Re: [gnome-db] gda_connection_execute_select_command and one column
- From: Piotr Pokora <piotrek pokora gmail com>
- To: Diego Zuccato <diego otello alma unibo it>
- Cc: gnome-db-list <gnome-db-list gnome org>
- Subject: Re: [gnome-db] gda_connection_execute_select_command and one column
- Date: Mon, 12 May 2008 21:56:08 +0200
Diego Zuccato writes:
> Piotr Pokora ha scritto:
>> Diego Zuccato writes:
>>>> There's no row in model but there's one column.
>>>> I suppose there should be at least one row if there's at least one
>>>> column.
>>> Why?
>> Because it's 2D model and must have x and y :)
> Nope. Y (rows) could be 0 even if X is greater than 0.
> Try "SELECT * FROM a_table_in_your_db WHERE false".
> It returns a model w/o rows. But column descriptions are there.
And GdaDataModel has 1 column because of this description?
> I did not say *gda* returns the header. I said the backend does.
> It first parses the query, gets the expected column names and their
> types, sends 'em to the frontend w/ other infos (like number of rows),
> then sends each row (more or less... it might even send NO data until
> further requests, to avoid sending all the records when the FE is not
> interested, or just send an iterator).
Yeah, I understand that backened may return additional headers, but
should I expect that GdaDataModel may also hold headers or any
additional data, which is not retrieved data?
GdaDataModel seems to be designed to return X.Y data model.
If there's no row and at least one column, then
gda_data_model_get_value_at seems to be a "bit of luck" function.
I just wonder what could trigger this weird issue.
And I am glad it is working :)
Piotras
[
Date Prev][
Date Next] [
Thread Prev][
Thread Next]
[
Thread Index]
[
Date Index]
[
Author Index]