For example, EXPLAIN SELECT 1, where this is currently identified as UNKNOWN. We should probably support this as it's somewhat common, and easy to support (just detect if first word in EXPLAIN.
Support for it would probably be adding a new command type (EXPLAIN) and execution type (EXPLAIN) as it doesn't really make sense to call it a LISTING or MODIFICATION.
This would improve the state of things for consumers where while the pg npm package will return EXPLAIN as a command type which we can leverage, sqlite on the other hand would just have UNKNOWN.
Any reason not to do this @rathboma, other than perhaps that different databases have different outputs for explain (e.g. sqlite vs postgresql).
For example,
EXPLAIN SELECT 1, where this is currently identified asUNKNOWN. We should probably support this as it's somewhat common, and easy to support (just detect if first word inEXPLAIN.Support for it would probably be adding a new command type (
EXPLAIN) and execution type (EXPLAIN) as it doesn't really make sense to call it aLISTINGorMODIFICATION.This would improve the state of things for consumers where while the
pgnpm package will returnEXPLAINas a command type which we can leverage, sqlite on the other hand would just haveUNKNOWN.Any reason not to do this @rathboma, other than perhaps that different databases have different outputs for explain (e.g. sqlite vs postgresql).