RE: DBI:CSV join ... flashback to 2002
- From: cleach@xxxxxxxxxxx (Curtis Leach)
- Date: Mon, 16 Jul 2007 09:24:45 -0500
Another way to remove the dot in the query is as follows.
$dbh->{'csv_tables'}->{table_a} = { 'file' => table_a.csv };
$dbh->{'csv_tables'}->{table_b} = { 'file' => table_b.csv };
$sth = $dbh->prepare("Select a.subject from table_a as a, table_b as b
where a.subject=b.subject");
I haven't verified it will resolve your issue, but if it works for you,
you don't have to rename all your files.
Good luck,
Curtis
-----Original Message-----
From: Robert Roggenbuck [mailto:rroggenb@xxxxxxxxxxxxxxxxxxxxxxxxxxxx]
Sent: Monday, July 16, 2007 8:25 AM
To: ballard.d@xxxxxxxxx
Cc: dbi-users@xxxxxxxx
Subject: Re: DBI:CSV join ... flashback to 2002
The problem seems to be the dot in the table names. The dot have special
meanings in SQL for table names and is not just an allowed character
(for example it is used as scheme separator 'scheme.tablename'). Just
try to rename Your CSV-files by cutting of the extension '.csv'. May be
all the problems will be gone then.
Best regrads
Robert
----
ballard.d@xxxxxxxxx schrieb:
Question about joining using DBI:CSV. I must be making a stupid
mistake somewhere.
Earlier posts (2002) state that there were problems using aliases in
joins. I've just installed the modules on a windows machine and am
having the same problems.
DBI (v1.58)
SQL-Statement(v1.15)
Text-CSV_XS(v.3)
DBD-CSV(v.22)
DBD-File(v.35)
Running a simple query with no aliases works.
When I run the same query with an alias i get an error message
(below). However, the error message is followed by the correct query
results.
$sth = $dbh->prepare("Select b.subject from table_b.csv as b"); $sth =
$dbh->prepare("Select a.subject from table_a.csv as a");unblessed reference at c:\perl\site\lib\SQL\Statement.pm line 610, <GEN>
SQL ERROR: Table 'CSV' referenced but not found in FROM list!
I then tried a simple join (using where a.subject=b.subject). I get
the error referenced in previous posts.
$sth = $dbh->prepare("Select a.subject from table_a.csv as a,
table_b.csv as b where a.subject=b.subject");
DBD::CSV::st execute failed: Can't call method "col_names" on
line 1.
[for Statement "Select a.subject from table_a.csv as a, table_b.csv as
b where a.subject=b.subject"] at temp.pl line 14
I then tried using the join syntax instead of the WHERE statement as
suggested in the previous post, but i get an error message
$sth = $dbh->prepare("Select * from table_a.csv natural join
table_b.csv");
SQL ERROR: Couldn't parse the explicit JOIN!
table_a.csv
-----------------
chromosome,snp,subject,xx
1,rs1203,102,A
1,rs1203,1025,A
1,rs1203,1034,A
1,rs1203,1078,A
table_b.csv
-----------------
subject
102
1025
1034
If anyone has any suggestions, please let me know. I would greatly
appreciate it.
cheers,
david
.
- References:
- DBI:CSV join ... flashback to 2002
- From: Ballard.D@xxxxxxxxx
- Re: DBI:CSV join ... flashback to 2002
- From: Robert Roggenbuck
- DBI:CSV join ... flashback to 2002
- Prev by Date: Re: DBI requires 5.6.0 or 5.6.1?
- Next by Date: RE: Trouble Installing ODBC 1.13 on Linux 64
- Previous by thread: Re: DBI:CSV join ... flashback to 2002
- Next by thread: DBI requires 5.6.0 or 5.6.1?
- Index(es):
Relevant Pages
|
