Thank you for the feedback. I ended up reading one row at a time,
changing the columns that needed to be changed, then updating as follows:
########################## perl
############################################################
$fetch = $db -> query('SELECT * FROM sardata ORDER BY server')
or die $db -> error;
###################### made all the changes I needed to the record
##############
while ($row = $fetch->hash)
{
$db->update('sardata',
{
datetime => $row->{datetime},
numcpu => $row->{numcpu},
cpuspeed => $row->{cpuspeed},
cpuunits => $row->{cpuunits},
osversion => $row->{osversion},
osrelease => $row->{osrelease},
oscode => $row->{oscode},
osname => $row->{osname},
availtm => $row->{availtm},
sartm => $row->{sartm}
},
{ id => $row->{id} })
}
################################################################### end
##########
Regards,
Uriel_Carrasquilla
Clark Christensen
<[EMAIL PROTECTED] To:
[email protected]
om> cc:
Subject: Re: [sqlite] Cursors
02/02/2005 02:46
PM
Please respond to
sqlite-users
--- [EMAIL PROTECTED] wrote:
>
>
>
>
>
> Hello list!
> I have run into problems trying to use SQLite via perl
> DBA.
> According to the SQLite.org web site, DECLARE CURSOR
> should be supported
> (actually, it is not in the list of the only SQL not
> supported).
> I get a error saying "error near DECLARE". I have tried
> to declare cursor
> in many ways but always get the same error.
> If "cursro" is not supported, does anybody have a
> suggestion on how to
> achieve the same end?
As I remember, the PERL DBI doesn't support cursors.
Instead, depending on how you requested the data, you
end-up with something like an array, and you iterate
through it. I leave it vague here, because, like with most
other things PERL, there are numerous ways to get what you
want :-)
I for small to moderate sized datasets, like the DBI
utility methods. $dbh->selectall_arrayref('select a, b, c
from t1'); returns an array (records) of arrays (columns in
record).
Otherwise, you can do the 3 or 4-step process of:
$sth = $dbh->prepare('select a, b, c from t1');
$sth->execute();
$sth->bind_columns(undef, \$col_a, \$col_b, \$col_c);
while ($sth->fetchrow_arrayref()) {
#do something with the row data
print "$col_a\t$col_b\t$col_c\n"; #perhaps
}
When you use the multi-step prepare/execute/fetch
algorithm, once a row is fetched, you can't back-up later
and fetch it again.
If you use the utility methods like
$dbh->selectall_arrayref() and its cousins, you iterate
through the array, and can re-set the iterator to jump
forward or back at any time.
Have a look at the DBI documentation. I always need more
explanation then those PERL module docs provide, and I
found the O'Reilly PERL DBI book very useful.
There's also some useful explanations and usage tips
available. Just do a Google search on "perl dbi".
Of particular interest for me was this presentation
http://search.cpan.org/src/TIMB/DBI_AdvancedTalk_2004/index.htm
-Clark
