-- Last updated 20100802 (for Windows)

Added truncate_table() stored procedure definition to create_tables.sql
file for carsales database.

-- Last updated 20100802 (for Windows)

One hundred and ten (110) queries have been added to the sql.zip file - 
55 running against brighthouse tables and 55 running against the MyISAM
tables (if created, loaded and indexed).  Descriptions of the queries can
be found in the file query_descs.txt.

-- Last updated 20100802 (for Windows)

Added explicit CHARSET and COLLATE clauses to all DDL to ensure that CMAP
Knowledge Nodes get created.

-- Last updated 20090802 (for Windows)

In Windows Explorer, if one does not exist, create a folder named c:\tmp

Copy the demo database ZIP file to this folder.

Unzip the <data>_carsales.zip file.

(If a different path, or paths, were used than above the LOAD script will need
to be edited to reflect the correct path.)

Create a carsales database schema in the mysql client

# mysql-ib 
-- splash info omittied --
mysql> create database carsales;

Select the new carsales database schema.

mysql> use carsales

Execute the DDL creation script.

mysql> source c:\tmp\carsales\create_tables.sql

Warnings are normal the first time you run this.  It contains DROP table 
statements.  Run it again and confirm that there are no warnings or errors.

A list of the tables created will be displayed at the end.

+--------------------+
| Tables_in_carsales |
+--------------------+
| _dim_cars          |
| _dim_dates         |
| _dim_dealers       |
| _dim_msa           |
| _dim_sales_area    |
| _dim_vehicles      |
| _fact_sales        |
| _fact_sales_wide   |
| dim_cars           |
| dim_dates          |
| dim_dealers        |
| dim_msa            |
| dim_sales_area     |
| dim_vehicles       |
| fact_sales         |
| fact_sales_wide    |
+--------------------+
16 rows in set (0.00 sec)

Load the data.

mysql> source c:\tmp\carsales\load_carsales.sql

NOTE: This loads the same data into identical Infobright and MyISAM tables for
comparison purposes.  It also loads a 10 million record fact table, a 9.3
million normalized version of the table and its companion 9.3 million record
dimension table.  This takes a litte time in Infobright and A LOT of time with
the MyISAM tables.  You can comment out loading the MyISAM tables if this is not
of interest to you or you don't have the space for it.

NOTE:  The fact_sales_wide table (and _fact_sales_wide in MyISAM) is a
denormalized version of the fact_sales table that includes a complete
denormalization of the dim_vehicles and dim_dates tables.  As there is a
one-to-one relationship between the events in the fact_sales table and the
vehicles in the dim_vehicles table, this is a reasonable design approach.  As
the dim_dates is very narrow, this also is a consideration one might make.