com.atlassian.confluence.content.render.xhtml.migration.exceptions.UnknownMacroMigrationException: The macro 'next_previous_links' is unknown.

Adding Multiple SQL Dialects

SQL Dialect support allows users to create multiple SQL queries for different SQL dialects by checking the driver. This helps maintain driver-specific functionality/key words in SQL statements of a single data service query.

For example, to determine the data length retrieved from a column, there can be different ways to write the SQL query depending on the SQL driver.

On MySQL and PostgreSQL:

SELECT OCTET_LENGTH(employeeNumber)
FROM Employees;

On Microsoft SQL Server:

SELECT DATALENGTH(employeeNumber)
FROM Employees

On Oracle:

SELECT LENGTH(employeeNumber)
FROM Employees

To avoid writing different data service queries for the same operation depending on the configuration, write all three SQL queries in the same data service query as shown below.

Follow the steps below to add SQL dialects to a query.

  1. The SQL dialect option appears when adding a query to datasources such as RDBMS. For example,
  2. Click Add New SQL Dialect to open the SQL Dialect window. Select the required SQL driver and define the SQL statement as shown below.
     

  3. If the SQL statement should be the same for multiple drivers, select all drivers (e.g., MySQL, PostgreSQL and H2) at once and define the statement as shown below.

     
    To define an SQL dialect for a driver other than the ones listed in the supported drivers list, provide the driver prefix in the text field and define the SQL query as shown above.
com.atlassian.confluence.content.render.xhtml.migration.exceptions.UnknownMacroMigrationException: The macro 'next_previous_links2' is unknown.