Subject | findkey/seek on TIB_Query |
---|---|
Author | Boris Schlüszler |
Post date | 2001-12-04T16:08:07Z |
Hello!
Is there a way to accomplish a Seek/FindKey on a TIB_Query?
I have a large dataset which is displayed in a TIB_Grid ("select * from
pipapo").
In this dataset I'd like to make one particular row active. Maybe the
last one.
The column is indexed. I have great respone times with "select * from
pipapo where IDXColumn=:search'").
Is it possible, to Locate in the dataset, without fetching all the rows?
I could imagine something like this:
1.) The buffer is load with as many values as needed for display. (As
indeed is, as long as AutoFetchAll is set to False): "select * from
table"
2.) Search value is given. On Interbase a Cursor with the search values
is declared. "select * from table where Search=:search"
3.) Fetch (and add) as many rows from the Cursor into the current
dataset's buffer
as needed for display. While doing this, suppress duplicates by checking
KeyLinks.
Then I could overcome longer times of loading column values for
scannning all the rows
into the buffer.
Maybe this is implemented and I do not understand it?
Regards, Boris
Is there a way to accomplish a Seek/FindKey on a TIB_Query?
I have a large dataset which is displayed in a TIB_Grid ("select * from
pipapo").
In this dataset I'd like to make one particular row active. Maybe the
last one.
The column is indexed. I have great respone times with "select * from
pipapo where IDXColumn=:search'").
Is it possible, to Locate in the dataset, without fetching all the rows?
I could imagine something like this:
1.) The buffer is load with as many values as needed for display. (As
indeed is, as long as AutoFetchAll is set to False): "select * from
table"
2.) Search value is given. On Interbase a Cursor with the search values
is declared. "select * from table where Search=:search"
3.) Fetch (and add) as many rows from the Cursor into the current
dataset's buffer
as needed for display. While doing this, suppress duplicates by checking
KeyLinks.
Then I could overcome longer times of loading column values for
scannning all the rows
into the buffer.
Maybe this is implemented and I do not understand it?
Regards, Boris