SimpleSQLGenerator

Returns the DatabaseType.TypeValue object that represents the database type. Each SQLGenerator is instantiated for a particular database type.

Converts a double to String representation. For most database vendors, this is just invoking toString on the double object. For Teradata, this will format the number of digits to no more than 15.

Returns the expression for the correlation function to be used for this database. Most RDBMSs have a CORR(X,Y) function, but some do not. For those that do not have a built-in function, we substitute it with the expression (AVG(XY)-AVG(X)AVG(Y))/(STDDEV_POP(X)*STDDEV_POP(Y)) where both X and Y are NOT NULL.

Returns a SQL DDL statement to generate a table based on a SELECT query, not necessarily from any particular table. The entire SELECT query must be supplied to generate something like "CREATE TABLE destinationTable AS selectQuery". This method is useful for calls to UDFs or stored procedures that might be database-specific and don't conform to selecting columns from a source table.

NOTE: Currently, this works only on certain databases like PostgreSQL, Greenplum, Oracle, MySQL, Teradata, but not MSSQL, which does not support CREATE TABLE ... AS SELECT ....

Table name should be fully qualified and delimited, if necessary (i.e. "schemaname"."tablename").

selectQuery is not necessarily a SELECT query, but any query that can feed the CREATE TABLE statement.

Returns a SQL DDL statement to generate a table based on a SELECT query from an existing table. For most databases, this is something like "CREATE TABLE destinationTable AS SELECT columns FROM sourceTable"

Table names should be fully qualified and delimited, if necessary (i.e. "schemaname"."tablename"). it is acceptable to specify a table join as the source table (i.e. table1 INNER JOIN table2 ...), but in such cases, column names might not be unique and full qualification and aliasing may be required (i.e. "schemaname"."tablename"."columnname" AS "aliasname")

Returns a SQL DDL statement to generate a table based on a SELECT query from an existing table. For most databases, this is something like "CREATE TABLE destinationTable AS SELECT columns FROM sourceTable whereClause"

Table names should be fully qualified and delimited, if necessary (i.e. "schemaname"."tablename"). it is acceptable to specify a table join as the source table (i.e. table1 INNER JOIN table2 ...), but in such cases, column names might not be unique and full qualification and aliasing may be required (i.e. "schemaname"."tablename"."columnname" AS "aliasname")

The specified WHERE clause should include the "WHERE" keyword. Note that this is simply SQL that is appended to the query as "SELECT columns FROM sourceTable whereClause" and can include other SQL outside of just a WHERE clause, such as GROUP BY, ORDER BY, etc.

Returns a SQL DDL statement to generate a table or view based on a SELECT query, not necessarily from any particular table. The entire SELECT query must be supplied to generate something like "CREATE TABLE destinationTable AS selectQuery" or "CREATE VIEW destinationTable AS selectQuery", depending on the isView parameter. This method is useful for calls to UDFs or stored procedures that might be database-specific and don't conform to selecting columns from a source table.

NOTE: Currently, creating a table works only on certain databases but not MSSQL, which does not support CREATE TABLE ... AS SELECT ....

Table or view name should be fully qualified and delimited, if necessary (i.e. "schemaname"."tablename" or "schemaname"."viewname").

selectQuery is not necessarily a SELECT query, but any query that can feed the CREATE TABLE statement.

Returns a SQL DDL statement to generate a table or view based on a SELECT query from an existing table. For most databases, this is something like "CREATE TABLE destinationTable AS SELECT columns FROM sourceTable" or "CREATE VIEW destinationTable AS SELECT columns FROM sourceTable", depending on the isView parameter

Table and view names should be fully qualified and delimited, if necessary (i.e. "schemaname"."tablename", "schemaname"."viewname"). it is acceptable to specify a table join as the source table (i.e. table1 INNER JOIN table2 ...), but in such cases, column names might not be unique and full qualification and aliasing may be required (i.e. "schemaname"."tablename"."columnname" AS "aliasname")

Returns a SQL DDL statement to generate a table or view based on a SELECT query from an existing table. For most databases, this is something like "CREATE TABLE destinationTable AS SELECT columns FROM sourceTable whereClause" or "CREATE VIEW destinationTable AS SELECT columns FROM sourceTable whereClause", depending on the isView parameter

Table and view names should be fully qualified and delimited, if necessary (i.e. "schemaname"."tablename", "schemaname"."viewname"). it is acceptable to specify a table join as the source table (i.e. table1 INNER JOIN table2 ...), but in such cases, column names might not be unique and full qualification and aliasing may be required (i.e. "schemaname"."tablename"."columnname" AS "aliasname")

The specified WHERE clause should include the "WHERE" keyword. Note that this is simply SQL that is appended to the query as "SELECT columns FROM sourceTable whereClause" and can include other SQL outside of just a WHERE clause, such as GROUP BY, ORDER BY, etc.

Returns a SQL DDL statement to generate a temporary table based on a SELECT query, not necessarily from any particular table. The entire SELECT query must be supplied to generate something like "CREATE TEMP TABLE destinationTable AS selectQuery". This method is useful for calls to UDFs or stored procedures that might be database-specific and don't conform to selecting columns from a source table.

NOTE: Currently, this works only on certain databases that support temporary tables.

Table name should be fully qualified and delimited, if necessary (i.e. "schemaname"."tablename").

selectQuery is not necessarily a SELECT query, but any query that can feed the CREATE TABLE statement.

Returns a SQL DDL statement to generate a view based on a SELECT query, not necessarily from any particular table. The entire SELECT query must be supplied to generate something like "CREATE VIEW destinationTable AS selectQuery". This method is useful for calls to UDFs or stored procedures that might be database-specific and don't conform to selecting columns from a source table.

Table name should be fully qualified and delimited, if necessary (i.e. "schemaname"."tablename").

selectQuery is not necessarily a SELECT query, but any query that can feed the CREATE TABLE statement.

Returns a SQL DDL statement to generate a view based on a SELECT query from an existing table. For most databases, this is something like "CREATE VIEW destinationTable AS SELECT columns FROM sourceTable"

View names should be fully qualified and delimited, if necessary (i.e. "schemaname"."viewname"). it is acceptable to specify a table join as the source table (i.e. table1 INNER JOIN table2 ...), but in such cases, column names might not be unique and full qualification and aliasing may be required (i.e. "schemaname"."tablename"."columnname" AS "aliasname")

Returns a SQL DDL statement to generate a view based on a SELECT query from an existing table. For most databases, this is something like "CREATE VIEW destinationTable AS SELECT columns FROM sourceTable [whereClause]"

View names should be fully qualified and delimited, if necessary (i.e. "schemaname"."viewname"). it is acceptable to specify a table join as the source table (i.e. table1 INNER JOIN table2 ...), but in such cases, column names might not be unique and full qualification and aliasing may be required (i.e. "schemaname"."tablename"."columnname" AS "aliasname")

The specified WHERE clause should include the "WHERE" keyword. Note that this is simply SQL that is appended to the query as "SELECT columns FROM sourceTable whereClause" and can include other SQL outside of just a WHERE clause, such as GROUP BY, ORDER BY, etc.

Returns a SQL DDL statement to drop a table if it exists, and optionally, cascade the drop to dependent tables. For most databases, this is something like "DROP TABLE IF EXISTS tableName [CASCADE]"

NOTE: Not all databases support CASCADE DROP

Returns a SQL DDL statement to drop a view if it exists, and optionally, cascade the drop to dependent views. For most databases, this is something like "DROP VIEW IF EXISTS viewName [CASCADE]"

NOTE: Not all databases support CASCADE DROP

Returns the expression for modulo division for this database. This differs widely among databases. For example, in PostgreSQL, this is "dividend % divisor". But in Oracle, this is "MODULO(dividend, divisor)". For this reason, we must pass in the expressions used for dividend and divisor.

Returns the name of the entire population standard deviation function to be used for this database. This is defined as σ=SQRT((1/N)Σ_i=1^N (x_i)-μ)²). Note that this is a different function from the sample standard deviation function. Different RDBMSs have different names:
PostgreSQL/GreenPlum: STDDEV_POP() MSSQL: STDEVP() MySQL/MariaDB: STD(), STDDEV(), or STDDEV_POP() Oracle: STDDEV_POP() Teradata: STDDEV_POP() Hive: STDDEV_POP() DB2: STDDEV() Sybase: STDDEV_POP()

Returns the name of the sample standard deviation function to be used for this database. This is defined as s=SQRT((1/(n-1))Σ_i=1^N (x_i-xˉ)²). Note that this is a different function from the entire population standard deviation function. Different RDBMSs have different names:
PostgreSQL/GreenPlum: STDDEV_SAMP() MSSQL: STDEV() MySQL/MariaDB: STDDEV_SAMP() Oracle: STDDEV_SAMP(), STDDEV() Teradata: STDDEV_SAMP() Hive: STDDEV_SAMP() DB2: STDDEV_SAMP() Sybase: STDDEV_SAMP(), STDDEVIATION(), STDDEV()

Returns the name of the entire population standard deviation function to be used for this database. This is defined as σ²=(1/N)Σ_i=1^N (x_i)-μ)². Note that this is a different function from the sample standard deviation function. Different RDBMSs have different names:
PostgreSQL/GreenPlum: VAR_POP() MSSQL: VARP() MySQL/MariaDB: VARIANCE(), VAR_POP() Oracle: VAR_POP() Teradata: VAR_POP() Hive: VARIANCE(), VAR_POP() DB2: VARIANCE() Sybase: VAR_POP()

Returns the name of the sample variance function to be used for this database. This is defined as s²=(1/(n-1))Σ(i=1..n (x_i)-xˉ))²). Note that this is a different function from the entire population standard deviation function. Different RDBMSs have different names:
PostgreSQL/GreenPlum: , "VAR_POP"(), VARIANCE() MSSQL: VAR() MySQL/MariaDB: VAR_SAMP() Oracle: VAR_SAMP(), VARIANCE() Teradata: VAR_SAMP() Hive: VAR_SAMP() DB2: VARIANCE_SAMP() Sybase: VAR_SAMP()

Character or String to use when quoting identifiers. Typically, this is a double-quote (for PostgreSQL and other databases), but can be other characters or even Strings for other platforms.

Wraps an identifier in the appropriate quote character to preserve case and special characters. If there is no quoting mechanism for this database type, just return the argument.

Wraps a schema and table or view name in the appropriate quote character to preserve case and special characters. If schemaName is unspecified, quote only the objectName. Uses quoteIdentifier() for quoting.

If true, subqueries must be aliased. Some databases, such as Oracle, require subqueries (queries that appear in the FROM clause) to be aliased. For example:

SELECT ... FROM ( SELECT ... FROM foo WHERE ... ) AS fooalias WHERE ...

Wraps a column name in the appropriate quote character to preserve case and special characters. If there is no quoting mechanism for this database type, just return the argument.

Note: Deprecated -- please use quoteIdentifier instead.

Returns the name of the standard deviation function to be used for this database. For example, in PostgreSQL, this is "stddev". Note that this is just the function name, and does not include the argument. A typical SQL call would be "stddev(expression)".

Returns the name of the variance function to be used for this database. For example, in PostgreSQL, this is "variance". Note that this is just the function name, and does not include the argument. A typical call would be "variance(expression)".