Re: [Codenarc-user] LoggerWithWrongModifiersRule supporting subclasses

[R. S. <ren...@go...>]

It's not about the "protected" (this is just required).

The interesting part is the non-static one. Because of it, a superclass
uses the logger of the derived class based on "getLogger(this.class)".

With such a logger it's for example possible to write every log message
of a class (including its superclass) to a separate file because they
are using the same logger.

Regards,
René Scheibe

On 05/31/2011 09:55 AM, Hamlet DArcy wrote:
> Personally, I would require loggers to be static, private, and final. 
> If someone really needed a protected logger then i would still make it static and final, just protected. 
> 
> So I think we need both these settings. 
> 
> 
> 
> ----- Original Message -----
>> That's a great question.
>>
>> I do use that pattern in framework superclasses:
>> 	protected final LOG = Logger.getLogger(this.class)
>>
>> to provide a built-in logger for all subclasses. And so I have to
>> configure
>> exceptions for this rule for those superclasses.
>>
>> Is there any value in splitting up the "exception" configuration into
>> two
>> properties, if we do allow it?
>>   * allowNonStaticLogger
>>   * allowProtectedLogger
>>
>> Are there use cases we can envision where we (or anyone) would choose
>> to
>> allow non-static but not protected, or vice versa?
>>
>> Chris
>> -----Original Message-----
>> From: Hamlet D'Arcy [mailto:ham...@gm...]
>> Sent: Monday, May 30, 2011 2:50 AM
>> To: René Scheibe; Cod...@li...
>> Subject: [Codenarc-user] LoggerWithWrongModifiersRule supporting
>> subclasses
>>
>> Hi everyone,
>>
>> This email is about the idea about the LoggerWithWrongModifiersRule
>> supporting subclasses.
>>
>> René filed an issue (and supplied a patch) describing the issue:
>> https://sourceforge.net/tracker/index.php?func=detail&aid=3308930&group_id=2
>> 50145&atid=1126575#
>>
>> Basically, the question is how subclasses should be allowed to access
>> a
>> Logger in a parent class.
>>
>> Normally Loggers are private, static, and final. This patch allows
>> you to
>> accept this as valid code when you turn a certain property on:
>>
>>     protected final LOG = Logger.getLogger(this.class)
>>
>> I'd like to discuss this just a little more before committing the
>> patch.
>>
>> With this change, it is acceptable to have a non-static logger. So
>> each
>> instance of a subclass makes and creates a logger. This does not seem
>> right,
>> and I would reject it in a code review as improper logging. I would
>> accept
>> this form though:
>>
>>     protected static final LOG = Logger.getLogger(MyClass.class)
>>
>> Do we really want to allow non-static Loggers?
>>
>> if so, I recommend two new properties instead of just one:
>>   * allowNonStaticLogger
>>   * allowProtectedLogger
>>
>> Then the user can control the behavior a little better.
>>
>>
>>
>> --
>> Hamlet D'Arcy
>> ham...@gm...
>>
>> ----------------------------------------------------------------------------
>> --
>> vRanger cuts backup time in half-while increasing security.
>> With the market-leading solution for virtual backup and recovery, you
>> get
>> blazing-fast, flexible, and affordable data protection.
>> Download your free trial now.
>> http://p.sf.net/sfu/quest-d2dcopy1
>> _______________________________________________
>> Codenarc-user mailing list
>> Cod...@li...
>> https://lists.sourceforge.net/lists/listinfo/codenarc-user
>>
>>
>> ------------------------------------------------------------------------------
>> Simplify data backup and recovery for your virtual environment with
>> vRanger.
>> Installation's a snap, and flexible recovery options mean your data
>> is safe,
>> secure and there when you need it. Data protection magic?
>> Nope - It's vRanger. Get your free trial download today.
>> http://p.sf.net/sfu/quest-sfdev2dev
>> _______________________________________________
>> Codenarc-user mailing list
>> Cod...@li...
>> https://lists.sourceforge.net/lists/listinfo/codenarc-user
>>