used to do a run-time
1204 conversion, it will internally invoke a registered function to
1205 perform the conversion. By convention, these conversion functions
1206 have the same name as their output type, and thus the <quote>function-like
1207 syntax</> is nothing more than a direct invocation of the underlying
1208 conversion function. Obviously, this is not something that a portable
1209 application should rely on.)
1213 <sect2 id="sql-syntax-scalar-subqueries">
1214 <title>Scalar Subqueries</title>
1217 <primary>subquery</primary>
1221 A scalar subquery is an ordinary
1222 <command>SELECT</command> query in parentheses that returns exactly one
1223 row with one column. (See <xref linkend="queries"> for information about writing queries.)
1224 The <command>SELECT</command> query is executed
1225 and the single returned value is used in the surrounding value expression.
1226 It is an error to use a query that
1227 returns more than one row or more than one column as a scalar subquery.
1228 (But if, during a particular execution, the subquery returns no rows,
1229 there is no error; the scalar result is taken to be null.)
1230 The subquery can refer to variables from the surrounding query,
1231 which will act as constants during any one evaluation of the subquery.
1232 See also <xref linkend="functions-subquery"> for other expressions involving subqueries.
1236 For example, the following finds the largest city population in each
1239 SELECT name, (SELECT max(pop) FROM cities WHERE cities.state = states.name)
1245 <sect2 id="sql-syntax-array-constructors">
1246 <title>Array Constructors</title>
1249 <primary>array</primary>
1250 <secondary>constructor</secondary>
1254 An array constructor is an expression that builds an
1255 array value from values for its member elements. A simple array
1257 consists of the key word <literal>ARRAY</literal>, a left square bracket
1258 <literal>[</>, one or more expressions (separated by commas) for the
1259 array element values, and finally a right square bracket <literal>]</>.
1262 SELECT ARRAY[1,2,3+4];
1268 The array element type is the common type of the member expressions,
1269 determined using the same rules as for <literal>UNION</> or
1270 <literal>CASE</> constructs (see <xref linkend="typeconv-union-case">).
1274 Multidimensional array values can be built by nesting array
1276 In the inner constructors, the key word <literal>ARRAY</literal> may
1277 be omitted. For example, these produce the same result:
1280 SELECT ARRAY[ARRAY[1,2], ARRAY[3,4]];
1286 SELECT ARRAY[[1,2],[3,4]];
1293 Since multidimensional arrays must be rectangular, inner constructors
1294 at the same level must produce sub-arrays of identical dimensions.
1298 Multidimensional array constructor elements can be anything yielding
1299 an array of the proper kind, not only a sub-<literal>ARRAY</> construct.
1302 CREATE TABLE arr(f1 int[], f2 int[]);
1304 INSERT INTO arr VALUES (ARRAY[[1,2],[3,4]], ARRAY[[5,6],[7,8]]);
1306 SELECT ARRAY[f1, f2, '{{9,10},{11,12}}'::int[]] FROM arr;
1308 ------------------------------------------------
1309 {{{1,2},{3,4}},{{5,6},{7,8}},{{9,10},{11,12}}}
1315 It is also possible to construct an array from the results of a
1316 subquery. In this form, the array constructor is written with the
1317 key word <literal>ARRAY</literal> followed by a parenthesized (not
1318 bracketed) subquery. For example:
1320 SELECT ARRAY(SELECT oid FROM pg_proc WHERE proname LIKE 'bytea%');
1322 -------------------------------------------------------------
1323 {2011,1954,1948,1952,1951,1244,1950,2005,1949,1953,2006,31}
1326 The subquery must return a single column. The resulting
1327 one-dimensional array will have an element for each row in the
1328 subquery result, with an element type matching that of the
1329 subquery's output column.
1333 The subscripts of an array value built with <literal>ARRAY</literal>
1334 always begin with one. For more information about arrays, see
1335 <xref linkend="arrays">.
1340 <sect2 id="syntax-express-eval">
1341 <title>Expression Evaluation Rules</title>
1344 <primary>expression</primary>
1345 <secondary>order of evaluation</secondary>
1349 The order of evaluation of subexpressions is not defined. In
1350 particular, the inputs of an operator or function are not necessarily
1351 evaluated left-to-right or in any other fixed order.
1355 Furthermore, if the result of an expression can be determined by
1356 evaluating only some parts of it, then other subexpressions
1357 might not be evaluated at all. For instance, if one wrote
1359 SELECT true OR somefunc();
1361 then <literal>somefunc()</literal> would (probably) not be called
1362 at all. The same would be the case if one wrote
1364 SELECT somefunc() OR true;
1366 Note that this is not the same as the left-to-right
1367 <quote>short-circuiting</quote> of Boolean operators that is found
1368 in some programming languages.
1372 As a consequence, it is unwise to use functions with side effects
1373 as part of complex expressions. It is particularly dangerous to
1374 rely on side effects or evaluation order in <literal>WHERE</> and <literal>HAVING</> clauses,
1375 since those clauses are extensively reprocessed as part of
1376 developing an execution plan. Boolean
1377 expressions (<literal>AND</>/<literal>OR</>/<literal>NOT</> combinations) in those clauses may be reorganized
1378 in any manner allowed by the laws of Boolean algebra.
1382 When it is essential to force evaluation order, a <literal>CASE</>
1383 construct (see <xref linkend="functions-conditional">) may be
1384 used. For example, this is an untrustworthy way of trying to
1385 avoid division by zero in a <literal>WHERE</> clause:
1387 SELECT ... WHERE x <> 0 AND y/x > 1.5;
1391 SELECT ... WHERE CASE WHEN x <> 0 THEN y/x > 1.5 ELSE false END;
1393 A <literal>CASE</> construct used in this fashion will defeat optimization
1394 attempts, so it should only be done when necessary. (In this particular
1395 example, it would doubtless be best to sidestep the problem by writing
1396 <literal>y > 1.5*x</> instead.)
1403 <!-- Keep this comment at the end of the file
1408 sgml-minimize-attributes:nil
1409 sgml-always-quote-attributes:t
1412 sgml-parent-document:nil
1413 sgml-default-dtd-file:"./reference.ced"
1414 sgml-exposed-tags:nil
1415 sgml-local-catalogs:("/usr/lib/sgml/catalog")
1416 sgml-local-ecat-files:nil