From kou at cozmixng.org  Tue Nov  7 21:59:00 2006
From: kou at cozmixng.org (Kouhei Sutou)
Date: Wed, 8 Nov 2006 11:59:00 +0900
Subject: [Ruby-activeldap-discuss] ActiveLDAP -> ActiveLdap
Message-ID: <a799fada0611071859kb3a8addmbd909f8345bcd346@mail.gmail.com>

Hi,

Can I change namespace of ActiveLDAP to ActiveLdap and rename
lib/activeldap to lib/active_ldap?
This changes support auto-require by ActiveSupport.

ActiveLDAP on repository has many changes and no full backward
compatibility of ActiveLDAP 0.7.4. This changes will help decisiving break
with old ActiveLDAP and will not confuse users of old ActiveLDAP by
difference of new ActiveLDAP API and old ActiveLDAP API (those APIs are
similar but not same).


Thanks,
--
kou

From kou at cozmixng.org  Wed Nov  8 20:49:52 2006
From: kou at cozmixng.org (Kouhei Sutou)
Date: Thu, 9 Nov 2006 10:49:52 +0900
Subject: [Ruby-activeldap-discuss] removing old API
Message-ID: <a799fada0611081749u7d8d3411ke1613f386f49761c@mail.gmail.com>

Hi,

Now, ActiveLDAP is renamed to ActiveLdap and breaked away from
old API. So, I want to remove old APIs like the following:

* ActiveLdap::Base.connect
* ActiveLdap::Base.find_all
* ActiveLdap::Base.dnattr (is dn_attribute now)
* ActiveLdap::Base#dnattr (is dn_attribute now)
* ActiveLdap::Base#write

Can I remove them?

And I want to invert not_array argument semantic of #get_attribute.
It seems that *not* is very confusable. Can I change this like the
following?

to:

  def get_attribute(name, force_array=false)

from:

  def get_attribute(name, not_array=true)


Thanks,
--
kou

From will at alum.bu.edu  Thu Nov  9 04:54:54 2006
From: will at alum.bu.edu (=?WINDOWS-1252?Q?Will_Drewry=99?=)
Date: Thu, 9 Nov 2006 09:54:54 +0000
Subject: [Ruby-activeldap-discuss] removing old API
In-Reply-To: <a799fada0611081749u7d8d3411ke1613f386f49761c@mail.gmail.com>
References: <a799fada0611081749u7d8d3411ke1613f386f49761c@mail.gmail.com>
Message-ID: <abc4304b0611090154nb53c7d3w328c6fc0871b939b@mail.gmail.com>

On 11/9/06, Kouhei Sutou <kou at cozmixng.org> wrote:
> Hi,
>
> Now, ActiveLDAP is renamed to ActiveLdap and breaked away from
> old API. So, I want to remove old APIs like the following:
>
> * ActiveLdap::Base.connect
> * ActiveLdap::Base.find_all
> * ActiveLdap::Base.dnattr (is dn_attribute now)
> * ActiveLdap::Base#dnattr (is dn_attribute now)
> * ActiveLdap::Base#write
>
> Can I remove them?

That sounds perfect.  The whole goal of the 0.8.0 release is to break
away - so let's do it :-)


>
> And I want to invert not_array argument semantic of #get_attribute.
> It seems that *not* is very confusable. Can I change this like the
> following?
>
> to:
>
>   def get_attribute(name, force_array=false)
>
> from:
>
>   def get_attribute(name, not_array=true)
>

Absolutely!  The original approach was due to a legacy bug. I
introduced a force array style semantic but had inverted its function
and led users to have code dependent on it.  It's definitely fine to
change this over.

thanks!
will

From kou at cozmixng.org  Thu Nov  9 08:21:30 2006
From: kou at cozmixng.org (Kouhei Sutou)
Date: Thu, 09 Nov 2006 22:21:30 +0900 (JST)
Subject: [Ruby-activeldap-discuss] removing old API
In-Reply-To: <abc4304b0611090154nb53c7d3w328c6fc0871b939b@mail.gmail.com>
References: <a799fada0611081749u7d8d3411ke1613f386f49761c@mail.gmail.com>
	<abc4304b0611090154nb53c7d3w328c6fc0871b939b@mail.gmail.com>
Message-ID: <20061109.222130.07585575.kou@cozmixng.org>

Hi,

Thanks for accepting my request. I'll do tomorrow.

Regards,
--
kou

In <abc4304b0611090154nb53c7d3w328c6fc0871b939b at mail.gmail.com>
  "Re: [Ruby-activeldap-discuss] removing old API" on Thu, 9 Nov 2006 09:54:54 +0000,
  "Will_Drewry" <will at alum.bu.edu> wrote:

> On 11/9/06, Kouhei Sutou <kou at cozmixng.org> wrote:
> > Hi,
> >
> > Now, ActiveLDAP is renamed to ActiveLdap and breaked away from
> > old API. So, I want to remove old APIs like the following:
> >
> > * ActiveLdap::Base.connect
> > * ActiveLdap::Base.find_all
> > * ActiveLdap::Base.dnattr (is dn_attribute now)
> > * ActiveLdap::Base#dnattr (is dn_attribute now)
> > * ActiveLdap::Base#write
> >
> > Can I remove them?
> 
> That sounds perfect.  The whole goal of the 0.8.0 release is to break
> away - so let's do it :-)
> 
> 
> >
> > And I want to invert not_array argument semantic of #get_attribute.
> > It seems that *not* is very confusable. Can I change this like the
> > following?
> >
> > to:
> >
> >   def get_attribute(name, force_array=false)
> >
> > from:
> >
> >   def get_attribute(name, not_array=true)
> >
> 
> Absolutely!  The original approach was due to a legacy bug. I
> introduced a force array style semantic but had inverted its function
> and led users to have code dependent on it.  It's definitely fine to
> change this over.
> 
> thanks!
> will

From kou at cozmixng.org  Thu Nov  9 08:24:09 2006
From: kou at cozmixng.org (Kouhei Sutou)
Date: Thu, 09 Nov 2006 22:24:09 +0900 (JST)
Subject: [Ruby-activeldap-discuss] ruby-activeldap-commit ML doesn't work
Message-ID: <20061109.222409.45026649.kou@cozmixng.org>

Hi,

I have committed some changes but I haven't receive any
commit mail. Will, could you check configuration of
ruby-activeldap-commit ML and/or code.google.com?

Thanks,
--
kou

From kou at cozmixng.org  Thu Nov  9 08:49:09 2006
From: kou at cozmixng.org (Kouhei Sutou)
Date: Thu, 09 Nov 2006 22:49:09 +0900 (JST)
Subject: [Ruby-activeldap-discuss] Release schedule
Message-ID: <20061109.224909.53334837.kou@cozmixng.org>

Hi,

I'm writing test codes for scripts under examples/
directory. There are some ToDo tasks(*) but ActiveLdap can
be released after I completed test codes for example
scripts. And I will finish the work tomorrow.

It's true that there are some important tasks in
ToDo. (e.g. providing full documentation) So, I suggest that
we version next release as 0.9.0. And we release 1.0.0 after
all ToDo tasks are completed.

Any opinion?

(*)
  * provide full documentation for new API.
  * handle all exception raised from Ruby/LDAP and wrap as
    ActiveLdap exception. I think we need to develop an
    application using ActiveLdap.
  * support Reloadable::Subclasses. I think we need to
    develop a Rails + ActiveLdap application to improve
    Rails support.
  * support Ruby/GetText.
  * support Net::LDAP as LDAP backend after Net::LDAP
    supports START_TLS. (I made a patch and submitted to the
    bug tracker of Net::LDAP)


Thanks,
--
kou

P.S. I want to release my library in next week but this is
only my work. Will, you don't care about this. It seems that
you are still busy. :-)

From will at alum.bu.edu  Thu Nov  9 09:14:50 2006
From: will at alum.bu.edu (=?WINDOWS-1252?Q?Will_Drewry=99?=)
Date: Thu, 9 Nov 2006 14:14:50 +0000
Subject: [Ruby-activeldap-discuss] ruby-activeldap-commit ML doesn't work
In-Reply-To: <20061109.222409.45026649.kou@cozmixng.org>
References: <20061109.222409.45026649.kou@cozmixng.org>
Message-ID: <abc4304b0611090614n2f1344c8ke9eabe682272ed6a@mail.gmail.com>

I'm attempting to whitelist the google.com address the automated mails
come from.  Hopefully that'll do it!

On 11/9/06, Kouhei Sutou <kou at cozmixng.org> wrote:
> Hi,
>
> I have committed some changes but I haven't receive any
> commit mail. Will, could you check configuration of
> ruby-activeldap-commit ML and/or code.google.com?
>
> Thanks,
> --
> kou
> _______________________________________________
> Ruby-activeldap-discuss mailing list
> Ruby-activeldap-discuss at rubyforge.org
> http://rubyforge.org/mailman/listinfo/ruby-activeldap-discuss
>

From kou at cozmixng.org  Thu Nov  9 09:26:42 2006
From: kou at cozmixng.org (Kouhei Sutou)
Date: Thu, 09 Nov 2006 23:26:42 +0900 (JST)
Subject: [Ruby-activeldap-discuss] ruby-activeldap-commit ML doesn't work
In-Reply-To: <abc4304b0611090614n2f1344c8ke9eabe682272ed6a@mail.gmail.com>
References: <20061109.222409.45026649.kou@cozmixng.org>
	<abc4304b0611090614n2f1344c8ke9eabe682272ed6a@mail.gmail.com>
Message-ID: <20061109.232642.119668508.kou@cozmixng.org>

Hi,

In <abc4304b0611090614n2f1344c8ke9eabe682272ed6a at mail.gmail.com>
  "Re: [Ruby-activeldap-discuss] ruby-activeldap-commit ML doesn't work" on Thu, 9 Nov 2006 14:14:50 +0000,
  "Will_Drewry" <will at alum.bu.edu> wrote:

> I'm attempting to whitelist the google.com address the automated mails
> come from.  Hopefully that'll do it!

It seems that it works. Thanks!

I have a request if you have a time. Could you configure
commit mail to include diff?


Thanks,
--
kou

From kou at cozmixng.org  Fri Nov 10 02:58:33 2006
From: kou at cozmixng.org (Kouhei Sutou)
Date: Fri, 10 Nov 2006 16:58:33 +0900
Subject: [Ruby-activeldap-discuss] Release schedule
In-Reply-To: <20061109.224909.53334837.kou@cozmixng.org>
References: <20061109.224909.53334837.kou@cozmixng.org>
Message-ID: <a799fada0611092358g3368d885o1205730148e4110d@mail.gmail.com>

Hi,

2006/11/9, Kouhei Sutou <kou at cozmixng.org>:

> I'm writing test codes for scripts under examples/
> directory. There are some ToDo tasks(*) but ActiveLdap can
> be released after I completed test codes for example
> scripts. And I will finish the work tomorrow.

I have finished. Could you confirm my work in last 2 weeks?


Thanks,
--
kou

From will at alum.bu.edu  Fri Nov 10 03:47:50 2006
From: will at alum.bu.edu (=?WINDOWS-1252?Q?Will_Drewry=99?=)
Date: Fri, 10 Nov 2006 08:47:50 +0000
Subject: [Ruby-activeldap-discuss] Release schedule
In-Reply-To: <a799fada0611092358g3368d885o1205730148e4110d@mail.gmail.com>
References: <20061109.224909.53334837.kou@cozmixng.org>
	<a799fada0611092358g3368d885o1205730148e4110d@mail.gmail.com>
Message-ID: <abc4304b0611100047l5acfdc9bod11972af15583b46@mail.gmail.com>

On 11/10/06, Kouhei Sutou <kou at cozmixng.org> wrote:
> Hi,
>
> 2006/11/9, Kouhei Sutou <kou at cozmixng.org>:
>
> > I'm writing test codes for scripts under examples/
> > directory. There are some ToDo tasks(*) but ActiveLdap can
> > be released after I completed test codes for example
> > scripts. And I will finish the work tomorrow.
>
> I have finished. Could you confirm my work in last 2 weeks?

Excellent!  I will start going through all of your changes after work
tonight (I'm in UTC).


Thanks!!

From will at alum.bu.edu  Sun Nov 19 18:17:35 2006
From: will at alum.bu.edu (=?WINDOWS-1252?Q?Will_Drewry=99?=)
Date: Sun, 19 Nov 2006 23:17:35 +0000
Subject: [Ruby-activeldap-discuss] Release schedule
In-Reply-To: <a799fada0611092358g3368d885o1205730148e4110d@mail.gmail.com>
References: <20061109.224909.53334837.kou@cozmixng.org>
	<a799fada0611092358g3368d885o1205730148e4110d@mail.gmail.com>
Message-ID: <abc4304b0611191517n6b9d1107t1bb0e8a26022c92@mail.gmail.com>

Hi kou!

I've just gone through the massive amount of work you've done, and it
looks fantastic. I'm happy to release what is in svn now as 0.8.0
(since there hasn't been an 0.8 release yet). Then, we can proceed to
knock out the remaining items in the TODO file.


After that, we can either jump straight to 1.0.0 or spend a bit of
time with 0.9.x as release candidates for 1.0.0 or just make 1.0.0rcX
until we feel it is robust and stable.  That would also give me some
time to start pushing the new API into use with a few
commercial/enterprise-scale users and see what bugs/problems they hit.

What do you think?

(Note, I've also changed your status on rubyforge which will make it
easier to do quick bugfix releaes, etc.)

If you're ready, I'll tag and release the current subversion contents
as soon as I hear back.

Thanks!!
will

From kou at cozmixng.org  Sun Nov 19 20:37:41 2006
From: kou at cozmixng.org (Kouhei Sutou)
Date: Mon, 20 Nov 2006 10:37:41 +0900
Subject: [Ruby-activeldap-discuss] Release schedule
In-Reply-To: <abc4304b0611191517n6b9d1107t1bb0e8a26022c92@mail.gmail.com>
References: <20061109.224909.53334837.kou@cozmixng.org>
	<a799fada0611092358g3368d885o1205730148e4110d@mail.gmail.com>
	<abc4304b0611191517n6b9d1107t1bb0e8a26022c92@mail.gmail.com>
Message-ID: <a799fada0611191737r33bd87b0hb8e6116d70d941a9@mail.gmail.com>

Hi,

2006/11/20, Will Drewry? <will at alum.bu.edu>:
> I've just gone through the massive amount of work you've done, and it
> looks fantastic. I'm happy to release what is in svn now as 0.8.0
> (since there hasn't been an 0.8 release yet). Then, we can proceed to
> knock out the remaining items in the TODO file.
>
>
> After that, we can either jump straight to 1.0.0 or spend a bit of
> time with 0.9.x as release candidates for 1.0.0 or just make 1.0.0rcX
> until we feel it is robust and stable.  That would also give me some
> time to start pushing the new API into use with a few
> commercial/enterprise-scale users and see what bugs/problems they hit.
>
> What do you think?

I think 1.0.0rcX is too early now because RC release means 'API stable
but not yet robust'. I can't declare that 'API stable' yet.
So, it's better that we release 0.9.x series before 1.0.0 to improve API.

Here are some AL's ToDo I think:

APIs to improve:
* add filter value sanitizing API like DBI's ["XXX ? YYY", "VALUE"]
  syntax.
* add required_clases to default filter of Base.search.
* ...

For robust:
* handle all Ruby/LDAP exception in adaptor/ldap.rb: I just handle
  a bit of exceptions which occurred my environment.
* sanitize filter value.
* ...

To resolve the above tasks, it's good that we release 0.9.x and get
feedback from users.

> (Note, I've also changed your status on rubyforge which will make it
> easier to do quick bugfix releaes, etc.)
>
> If you're ready, I'll tag and release the current subversion contents
> as soon as I hear back.

Thanks for your work!


Regards,
--
kou

From will at alum.bu.edu  Mon Nov 20 06:08:58 2006
From: will at alum.bu.edu (=?WINDOWS-1252?Q?Will_Drewry=99?=)
Date: Mon, 20 Nov 2006 11:08:58 +0000
Subject: [Ruby-activeldap-discuss] Release schedule
In-Reply-To: <a799fada0611191737r33bd87b0hb8e6116d70d941a9@mail.gmail.com>
References: <20061109.224909.53334837.kou@cozmixng.org>
	<a799fada0611092358g3368d885o1205730148e4110d@mail.gmail.com>
	<abc4304b0611191517n6b9d1107t1bb0e8a26022c92@mail.gmail.com>
	<a799fada0611191737r33bd87b0hb8e6116d70d941a9@mail.gmail.com>
Message-ID: <abc4304b0611200308n165585d9n6f7ddc9df2257559@mail.gmail.com>

On 11/20/06, Kouhei Sutou <kou at cozmixng.org> wrote:
> I think 1.0.0rcX is too early now because RC release means 'API stable
> but not yet robust'. I can't declare that 'API stable' yet.
> So, it's better that we release 0.9.x series before 1.0.0 to improve API.

Ah - I think we're talking at cross purposes - I wasn't thinking about
going straight to 1.0.0rcX from what's in svn now.  I was trying to
decide if we should version past 0.8 to 0.9 (given that the last
release is 0.7.x).

>
> Here are some AL's ToDo I think:
>
> APIs to improve:
> * add filter value sanitizing API like DBI's ["XXX ? YYY", "VALUE"]
>   syntax.
> * add required_clases to default filter of Base.search.
> * ...
>
> For robust:
> * handle all Ruby/LDAP exception in adaptor/ldap.rb: I just handle
>   a bit of exceptions which occurred my environment.
> * sanitize filter value.
> * ...
>
> To resolve the above tasks, it's good that we release 0.9.x and get
> feedback from users.

This sounds exactly right to me.  With this in mind, making a 0.8.x
release now followed with some API additions moving us to 0.9.x makes
sense.  Then we can move the 0.9.x feature complete and bug fixed
version to a 1.0.0rcX and move from there.


Are you happy with what's in SVN? The only thing I noticed that I
haven't quite worked out yet is -
is it still possible to call associations with additional options to
avoid loading entire objects when you only want the dn (or a few
select attributes)?  If not, I'll start to add a find_options hash
back into the associations and add @loaded caching.  This was a pretty
heavily demanded feature in earlier release.

Thanks!
will

From kou at cozmixng.org  Mon Nov 20 06:57:15 2006
From: kou at cozmixng.org (Kouhei Sutou)
Date: Mon, 20 Nov 2006 20:57:15 +0900
Subject: [Ruby-activeldap-discuss] Release schedule
In-Reply-To: <abc4304b0611200308n165585d9n6f7ddc9df2257559@mail.gmail.com>
References: <20061109.224909.53334837.kou@cozmixng.org>
	<a799fada0611092358g3368d885o1205730148e4110d@mail.gmail.com>
	<abc4304b0611191517n6b9d1107t1bb0e8a26022c92@mail.gmail.com>
	<a799fada0611191737r33bd87b0hb8e6116d70d941a9@mail.gmail.com>
	<abc4304b0611200308n165585d9n6f7ddc9df2257559@mail.gmail.com>
Message-ID: <a799fada0611200357g58282b19w911369e4d2668d5c@mail.gmail.com>

Hi,

2006/11/20, Will Drewry? <will at alum.bu.edu>:

> select attributes)?  If not, I'll start to add a find_options hash
> back into the associations and add @loaded caching.  This was a pretty
> heavily demanded feature in earlier release.

If you want to get only dn and attributes, it's better that using
Base.search instead of adding new option to Base.find.


Thanks,
--
kou

From kou at cozmixng.org  Mon Nov 20 06:42:46 2006
From: kou at cozmixng.org (Kouhei Sutou)
Date: Mon, 20 Nov 2006 20:42:46 +0900 (JST)
Subject: [Ruby-activeldap-discuss] Release schedule
In-Reply-To: <abc4304b0611200308n165585d9n6f7ddc9df2257559@mail.gmail.com>
References: <abc4304b0611191517n6b9d1107t1bb0e8a26022c92@mail.gmail.com>
	<a799fada0611191737r33bd87b0hb8e6116d70d941a9@mail.gmail.com>
	<abc4304b0611200308n165585d9n6f7ddc9df2257559@mail.gmail.com>
Message-ID: <20061120.204246.75727136.kou@cozmixng.org>

Hi,

In <abc4304b0611200308n165585d9n6f7ddc9df2257559 at mail.gmail.com>
  "Re: [Ruby-activeldap-discuss] Release schedule" on Mon, 20 Nov 2006 11:08:58 +0000,
  "Will_Drewry" <will at alum.bu.edu> wrote:

> > I think 1.0.0rcX is too early now because RC release means 'API stable
> > but not yet robust'. I can't declare that 'API stable' yet.
> > So, it's better that we release 0.9.x series before 1.0.0 to improve API.
> 
> Ah - I think we're talking at cross purposes - I wasn't thinking about
> going straight to 1.0.0rcX from what's in svn now.  I was trying to
> decide if we should version past 0.8 to 0.9 (given that the last
> release is 0.7.x).

Oh, I'm sorry. I think 0.8.X -> 1.0.0rcX is good unless big
API changes are happened (not include adding API). If the
event is occurred we should release 0.9.X between 0.8.X and
1.0.0rcX.

> This sounds exactly right to me.  With this in mind, making a 0.8.x
> release now followed with some API additions moving us to 0.9.x makes
> sense.  Then we can move the 0.9.x feature complete and bug fixed
> version to a 1.0.0rcX and move from there.

OK. I agree with you.

> Are you happy with what's in SVN?

* LDAP filter sanitizing.
* acts_as_tree related association.

But I don't have enough time to do the above things now...

>                                   The only thing I noticed that I
> haven't quite worked out yet is -
> is it still possible to call associations with additional options to
> avoid loading entire objects when you only want the dn (or a few
> select attributes)?  If not, I'll start to add a find_options hash
> back into the associations and add @loaded caching.  This was a pretty
> heavily demanded feature in earlier release.

No. Associations (has_many and belongs_to) always retrieve
objects. But retrieved objects are cached per object.


Thanks,
--
kou

From will at alum.bu.edu  Mon Nov 20 16:57:08 2006
From: will at alum.bu.edu (=?WINDOWS-1252?Q?Will_Drewry=99?=)
Date: Mon, 20 Nov 2006 21:57:08 +0000
Subject: [Ruby-activeldap-discuss] Release schedule
In-Reply-To: <a799fada0611200357g58282b19w911369e4d2668d5c@mail.gmail.com>
References: <20061109.224909.53334837.kou@cozmixng.org>
	<a799fada0611092358g3368d885o1205730148e4110d@mail.gmail.com>
	<abc4304b0611191517n6b9d1107t1bb0e8a26022c92@mail.gmail.com>
	<a799fada0611191737r33bd87b0hb8e6116d70d941a9@mail.gmail.com>
	<abc4304b0611200308n165585d9n6f7ddc9df2257559@mail.gmail.com>
	<a799fada0611200357g58282b19w911369e4d2668d5c@mail.gmail.com>
Message-ID: <abc4304b0611201357l4b7203a0w89794cd67311fd73@mail.gmail.com>

On 11/20/06, Kouhei Sutou <kou at cozmixng.org> wrote:
> Hi,
>
> 2006/11/20, Will Drewry? <will at alum.bu.edu>:
>
> > select attributes)?  If not, I'll start to add a find_options hash
> > back into the associations and add @loaded caching.  This was a pretty
> > heavily demanded feature in earlier release.
>
> If you want to get only dn and attributes, it's better that using
> Base.search instead of adding new option to Base.find.

Ah of course! I've added to the example/objects/user.rb an example of
how to implement the old-style API with the new API.

I've also gone ahead and tagged this release as 0.8.0.  The release
numbering system I've been using is as follows:

major.minor.subminor:
- major: for full stable release - indicate a stable API
- minor: increment every time the API changes leading to the next stable API
- subminor: bug fixes and additive API changes (nothing that affects
compatibility)

If you think we should adopt a different style, I'm all ears!   (e.g.
the odd/even dev/stable approach, etc). acts_as_tree and LDAP filter
sanitization can be added (mostly) with only subminor changes. (If
sanitization is more invasiive, we may want to move to 0.9 once it is
in.)

Thanks!
will

From daniel at pfile.net  Thu Nov 30 10:35:07 2006
From: daniel at pfile.net (Daniel Pfile)
Date: Thu, 30 Nov 2006 09:35:07 -0600
Subject: [Ruby-activeldap-discuss] PATCH: Fedora custom schemas
Message-ID: <456EFA2B.10501@pfile.net>

Fedora's directory server (aka netscape) uses a string followed by a
-oid (someobject-oid) to name it's custom objects instead of a proper
number (1.3.6....) . I'm working on getting some of the custom
extensions we've done working with your library. This is my first time
using ruby and ActiveLdap for anything other than a basic tutorial, so
be gentle. Here's my patch, not against SVN, but just a copy of the file
out of the 0.8 gem. I think I tore the regex apart correctly, but feel
free to correct it.

--- schema.rb   2006-11-30 10:03:49.000000000 -0500
+++ schema-new.rb       2006-11-30 10:06:10.000000000 -0500
@@ -42,7 +42,7 @@
       return ids[id] if ids.has_key?(id)
 
       while schema = @entries[group].shift
-        next unless /\A\s*\(\s*([\d\.]+)\s*(.*)\s*\)\s*\z/ =~ schema
+        next unless /\A\s*\(\s*([\d\.]+|[\w]+-oid)\s*(.*)\s*\)\s*\z/ =~
schema
         schema_id = $1
         rest = $2
         next if ids.has_key?(schema_id)

Thanks for the software, keep up the great work.

-- Daniel

From kou at cozmixng.org  Thu Nov 30 20:00:18 2006
From: kou at cozmixng.org (Kouhei Sutou)
Date: Fri, 1 Dec 2006 10:00:18 +0900
Subject: [Ruby-activeldap-discuss] PATCH: Fedora custom schemas
In-Reply-To: <456EFA2B.10501@pfile.net>
References: <456EFA2B.10501@pfile.net>
Message-ID: <a799fada0611301700i435e4169tb44eaa42fe87caef@mail.gmail.com>

Hi,

2006/12/1, Daniel Pfile <daniel at pfile.net>:

> Fedora's directory server (aka netscape) uses a string followed by a
> -oid (someobject-oid) to name it's custom objects instead of a proper
> number (1.3.6....) . I'm working on getting some of the custom
> extensions we've done working with your library. This is my first time
> using ruby and ActiveLdap for anything other than a basic tutorial, so
> be gentle. Here's my patch, not against SVN, but just a copy of the file
> out of the 0.8 gem. I think I tore the regex apart correctly, but feel
> free to correct it.

Thanks. I'll be merge your work.

By the way, could you send us the schema? I want to add the schema
to our test suites. You will be able to get the schema with the following
patch:

Index: lib/active_ldap/adaptor/ldap.rb
===================================================================
--- lib/active_ldap/adaptor/ldap.rb     (revision 97)
+++ lib/active_ldap/adaptor/ldap.rb     (working copy)
@@ -87,6 +87,7 @@
           result = @connection.search2(base, LDAP::LDAP_SCOPE_BASE,
                                        '(objectClass=subschema)', attrs, false,

                                        sec, usec).first
+          p result
           Schema.new(result)
         end
 #       rescue



Regards,
--
kou