Re: [Springframework-developer] Summary of pending points

[<jue...@we...>]

Jean-Pierre, Thomas, Rod, everybody,

I suggest to address the issue by offering overloaded execute and mapRow =
methods with a "Map context" parameter (instead of "Object context"). =
Additionally, I suggest to offer just one single MappingSqlQuery class =
with 3 mapRow hooks:

- mapRow(ResultSet rs, int rowNum, Map context, Object[] parameters)
- mapRow(ResultSet rs, int rowNum, Map context)
- mapRow(ResultSet rs, int rowNum)

The default implementations of the former two could simply cascade =
towards the latter. A subclass can choose to override any of those. This =
mechanism is inspired by our SimpleFormController that offers similar =
cascading methods for onSubmit.

This would effectively mean dissolving MappingSqlQueryWithParameters and =
merging it with MappingSqlQuery. As I assume that most people use =
MappingSqlQuery anyway, this shouldn't pose a problem. All former =
MappingSqlQueryWithParameters subclasses would simply have to extend the =
merged MappingSqlQuery then.

As I consider both a context map and a consolidation of all overloaded =
mapRow methods into MappingSqlQuery sensible and straightforward =
changes, I suggest to implement them promptly. I'm even willing to =
include this in 0.9.1, to reflect the merging of =
MappingSqlQueryWithParameters and MappingSqlQuery as early as possible.

What do you think?

Juergen


-----Original Message-----
From: jp....@ti... [mailto:jp....@ti...]
Sent: Wednesday, July 23, 2003 10:08 AM
To: j=FCrgen h=F6ller [werk3AT]
Cc: springframework-developer
Subject: Re: [Springframework-developer] Summary of pending points


Hi Juergen,

It's the second case. In fact, the locale is or is not used in the =
request itself, depending if the root object uses it directly or not.

The simpliest example is this one. Admitting a user or an address that =
is classicaly not localized in the database. If this record contains a =
Country indication, only the country code will still be stored in the =
database. When mapRow will create the Country object with the code from =
the database, it will need the locale, not of the record, but of the =
user to serve. To allow the view displaying the name of the country =
correctly.

As we said both, the locale cannot be a current execute parameter in =
this case.

The core of rewriting is done. Just the doc is not changed accordingly =
and for the new feature, I have not rewrited convenience methods.=20

I will, when at home, put my current version of the three involved =
classes in the sandbox. So you will be able to see what changes I've =
made and if this way has to be changed.

The changes are completely backward compatible. Currently only one test =
breaks, but its only a lack of a constructor signature.

Note: I have also attached these files in this mail.

Regards,
Jean-Pierre



---------- Initial Header -----------

From      : j=FCrgen h=F6ller [werk3AT] <jue...@we...>
To          : "JP Pawlak" =
<jp....@ti...>,<spr...@li...>
Cc          :=20
Date      : Wed, 23 Jul 2003 09:27:53 +0200
Subject : Re: [Springframework-developer] Summary of pending points

Hi Jean-Pierre,
=20
<quote>
>>> Ad 3) I haven't thought about that in depth, as I'm mainly using
Hibernate these days. But it seems to me that simply using a
MappingSqlQuery instance per request should solve the problem. It's not
really about creating many such instances I guess, as they are pretty
lightweight. Do you have a particular reason for wanting to reuse
MappingSqlQuery instances in this case?

You're right that using a new instance per request solves the problem.

Nevertheless, it's not about a few particular requests, but almost all.
In addition to the class creation, it's also the prepared statement
re-creation and re-compilation. It's possible anyway that the gain is
not so high. But as the job is done for me, I prefer continue to reuse
as so it can only be faster and has no additional risk.
</quote>
=20
OK, I see that there is value in reusing the prepared statements.
=20
What exactly are you doing with the Locale inside the mapRow =
implementation? If you'd use it as an argument for the prepared =
statement too, it could be a parameter of the execute method, but it =
probably won't go into the statement as Locale object but rather as some =
language code. So are you setting the Locale to the mapped object but =
not using it directly for the statement? Then it couldn't be an execute =
parameter, so you'd have to get it there in some other way.
=20
That would require adding a new context parameter of type Object or Map =
to execute, newResultReader, and mapRow. There should be a solution to =
do this as backward compatible as possible, maybe by introducing a super =
class of MappingSqlQueryWithParameters called =
MappingSqlQueryWithParametersAndContext? The two subclasses could still =
offer the same mapRow implementations as before.
=20
Regards,
Juergen


********** L'ADSL A 20 EUR/MOIS**********
Tiscali propose l'ADSL le moins cher du march=E9 : 20 EUR/mois et le =
modem ADSL offert !=20
Pour profiter de cette offre exceptionnelle, cliquez ici : =
http://register.tiscali.fr/adsl/
Offre soumise =E0 conditions.