|
|
|
@ -1,8 +1,8 @@ |
|
|
|
<!-- |
|
|
|
<!-- |
|
|
|
$Header: /cvsroot/pgsql/doc/src/sgml/syntax.sgml,v 1.31 2000/12/22 18:57:50 petere Exp $ |
|
|
|
$Header: /cvsroot/pgsql/doc/src/sgml/syntax.sgml,v 1.32 2001/01/06 11:58:56 petere Exp $ |
|
|
|
--> |
|
|
|
--> |
|
|
|
|
|
|
|
|
|
|
|
<chapter id="syntax"> |
|
|
|
<chapter id="sql-syntax"> |
|
|
|
<title>SQL Syntax</title> |
|
|
|
<title>SQL Syntax</title> |
|
|
|
|
|
|
|
|
|
|
|
<abstract> |
|
|
|
<abstract> |
|
|
|
@ -11,558 +11,548 @@ $Header: /cvsroot/pgsql/doc/src/sgml/syntax.sgml,v 1.31 2000/12/22 18:57:50 pete |
|
|
|
</para> |
|
|
|
</para> |
|
|
|
</abstract> |
|
|
|
</abstract> |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
<sect1 id="sql-syntax-lexical"> |
|
|
|
|
|
|
|
<title>Lexical Structure</title> |
|
|
|
|
|
|
|
|
|
|
|
<para> |
|
|
|
<para> |
|
|
|
<acronym>SQL</acronym> manipulates sets of data. The language is |
|
|
|
SQL input consists of a sequence of |
|
|
|
composed of various <firstterm>key words</firstterm>. Arithmetic |
|
|
|
<firstterm>commands</firstterm>. A command is composed of a |
|
|
|
and procedural expressions are allowed. We will cover these topics |
|
|
|
sequence of <firstterm>tokens</firstterm>, which depend on the |
|
|
|
in this chapter; subsequent chapters will include details on data |
|
|
|
syntax of the particular command, terminated by a semicolon |
|
|
|
types, functions, and operators. |
|
|
|
(<quote>;</quote>). The end of the input stream also terminates a |
|
|
|
|
|
|
|
command. |
|
|
|
</para> |
|
|
|
</para> |
|
|
|
|
|
|
|
|
|
|
|
<sect1 id="sql-keywords"> |
|
|
|
<para> |
|
|
|
<title>Key Words</title> |
|
|
|
A token can be a <firstterm>key word</firstterm>, an |
|
|
|
|
|
|
|
<firstterm>identifier</firstterm>, a <firstterm>quoted |
|
|
|
|
|
|
|
identifier</firstterm>, a <firstterm>literal</firstterm> (or |
|
|
|
|
|
|
|
constant), or a special character symbol. Tokens are normally |
|
|
|
|
|
|
|
separated by whitespace (space, tab, newline), but need not be if |
|
|
|
|
|
|
|
there is no ambiguity (which is generally only the case if a |
|
|
|
|
|
|
|
special character is adjacent to some other token type). |
|
|
|
|
|
|
|
</para> |
|
|
|
|
|
|
|
|
|
|
|
<para> |
|
|
|
<para> |
|
|
|
<acronym>SQL92</acronym> defines <firstterm>key words</firstterm> |
|
|
|
Additionally, <firstterm>comments</firstterm> can occur in SQL |
|
|
|
for the language |
|
|
|
input. They are not tokens, they are effectively equivalent to |
|
|
|
which have specific meaning. Some key words are |
|
|
|
whitespace. |
|
|
|
<firstterm>reserved</firstterm>, which indicates that they are |
|
|
|
|
|
|
|
restricted to appear in only certain contexts. Other key words are |
|
|
|
|
|
|
|
<firstterm>not restricted</firstterm>, which indicates that in certain |
|
|
|
|
|
|
|
contexts they |
|
|
|
|
|
|
|
have a specific meaning but are not otherwise constrained. |
|
|
|
|
|
|
|
</para> |
|
|
|
</para> |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
<informalexample id="sql-syntax-ex-commands"> |
|
|
|
<para> |
|
|
|
<para> |
|
|
|
<productname>Postgres</productname> implements an extended subset of the |
|
|
|
For example, the following is (lexically) valid SQL input: |
|
|
|
<acronym>SQL92</acronym> and <acronym>SQL3</acronym> languages. Some language |
|
|
|
<programlisting> |
|
|
|
elements are not as restricted in this implementation as is |
|
|
|
SELECT * FROM MY_TABLE; |
|
|
|
called for in the language standards, in part due |
|
|
|
UPDATE MY_TABLE SET A = 5; |
|
|
|
to the extensibility features of <productname>Postgres</productname>. |
|
|
|
INSERT INTO MY_TABLE VALUES (3, 'hi there'); |
|
|
|
|
|
|
|
</programlisting> |
|
|
|
|
|
|
|
This is a sequence of three commands, one per line (although this |
|
|
|
|
|
|
|
is not required; more than one command can be on a line, and |
|
|
|
|
|
|
|
commands can be usefully split across lines). |
|
|
|
</para> |
|
|
|
</para> |
|
|
|
|
|
|
|
</informalexample> |
|
|
|
|
|
|
|
|
|
|
|
<para> |
|
|
|
<para> |
|
|
|
Information on <acronym>SQL92</acronym> and <acronym>SQL3</acronym> key words |
|
|
|
The SQL syntax is not very consistent regarding what tokens |
|
|
|
is derived from <xref linkend="DATE97" endterm="DATE97">. |
|
|
|
identify commands and which are operands or parameters. The first |
|
|
|
|
|
|
|
few tokens are generally the command name, so in the above example |
|
|
|
|
|
|
|
we would usually speak of a <quote>SELECT</quote>, an |
|
|
|
|
|
|
|
<quote>UPDATE</quote>, and an <quote>INSERT</quote> command. But |
|
|
|
|
|
|
|
for instance the <command>UPDATE</command> command always requires |
|
|
|
|
|
|
|
a <token>SET</token> token to appear in a certain position, and |
|
|
|
|
|
|
|
this particular variation of <command>INSERT</command> also |
|
|
|
|
|
|
|
requires a <token>VALUES</token> in order to be complete. The |
|
|
|
|
|
|
|
precise syntax rules for each command are described in the |
|
|
|
|
|
|
|
<citetitle>Reference Manual</citetitle>. |
|
|
|
</para> |
|
|
|
</para> |
|
|
|
|
|
|
|
|
|
|
|
<sect2> |
|
|
|
<sect2 id="sql-syntax-identifiers"> |
|
|
|
<title>Reserved Key Words</title> |
|
|
|
<title>Identifiers and Key Words</title> |
|
|
|
|
|
|
|
|
|
|
|
<para> |
|
|
|
<para> |
|
|
|
<acronym>SQL92</acronym> and <acronym>SQL3</acronym> have |
|
|
|
Tokens such as <token>SELECT</token>, <token>UPDATE</token>, or |
|
|
|
<firstterm>reserved key words</firstterm> which are not allowed |
|
|
|
<token>VALUES</token> in the example above are examples of |
|
|
|
as identifiers and not allowed in any usage other than as fundamental |
|
|
|
<firstterm>key words</firstterm>, that is, words that have a fixed |
|
|
|
tokens in <acronym>SQL</acronym> statements. |
|
|
|
meaning in the SQL language. The tokens <token>MY_TABLE</token> |
|
|
|
<productname>Postgres</productname> has additional key words |
|
|
|
and <token>A</token> are examples of |
|
|
|
which have similar restrictions. In particular, these key words |
|
|
|
<firstterm>identifiers</firstterm>. They identify names of |
|
|
|
are not allowed as column or table names, though in some cases |
|
|
|
tables, columns, or other database objects, depending on the |
|
|
|
they are allowed to be column labels (i.e. in AS clauses). |
|
|
|
command they are used in. Therefore they are sometimes simply |
|
|
|
|
|
|
|
called <quote>names</quote>. Key words and identifiers have the |
|
|
|
|
|
|
|
same lexical structure, meaning that one cannot know whether a |
|
|
|
|
|
|
|
token is an identifier or a key word without knowing the language. |
|
|
|
|
|
|
|
A complete list of key words can be found in <xref |
|
|
|
|
|
|
|
linkend="sql-keywords-appendix">. |
|
|
|
</para> |
|
|
|
</para> |
|
|
|
|
|
|
|
|
|
|
|
<tip> |
|
|
|
|
|
|
|
<para> |
|
|
|
<para> |
|
|
|
Any string can be used as an identifier if surrounded by |
|
|
|
SQL identifiers and key words must begin with a letter |
|
|
|
double quotes (<quote>like this!</quote>). Some care is required since |
|
|
|
(<literal>a</literal>-<literal>z</literal>) or underscore |
|
|
|
such an identifier will be case sensitive |
|
|
|
(<literal>_</literal>). Subsequent characters in an identifier or |
|
|
|
and will retain embedded whitespace and most other special characters. |
|
|
|
key word can be letters, digits |
|
|
|
|
|
|
|
(<literal>0</literal>-<literal>9</literal>), or underscores, |
|
|
|
|
|
|
|
although the SQL standard will not define a key word that contains |
|
|
|
|
|
|
|
digits or start or ends with an underscore. |
|
|
|
</para> |
|
|
|
</para> |
|
|
|
</tip> |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
<para> |
|
|
|
<para> |
|
|
|
The following are <productname>Postgres</productname> |
|
|
|
The system uses no more than <symbol>NAMEDATALEN</symbol>-1 |
|
|
|
reserved words that are neither <acronym>SQL92</acronym> |
|
|
|
characters of an identifier; longer names can be written in |
|
|
|
nor <acronym>SQL3</acronym> reserved words. These are allowed |
|
|
|
commands, but they will be truncated. By default, |
|
|
|
to be present as column labels, but not as identifiers: |
|
|
|
<symbol>NAMEDATALEN</symbol> is 32 so the maximum identifier length |
|
|
|
|
|
|
|
is 31 (but at the time the system is built, |
|
|
|
<programlisting> |
|
|
|
<symbol>NAMEDATALEN</symbol> can be changed in |
|
|
|
ABORT ANALYZE |
|
|
|
<filename>src/include/postgres_ext.h</filename>). |
|
|
|
BINARY |
|
|
|
|
|
|
|
CLUSTER CONSTRAINT COPY |
|
|
|
|
|
|
|
DO |
|
|
|
|
|
|
|
EXPLAIN EXTEND |
|
|
|
|
|
|
|
LISTEN LOAD LOCK |
|
|
|
|
|
|
|
MOVE |
|
|
|
|
|
|
|
NEW NONE NOTIFY |
|
|
|
|
|
|
|
OFFSET |
|
|
|
|
|
|
|
RESET |
|
|
|
|
|
|
|
SETOF SHOW |
|
|
|
|
|
|
|
UNLISTEN UNTIL |
|
|
|
|
|
|
|
VACUUM VERBOSE |
|
|
|
|
|
|
|
</programlisting> |
|
|
|
|
|
|
|
</para> |
|
|
|
</para> |
|
|
|
|
|
|
|
|
|
|
|
<para> |
|
|
|
<para> |
|
|
|
The following are <productname>Postgres</productname> |
|
|
|
Identifier and key word names are case insensitive. Therefore |
|
|
|
reserved words that are also <acronym>SQL92</acronym> |
|
|
|
|
|
|
|
or <acronym>SQL3</acronym> reserved words, and that |
|
|
|
|
|
|
|
are allowed to be present as column labels, but not as identifiers: |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
<programlisting> |
|
|
|
<programlisting> |
|
|
|
ALL ANY ASC BETWEEN BIT BOTH |
|
|
|
UPDATE MY_TABLE SET A = 5; |
|
|
|
CASE CAST CHAR CHARACTER CHECK COALESCE COLLATE COLUMN |
|
|
|
|
|
|
|
CONSTRAINT CROSS CURRENT CURRENT_DATE CURRENT_TIME |
|
|
|
|
|
|
|
CURRENT_TIMESTAMP CURRENT_USER |
|
|
|
|
|
|
|
DEC DECIMAL DEFAULT DESC DISTINCT |
|
|
|
|
|
|
|
ELSE END EXCEPT EXISTS EXTRACT |
|
|
|
|
|
|
|
FALSE FLOAT FOR FOREIGN FROM FULL |
|
|
|
|
|
|
|
GLOBAL GROUP |
|
|
|
|
|
|
|
HAVING |
|
|
|
|
|
|
|
IN INNER INTERSECT INTO IS |
|
|
|
|
|
|
|
JOIN |
|
|
|
|
|
|
|
LEADING LEFT LIKE LOCAL |
|
|
|
|
|
|
|
NATURAL NCHAR NOT NULL NULLIF NUMERIC |
|
|
|
|
|
|
|
ON OR ORDER OUTER OVERLAPS |
|
|
|
|
|
|
|
POSITION PRECISION PRIMARY PUBLIC |
|
|
|
|
|
|
|
REFERENCES RIGHT |
|
|
|
|
|
|
|
SELECT SESSION_USER SOME SUBSTRING |
|
|
|
|
|
|
|
TABLE THEN TO TRANSACTION TRIM TRUE |
|
|
|
|
|
|
|
UNION UNIQUE USER |
|
|
|
|
|
|
|
VARCHAR |
|
|
|
|
|
|
|
WHEN WHERE |
|
|
|
|
|
|
|
</programlisting> |
|
|
|
</programlisting> |
|
|
|
|
|
|
|
can equivalently be written as |
|
|
|
The following are <productname>Postgres</productname> |
|
|
|
|
|
|
|
reserved words that are also <acronym>SQL92</acronym> |
|
|
|
|
|
|
|
or <acronym>SQL3</acronym> reserved words: |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
<programlisting> |
|
|
|
<programlisting> |
|
|
|
ADD ALTER AND AS |
|
|
|
uPDaTE my_TabLE SeT a = 5; |
|
|
|
BEGIN BY |
|
|
|
|
|
|
|
CASCADE CLOSE COMMIT CREATE CURSOR |
|
|
|
|
|
|
|
DECLARE DEFAULT DELETE DESC DISTINCT DROP |
|
|
|
|
|
|
|
EXECUTE EXISTS EXTRACT |
|
|
|
|
|
|
|
FETCH FLOAT FOR FROM FULL |
|
|
|
|
|
|
|
GRANT |
|
|
|
|
|
|
|
HAVING |
|
|
|
|
|
|
|
IN INNER INSERT INTERVAL INTO INOUT IS |
|
|
|
|
|
|
|
JOIN |
|
|
|
|
|
|
|
LEADING LEFT LIKE LOCAL |
|
|
|
|
|
|
|
NAMES NATIONAL NATURAL NCHAR NO NOT NULL |
|
|
|
|
|
|
|
ON OR OUT OUTER |
|
|
|
|
|
|
|
PARTIAL PRIMARY PRIVILEGES PROCEDURE PUBLIC |
|
|
|
|
|
|
|
REFERENCES REVOKE RIGHT ROLLBACK |
|
|
|
|
|
|
|
SELECT SESSION SET SUBSTRING |
|
|
|
|
|
|
|
TO TRAILING TRIM |
|
|
|
|
|
|
|
UNION UNIQUE UPDATE USING |
|
|
|
|
|
|
|
VALUES VARCHAR VARYING VIEW |
|
|
|
|
|
|
|
WHERE WITH WITHOUT WORK |
|
|
|
|
|
|
|
</programlisting> |
|
|
|
</programlisting> |
|
|
|
</para> |
|
|
|
A good convention to adopt is perhaps to write key words in upper |
|
|
|
|
|
|
|
case and names in lower case, e.g., |
|
|
|
<para> |
|
|
|
|
|
|
|
The following are <acronym>SQL92</acronym> reserved key words that |
|
|
|
|
|
|
|
are not <productname>Postgres</productname> reserved key words, but that |
|
|
|
|
|
|
|
if used as function names are always translated into the function |
|
|
|
|
|
|
|
<function>CHAR_LENGTH</function>: |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
<programlisting> |
|
|
|
<programlisting> |
|
|
|
CHARACTER_LENGTH |
|
|
|
UPDATE my_table SET a = 5; |
|
|
|
</programlisting> |
|
|
|
</programlisting> |
|
|
|
</para> |
|
|
|
</para> |
|
|
|
|
|
|
|
|
|
|
|
<para> |
|
|
|
<para> |
|
|
|
The following are <acronym>SQL92</acronym> or <acronym>SQL3</acronym> |
|
|
|
There is a second kind of identifier: the <firstterm>delimited |
|
|
|
reserved key words that |
|
|
|
identifier</firstterm> or <firstterm>quoted |
|
|
|
are not <productname>Postgres</productname> reserved key words, but |
|
|
|
identifier</firstterm>. It is formed by enclosing an arbitrary |
|
|
|
if used as type names are always translated into an alternate, native type: |
|
|
|
sequence of characters in double-quotes |
|
|
|
|
|
|
|
(<literal>"</literal>). <!-- " font-lock mania --> A delimited |
|
|
|
|
|
|
|
identifier is always an identifier, never a key word. So |
|
|
|
|
|
|
|
<literal>"select"</literal> could be used to refer to a column or |
|
|
|
|
|
|
|
table named <quote>select</quote>, whereas an unquoted |
|
|
|
|
|
|
|
<literal>select</literal> would be taken as part of a command and |
|
|
|
|
|
|
|
would therefore provoke a parse error when used where a table or |
|
|
|
|
|
|
|
column name is expected. The example can be written with quoted |
|
|
|
|
|
|
|
identifiers like so: |
|
|
|
<programlisting> |
|
|
|
<programlisting> |
|
|
|
BOOLEAN DOUBLE FLOAT INT INTEGER INTERVAL REAL SMALLINT |
|
|
|
UPDATE "my_table" SET "a" = 5; |
|
|
|
</programlisting> |
|
|
|
</programlisting> |
|
|
|
</para> |
|
|
|
</para> |
|
|
|
|
|
|
|
|
|
|
|
<para> |
|
|
|
<para> |
|
|
|
The following are not keywords of any kind, but when used in the |
|
|
|
Quoted identifiers can contain any character other than a double |
|
|
|
context of a type name are translated into a native |
|
|
|
quote itself. This allows constructing table or column names that |
|
|
|
<productname>Postgres</productname> type, and when used in the |
|
|
|
would otherwise not be possible, such as ones containing spaces or |
|
|
|
context of a function name are translated into a native function: |
|
|
|
ampersands. The length limitation still applies. |
|
|
|
|
|
|
|
|
|
|
|
<programlisting> |
|
|
|
|
|
|
|
DATETIME TIMESPAN |
|
|
|
|
|
|
|
</programlisting> |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
(translated to <type>TIMESTAMP</type> and <type>INTERVAL</type>, |
|
|
|
|
|
|
|
respectively). This feature is intended to help with |
|
|
|
|
|
|
|
transitioning to version 7.0, and will be removed in a future release. |
|
|
|
|
|
|
|
</para> |
|
|
|
</para> |
|
|
|
|
|
|
|
|
|
|
|
<para> |
|
|
|
<para> |
|
|
|
The following are either <acronym>SQL92</acronym> |
|
|
|
Quoting an identifier also makes it case-sensitive, whereas |
|
|
|
or <acronym>SQL3</acronym> reserved key words |
|
|
|
unquoted names are always folded to lower case. For example, the |
|
|
|
that are not key words in <productname>Postgres</productname>. |
|
|
|
identifiers <literal>FOO</literal>, <literal>foo</literal> and |
|
|
|
These have no proscribed usage in <productname>Postgres</productname> |
|
|
|
<literal>"foo"</literal> are considered the same by |
|
|
|
at the time of writing (version 7.0) but may become reserved key words in the |
|
|
|
<productname>Postgres</productname>, but <literal>"Foo"</literal> |
|
|
|
future: |
|
|
|
and <literal>"FOO"</literal> are different from these three and |
|
|
|
|
|
|
|
each other. |
|
|
|
<note> |
|
|
|
<footnote> |
|
|
|
<para> |
|
|
|
<para> |
|
|
|
Some of these key words represent functions in <acronym>SQL92</acronym>. |
|
|
|
This is incompatible with SQL, where unquoted names are folded to |
|
|
|
These functions are defined in <productname>Postgres</productname>, |
|
|
|
upper case. Thus, <literal>foo</literal> is equivalent to |
|
|
|
but the parser does not consider the names to be key words and they are allowed |
|
|
|
<literal>"FOO"</literal>. If you want to write portable |
|
|
|
in other contexts. |
|
|
|
applications you are advised to always quote a particular name or |
|
|
|
|
|
|
|
never quote it. |
|
|
|
</para> |
|
|
|
</para> |
|
|
|
</note> |
|
|
|
</footnote> |
|
|
|
|
|
|
|
|
|
|
|
<programlisting> |
|
|
|
|
|
|
|
ALLOCATE ARE ASSERTION AT AUTHORIZATION AVG |
|
|
|
|
|
|
|
BIT_LENGTH |
|
|
|
|
|
|
|
CASCADED CATALOG CHAR_LENGTH CHARACTER_LENGTH COLLATION |
|
|
|
|
|
|
|
CONNECT CONNECTION CONTINUE CONVERT CORRESPONDING COUNT |
|
|
|
|
|
|
|
CURRENT_SESSION |
|
|
|
|
|
|
|
DATE DEALLOCATE DEC DESCRIBE DESCRIPTOR |
|
|
|
|
|
|
|
DIAGNOSTICS DISCONNECT DOMAIN |
|
|
|
|
|
|
|
ESCAPE EXCEPT EXCEPTION EXEC EXTERNAL |
|
|
|
|
|
|
|
FIRST FOUND |
|
|
|
|
|
|
|
GET GO GOTO |
|
|
|
|
|
|
|
IDENTITY INDICATOR INPUT INTERSECT |
|
|
|
|
|
|
|
LAST LOWER |
|
|
|
|
|
|
|
MAX MIN MODULE |
|
|
|
|
|
|
|
OCTET_LENGTH OPEN OUTPUT OVERLAPS |
|
|
|
|
|
|
|
PREPARE PRESERVE |
|
|
|
|
|
|
|
ROWS |
|
|
|
|
|
|
|
SCHEMA SECTION SESSION SIZE SOME |
|
|
|
|
|
|
|
SQL SQLCODE SQLERROR SQLSTATE SUM SYSTEM_USER |
|
|
|
|
|
|
|
TEMPORARY TRANSLATE TRANSLATION |
|
|
|
|
|
|
|
UNKNOWN UPPER USAGE |
|
|
|
|
|
|
|
VALUE |
|
|
|
|
|
|
|
WHENEVER WRITE |
|
|
|
|
|
|
|
</programlisting> |
|
|
|
|
|
|
|
</para> |
|
|
|
</para> |
|
|
|
</sect2> |
|
|
|
</sect2> |
|
|
|
|
|
|
|
|
|
|
|
<sect2> |
|
|
|
|
|
|
|
<title>Non-reserved Keywords</title> |
|
|
|
<sect2 id="sql-syntax-constants"> |
|
|
|
|
|
|
|
<title>Constants</title> |
|
|
|
|
|
|
|
|
|
|
|
<para> |
|
|
|
<para> |
|
|
|
<acronym>SQL92</acronym> and <acronym>SQL3</acronym> have |
|
|
|
There are four kinds of <firstterm>implicitly typed |
|
|
|
<firstterm>non-reserved keywords</firstterm> which have |
|
|
|
constants</firstterm> in <productname>Postgres</productname>: |
|
|
|
a prescribed meaning in the language but which are also allowed |
|
|
|
strings, bit strings, integers, and floating point numbers. |
|
|
|
as identifiers. |
|
|
|
Constants can also be specified with explicit types, which can |
|
|
|
<productname>Postgres</productname> has additional keywords |
|
|
|
enable more accurate representation and more efficient handling by |
|
|
|
which allow similar unrestricted usage. |
|
|
|
the system. The implicit constants are described below; explicit |
|
|
|
In particular, these keywords |
|
|
|
constants are discussed afterwards. |
|
|
|
are allowed as column or table names. |
|
|
|
|
|
|
|
</para> |
|
|
|
</para> |
|
|
|
|
|
|
|
|
|
|
|
<para> |
|
|
|
<sect3> |
|
|
|
The following are <productname>Postgres</productname> |
|
|
|
<title>String Constants</title> |
|
|
|
non-reserved key words that are neither <acronym>SQL92</acronym> |
|
|
|
|
|
|
|
nor <acronym>SQL3</acronym> non-reserved key words: |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
<programlisting> |
|
|
|
<para> |
|
|
|
ACCESS AFTER AGGREGATE |
|
|
|
A string constant in SQL is an arbitrary sequence of characters |
|
|
|
BACKWARD BEFORE |
|
|
|
bounded by single quotes (<quote>'</quote>), e.g., <literal>'This |
|
|
|
CACHE COMMENT CREATEDB CREATEUSER CYCLE |
|
|
|
is a string'</literal>. SQL allows single quotes to be embedded |
|
|
|
DATABASE DELIMITERS |
|
|
|
in strings by typing two adjacent single quotes (e.g., |
|
|
|
EACH ENCODING EXCLUSIVE |
|
|
|
<literal>'Dianne''s horse'</literal>). In |
|
|
|
FORCE FORWARD FUNCTION |
|
|
|
<productname>Postgres</productname> single quotes may |
|
|
|
HANDLER |
|
|
|
alternatively be escaped with a backslash (<quote>\</quote>, |
|
|
|
INCREMENT INDEX INHERITS INSENSITIVE INSTEAD ISNULL |
|
|
|
e.g., <literal>'Dianne\'s horse'</literal>). |
|
|
|
LANCOMPILER LOCATION |
|
|
|
|
|
|
|
MAXVALUE MINVALUE MODE |
|
|
|
|
|
|
|
NOCREATEDB NOCREATEUSER NOTHING NOTIFY NOTNULL |
|
|
|
|
|
|
|
OIDS OPERATOR |
|
|
|
|
|
|
|
PASSWORD PROCEDURAL |
|
|
|
|
|
|
|
RECIPE REINDEX RENAME RETURNS ROW RULE |
|
|
|
|
|
|
|
SEQUENCE SERIAL SHARE START STATEMENT STDIN STDOUT |
|
|
|
|
|
|
|
TEMP TRUSTED |
|
|
|
|
|
|
|
UNLISTEN UNTIL |
|
|
|
|
|
|
|
VALID VERSION |
|
|
|
|
|
|
|
</programlisting> |
|
|
|
|
|
|
|
</para> |
|
|
|
</para> |
|
|
|
|
|
|
|
|
|
|
|
<para> |
|
|
|
<para> |
|
|
|
The following are <productname>Postgres</productname> |
|
|
|
C-style backslash escapes are also available: |
|
|
|
non-reserved key words that are <acronym>SQL92</acronym> |
|
|
|
<literal>\b</literal> is a backspace, <literal>\f</literal> is a |
|
|
|
or <acronym>SQL3</acronym> reserved key words: |
|
|
|
form feed, <literal>\n</literal> is a newline, |
|
|
|
|
|
|
|
<literal>\r</literal> is a carriage return, <literal>\t</literal> |
|
|
|
<programlisting> |
|
|
|
is a tab, and <literal>\<replaceable>xxx</replaceable></literal>, |
|
|
|
ABSOLUTE ACTION |
|
|
|
where <replaceable>xxx</replaceable> is an octal number, is the |
|
|
|
CHARACTERISTICS CONSTRAINTS |
|
|
|
character with the corresponding ASCII code. Any other character |
|
|
|
DAY DEFERRABLE DEFERRED |
|
|
|
following a backslash is taken literally. Thus, to include a |
|
|
|
HOUR |
|
|
|
backslash in a string constant, type two backslashes. |
|
|
|
IMMEDIATE INITIALLY INSENSITIVE ISOLATION |
|
|
|
|
|
|
|
KEY |
|
|
|
|
|
|
|
LANGUAGE LEVEL |
|
|
|
|
|
|
|
MATCH MINUTE MONTH |
|
|
|
|
|
|
|
NEXT |
|
|
|
|
|
|
|
OF ONLY OPTION |
|
|
|
|
|
|
|
PATH PENDANT PRIOR PRIVILEGES |
|
|
|
|
|
|
|
READ RELATIVE RESTRICT |
|
|
|
|
|
|
|
SCHEMA SCROLL SECOND |
|
|
|
|
|
|
|
TIME TIMESTAMP TIMEZONE_HOUR TIMEZONE_MINUTE TRIGGER |
|
|
|
|
|
|
|
YEAR |
|
|
|
|
|
|
|
ZONE |
|
|
|
|
|
|
|
</programlisting> |
|
|
|
|
|
|
|
</para> |
|
|
|
</para> |
|
|
|
|
|
|
|
|
|
|
|
<para> |
|
|
|
<para> |
|
|
|
The following are <productname>Postgres</productname> |
|
|
|
The character with the code zero cannot be in a string constant. |
|
|
|
non-reserved key words that are also either <acronym>SQL92</acronym> |
|
|
|
</para> |
|
|
|
or <acronym>SQL3</acronym> non-reserved key words: |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
<para> |
|
|
|
|
|
|
|
Two string constants that are only separated by whitespace |
|
|
|
|
|
|
|
<emphasis>with at least one newline</emphasis> are concatenated |
|
|
|
|
|
|
|
and effectively treated as if the string had been written in one |
|
|
|
|
|
|
|
constant. For example: |
|
|
|
|
|
|
|
<programlisting> |
|
|
|
|
|
|
|
SELECT 'foo' |
|
|
|
|
|
|
|
'bar'; |
|
|
|
|
|
|
|
</programlisting> |
|
|
|
|
|
|
|
is equivalent to |
|
|
|
|
|
|
|
<programlisting> |
|
|
|
|
|
|
|
SELECT 'foobar'; |
|
|
|
|
|
|
|
</programlisting> |
|
|
|
|
|
|
|
but |
|
|
|
<programlisting> |
|
|
|
<programlisting> |
|
|
|
COMMITTED SERIALIZABLE TYPE |
|
|
|
SELECT 'foo' 'bar'; |
|
|
|
</programlisting> |
|
|
|
</programlisting> |
|
|
|
|
|
|
|
is not valid syntax. |
|
|
|
</para> |
|
|
|
</para> |
|
|
|
|
|
|
|
</sect3> |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
<sect3> |
|
|
|
|
|
|
|
<title>Bit String Constants</title> |
|
|
|
|
|
|
|
|
|
|
|
<para> |
|
|
|
<para> |
|
|
|
The following are either <acronym>SQL92</acronym> |
|
|
|
Bit string constants look like string constants with a |
|
|
|
or <acronym>SQL3</acronym> non-reserved key words that are not |
|
|
|
<literal>B</literal> (upper or lower case) immediately before the |
|
|
|
key words of any kind in <productname>Postgres</productname>: |
|
|
|
opening quote (no intervening whitespace), e.g., |
|
|
|
|
|
|
|
<literal>B'1001'</literal>. The only characters allowed within |
|
|
|
|
|
|
|
bit string constants are <literal>0</literal> and |
|
|
|
|
|
|
|
<literal>1</literal>. Bit strings constants can be continued |
|
|
|
|
|
|
|
across lines in the same way as regular string constants. |
|
|
|
|
|
|
|
</para> |
|
|
|
|
|
|
|
</sect3> |
|
|
|
|
|
|
|
|
|
|
|
<programlisting> |
|
|
|
<sect3> |
|
|
|
ADA |
|
|
|
<title>Integer Constants</title> |
|
|
|
C CATALOG_NAME CHARACTER_SET_CATALOG CHARACTER_SET_NAME |
|
|
|
|
|
|
|
CHARACTER_SET_SCHEMA CLASS_ORIGIN COBOL COLLATION_CATALOG |
|
|
|
<para> |
|
|
|
COLLATION_NAME COLLATION_SCHEMA COLUMN_NAME |
|
|
|
Integer constants in SQL are sequences of decimal digits (0 |
|
|
|
COMMAND_FUNCTION CONDITION_NUMBER |
|
|
|
though 9) with no decimal point. The range of legal values |
|
|
|
CONNECTION_NAME CONSTRAINT_CATALOG CONSTRAINT_NAME |
|
|
|
depends on which integer data type is used, but the plain |
|
|
|
CONSTRAINT_SCHEMA CURSOR_NAME |
|
|
|
<type>integer</type> type accepts values ranging from -2147483648 |
|
|
|
DATA DATE_TIME_INTERVAL_CODE DATE_TIME_INTERVAL_PRECISION |
|
|
|
to +2147483647. (The optional plus or minus sign is actually a |
|
|
|
DYNAMIC_FUNCTION |
|
|
|
separate unary operator and not part of the integer constant.) |
|
|
|
FORTRAN |
|
|
|
|
|
|
|
LENGTH |
|
|
|
|
|
|
|
MESSAGE_LENGTH MESSAGE_OCTET_LENGTH MORE MUMPS |
|
|
|
|
|
|
|
NAME NULLABLE NUMBER |
|
|
|
|
|
|
|
PAD PASCAL PLI |
|
|
|
|
|
|
|
REPEATABLE RETURNED_LENGTH RETURNED_OCTET_LENGTH |
|
|
|
|
|
|
|
RETURNED_SQLSTATE ROW_COUNT |
|
|
|
|
|
|
|
SCALE SCHEMA_NAME SERVER_NAME SPACE SUBCLASS_ORIGIN |
|
|
|
|
|
|
|
TABLE_NAME |
|
|
|
|
|
|
|
UNCOMMITTED UNNAMED |
|
|
|
|
|
|
|
</programlisting> |
|
|
|
|
|
|
|
</para> |
|
|
|
</para> |
|
|
|
</sect2> |
|
|
|
</sect3> |
|
|
|
</sect1> |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
<sect1 id="sql-comments"> |
|
|
|
<sect3> |
|
|
|
<title>Comments</title> |
|
|
|
<title>Floating Point Constants</title> |
|
|
|
|
|
|
|
|
|
|
|
<para> |
|
|
|
<para> |
|
|
|
A <firstterm>comment</firstterm> |
|
|
|
Floating point constants are accepted in these general forms: |
|
|
|
is an arbitrary sequence of characters beginning with double dashes |
|
|
|
<synopsis> |
|
|
|
and extending to the end of the line, e.g.: |
|
|
|
<replaceable>digits</replaceable>.<optional><replaceable>digits</replaceable></optional><optional>e<optional>+-</optional><replaceable>digits</replaceable></optional> |
|
|
|
|
|
|
|
<optional><replaceable>digits</replaceable></optional>.<replaceable>digits</replaceable><optional>e<optional>+-</optional><replaceable>digits</replaceable></optional> |
|
|
|
|
|
|
|
<replaceable>digits</replaceable>e<optional>+-</optional><replaceable>digits</replaceable> |
|
|
|
|
|
|
|
</synopsis> |
|
|
|
|
|
|
|
where <replaceable>digits</replaceable> is one or more decimal |
|
|
|
|
|
|
|
digits. At least one digit must be before or after the decimal |
|
|
|
|
|
|
|
point and after the <literal>e</literal> if you use that option. |
|
|
|
|
|
|
|
Thus, a floating point constant is distinguished from an integer |
|
|
|
|
|
|
|
constant by the presence of either the decimal point or the |
|
|
|
|
|
|
|
exponent clause (or both). There must not be a space or other |
|
|
|
|
|
|
|
characters embedded in the constant. |
|
|
|
|
|
|
|
</para> |
|
|
|
|
|
|
|
|
|
|
|
<programlisting> |
|
|
|
<informalexample> |
|
|
|
-- This is a standard SQL comment |
|
|
|
<para> |
|
|
|
</programlisting> |
|
|
|
These are some examples of valid floating point constants: |
|
|
|
|
|
|
|
<literallayout> |
|
|
|
|
|
|
|
3.5 |
|
|
|
|
|
|
|
4. |
|
|
|
|
|
|
|
.001 |
|
|
|
|
|
|
|
5e2 |
|
|
|
|
|
|
|
1.925e-3 |
|
|
|
|
|
|
|
</literallayout> |
|
|
|
</para> |
|
|
|
</para> |
|
|
|
|
|
|
|
</informalexample> |
|
|
|
|
|
|
|
|
|
|
|
<para> |
|
|
|
<para> |
|
|
|
We also support C-style block comments, e.g.: |
|
|
|
Floating point constants are of type <type>DOUBLE |
|
|
|
|
|
|
|
PRECISION</type>. <type>REAL</type> can be specified explicitly |
|
|
|
|
|
|
|
by using <acronym>SQL</acronym> string notation or |
|
|
|
|
|
|
|
<productname>Postgres</productname> type notation: |
|
|
|
|
|
|
|
|
|
|
|
<programlisting> |
|
|
|
<programlisting> |
|
|
|
/* multi-line comment |
|
|
|
REAL '1.23' -- string style |
|
|
|
* with nesting: /* nested block comment */ |
|
|
|
'1.23'::REAL -- Postgres (historical) style |
|
|
|
*/ |
|
|
|
|
|
|
|
</programlisting> |
|
|
|
</programlisting> |
|
|
|
|
|
|
|
|
|
|
|
where the comment begins with "<literal>/*</literal>" and extends |
|
|
|
|
|
|
|
to the matching occurrence of "<literal>*/</literal>". These block |
|
|
|
|
|
|
|
comments nest, as specified in SQL99, so that one can comment out |
|
|
|
|
|
|
|
larger blocks of code that may contain existing block comments. |
|
|
|
|
|
|
|
</para> |
|
|
|
</para> |
|
|
|
</sect1> |
|
|
|
</sect3> |
|
|
|
|
|
|
|
|
|
|
|
<sect1 id="sql-names"> |
|
|
|
<sect3> |
|
|
|
<title>Names</title> |
|
|
|
<title>Constants of Other Types</title> |
|
|
|
|
|
|
|
|
|
|
|
<para> |
|
|
|
<para> |
|
|
|
Names in SQL must begin with a letter |
|
|
|
A constant of an <emphasis>arbitrary</emphasis> type can be |
|
|
|
(<literal>a</literal>-<literal>z</literal>) or underscore |
|
|
|
entered using any one of the following notations: |
|
|
|
(<literal>_</literal>). |
|
|
|
<synopsis> |
|
|
|
Subsequent characters in a name can be letters, digits |
|
|
|
<replaceable>type</replaceable> '<replaceable>string</replaceable>' |
|
|
|
(<literal>0</literal>-<literal>9</literal>), |
|
|
|
'<replaceable>string</replaceable>'::<replaceable>type</replaceable> |
|
|
|
or underscores. The system uses no more than NAMEDATALEN-1 characters |
|
|
|
CAST ( '<replaceable>string</replaceable>' AS <replaceable>type</replaceable> ) |
|
|
|
of a name; longer names can be written in queries, but they will be |
|
|
|
</synopsis> |
|
|
|
truncated. |
|
|
|
The value inside the string is passed to the input conversion |
|
|
|
By default, NAMEDATALEN is 32 so the maximum name length is 31 (but |
|
|
|
routine for the type called <replaceable>type</replaceable>. The |
|
|
|
at the time the system is built, NAMEDATALEN can be changed in |
|
|
|
result is a constant of the indicated type. The explicit type |
|
|
|
<filename>src/include/postgres_ext.h</filename>). |
|
|
|
cast may be omitted if there is no ambiguity as to the type the |
|
|
|
|
|
|
|
constant must be (for example, when it is passed as an argument |
|
|
|
|
|
|
|
to a non-overloaded function), in which case it is automatically |
|
|
|
|
|
|
|
coerced. |
|
|
|
</para> |
|
|
|
</para> |
|
|
|
|
|
|
|
|
|
|
|
<para> |
|
|
|
<para> |
|
|
|
Names containing other characters may be formed by surrounding them |
|
|
|
It is also possible to specify a type coercion using a function-like |
|
|
|
with double quotes (<literal>"</literal>). For example, table or column |
|
|
|
syntax: |
|
|
|
names may contain |
|
|
|
<synopsis> |
|
|
|
otherwise disallowed characters such as spaces, ampersands, etc. if |
|
|
|
<replaceable>typename</replaceable> ( <replaceable>value</replaceable> ) |
|
|
|
quoted. Quoting a name also makes it case-sensitive, |
|
|
|
</synopsis> |
|
|
|
whereas unquoted names are always folded to lower case. For example, |
|
|
|
although this only works for types whose names are also valid as |
|
|
|
the names <literal>FOO</literal>, <literal>foo</literal> |
|
|
|
function names. (For example, <literal>double precision</literal> |
|
|
|
and <literal>"foo"</literal> are |
|
|
|
can't be used this way --- but the equivalent <literal>float8</literal> |
|
|
|
considered the same by <productname>Postgres</productname>, but |
|
|
|
can.) |
|
|
|
<literal>"Foo"</literal> is a different name. |
|
|
|
|
|
|
|
</para> |
|
|
|
</para> |
|
|
|
|
|
|
|
|
|
|
|
<para> |
|
|
|
<para> |
|
|
|
Double quotes can also be used to protect a name that would otherwise |
|
|
|
The <literal>::</literal>, <literal>CAST()</literal>, and |
|
|
|
be taken to be an SQL keyword. For example, <literal>IN</literal> |
|
|
|
function-call syntaxes can also be used to specify the type of |
|
|
|
is a keyword but <literal>"IN"</literal> is a name. |
|
|
|
arbitrary expressions, but the form |
|
|
|
|
|
|
|
<replaceable>type</replaceable> |
|
|
|
|
|
|
|
'<replaceable>string</replaceable>' can only be used to specify |
|
|
|
|
|
|
|
the type of a literal constant. |
|
|
|
</para> |
|
|
|
</para> |
|
|
|
</sect1> |
|
|
|
</sect3> |
|
|
|
|
|
|
|
|
|
|
|
<sect1 id="sql-constants"> |
|
|
|
<sect3> |
|
|
|
<title>Constants</title> |
|
|
|
<title>Array constants</title> |
|
|
|
|
|
|
|
|
|
|
|
<para> |
|
|
|
<para> |
|
|
|
There are three kinds of <firstterm>implicitly typed constants</firstterm> |
|
|
|
The general format of an array constant is the following: |
|
|
|
in <productname>Postgres</productname>: strings, integers, |
|
|
|
<synopsis> |
|
|
|
and floating point numbers. Constants can |
|
|
|
'{ <replaceable>val1</replaceable> <replaceable>delim</replaceable> <replaceable>val2</replaceable> <replaceable>delim</replaceable> ... }' |
|
|
|
also be specified with explicit types, which can enable more |
|
|
|
</synopsis> |
|
|
|
accurate representation and more efficient handling by the |
|
|
|
where <replaceable>delim</replaceable> is the delimiter character |
|
|
|
backend. The implicit constants are described below; explicit |
|
|
|
for the type, as recorded in its <literal>pg_type</literal> |
|
|
|
constants are discussed afterwards. |
|
|
|
entry. (For all built-in types, this is the comma character |
|
|
|
|
|
|
|
",".) Each <replaceable>val</replaceable> is either a constant |
|
|
|
|
|
|
|
of the array element type, or a sub-array. An example of an |
|
|
|
|
|
|
|
array constant is |
|
|
|
|
|
|
|
<programlisting> |
|
|
|
|
|
|
|
'{{1,2,3},{4,5,6},{7,8,9}}' |
|
|
|
|
|
|
|
</programlisting> |
|
|
|
|
|
|
|
This constant is a two-dimensional, 3 by 3 array consisting of three |
|
|
|
|
|
|
|
sub-arrays of integers. |
|
|
|
</para> |
|
|
|
</para> |
|
|
|
|
|
|
|
|
|
|
|
<sect2> |
|
|
|
|
|
|
|
<title>String Constants</title> |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
<para> |
|
|
|
<para> |
|
|
|
<firstterm>Strings</firstterm> |
|
|
|
Individual array elements can be placed between double-quote |
|
|
|
in SQL are arbitrary sequences of ASCII characters bounded by single |
|
|
|
marks (<literal>"</literal>) <!-- " --> to avoid ambiguity |
|
|
|
quotes ("'", e.g. <literal>'This is a string'</literal>). |
|
|
|
problems with respect to white space. Without quote marks, the |
|
|
|
SQL92 allows single quotes to be embedded in strings by typing two |
|
|
|
array-value parser will skip leading white space. |
|
|
|
adjacent single quotes (e.g. <literal>'Dianne''s horse'</literal>). |
|
|
|
|
|
|
|
In <productname>Postgres</productname> single quotes may alternatively |
|
|
|
|
|
|
|
be escaped with a backslash ("\", e.g. |
|
|
|
|
|
|
|
<literal>'Dianne\'s horse'</literal>). To include a |
|
|
|
|
|
|
|
backslash in a string constant, type two backslashes. |
|
|
|
|
|
|
|
Non-printing characters may also be embedded within strings by |
|
|
|
|
|
|
|
prepending them with a backslash |
|
|
|
|
|
|
|
(e.g. <literal>'\<replaceable>tab</replaceable>'</literal>). |
|
|
|
|
|
|
|
</para> |
|
|
|
</para> |
|
|
|
|
|
|
|
|
|
|
|
</sect2> |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
<sect2> |
|
|
|
|
|
|
|
<title>Integer Constants</title> |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
<para> |
|
|
|
<para> |
|
|
|
<firstterm>Integer constants</firstterm> |
|
|
|
(Array constants are actually only a special case of the generic |
|
|
|
in SQL are sequences of ASCII digits with no decimal point. |
|
|
|
type constants discussed in the previous section. The constant |
|
|
|
The range of legal values depends on which integer datatype is |
|
|
|
is initially treated as a string and passed to the array input |
|
|
|
used, but the plain <literal>integer</literal> type accepts values |
|
|
|
conversion routine. An explicit type specification might be |
|
|
|
ranging from -2147483648 to +2147483647. |
|
|
|
necessary.) |
|
|
|
</para> |
|
|
|
</para> |
|
|
|
|
|
|
|
</sect3> |
|
|
|
</sect2> |
|
|
|
</sect2> |
|
|
|
|
|
|
|
|
|
|
|
<sect2> |
|
|
|
|
|
|
|
<title>Floating Point Constants</title> |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
<para> |
|
|
|
<sect2 id="sql-syntax-operators"> |
|
|
|
<firstterm>Floating point constants</firstterm> |
|
|
|
<title>Operators</title> |
|
|
|
consist of an integer part, a decimal point, and a fraction part or |
|
|
|
|
|
|
|
scientific notation of the following format: |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
<synopsis> |
|
|
|
<para> |
|
|
|
{<replaceable>dig</replaceable>}.{<replaceable>dig</replaceable>} [e [+-] {<replaceable>dig</replaceable>}] |
|
|
|
An operator is a sequence of up to <symbol>NAMEDATALEN</symbol>-1 |
|
|
|
</synopsis> |
|
|
|
(31 by default) characters from the following list: |
|
|
|
|
|
|
|
<literallayout> |
|
|
|
|
|
|
|
+ - * / < > = ~ ! @ # % ^ & | ` ? $ |
|
|
|
|
|
|
|
</literallayout> |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
There are a few restrictions on operator names, however: |
|
|
|
|
|
|
|
<itemizedlist> |
|
|
|
|
|
|
|
<listitem> |
|
|
|
|
|
|
|
<para> |
|
|
|
|
|
|
|
"$" (dollar) cannot be a single-character operator, although it |
|
|
|
|
|
|
|
can be part of a multi-character operator name. |
|
|
|
|
|
|
|
</para> |
|
|
|
|
|
|
|
</listitem> |
|
|
|
|
|
|
|
|
|
|
|
where <replaceable>dig</replaceable> is one or more digits. |
|
|
|
<listitem> |
|
|
|
You must include at least one <replaceable>dig</replaceable> after the |
|
|
|
<para> |
|
|
|
period and after the [+-] if you use those options. An exponent with |
|
|
|
<literal>--</literal> and <literal>/*</literal> cannot appear |
|
|
|
a missing mantissa has a mantissa of 1 inserted. There may be no |
|
|
|
anywhere in an operator name, since they will be taken as the |
|
|
|
extra characters embedded in the string. |
|
|
|
start of a comment. |
|
|
|
</para> |
|
|
|
</para> |
|
|
|
|
|
|
|
</listitem> |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
<listitem> |
|
|
|
<para> |
|
|
|
<para> |
|
|
|
Floating point constaints are of type |
|
|
|
A multi-character operator name cannot end in "+" or "-", |
|
|
|
<type>float8</type>. <type>float4</type> can be specified |
|
|
|
unless the name also contains at least one of these characters: |
|
|
|
explicitly by using <acronym>SQL92</acronym> string notation or |
|
|
|
<literallayout> |
|
|
|
<productname>Postgres</productname> type notation: |
|
|
|
~ ! @ # % ^ & | ` ? $ |
|
|
|
|
|
|
|
</literallayout> |
|
|
|
|
|
|
|
For example, <literal>@-</literal> is an allowed operator name, |
|
|
|
|
|
|
|
but <literal>*-</literal> is not. This restriction allows |
|
|
|
|
|
|
|
<productname>Postgres</productname> to parse SQL-compliant |
|
|
|
|
|
|
|
queries without requiring spaces between tokens. |
|
|
|
|
|
|
|
</para> |
|
|
|
|
|
|
|
</listitem> |
|
|
|
|
|
|
|
</itemizedlist> |
|
|
|
|
|
|
|
</para> |
|
|
|
|
|
|
|
|
|
|
|
<programlisting> |
|
|
|
<para> |
|
|
|
float4 '1.23' -- string style |
|
|
|
When working with non-SQL-standard operator names, you will usually |
|
|
|
'1.23'::float4 -- Postgres (historical) style |
|
|
|
need to separate adjacent operators with spaces to avoid ambiguity. |
|
|
|
</programlisting> |
|
|
|
For example, if you have defined a left-unary operator named "@", |
|
|
|
|
|
|
|
you cannot write <literal>X*@Y</literal>; you must write |
|
|
|
|
|
|
|
<literal>X* @Y</literal> to ensure that |
|
|
|
|
|
|
|
<productname>Postgres</productname> reads it as two operator names |
|
|
|
|
|
|
|
not one. |
|
|
|
</para> |
|
|
|
</para> |
|
|
|
</sect2> |
|
|
|
</sect2> |
|
|
|
|
|
|
|
|
|
|
|
<sect2> |
|
|
|
<sect2> |
|
|
|
<title>Constants of Postgres User-Defined Types</title> |
|
|
|
<title>Special Characters</title> |
|
|
|
|
|
|
|
|
|
|
|
<para> |
|
|
|
<para> |
|
|
|
A constant of an |
|
|
|
Some characters that are not alphanumeric have a special meaning |
|
|
|
<emphasis>arbitrary</emphasis> |
|
|
|
that is different from being an operator. Details on the usage can |
|
|
|
type can be entered using any one of the following notations: |
|
|
|
be found at the location where the respective syntax element is |
|
|
|
|
|
|
|
described. This section only exists to advise the existence and |
|
|
|
|
|
|
|
summarize the purposes of these characters. |
|
|
|
|
|
|
|
|
|
|
|
<synopsis> |
|
|
|
<itemizedlist> |
|
|
|
<replaceable>type</replaceable> '<replaceable>string</replaceable>' |
|
|
|
<listitem> |
|
|
|
'<replaceable>string</replaceable>'::<replaceable>type</replaceable> |
|
|
|
<para> |
|
|
|
CAST ( '<replaceable>string</replaceable>' AS <replaceable>type</replaceable> ) |
|
|
|
A dollar sign (<literal>$</literal>) followed by digits is used |
|
|
|
</synopsis> |
|
|
|
to represent the positional parameters in the body of a function |
|
|
|
|
|
|
|
definition. In other contexts the dollar sign may be part of an |
|
|
|
|
|
|
|
operator name. |
|
|
|
|
|
|
|
</para> |
|
|
|
|
|
|
|
</listitem> |
|
|
|
|
|
|
|
|
|
|
|
The value inside the string is passed to the input |
|
|
|
<listitem> |
|
|
|
conversion routine for the type called |
|
|
|
<para> |
|
|
|
<replaceable>type</replaceable>. The result is a |
|
|
|
Parentheses (<literal>()</literal>) have their usual meaning to |
|
|
|
constant of the indicated type. The explicit typecast may be omitted |
|
|
|
group expressions and enforce precedence. In some cases |
|
|
|
if there is no ambiguity as to the type the constant must be, in which |
|
|
|
parentheses are required as part of the fixed syntax of a |
|
|
|
case it is automatically coerced. |
|
|
|
particular SQL command. |
|
|
|
</para> |
|
|
|
</para> |
|
|
|
|
|
|
|
</listitem> |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
<listitem> |
|
|
|
<para> |
|
|
|
<para> |
|
|
|
It is also possible to specify a type coercion using a function-like |
|
|
|
Brackets (<literal>[]</literal>) are used to select the elements |
|
|
|
syntax: |
|
|
|
of an array. See <xref linkend="arrays"> for more information |
|
|
|
|
|
|
|
on arrays. |
|
|
|
|
|
|
|
</para> |
|
|
|
|
|
|
|
</listitem> |
|
|
|
|
|
|
|
|
|
|
|
<synopsis> |
|
|
|
<listitem> |
|
|
|
<replaceable>typename</replaceable> ( <replaceable>value</replaceable> ) |
|
|
|
<para> |
|
|
|
</synopsis> |
|
|
|
Commas (<literal>,</literal>) are used in some syntactical |
|
|
|
|
|
|
|
constructs to separate the elements of a list. |
|
|
|
|
|
|
|
</para> |
|
|
|
|
|
|
|
</listitem> |
|
|
|
|
|
|
|
|
|
|
|
although this only works for types whose names are also valid as |
|
|
|
<listitem> |
|
|
|
function names. (For example, <literal>double precision</literal> |
|
|
|
<para> |
|
|
|
can't be used this way --- but the equivalent <literal>float8</literal> |
|
|
|
The semicolon (<literal>;</literal>) terminates an SQL command. |
|
|
|
can.) |
|
|
|
It cannot appear anywhere within a command, except when quoted |
|
|
|
|
|
|
|
as a string constant or identifier. |
|
|
|
</para> |
|
|
|
</para> |
|
|
|
|
|
|
|
</listitem> |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
<listitem> |
|
|
|
<para> |
|
|
|
<para> |
|
|
|
The <literal>::</literal>, <literal>CAST()</literal>, and function-call |
|
|
|
The colon (<literal>:</literal>) is used to select |
|
|
|
syntaxes can also be used to specify run-time type conversions. But |
|
|
|
<quote>slices</quote> from arrays. (See <xref |
|
|
|
the form <replaceable>type</replaceable> |
|
|
|
linkend="arrays">.) In certain SQL dialects (such as Embedded |
|
|
|
'<replaceable>string</replaceable>' can only be used to specify the |
|
|
|
SQL), the colon is used to prefix variable names. |
|
|
|
type of a literal constant. |
|
|
|
|
|
|
|
</para> |
|
|
|
</para> |
|
|
|
</sect2> |
|
|
|
</listitem> |
|
|
|
|
|
|
|
|
|
|
|
<sect2> |
|
|
|
<listitem> |
|
|
|
<title>Array constants</title> |
|
|
|
<para> |
|
|
|
|
|
|
|
The asterisk (<literal>*</literal>) has a special meaning when |
|
|
|
|
|
|
|
used in the <command>SELECT</command> command or with the |
|
|
|
|
|
|
|
<function>COUNT</function> aggregate function. |
|
|
|
|
|
|
|
</para> |
|
|
|
|
|
|
|
</listitem> |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
<listitem> |
|
|
|
<para> |
|
|
|
<para> |
|
|
|
<firstterm>Array constants</firstterm> |
|
|
|
The period (<literal>.</literal>) is used in floating point |
|
|
|
are n-dimensional arrays of any Postgres datatype. |
|
|
|
constants, and to separate table and column names. |
|
|
|
The general format of an array constant is the following: |
|
|
|
</para> |
|
|
|
|
|
|
|
</listitem> |
|
|
|
|
|
|
|
</itemizedlist> |
|
|
|
|
|
|
|
|
|
|
|
<synopsis> |
|
|
|
</para> |
|
|
|
{ <replaceable>val1</replaceable> <replaceable>delim</replaceable> <replaceable>val2</replaceable> <replaceable>delim</replaceable> ... } |
|
|
|
</sect2> |
|
|
|
</synopsis> |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
where <replaceable>delim</replaceable> |
|
|
|
<sect2 id="sql-syntax-comments"> |
|
|
|
is the delimiter character for the type, as recorded in its |
|
|
|
<title>Comments</title> |
|
|
|
<literal>pg_type</literal> class entry. |
|
|
|
|
|
|
|
(For all built-in types, this is the comma character ",".) |
|
|
|
|
|
|
|
Each <replaceable>val</replaceable> is either a constant |
|
|
|
|
|
|
|
of the array element type, or a sub-array. |
|
|
|
|
|
|
|
An example of an array constant is |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
<para> |
|
|
|
|
|
|
|
A comment is an arbitrary sequence of characters beginning with |
|
|
|
|
|
|
|
double dashes and extending to the end of the line, e.g.: |
|
|
|
<programlisting> |
|
|
|
<programlisting> |
|
|
|
{{1,2,3},{4,5,6},{7,8,9}} |
|
|
|
-- This is a standard SQL92 comment |
|
|
|
</programlisting> |
|
|
|
</programlisting> |
|
|
|
|
|
|
|
</para> |
|
|
|
|
|
|
|
|
|
|
|
This constant is a two-dimensional, 3 by 3 array consisting of three |
|
|
|
<para> |
|
|
|
sub-arrays of integers. |
|
|
|
Alternatively, C-style block comments can be used: |
|
|
|
|
|
|
|
<programlisting> |
|
|
|
|
|
|
|
/* multi-line comment |
|
|
|
|
|
|
|
* with nesting: /* nested block comment */ |
|
|
|
|
|
|
|
*/ |
|
|
|
|
|
|
|
</programlisting> |
|
|
|
|
|
|
|
where the comment begins with <literal>/*</literal> and extends to |
|
|
|
|
|
|
|
the matching occurrence of <literal>*/</literal>. These block |
|
|
|
|
|
|
|
comments nest, as specified in SQL99 but unlike C, so that one can |
|
|
|
|
|
|
|
comment out larger blocks of code that may contain existing block |
|
|
|
|
|
|
|
comments. |
|
|
|
</para> |
|
|
|
</para> |
|
|
|
|
|
|
|
|
|
|
|
<para> |
|
|
|
<para> |
|
|
|
Individual array elements can be placed between double-quote |
|
|
|
A comment is removed from the input stream before further syntax |
|
|
|
marks (<literal>"</literal>) to avoid ambiguity problems with respect to |
|
|
|
analysis and is effectively replaced by whitespace. |
|
|
|
white space. |
|
|
|
|
|
|
|
Without quote marks, the array-value parser will skip leading white space. |
|
|
|
|
|
|
|
</para> |
|
|
|
</para> |
|
|
|
</sect2> |
|
|
|
</sect2> |
|
|
|
</sect1> |
|
|
|
</sect1> |
|
|
|
|
|
|
|
|
|
|
|
<sect1 id="sql-columns"> |
|
|
|
|
|
|
|
|
|
|
|
<sect1 id="sql-syntax-columns"> |
|
|
|
<title>Fields and Columns</title> |
|
|
|
<title>Fields and Columns</title> |
|
|
|
|
|
|
|
|
|
|
|
<sect2> |
|
|
|
<sect2> |
|
|
|
@ -664,18 +654,6 @@ CAST ( '<replaceable>string</replaceable>' AS <replaceable>type</replaceable> ) |
|
|
|
</sect2> |
|
|
|
</sect2> |
|
|
|
</sect1> |
|
|
|
</sect1> |
|
|
|
|
|
|
|
|
|
|
|
<sect1 id="sql-operators"> |
|
|
|
|
|
|
|
<title>Operators</title> |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
<para> |
|
|
|
|
|
|
|
Any built-in or user-defined operator may be used in SQL. |
|
|
|
|
|
|
|
For the list of built-in operators consult <xref linkend="functions">. |
|
|
|
|
|
|
|
For a list of user-defined operators consult your system administrator |
|
|
|
|
|
|
|
or run a query on the <literal>pg_operator</literal> class. |
|
|
|
|
|
|
|
Parentheses may be used for arbitrary grouping of operators in expressions. |
|
|
|
|
|
|
|
</para> |
|
|
|
|
|
|
|
</sect1> |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
<sect1 id="sql-expressions"> |
|
|
|
<sect1 id="sql-expressions"> |
|
|
|
<title>Expressions</title> |
|
|
|
<title>Expressions</title> |
|
|
|
|
|
|
|
|
|
|
|
|
Peter Eisentraut
25 years ago