Omnidex SQL Quicklinks | ||||
---|---|---|---|---|
Overview | Statements | Functions | Examples | Quick Reference |
The CREATE DATABASE statement declares a database in an Omnidex Environment File. Once one or more databases and their respective tables have been declared in an environment, they can be indexed and retrieved using SQL statements.
An Omnidex database can point to an underlying Oracle, SQL Server or MySQL database, or any other database that can be accessed using ODBC. An Omnidex database can also point to a collection of raw data files that combine together to form a database. These databases are accessible using the Omnidex SQL engines, available through ODBC and JDBC interfaces.
The CREATE DATABASE statement records information about the underlying database in the Omnidex Environment File, but it does not actually create an underlying relational database. The physical database should be created using the relational database's tools. In the case of raw data files, the files are usually received from a data provider, and application or an ETL tool. The CREATE DATABASE statement simply stores meta-data about the database in the Omnidex Environment File.
Omnidex allows multiple databases to be declared in an environment. When an application connects to an Omnidex Environment File, then SQL statements can be issued that access all of the databases in the environment. Databases can be of different types as well, and this allows applications to issue SELECT statements that join tables from different database types, including raw data files.
CREATE DATABASE database_name [NODE grid_node] TYPE database_type [SUBTYPE odbc_subtype] [INDEX_DIRECTORY “directory”] [VERSION “rdbms_version”] [PHYSICAL “rdbms_string”] [USER “rdbms_user” PASSWORD “user_password”] [INDEX_MAINTENANCE index_maintenance_type] IN “environment_filespec” /* must be after all previous clauses */
The database_name uniquely identifies the database within this environment. It may be up to 32 characters long, and can contain letters, digits and the following special characters: ! @ # $ % ^ _. It must begin with a letter and is case-insensitive. The database_name must be unique within the environment. If database_name is a SQL reserved word, enclose it in double quotation marks.
Database_name is used to qualify table and index names when the same table or index name occurs more than once within an Omnidex Environment File.
Database_name is also used as the prefix for index files located in the index_directory as specified in the INDEX_DIRECTORY setting.
CREATE DATABASE SALES IN "SALES.XML"; ... UPDATE INDEXES exit os> dir SALES0001, SALES0002, SALES0003, SALES.XML
The optional NODE clause is used to configure an Omnidex Grid. Nodes that were previously declared in the CREATE ENVIRONMENT statement can be referenced here, and the database configuration can be named for each node. The NODE clause is only required for databases within an Omnidex Grid that are partitioned and distributed across multiple nodes. If not using an Omnidex Grid, the NODE clause can be skipped and the underlying configuration can be specified for the entire database.
For simple environments, the database configuration does not require NODE clauses:
create environment in "list.xml"; create database "LIST" type FILE index_directory "idx" in "list.xml";
For Omnidex Grid environments, the database requires NODE clauses:
create environment max_threads 2 node "GRID01" partitioned node "GRID02" partitioned node "GRID03" partitioned node "GRID04" partitioned node "GRID05" unpartitioned in "list.xml"; create database "LIST" node "GRID01" type FILE index_directory "idx\grid01" node "GRID02" type FILE index_directory "idx\grid02" node "GRID03" type FILE index_directory "idx\grid03" node "GRID04" type FILE index_directory "idx\grid04" node "GRID05" type FILE index_directory "idx\grid05" in "list.xml";
The TYPE clause specifies the type of database being declared. The database_type is required.
Valid database types are:
When a database is spread across multiple nodes of an Omnidex Grid, the PHYSICAL clause is declared for each node. At present, the database_type must be the same across all nodes; however, support for different database types for each node may be supported in the future.
The SUBTYPE setting within the TYPE clause is used to specify a subtype for ODBC databases. Valid subtypes are:
Omnidex can access other databases through ODBC, even if the subtype is not declared. Note that there can be differences in the way that each underlying database implements ODBC. Because of this, some Omnidex functionality may not be available.
The INDEX_DIRECTORY is an optional clause specifies the operating system location of the Omnidex index files. The directory can either be an absolute or relative path. Relative paths are relative to the location of the Omnidex Environment File. Relative locations are recommended when feasible.
If no directory is declared, then the Omnidex index files will reside in the same directory as the Omnidex Environment File.
The directory can also be specified with an environment variable using curly braces.
create database "LIST" type FILE index_directory {$LIST_DIR}
The VERSION clause is used to by Omnidex to determine which syntax and API to use with the underlying relational database.
For Oracle, the choices are:
Version | Description |
---|---|
10 | All versions of Oracle 10 |
11 | All versions of Oracle 11 |
SQLNET10 | All versions of Oracle 10, using SQL*Net to access the database |
SQLNET11 | All versions of Oracle 11, using SQL*Net to access the database |
For SQL Server, the choices are:
Version | Description |
---|---|
8 | All versions of SQL Server 2000 |
9 | All versions of SQL Server 2005 |
The PHYSICAL rdbms_string declares RDBMS connection information and is only used with relational databases and ODBC. The rdbms_string varies according to database type as follows:
Database Type | Description |
---|---|
FILE | ignored |
ORACLE | The PHYSICAL clause is ignored and Oracle will use the default database for the specified user. |
ORACLE w/SQLNET | if using Oracle SQLNET, use the name of the SQLNET service and specify “SQLNETx” in the VERSION clause. |
SQLSERVER | The PHYSICAL clause contains the ODBC connection information. For File DSNs, use “FILEDSN=file_dsn_filespec”. For System and User DSNs, use “DSN=dsn_name” or just “dsn_name” where dsn_name is the data source name set up in the Windows ODBC Administrator program. |
ODBC | The PHYSICAL clause contains the ODBC connection information. For File DSNs, use “FILEDSN=file_dsn_filespec”. For System and User DSNs, use “DSN=dsn_name” or just “dsn_name” where dsn_name is the data source name set up in the Windows ODBC Administrator program. |
The USER and PASSWORD clauses are used to provide a valid RDBMS database user and password to be used for Omnidex to access the underlying RDBMS data source. The USER and PASSWORD clauses are required for Oracle, and are otherwise optional.
All connections to the underlying database will use this rdbms_user and user_password. Omnidex does not have a supplemental security facility that allows sending different users and passwords.
The INDEX_MAINTENANCE clause is an optional clause that specifies how Omnidex indexes will be updated. The index_maintenance_type can be either API or DBMS.
This setting applies to the entire index installation for this database. It can be overridden on a table-by-table basis by explicitly defining the index_maintenance setting on an individual table.
The IN clause is required and specifies the location of the physical XML Environment File. The IN clause must be specified after all other clauses.
Environment_filespec can also include either an absolute or relative path. When using a relative file path , it will be relative to the current working directory for the program that is being run. Typically in a production environment, absolute paths are used. When using Omnidex's Network Services and the OdxNet process, absolute paths should always be used.
The environment_filespec must be enclosed within double quotes.
create environment in "env1.xml" with delete; create database db1 type flatfile in "env1.xml";
create environment in "env1.xml" with delete; create database db1 type flatfile in "env1.xml"; create database db2 type flatfile in "env1.xml"; create table db1.table1 physical "db1table1" /* Note the db1 database qualification */ ( pk integer, mytext char(60) quicktext ) in "env1.xml"; create table "db2"."table1" physical "db2table1" /* Note the db2 database qualification */ ( pk integer, mytext char(60) quicktext ) in "env1.xml"; create file db1table1; create file db2table1; connect env1 insert into db1.table1 values (1, 'db1 row 1'); insert into db1.table1 values (2, 'db1 row 2'); select * from db1.table1; insert into db2.table1 values (1, 'db2 row 1'); insert into db2.table1 values (2, 'db2 row 2'); select * from db2.table1; select * from db1.table1 union select * from db2.table1; disconnect
See also:
Articles: