Cancel Manual.txi edits as it produces too much merge problems (49a44d22) · Commits · Software / OSDI20 Artifacts / mariadb

Docs/manual.texi

+6 −184

Original line number	Diff line number	Diff line
		@@ -31867,24 +31867,9 @@ mysql> SELECT EXP(2);
		mysql> SELECT EXP(-2);
		-> 0.135335
		@end example

		@findex LN()
		@item LN(X)
		Returns the natural logarithm of @code{X}:
		@example
		mysql> SELECT LN(2);
		-> 0.693147
		mysql> SELECT LN(-2);
		-> NULL
		@end example

		This function was added in MySQL version 4.0.3.
		It is synonymous with @code{LOG(X)} in MySQL.

		@findex LOG()
		@item LOG([B,]X)
		When called with one parameter, this function returns the natural logarithm
		of @code{X}:
		@item LOG(X)
		Returns the natural logarithm of @code{X}:
		@example
		mysql> SELECT LOG(2);
		-> 0.693147
		@@ -31892,33 +31877,7 @@ mysql> SELECT LOG(-2);
		-> NULL
		@end example
		If you want the log of a number @code{X} to some arbitrary base @code{B}, use

		When called with two parameters, this function returns the logarithm of
		@code{X} for an arbitary base @code{B}:
		@example
		mysql> SELECT LOG(2,65536);
		-> 16.000000
		mysql> SELECT LOG(1,100);
		-> NULL
		@end example

		The arbitrary base option was added in MySQL version 4.0.3.
		@code{LOG(B,X)} is equivalent to @code{LOG(X)/LOG(B)}.

		@findex LOG2()
		@item LOG2(X)
		Returns the base-2 logarithm of @code{X}:
		@example
		mysql> SELECT LOG2(65536);
		-> 16.000000
		mysql> SELECT LOG2(-100);
		-> NULL
		@end example

		@code{LOG2()} is useful for finding out how many bits a number would
		require for storage.
		This function was added in MySQL version 4.0.3.
		In earlier versions, you can use @code{LOG(X)/LOG(2)} instead.
		the formula @code{LOG(X)/LOG(B)}.

		@findex LOG10()
		@item LOG10(X)
		@@ -49880,148 +49839,11 @@ Our TODO section contains what we plan to have in 4.0. @xref{TODO MySQL 4.0}.
		@itemize @bullet
		@item
		Fixed a bug that made the pager option in the mysql client non-functional.
		@end itemize
		@item
		Added full @code{AUTO_INCREMENT} support to @code{MERGE} tables.
		Added OLAP functionality.

		@c Arjen , please add the following text somewhere appropriate in the
		text above:

		-------------------------------------------------------------------------

		Documentation for OLAP extension

		Introduction
		------------

		MySQL will first support CUBE and ROLLUP operators from entire OLAP
		functionality.

		The CUBE and ROLLUP extensions to SQL make querying and reporting
		easier in data warehousing environments. ROLLUP creates subtotals at
		increasing levels of aggregation, from the most detailed up to a grand
		total. CUBE is an extension similar to ROLLUP, enabling a single
		statement to calculate all possible combinations of subtotals.


		Syntax:
		------

		The syntax supported by the enhanced mysql for CUBE and ROLLUP
		operators.

		CUBE
		----

		SELECT field1, field2, ... AGGR(fieldn) FROM
		table GROUP BY field1, field2 field(n-1) WITH CUBE

		This would generate the aggregates with group bys of all the
		combinations of dimensions.

		ROLLUP:
		-----

		SELECT field1, field2, ... AGGR(fieldn) FROM table
		GROUP BY field1, field2 field(n-1) WITH ROLLUP

		Example:
		-------

		mysql> select * from sales;
		+------------+---------------+------+--------+
		\| product \| country \| year \| profit \|
		+------------+---------------+------+--------+
		\| Computer \| India \| 2000 \| 1200 \|
		\| TV \| United States \| 1999 \| 150 \|
		\| Calculator \| United States \| 1999 \| 50 \|
		\| Computer \| United States \| 1999 \| 1500 \|
		\| Computer \| United States \| 2000 \| 1500 \|
		\| TV \| United States \| 2000 \| 150 \|
		\| TV \| India \| 2000 \| 100 \|
		\| TV \| India \| 2000 \| 100 \|
		\| Calculator \| United States \| 2000 \| 75 \|
		\| Calculator \| India \| 2000 \| 75 \|
		\| TV \| India \| 1999 \| 100 \|
		\| Computer \| India \| 1999 \| 1200 \|
		\| Computer \| United States \| 2000 \| 1500 \|
		\| Calculator \| United States \| 2000 \| 75 \|
		+------------+---------------+------+--------+
		14 rows in set (0.00 sec)



		mysql> Select sales.product, sales.country , sales.year, sum(profit) from
		sales group by product, country, year with cube;

		+------------+---------------+------+-------------+
		\| product \| country \| year \| sum(profit) \|
		+------------+---------------+------+-------------+
		\| Calculator \| India \| 2000 \| 75 \|
		\| Calculator \| United States \| 1999 \| 50 \|
		\| Calculator \| United States \| 2000 \| 150 \|
		\| Computer \| India \| 1999 \| 1200 \|
		\| Computer \| India \| 2000 \| 1200 \|
		\| Computer \| United States \| 1999 \| 1500 \|
		\| Computer \| United States \| 2000 \| 3000 \|
		\| TV \| India \| 1999 \| 100 \|
		\| TV \| India \| 2000 \| 200 \|
		\| TV \| United States \| 1999 \| 150 \|
		\| TV \| United States \| 2000 \| 150 \|
		\| ALL \| India \| 1999 \| 1300 \|
		\| ALL \| India \| 2000 \| 1475 \|
		\| ALL \| United States \| 1999 \| 1700 \|
		\| ALL \| United States \| 2000 \| 3300 \|
		\| Calculator \| ALL \| 1999 \| 50 \|
		\| Calculator \| ALL \| 2000 \| 225 \|
		\| Computer \| ALL \| 1999 \| 2700 \|
		\| Computer \| ALL \| 2000 \| 4200 \|
		\| TV \| ALL \| 1999 \| 250 \|
		\| TV \| ALL \| 2000 \| 350 \|
		\| Calculator \| India \| 0 \| 75 \|
		\| Calculator \| United States \| 0 \| 200 \|
		\| Computer \| India \| 0 \| 2400 \|
		\| Computer \| United States \| 0 \| 4500 \|
		\| TV \| India \| 0 \| 300 \|
		\| TV \| United States \| 0 \| 300 \|
		\| ALL \| ALL \| 1999 \| 3000 \|
		\| ALL \| ALL \| 2000 \| 4775 \|
		\| ALL \| India \| 0 \| 2775 \|
		\| ALL \| United States \| 0 \| 5000 \|
		\| Calculator \| ALL \| 0 \| 275 \|
		\| Computer \| ALL \| 0 \| 6900 \|
		\| TV \| ALL \| 0 \| 600 \|
		\| ALL \| ALL \| 0 \| 7775 \|
		+------------+---------------+------+-------------+
		35 rows in set (0.00 sec)


		MySQL supports now CUBE and ROLLUP extensions, with all functions
		that one wishes to use with them.

		Those extensions already work in all tested combinations. They work
		with UNION's, should work with sub-selects in 4.1 and derived tables.

		TODO
		----

		For the moment, ORDER and LIMIT are disabled for CUBE and ROLLUP. This
		however remains to be added later.

		Another feature that has to be added later is grouping of select list
		itmes in order to alleviate user errors. For the moment, missing
		(hidden) columns are not used at all.

		Third feature to be done is to make SQL SET OPTION to set values for
		all values of the column optionally to:

		* ALL (as it is now)
		* NULL
		* blank

		-------------------------------------------------------------------------

		@item
		Added support for @code{CUBE} and @code{ROLLUP}.
		@end itemize

		@node News-4.0.2, News-4.0.1, News-4.0.3, News-4.0.x
		@appendixsubsec Changes in release 4.0.2 (01 July 2002)