SQLite Implementation

The SQLite engine applies the second-level query to the data fetched by the data filter. This query is a SQL-compliant one with some restrictions enforced by the DataHub query engine:

• Only a single query is supported, no multiple comma-separated queries
• No SQL statement other than SELECT ... FROM Data is allowed
• The query may only use Data table as source table
• Common Table Expressions (CTE) are not supported
• Join clauses are not supported

Input table

The only table that you can query is called Data and has the following schema:

CREATE TABLE Data (
    variableId INT,
    sourceId INT,
    value DOUBLE,
    timeStamp TIMESTAMP)

This means that you can only specify VariableId, SourceId, Value and TimeStamp as source columns.

Note

You can select a constant (SELECT 'Some constant', ... FROM Data) or computation result (SELECT 1 + 2, ... as Result FROM Data). Although being useless, if you do not select any column from Data table, you still need to provide the source table, which would return 1 constant/result per row in Data table. Use DISTINCT to get a single result: SELECT DISTINCT 'Some constant' FROM Data

Result set

The structure and content of the result set is driven by the query which drives the following result set characteristics:

• the number of columns per row
• the name of each column (use SQL aliases to customize a column name)
• the type of each column (use SQL CAST to customize a column type )