Re: SQL queries and their results

[Murray Cumming <murrayc murrayc com>]

On Mon, 2010-04-26 at 22:38 -0400, Scott Nettleton wrote:
> First of all, let me start by saying that I love Glom. I've tried a lot
> of other database softwares, but haven't found anything approaching the
> ease of use and versatility that Glom offers.
>
> Unfortunately, its documentation is rather sparse. I'm familiar with SQL
> databases, and with the python language and syntax; but I'm accustomed
> to accessing SQL through PHP scripts.
>
> In PHP, if I wanted to perform and advanced database query, I would
> write something like:
> $query="SELECT * FROM tblCustomers WHERE fldId = $thisId AND
> fldFrequentCustomer = TRUE";

If you use Find (from the Mode menu), and enter values in two fields,
that will do an AND search.

However, I guess you want to do something in a button script.

> $result=mysql_query($query);
> while ($row=mysql_fetch_assoc($result)) {
>       echo $row["fldFirstName"]." ".$row["fldLastName"]."<br>";
> }
>
> It's useful to me because I can directly access the database without
> having to set up a relationship between tables in a GUI, and because I
> can specify the individual fields by name (fldFirstName, fldId, etc).
> The closest thing I've found to documentation on the subject is at the
> website:
> http://library.gnome.org/users/glom/unstable/sec-calculated-fields.html.en
> The bottom of the page shows an example utilizing an SQL query, but I'm
> not familiar with the functions used to access the information.

This is the pygda API. I must try to get its documentation online.
Sorry.

In the meantime you might have some luck with running this on the
command line:
pydoc -w gda
though the result is quite awful.

It's based on the C libgda API, which has better documentation:
http://library.gnome.org/devel/libgda/stable/

>  The only
> function I see that accomplishes direct data access is
> "data_model.get_value_at(col_index, row_index).get()." The problem is,
> the data must be specified by row index (not too big a problem, but also
> by a numerical column index (rather than the field name). This seems a
> bit convoluted to me, and I expect that there's a simpler way to
> accomplish it, but I haven't been able to find it in online
> documentation or even an Ubuntu forum post.

Yes, I would prefer to use a column name, but you do know the column
number because you wrote the SQL query, so it is doable, right?

I guess the idea is that it would affect performance to do an additional
lookup via the field name.

But if this is important, I can try to add the feature to libda and
pygda for use in a future version of GLom.



-- 
murrayc murrayc com
www.murrayc.com
www.openismus.com