On 12.02.2008 07:56 [EMAIL PROTECTED] wrote:
> I get a
>
> closing dbh with active statement handles at ...Rose/DB.pm line 832.
>
> warning at the end. As
>
>
> http://www.mail-archive.com/rose-db-object@lists.sourceforge.net/msg02234.html
>
> indicates, this was attributed to DBD::SQ
On 04.02.2008 21:41 John Siracusa wrote:
> More bug fixes that I didn't want to wait until the next major release.
I get this test failure with 0.7663:
Test Summary Report
---
t/db-object-loader-8.t (Wstat: 65280 Tests: 19 Failed: 16)
Failed test number(s): 4-19
On 16.12.2007 02:05 John Siracusa wrote:
> On Dec 15, 2007 10:43 AM, Michael Lackhoff <[EMAIL PROTECTED]> wrote:
>> I guess, what you are talking about is done here:
>> $form->init_fields_with_cgi($self->query).
>
> What is $self in that line?
My (CGI::)Ap
On 14.12.2007 19:12 John Siracusa wrote:
> I think your problem is at a level higher: it's the thing feeding your
> params() to the form. If you've decided that an empty text field
> should result in an undef field value, then the piece of code you have
> that converts from a query string like "a
On 14.12.2007 09:21 Darren Duncan wrote:
> If you have a place that holds a value, it should always hold a
> value. If you don't know a value for a place or a value isn't
> applicable there, then the place shouldn't exist at all.
>
> The more correct way to do this which SQL supports is throug
On 13.12.2007 14:03 John Siracusa wrote:
> On Dec 13, 2007 4:24 AM, Michael Lackhoff <[EMAIL PROTECTED]> wrote:
>> When I create a record with such a field, it is usually NULL but
>> after some editing with the help of RHTMLO the empty field value could
>> become '
Hello,
perhaps this is related to the new "undef_sets_null" but I am not sure.
The problem:
I have some fields where it is important if they are "empty" or not. Now
"emptyness" is not so easy in SQL as we all know, it could be '' or 0 or
NULL. When I create a record with such a field, it is usuall
Hello,
I did some more tests and it looks as if everything works as it should
but RDBO still gives an error.
This is from a DBI->trace:
<- fetch= [ '[EMAIL PROTECTED]' 'ZBMED' 'K002054024' 'mysecret' ]
row1 at Object.pm line 385
<- rows= '0E0' at Object.pm line 387
<- finish= 1 at Obje
Hello,
Shame on me, I had forgotten to add a mapping of the driver to the
Oracle::Proxy class, now it is working.
After I got the connection I first tried to use the loader to create all
the classes for me but that didn't work. I don't know if the reason is
the proxy or RDBO because I couldn't te
On 09.08.2007 13:42 John Siracusa wrote:
> On 8/9/07 6:13 AM, Michael Lackhoff wrote:
>> I get this error message:
>> Attempt to change driver from 'oracle' to 'proxy' detected. The driver
>> cannot be changed after object creation. at test02.pl line 8
On 17.08.2007 22:53 John Siracusa wrote:
>> So, if I understand it right, there are really two problems:
>> - no rollback when it should be (my do_transaction above)
>> - a rollback when there should be none (your warn)
>
> Once the database handle is reset to auto-commit, any subsequent
> rollba
On 17.08.2007 16:20 John Siracusa wrote:
> On 8/15/07, Michael Lackhoff <[EMAIL PROTECTED]> wrote:
>> my $db = My::DB->new();
>>
>> $db->do_transaction(
>> sub {
>> my $o = My::Thing->new();
>> $o->
On 15.08.2007 18:47 Perrin Harkins wrote:
> I have a guess. Looking at the new code in Apache::DBI, I see that it
> now tries to reset the state of the $dbh every time you call
> connect(). This probably breaks the begin_work() state.
>
> I'm not sure why this was added, since it seems like thi
On 14.08.2007 22:43 John Siracusa wrote:
> On 8/14/07, Michael Lackhoff <[EMAIL PROTECTED]> wrote:
>> On 14.08.2007 20:34 John Siracusa wrote:
>>> On 8/14/07, Michael Lackhoff <[EMAIL PROTECTED]> wrote:
>>>> $db->begin_work;
>>>> my
On 14.08.2007 20:37 Bill Moseley wrote:
> On Tue, Aug 14, 2007 at 08:28:40PM +0200, Michael Lackhoff wrote:
>> > Have you also looked at Roes::DB's do_transaction() ?
>>
>> I know that this is another possible variant but because there is so
>> much code I didn
On 14.08.2007 20:34 John Siracusa wrote:
> On 8/14/07, Michael Lackhoff <[EMAIL PROTECTED]> wrote:
>> $db->begin_work;
>> my $obj = Products->new(db => $db);
>> ...
>> $obj->save;
>> some_helper_that_uses_db(); # rollback here
>> my $obj2 =
On 14.08.2007 20:20 Bill Moseley wrote:
> On Tue, Aug 14, 2007 at 11:57:27AM +0200, Michael Lackhoff wrote:
>> my $db = MyApp::DB->new;
>> $self->my_db($db); # a new db for a new transaction
>> $db->begin_work; # Start transaction
>>
>> ... create two
On 14.08.2007 18:54 Perrin Harkins wrote:
> On 8/14/07, John Siracusa <[EMAIL PROTECTED]> wrote:
>> The simplest way to do this is to leave everything as per the
>> defaults, and just use Apache::DBI.
>
> Note that Apache::DBI also does the automatic rollback for you, but
> only if you have AutoC
On 14.08.2007 17:50 Perrin Harkins wrote:
>> do you have any pointer on how this is done (correctly)?
>
>$r->push_handlers(PerlCleanupHandler => sub {
>MyApp::DB->rollback();
>});
Thanks! That's exactly what I needed. I found already that there is a
PerlCleanupHandler directive i
On 14.08.2007 16:54 Perrin Harkins wrote:
>> $db->begin_work; # Start transaction
>> die "Does it roll back?";
>> $db->commit;
>
> I don't see any rollback there. It's a common practice to put an
Well, my understandig was that if perl dies within a transaction it (or
DBI) would do a rollback,
Hello,
I thought I understood how to organize several write operations within a
transaction but it still doesn't work the way I do it.
Here is what I have so far:
In my base class:
sub init_db {
my $self = shift;
my $db;
if (not $self->my_db) {
$self->my_db( MyApp::DB->new() )
Hello,
I am trying to use Rose::DB for a new project that uses an Oracle
database but the connection is done with DBI::Proxy (then I don't have
to install DBD::Oracle on every machine that accesses the database)
After reading the docs the best I could manage was this:
My::DB->register_db(
dom
On 01.08.2007 17:43 George Hartzell wrote:
> I think we may have gone over this, but I can't find it in the list.
>
> I'm using current versions of Rose::DB::Object and friends on a
> Freebsd -STABLE system. I'm using sqlite 3.3.17 and DBD::SQLite 1.13.
I cannot say anything about your particul
Hello,
after some problems with broken relationships I tried to use delete with
the cascade option.
I found two problems with cascade:
- if multiple relationships point to the same table, the delete is
executed more often than necessary (once for each relationship)
This is not really a problem
Hello,
I would like to have the possibility to link from one record to other records
in many other tables,
something like this:
in table dates:
id = 1
when = 2007-07-04 14:00
... some more fields
a mapping table would look like this:
date_id = 1
link_id = 5
link_table = customers
date_id = 1
John Siracusa wrote:
> Okay, I've made the change in SVN. RDBO 0.765 will have tN table
> aliasing on by default, even for queries that only involve a single
> table. The new table_aliases Manager parameter can be used to force
> the old behavior.
Thanks!
-Michael
On 30 Jun 2007 at 17:10, John Siracusa wrote:
> Tables are only aliased as tN when more than one table is involved in
> the query. This "more than one" count does not include literal
> subqueries like yours, however. I'm thinking of changing it in 0.765
> so that the tN aliases are always used,
Hello,
I just stumbled upon a changed behaviour that causes my app to crash
with RDBO 0.764 (works with 0.758).
The problem occurs in a method that should give me all the groups a
user (therapeut in my case) is not in:
sub notingroups {
my $self = shift;
my $nr = $self->nr; # user id
On 27 Jun 2007 at 9:45, John Siracusa wrote:
> > And if it works in principle do I have to stuff the $dbh manually into every
> > RDBO object or does it "just work" because Apache::DBI takes care that RDBO
> > gets the right (perhaps reused) handle even if RDBO always requests a fresh
> > one?
>
David,
> > If yes, my question is, how all the RDBO internals work with such a setup.
> > If I remember correctly
> > RDBO wraps complex saves or deletes within a transaction. Would this lead
> > to something like
> > nested transactions? Is this possible at all, with sqlite (I am using now)
Hello,
I read quite a lot recently about Apache::DBI, transactions and the problems
that might occur but I
must say that I am more confused than before as to what this all means for my
RDBO project.
It is a web app with lots of database activity (read and write). So performance
and data integ
On 22 Jun 2007 at 10:03, Perrin Harkins wrote:
> On 6/22/07, Michael Lackhoff <[EMAIL PROTECTED]> wrote:
> > Thanks for the hint! This almost did it. Some tests showed that I had
> > to use the connection-variant of pnotes in my setup:
> >
> > use Apache
On 20 Jun 2007 at 13:40, Perrin Harkins wrote:
> On 6/20/07, Michael Lackhoff <[EMAIL PROTECTED]> wrote:
> > Is this possible? I tried $r->dir_config(user_obj => $rdbo_user_obj); from
> > the mod_perl
> > documentation but that didn't work.
>
> I
I am using Apache2::AuthCookie for my authentification and authorisation. For
both I have to get
the user record from the database.
For authorisation I have to do an additional lookup for an ordinary item that
is requested by the
user. I need this record to check e.g. if the requested item and
Hello,
I had to change a relationship from a simple ID to ID relationship to
one that also includes a second ID.
Worse still the second ID is not a column in one of the relationship
partners but has to be taken from a record two steps up the hierarchy.
Perhaps I will change the table to include
On 31 Jan 2007 at 8:52, John Siracusa wrote:
> I don't use CGI::Application, so maybe I'm misunderstanding how it works,
> but how about a class method that returns the current CGI::Application
> object? e.g., MyWebSite->current_app(). You'd set it at the beginning of
> each request. Then, in y
On 30 Jan 2007 at 17:00, John Siracusa wrote:
> id is in $session, which the form knows nothing about. But in most
> web app frameworks, there's usually some way to get at "globally
> applicable" data like the session, in which case it's reasonable for
> customer_from_form() to return you an obje
Thanks guys for all the help, I'll have something to digest now and learned
already a lot!
Just a short reply to the ORM versus SQL issue:
Yes you are right, you have to pay for the abstraction but then, if you manage
to fit it all in, it pays off.
I can see it with lots of scripts I developed f
On 5 Dec 2006 at 11:02, Perrin Harkins wrote:
> You can usually do this kind of ++ thing in one INSERT...SELECT statement:
>
> INSERT INTO invoice (id) SELECT MAX(id)+1 FROM invoice;
That's clever! Though it seems to slightly defeat the point in using
RDBO: To have Perl, classes and methods ins
On 5 Dec 2006 at 10:32, John Siracusa wrote:
> >> sub generate_invoice_number
> >> {
> >> my $class = shift;
> >> my $db = My::DB->new; # Rose::DB subclass
> >
> > Why not just use the default here?
>
> What default?
The one that is inherited from my common base class (I fol
Hello John,
many thanks for your help, the only part I don't understand is why I need a new
DB object:
> sub generate_invoice_number
> {
> my $class = shift;
> my $db = My::DB->new; # Rose::DB subclass
Why not just use the default here?
> my $num;
>
> eval
>
On 5 Dec 2006 at 8:41, Sean Davis wrote:
> > Is there an idiom for this kind of task? And how can I do it with RDBO?
> > (I did a search for 'lock' in the docs and didn't find anything)
>
> Are you using postgresql? If so, could you use a sequence and set the
> sequence back to 1 at a specified
Hello,
I need some pseudo-unique numbers like invoice numbers. They start every
year with 1, so autoincrement won't work. My idea was to use a helper
table with just just two fields 'year' and 'last_used_number'. Is it
enough to create a two column unique key or would it be better to do
something
On 5 Sep 2006, John Siracusa wrote:
> As for the uniqueness issue, if the combination really is
> unique, then why not add a unique key to the database and the class?
> Then you can load() speculatively instead of doing a Manager query.
That's what I was looking for. Thanks for the hint, I didn't
Hello,
in my application I have a hierarchy of company, organisational unit
and sub-unit. Some data should be configurable on every level but not
compulsory on every level. If it is not defined for the sub-unit it
should be taken from the unit and if it is not there from the
company.
An examp
Hello,
I just wanted to be up to date again and noticed that the oracle test
files both have tests => 22 instead of 53, leading to a test failure.
With the numbers corrected all tests pass.
Cheers,
Michael
-
Using Tomcat
On 29 May 2006 at 13:21, Svilen Ivanov wrote:
> I would like to extend RDBO to support versioning of properties for each
> instance. To explain it better I will use analogy with CVS:
> * calling "->save" will make new revision of the object ("cvs commit")
> * "->load(id=>1)" will get
On 16 May 2006 at 9:28, John Siracusa wrote:
> Yeah, I've been debating requiring DBD::SQLite 1.11 or later. SQLite
> in general has problems correctly honoring multi-column ORDER BY
> clauses, even when using the sqlite3 command-line tool. I'm not sure
> how or why using a newer version of DBD:
Hi Ron,
> I have DBD::SQLite V 1.09 installed. V 1.12 won't install under Windows.
> After installing V 1.11, I re-ran the Rose::DB::Object install. The errors
> above
> disappeared.
You can get a ppm of v 1.12 from the Canadians:
ppm install http://theoryx5.uwinnipeg.ca/ppms/DBD-SQLite.ppd
Wor
Hi Ron,
> > You can get the free express edition here:
> > http://www.oracle.com/technology/software/products/database/xe/index.h
> > tml
>
> Thanx.
>
> Can you access this Oracle from Perl?
I didn't try yet. Privately I use SQLite or MySQL and at work we have a
"normal" version.
The reason,
On 10 May 2006 at 20:13, John Siracusa wrote:
> On 5/10/06 7:20 PM, LU, ZHENYUAN [AG/1000] wrote:
> > I am using Oracle.
>
> Unfortunately, Rose::DB does not yet support Oracle. What you're using is
> the "generic" interface that Rose::DB falls back to when it encounters an
> unknown database.
>
On 10 May 2006 at 17:39, John Siracusa wrote:
> On 5/10/06, Jesse Brown <[EMAIL PROTECTED]> wrote:
> > I guess the question I should ask is: Why NOT return the newly created
> > objects?
>
> I just like to know the reasoning before I change the API :) The code
> examples are exactly what I was
On 5 May 2006, John Siracusa wrote:
> There's no way around copying every field when making a copy! :) That's
> just what RDBO's (undocumented) clone() method does. It's not documented
> because I'm not sure if it does everything that a clone method should do.
> You can be the first to try it an
Hello John,
thanks for your help on my "negative groups" question. I used the
manager method as you suggested and it is working fine now!
But as you might have guessed by now I ran into a new problem. (By
the way, if I am asking too many questions just stop me. Then I will
only ask again when I
Hello,
I seem to have particular problems with negative thinking ;-)
Well, here is the problem: I use a many to many relationship to
define group membership of users. So I have three tables 'users',
'groups' and 'groupmappings' and a relationship definition that looks
like this (within the User
On 30 Apr 2006, John Siracusa wrote:
> I can't think of any reason this would behave any differently under
> mod_perl.
You are right. Sorry, it was another one of these silly mistakes. I
have two perl installations, one as a "normal" installation and
another one that came with apache, this is t
Hello,
I am trying to write an authorization handler in mod_perl and got a
very strange behaviour.
What I want to achieve is use the group memberships stored in a
database that is handled by RDBO to restrict access to some parts of
my application.
This script works perfectly well when run from
On 4 Apr 2006, John Siracusa wrote:
> Rose::DateTime:
>
> 0.52 (04.04.2006) - John Siracusa <[EMAIL PROTECTED]>
>
> * Allow negative and fractional epoch values in parse_date().
> * Fixed a bug in the -mm-dd date parsing that caused it to
> incorrectly match 10-digit seconds-s
Just for the record, I found it.
> >> 'NOT EXISTS (SELECT * FROM products p WHERE p.vendor_id = id)';
> >>
> >> $vendors = Vendor::Manager->get_vendors(clauses => [ $subselect ]);
> >
> > This doesn't work for me (result list is empty but should have two
> > entries). If I change the
Hello John,
I did some more testing, got it working but still don't understand
everything.
> This should work:
>
> $vendors =
> Vendor::Manager->get_vendors(
> with_objects => [ 'products' ],
> query=> [ 'products.vendor_id' => undef ]);
Yes, this works fine.
On 24 Mar 2006, John Siracusa wrote:
> Both of the above approaches should be possible in a relationship
> method by using the manager_args and query_args relationship
> parameters.
Many, many thanks! It is really great how all of you help a poor
little newbee like myself ;-)
Cheers,
Michael
On 24 Mar 2006, Cees Hek wrote:
> SELECT vendors.* FROM vendors LEFT JOIN products ON (vendors.id =
> products.vendor_id) WHERE products.id IS NULL;
Thanks a lot, that works great!
> A left join will always grab all matching records from the left table,
> even if it doesn't have matching record
Say, I have two tables: vendors and products.
Now I want a list of all the vendors that have no products yet.
This can be seen from the products table if there are no records with
the vendor_id of the vendor in question.
Is this possible with a 'normal' manager method or with a
relationship or d
On 15 Mar 2006, John Siracusa wrote:
> Whoops, that looks like a bug. Try the version in SVN and tell me if it
> fixes your problem:
>
> svn co https://svn.sourceforge.net/svnroot/rose/trunk/modules/Rose-DB-Object
Now everything seems broken:
First I got some test failures in db-object-loader-5
Hello again,
I tried to add a 'many to many' relationship between two of my
classes. I followed the tutorial and everything works as long as I
define foreign_keys for the mapping table:
__PACKAGE__->meta->foreign_keys(
therapeut =>
{
class => 'PVonline::Therapeuten',
On 14 Mar 2006, Uwe Voelker wrote:
> $meta->relationships
> (
> changes => {type => 'one to many',
> class=> 'Changes',
> column_map => {ID => 'AUFTRAG'},
> manager_args => {sort_by => 'CHANGES.DATUM DESC'},
>
On 14 Mar 2006, John Siracusa wrote:
> On 3/14/06 4:17 AM, Michael Lackhoff wrote:
> > Is it possible to either define a default sort_by clause in the
> > relationship definition
>
> See Uwe's earlier response for how to do this.
Yes, that works great. Thanks!
&g
I just discovered the power of the relationship definitions. It is so
easy to create a tree of one to many relationships.
There is only one problem: they are unsorted.
Is it possible to either define a default sort_by clause in the
relationship definition or when the relationship method is actual
You might remember my mistake with the definition of a field that
differed between the "create table..." statement for the database and
the RDBO-class. I think it is always error prone to have two such
definitions around.
So woudn't it be great if it was possible to maintain only the class
defi
On 28 Feb 2006 at 7:42, John Siracusa wrote:
> What I was talking about was a convenience method to fill a Rose::HTML::Form
> based on a CGI object. Then you'd use Rose::HTML::Form's existing
> object_from_form() method to fill in the arbitrary object. Something like:
Hmm, then I think it won'
On 21 Feb 2006, John Siracusa wrote:
Thanks for your answer and sorry for my late reply, I couldn't work
on my project for a few days.
> Ah, okay. For the record, I use an object-based approach to this kind of
> thing by using Rose::HTML::Objects to manage my forms. Anyway, getting a
> hashref
On 21 Feb 2006, John Siracusa wrote:
> SQLite really has no concept of SQL data types. Well, it has a very
> simplified concept, anyway. All that detail you put in your CREATE
> TABLE statements is pretty much ignored by SQLite. Instead, it stores
> everything as strings and integers (and maybe
Hello again,
I found out about my problem from yesterday. It was all (well almost)
my fault. In the database definition I had forgotten the 'INTEGER'
qualifier but in the class definition everything was all right. Now
these two didn't fit together and I got these mysterious effects.
I say 'almo
Hello again,
I just started a project with Rose::DB::Object and I must say I am
very impressed. It is really great!
But still, I run into one or the other problem. I hope it is o.k. to
ask these beginner questions.
At the moment something drives me crazy that can be destilled to
these few line
Hi,
I wonder if there is a shortcut for multi level relationships.
I have three tables in a relationship like 'cities', 'states' and
'countries'. Now I would like to say something like
print $city->country->name;
I know how to to it in two steps (first get the state then the
country from there)
75 matches
Mail list logo
