BNF Grammar for ISO/IEC 9075-2:2003 - Database Language SQL (SQL-2003) SQL/Foundation

Cross-Reference: rules
Cross-Reference: keywords

Derived from file sql-2003-2.bnf version 1.11 dated 2005/07/13 18:37:30
Generated on 2005-07-13 18:37:56+00:00 by file bnf2html.pl version 3.7 dated 2005/07/13 18:32:35

Information taken from the Final Committee Draft (FCD) of ISO/IEC 9075-2:2003. However, the page numbers and some section titles (9.14 through 9.23, for example) are from the final standard. This means there could be other as yet undiagnosed differences between the final standard and the notation in this document; you were warned!

The plain text version of this grammar is sql-2003-2.bnf .

Key SQL Statements and Fragments

ALTER DOMAIN <alter domain statement>

ALTER TABLE <alter table statement>

CLOSE cursor <close statement>

Column definition <column definition>

COMMIT WORK <commit statement>

CONNECT <connect statement>

CREATE ASSERTION <assertion definition>

CREATE CHARACTER SET <character set definition>

CREATE COLLATION <collation definition>

CREATE DOMAIN <domain definition>

CREATE FUNCTION <schema function>

CREATE PROCEDURE <schema procedure>

CREATE SCHEMA <schema definition>

CREATE TABLE <table definition>

CREATE TRANSLATION <translation definition>

CREATE TRIGGER <trigger definition>

CREATE VIEW <view definition>

Data type <data type>

DEALLOCATE PREPARE <deallocate prepared statement>

DECLARE cursor <declare cursor> <dynamic declare cursor>

DECLARE LOCAL TEMPORARY TABLE <temporary table declaration>

DELETE <delete statement: positioned> <delete statement: searched> <dynamic delete statement: positioned>

DESCRIBE <describe statement>

DESCRIPTOR statements <system descriptor statement>

DISCONNECT <disconnect statement>

EXECUTE <execute statement>

EXECUTE IMMEDIATE <execute immediate statement>

FETCH cursor <fetch statement>

FROM clause <from clause>

GET DIAGNOSTICS <get diagnostics statement>

GRANT <grant statement>

GROUP BY clause <group by clause>

HAVING clause <having clause>

INSERT <insert statement>

Literals <literal>

Keywords <key word>

MERGE <merge statement>

OPEN cursor <open statement>

ORDER BY clause <order by clause>

PREPARE <prepare statement>

REVOKE <revoke statement>

ROLLBACK WORK <rollback statement>

SAVEPOINT <savepoint statement>

Search condition <search condition> <regular expression>

SELECT <query specification>

SET CATALOG <set catalog statement>

SET CONNECTION <set connection statement>

SET CONSTRAINTS <set constraints mode statement>

SET NAMES <set names statement>

SET SCHEMA <set schema statement>

SET SESSION AUTHORIZATION <set session user identifier statement>

SET TIME ZONE <set local time zone statement>

SET TRANSACTION <set transaction statement>

SQL Client MODULE <SQL-client module definition>

UPDATE <update statement: positioned> <update statement: searched> <dynamic update statement: positioned>

Value expression <value expression>

WHERE clause <where clause>

Top

5 Lexical Elements

Basic definitions of characters used, tokens, symbols, etc. Most of this section would normally be handled within the lexical analyzer rather than in the grammar proper. Further, the original document does not quote the various single characters, which makes it hard to process automatically.

5.1 <SQL terminal character> (p151)

<simple Latin upper case letter> ::=
A | B | C | D | E | F | G | H | I | J | K | L | M | N | O | P | Q | R | S | T | U | V | W | X | Y | Z

<simple Latin lower case letter> ::=
a | b | c | d | e | f | g | h | i | j | k | l | m | n | o | p | q | r | s | t | u | v | w | x | y | z

<digit> ::= 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9

<space> ::= !! See the Syntax Rules

<double quote> ::= "

<percent> ::= %

<ampersand> ::= &

<quote> ::= '

<left paren> ::= (

<right paren> ::= )

<asterisk> ::= *

<plus sign> ::= +

<comma> ::= ,

<minus sign> ::= -

<period> ::= .

<solidus> ::= /

<colon> ::= :

<semicolon> ::= ;

<less than operator> ::= <

<equals operator> ::= =

<question mark> ::= ?

The trigraphs are new in SQL-2003.

<left bracket> ::= [

<left bracket trigraph> ::= ??(

<right bracket> ::= ]

<right bracket trigraph> ::= ??)

<circumflex> ::= ^

<underscore> ::= _

<vertical bar> ::= /* Nothing */ |

<left brace> ::= {

<right brace> ::= }

Top

5.2 <token> and <separator> (p134)

Specifying lexical units (tokens and separators) that participate in SQL language.

<identifier body> ::= <identifier start> [ <identifier part> ... ]

Previous standard said:

<identifier start> ::= <initial alphabetic character> | <ideographic character>

<identifier start> ::= !! See the Syntax Rules

<identifier extend> ::= !! See the Syntax Rules

<multiplier> ::= K | M | G

The productions for <Unicode delimited identifier> and so on are new in SQL-2003.

<Unicode delimited identifier>    ::=
         U <ampersand> <double quote> <Unicode delimiter body> ... <double quote>
         [ ESCAPE <escape character> ]

<Unicode delimiter body> ::= <Unicode identifier part> ...

<escape character> ::= !! See the Syntax Rules

<nondoublequote character> ::= !! See the Syntax Rules

The rule for <doublequote symbol> in the standard uses two adjacent literal double quotes rather than referencing <double quote>; the reasons are not clear. It is annotated '!! two consecutive double quote characters'.

The rules for <not equals operator> etc in the standard uses two adjacent literal characters rather than referencing <less than> and <greater than>; the reasons are not clear. Note that two characters must be adjacent with no intervening space, not a pair of characters separated by arbitrary white space.

<separator> ::= { <comment> | <white space> }...

<simple comment introducer> ::= <minus sign> <minus sign> [ <minus sign> ... ]

The <bracketed comment> rule included '!! See the Syntax Rules'. This probably says something about the <slash> <asterisk> and <asterisk> <slash> needing to be adjacent characters rather than adjacent tokens.

<bracketed comment contents> ::= [ { <comment character> | <separator> }... ]

<newline> ::= !! See the Syntax Rules

There was a surprising amount of movement of keywords between the reserved and non-reserved word classes. There is also room to think that much of the host language support moved out of Part 2 (SQL/Foundation).

Top

5.3 <literal> (p143)

<character string literal>    ::=
         [ <introducer> <character set specification> ]
         <quote> [ <character representation> ... ] <quote>
         [ { <separator> <quote> [ <character representation> ... ] <quote> }... ]

<nonquote character> ::= !! See the Syntax Rules

The <quote symbol> rule consists of two immediately adjacent <quote> marks with no spaces. As usual, this would be best handled in the lexical analyzer, not in the grammar.

<national character string literal>    ::=
         N <quote> [ <character representation> ... ] <quote>
         [ { <separator> <quote> [ <character representation> ... ] <quote> }... ]

<Unicode character string literal>    ::=
         [ <introducer> <character set specification> ]
         U <ampersand> <quote> [ <Unicode representation> ... ] <quote>
         [ { <separator> <quote> [ <Unicode representation> ... ] <quote> }... ]
         [ ESCAPE <escape character> ]