Re: [Zope-dev] [PATCH] Better error support in subunit output formatter

2010-04-10 Thread Tres Seaver
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

Tres Seaver wrote:
> Jonathan Lange wrote:
>> On Sat, Apr 10, 2010 at 7:16 PM, Tres Seaver  wrote:
>>> -BEGIN PGP SIGNED MESSAGE-
>>> Hash: SHA1
>>>
>>> Jonathan Lange wrote:
 Hello,

 The subunit output formatter that was added recently currently raises
 NotImplementedError when it's given a generic "error". The error()
 method on formatters is used to communicate user error (e.g. when the
 user selects incompatible options), internal framework errors (afaict,
 only when there's a subprocess read error) and layer setUp errors.

 I find the idea of masking layer setUp errors to be unconscionably
 awful, so I've attached the following patch, which changes the subunit
 formatter's error() method to print out whatever it gets. It doesn't
 have any tests, but it's simple enough, and none of the other
 formatters have tests for this method either.

 Ideally, I think there should be a new method on OutputFormatter to
 handle layer setup failure. Probably teardown failure too. If someone
 wants to push back on this simple patch in favour of a new method, I
 could probably be persuaded to do the work, provided I had clear
 backwards compatibility guidelines.
>>> Your patches look pretty reasonable to me:  can you kick them into
>>> Launchpad so they don't get lost in the mailing list archives?  There is
>>> a project page for zope.app.testing, but it doesn't (as of today) have
>>> the bug tracker turned on.
>>>
>>>  https://launchpad.net/zope.app.testing
>>>
>> Branches pushed, merge proposals created, bugs filed, branches linked.
> 
> Thank you!

I have merged you changes for zope.app.testing and made a relaese:  I
also closed LP #560222.

However, I haven't yet been able to pull your branch for the
zope.testing subunit fix.



Tres.
- --
===
Tres Seaver  +1 540-429-0999  tsea...@palladion.com
Palladion Software   "Excellence by Design"http://palladion.com
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.9 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iEYEARECAAYFAkvBF6QACgkQ+gerLs4ltQ4K9gCgiA6ufSDRZEAMBBN3mQY4zFXQ
uzsAnjM/7VwJvUUwg5M7dKxBbOJGMOv7
=7it9
-END PGP SIGNATURE-

___
Zope-Dev maillist  -  Zope-Dev@zope.org
https://mail.zope.org/mailman/listinfo/zope-dev
**  No cross posts or HTML encoding!  **
(Related lists - 
 https://mail.zope.org/mailman/listinfo/zope-announce
 https://mail.zope.org/mailman/listinfo/zope )


Re: [Zope-dev] [PATCH] Better error support in subunit output formatter

2010-04-10 Thread Tres Seaver
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

Jonathan Lange wrote:
> On Sat, Apr 10, 2010 at 7:16 PM, Tres Seaver  wrote:
>> -BEGIN PGP SIGNED MESSAGE-
>> Hash: SHA1
>>
>> Jonathan Lange wrote:
>>> Hello,
>>>
>>> The subunit output formatter that was added recently currently raises
>>> NotImplementedError when it's given a generic "error". The error()
>>> method on formatters is used to communicate user error (e.g. when the
>>> user selects incompatible options), internal framework errors (afaict,
>>> only when there's a subprocess read error) and layer setUp errors.
>>>
>>> I find the idea of masking layer setUp errors to be unconscionably
>>> awful, so I've attached the following patch, which changes the subunit
>>> formatter's error() method to print out whatever it gets. It doesn't
>>> have any tests, but it's simple enough, and none of the other
>>> formatters have tests for this method either.
>>>
>>> Ideally, I think there should be a new method on OutputFormatter to
>>> handle layer setup failure. Probably teardown failure too. If someone
>>> wants to push back on this simple patch in favour of a new method, I
>>> could probably be persuaded to do the work, provided I had clear
>>> backwards compatibility guidelines.
>> Your patches look pretty reasonable to me:  can you kick them into
>> Launchpad so they don't get lost in the mailing list archives?  There is
>> a project page for zope.app.testing, but it doesn't (as of today) have
>> the bug tracker turned on.
>>
>>  https://launchpad.net/zope.app.testing
>>
> 
> Branches pushed, merge proposals created, bugs filed, branches linked.

Thank you!

> FWIW, I sent to the mailing list first because last time I tried this
> I filed bugs, pushed up branches etc. and didn't get any response
> until I posted patches to the ML :)

We'll try to do better this time ;).


Tres.
- --
===
Tres Seaver  +1 540-429-0999  tsea...@palladion.com
Palladion Software   "Excellence by Design"http://palladion.com
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.9 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iEYEARECAAYFAkvBDkgACgkQ+gerLs4ltQ6UPQCfTbRslosIjKiAEPyecXpEvnLb
2G0AoNPLQ7Oegzo1yj0H0XGxIxt82lUB
=wE53
-END PGP SIGNATURE-

___
Zope-Dev maillist  -  Zope-Dev@zope.org
https://mail.zope.org/mailman/listinfo/zope-dev
**  No cross posts or HTML encoding!  **
(Related lists - 
 https://mail.zope.org/mailman/listinfo/zope-announce
 https://mail.zope.org/mailman/listinfo/zope )


Re: [Zope-dev] [PATCH] Better error support in subunit output formatter

2010-04-10 Thread Jonathan Lange
On Sat, Apr 10, 2010 at 7:16 PM, Tres Seaver  wrote:
> -BEGIN PGP SIGNED MESSAGE-
> Hash: SHA1
>
> Jonathan Lange wrote:
>> Hello,
>>
>> The subunit output formatter that was added recently currently raises
>> NotImplementedError when it's given a generic "error". The error()
>> method on formatters is used to communicate user error (e.g. when the
>> user selects incompatible options), internal framework errors (afaict,
>> only when there's a subprocess read error) and layer setUp errors.
>>
>> I find the idea of masking layer setUp errors to be unconscionably
>> awful, so I've attached the following patch, which changes the subunit
>> formatter's error() method to print out whatever it gets. It doesn't
>> have any tests, but it's simple enough, and none of the other
>> formatters have tests for this method either.
>>
>> Ideally, I think there should be a new method on OutputFormatter to
>> handle layer setup failure. Probably teardown failure too. If someone
>> wants to push back on this simple patch in favour of a new method, I
>> could probably be persuaded to do the work, provided I had clear
>> backwards compatibility guidelines.
>
> Your patches look pretty reasonable to me:  can you kick them into
> Launchpad so they don't get lost in the mailing list archives?  There is
> a project page for zope.app.testing, but it doesn't (as of today) have
> the bug tracker turned on.
>
>  https://launchpad.net/zope.app.testing
>

Branches pushed, merge proposals created, bugs filed, branches linked.

FWIW, I sent to the mailing list first because last time I tried this
I filed bugs, pushed up branches etc. and didn't get any response
until I posted patches to the ML :)

jml
___
Zope-Dev maillist  -  Zope-Dev@zope.org
https://mail.zope.org/mailman/listinfo/zope-dev
**  No cross posts or HTML encoding!  **
(Related lists - 
 https://mail.zope.org/mailman/listinfo/zope-announce
 https://mail.zope.org/mailman/listinfo/zope )


Re: [Zope-dev] [PATCH] Better error support in subunit output formatter

2010-04-10 Thread Tres Seaver
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

Jonathan Lange wrote:
> Hello,
> 
> The subunit output formatter that was added recently currently raises
> NotImplementedError when it's given a generic "error". The error()
> method on formatters is used to communicate user error (e.g. when the
> user selects incompatible options), internal framework errors (afaict,
> only when there's a subprocess read error) and layer setUp errors.
> 
> I find the idea of masking layer setUp errors to be unconscionably
> awful, so I've attached the following patch, which changes the subunit
> formatter's error() method to print out whatever it gets. It doesn't
> have any tests, but it's simple enough, and none of the other
> formatters have tests for this method either.
> 
> Ideally, I think there should be a new method on OutputFormatter to
> handle layer setup failure. Probably teardown failure too. If someone
> wants to push back on this simple patch in favour of a new method, I
> could probably be persuaded to do the work, provided I had clear
> backwards compatibility guidelines.

Your patches look pretty reasonable to me:  can you kick them into
Launchpad so they don't get lost in the mailing list archives?  There is
a project page for zope.app.testing, but it doesn't (as of today) have
the bug tracker turned on.

 https://launchpad.net/zope.app.testing

For now, the Zope3 tracker is the right place for zope.app.* issues:

 https://bugs.launchpad.net/zope3



Tres.
- --
===
Tres Seaver  +1 540-429-0999  tsea...@palladion.com
Palladion Software   "Excellence by Design"http://palladion.com
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.9 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iEYEARECAAYFAkvAwH0ACgkQ+gerLs4ltQ57tACeP9VzumWF++MBaD6P6SohME9w
e6MAn2MYPBF3LK5TfFKRjJkNGAQfjH+w
=nYkx
-END PGP SIGNATURE-
___
Zope-Dev maillist  -  Zope-Dev@zope.org
https://mail.zope.org/mailman/listinfo/zope-dev
**  No cross posts or HTML encoding!  **
(Related lists - 
 https://mail.zope.org/mailman/listinfo/zope-announce
 https://mail.zope.org/mailman/listinfo/zope )


[Zope-dev] [PATCH] Better error support in subunit output formatter

2010-04-10 Thread Jonathan Lange
Hello,

The subunit output formatter that was added recently currently raises
NotImplementedError when it's given a generic "error". The error()
method on formatters is used to communicate user error (e.g. when the
user selects incompatible options), internal framework errors (afaict,
only when there's a subprocess read error) and layer setUp errors.

I find the idea of masking layer setUp errors to be unconscionably
awful, so I've attached the following patch, which changes the subunit
formatter's error() method to print out whatever it gets. It doesn't
have any tests, but it's simple enough, and none of the other
formatters have tests for this method either.

Ideally, I think there should be a new method on OutputFormatter to
handle layer setup failure. Probably teardown failure too. If someone
wants to push back on this simple patch in favour of a new method, I
could probably be persuaded to do the work, provided I had clear
backwards compatibility guidelines.

jml
=== modified file 'src/zope/testing/testrunner/formatter.py'
--- src/zope/testing/testrunner/formatter.py	2010-04-06 03:02:54 +
+++ src/zope/testing/testrunner/formatter.py	2010-04-10 15:27:41 +
@@ -795,8 +795,7 @@
 """Report an error."""
 # XXX: Mostly used for user errors, sometimes used for errors in the
 # test framework, sometimes used to record layer setUp failure (!!!).
-# How should this be encoded?
-raise NotImplementedError(self.error)
+self._stream.write('%s\n' % (message,))
 
 def error_with_banner(self, message):
 """Report an error with a big ASCII banner."""

___
Zope-Dev maillist  -  Zope-Dev@zope.org
https://mail.zope.org/mailman/listinfo/zope-dev
**  No cross posts or HTML encoding!  **
(Related lists - 
 https://mail.zope.org/mailman/listinfo/zope-announce
 https://mail.zope.org/mailman/listinfo/zope )