Re^4: Combine data in table

[DrAxeman]

in reply to Re^3: Combine data in table
in thread Combine data in table

I got the Perl loop method to work (mostly for the experience). But now I'm working on the SQL method. My code looks like:

# Connect to the database, (the directory containing our csv file(s))
my $dbh = DBI->connect("DBI:CSV:f_dir=.;csv_eol=\n;");

# Associate our csv file with the table name 'results'
$dbh->{'csv_tables'}->{'results'} = { 'file' => "psinfooutputdiskfile.
+csv"};

my @rows = $dbh->selectrow_arrayref("SELECT IP, SUM(PartitionSize), SU
+M(PartitionFree) FROM results GROUP BY IP");

print ("@rows\n"); # Just here to view output

I'm getting

SQL ERROR: Can't find table names in FROM clause!


Execution ERROR: No command found!.

Not sure what is wrong here.

it I modify the SELECT statement to:

my @rows = $dbh->selectrow_arrayref("SELECT IP FROM results");

It works. But if I try to include a GROUP BY or SUM I get the error.

[davidrw]

Note first that you probably want my $rows = $dbh->selectall_arrayref( ... ) instead of selectrow_arrayref()

are those column names right? should they be "colIP", "colPartitionSize", "colPartitionFree" instead?

Besides that, the SQL looks fine .. do the statements "SELECT IP FROM results" and "SELECT IP FROM results GROUP BY IP" and "SELECT SUM(PartitionSize), SUM(PartitionFree) FROM results" work?

[DrAxeman]

#!/usr/bin/perl
    use strict;
    use warnings;
    use DBI;
    use Data::Dumper;

# Connect to the database, (the directory containing our csv file(s))
my $dbh = DBI->connect("DBI:CSV:f_dir=.;csv_eol=\n;");

# Associate our csv file with the table name 'results'
$dbh->{'csv_tables'}->{'results'} = { 'file' => "psinfooutputdiskfile.
+csv"};

my @rows = $dbh->selectall_arrayref("SELECT IP FROM results GROUP BY I
+P"); 
print Dumper @rows;

Fails. If I take the GROUP BY off, everything is OK. Same with the SUM statements.

"SELECT IP FROM results" works
"SELECT IP FROM results GROUP BY IP" and "SELECT SUM(PartitionSize), SUM(PartitionFree) FROM results" fail.

[davidrw]

doh. after all this, 'GROUP BY' isn't supported by DBD::CSV ... following is from the cb:

davidrw 2005-08-07 16:03:24-04
    jZed (or anyone) did you see 481726 and replies? Any reason that a
+ "GROUP BY" sql statement wouldn't work w/DBD::CSV ?
bart 2005-08-07 16:05:53-04
    Because it didn't get implemented?
castaway 2005-08-07 16:06:22-04
    always a good reason ;)
davidrw 2005-08-07 16:08:10-04
    not implemented? doh! is that a guess or a fact?
davidrw 2005-08-07 16:08:59-04
    davidrw goes to skim DBD::CSV docs for known bugs/limitations
davidrw 2005-08-07 16:10:24-04
    davidrw doesn't find anything obvious in DBD::CSV or DBD::File pod
Zaxo 2005-08-07 16:12:43-04
    since there isn't a db daemon managing things, constructs like tha
+t have to be written from scratch in DBD::CSV
castaway 2005-08-07 16:13:25-04
    right
Corion 2005-08-07 16:13:41-04
    Hmmm. "Group By" would likely be a hash filter sitting between the
+ row-scanner and the aggregator function(s)
Corion 2005-08-07 16:14:03-04
    ... so it "shouldn't be hard". But then, what do I know about DBDs
+ :)
davidrw 2005-08-07 16:14:24-04
    hmm.. true.
davidrw 2005-08-07 16:15:15-04
    i wonder if any other non-db-daemon DBD:: modules implement GROUP 
+BY?
Zaxo 2005-08-07 16:16:32-04
    Berkeley might, but I don't know
Corion 2005-08-07 16:17:55-04
    davidrw: SQLite does. :)
davidrw 2005-08-07 16:18:42-04
    ah. y, i knew that -- i use it a lot :)
davidrw 2005-08-07 16:18:52-04
    it's been a slow lazy sunday ...