py.test

[percious]

Does anyone have a problem with removing the py.test requirement in
favor of nose? Im not sure why py is even required for installation,
it should only be needed for development...

[Mark Ramm]

I would greatly apriciate this as a consumer of migrate. In TG2 py
is included as a dependency just because of migrate, and it causes
some installation problems for some users that I'd love be be able to
avoid.

--Mark Ramm

--
Mark Ramm-Christensen
email: mark at compoundthinking dot com
blog: www.compoundthinking.com/blog

[Nate Lowrie]

On Wed, Nov 26, 2008 at 7:39 PM, percious <ch...@percious.com> wrote:

Go ahead. I would be glad to see it go. I always wondered why I
needed it for installation...

Nate L.

[percious]

woo hoo!

[Jan Dittberner]

+1

I'd love to get rid of py.test too but didn't have the time to make
the change myself.

Regards
Jan

[percious]

The initial work on this has started. I am not sure if there were
broken tests before, but here is where we stand:

$ nosetests
...
Ran 130 tests in 51.826s

FAILED (errors=36, failures=6)

not too shabby, I had to modify the usedb decorator a little to get it
working, as well as assertRaises

cheers.
-chris


On Nov 27, 3:53 am, "Jan Dittberner" <jan.dittber...@googlemail.com>
wrote:

[percious]

so, we are down to four failing tests. I am running postgres, mysql,
and sqlite. I have a limited time to look at the failures today, so
if someone else wants to take a stab at it, feel free. Also, if
someone can run the tests on oracle, that would be good.

cheers.
-chris

Ran 77 tests in 32.432s

FAILED (failures=4)

[Graham Higgins]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 2 Dec 2008, at 14:55, percious wrote:

>
> so, we are down to four failing tests. I am running postgres, mysql,
> and sqlite. I have a limited time to look at the failures today, so
> if someone else wants to take a stab at it, feel free. Also, if
> someone can run the tests on oracle, that would be good.
>
> cheers.
> -chris
>
> Ran 77 tests in 32.432s
>
> FAILED (failures=4)

Slight difference here, Python 2.5.1, OS X 10.5, running from trunk
with SQLite 3.4.0, PostgreSQL 8.2.5 and MySQL 5.0.67.

> bash-3.2$ nosetests
> ......../Library/Python/2.5/site-packages/SQLAlchemy-0.5.0rc4-
> py2.5.egg/sqlalchemy/databases/mysql.py:1540: Warning: Field 'data'
> doesn't have a default value
> cursor.execute(statement, parameters)
> ...........................................................
> ----------------------------------------------------------------------
> Ran 67 tests in 9.379s
>
> OK

Useful?

Cheers,

Graham

-----BEGIN PGP SIGNATURE-----

iEYEARECAAYFAkk1T3oACgkQOsmLt1Nhivy3pQCggxOYJACVheaVwfmx1uUdDXex
+2EAoOJxN1yKg2Ksi4zFMx/j7uZ43tf4iQCVAgUBSTVPelnrWVZ7aXD1AQLoNQQA
iStwpDMJvMCEFOE6Noyl1q6obEqCPn6JaZFmScnPqk6exehOM9hJ4Yi4EBa3AkTK
oFMqwmvndUlXl0RSiP1mAPS3PE7SX282ZZVlkOa/67bamxZwBEecvPgqbpOSPY2V
DiB3rFmgU9aS1weS2bFefXwUGVLsL19JfkbonDSfLy4=
=KWe0
-----END PGP SIGNATURE-----

[percious]

very useful graham. Thanks.

I don't know if you are running out of trunk, but there should be 77
tests now. Also, which databases are you using to test?

I am down to 2 failures, both of which are MYSQL related.

cheers.
-chris

[percious]

whoops, i missed that you said which DBs you are running

On Dec 2, 8:08 am, Graham Higgins <gjhigg...@googlemail.com> wrote:

[percious]

Ok, so the way I was testing each database was wrong, only one at a
time was running. This is now fixed, so you should have 129 tests,
and 3 failures... (using pg, mysql, sqlite)

cheers.
-chris

[percious]

Ok,

So there are 3 remaining failures that I cannot get to this morning.
They are all mysql related, and it does not make sense for me to work
on them right now since I am not using migrate with mysql at my
current employer(s). I suggest we do a soft release of 0.5a1 soon,
and have someone with mysql expertise (who needs it working) look into
the mysql bugs. I removed deprecated functionality, as it should have
been deprecated from 0.2 and now should cause errors if people try and
use that functionality.

cheers.
-chris

[Graham Higgins]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 2 Dec 2008, at 16:02, percious wrote:

> So there are 3 remaining failures that I cannot get to this morning.
> They are all mysql related, and it does not make sense for me to work
> on them right now since I am not using migrate with mysql at my
> current employer(s).

The MySQL issue may be version-dependent:

> (bel)bash-3.2$ nosetests
> ..............................................................................................................................
> ----------------------------------------------------------------------
> Ran 126 tests in 28.741s
>
> OK

Ave percious!

Cheers,

Graham

-----BEGIN PGP SIGNATURE-----

iEYEARECAAYFAkk1Yy8ACgkQOsmLt1NhivwARQCg+wC8YFb8SIHrfHarkt6FyvSf
QB8AniQkKxSI9RMQ2FuYoKRahSh0vpebiQCVAgUBSTVjL1nrWVZ7aXD1AQIF6wQA
prkw4bmU/I3fcYdiXPzJOmn/tg+LkI9LEcvn/78UddSxTxDABXSYZ38RqQkGHdBz
6WjXGlqaHrNqCf8uHzFs8itavL/kaEKXZZIzQK1rpVynx3AsFZjww81KE/B4K4rP
pyFjxY8hNr3fn1JB36SeXrgRiavrIEEL0KUcBNHB+Bg=
=klAc
-----END PGP SIGNATURE-----

[percious]

$ mysql
Welcome to the MySQL monitor. Commands end with ; or \g.
Your MySQL connection id is 535
Server version: 5.0.67 Source distribution


weird... Perhaps I need to upgrade my version of SA....

cheers.
-chris

[percious]

hmp...
python 2.6 SA 0.5rc4 pg 8.3.3 mysql 5.0.67 ...
----------------------------------------------------------------------
Ran 126 tests in 11.134s

FAILED (failures=8)

(mysql looks to be failing still, as is a few postgres tests...)

[percious]

ok, down to two failures again, python 2.6 though... (that's not a
typo) and 2.6 is FAST.

----------------------------------------------------------------------
Ran 126 tests in 23.932s

FAILED (failures=2)

[Graham Higgins]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 2 Dec 2008, at 20:25, percious wrote:

>
> ok, down to two failures again, python 2.6 though... (that's not a
> typo) and 2.6 is FAST.
>
> ----------------------------------------------------------------------
> Ran 126 tests in 23.932s
>
> FAILED (failures=2)

For 2.6 I have:

> ----------------------------------------------------------------------
> Ran 126 tests in 15.952s
>
> FAILED (errors=2, failures=1)

Couple of "Database already controlled" errors and an assertion
failure in test_schemadiff.py

and one's appeared in a 2.5.1 execution ...

> AssertionError: create() did not raise ControlledSchemaError
>
> ----------------------------------------------------------------------
> Ran 126 tests in 29.174s
>
> FAILED (failures=1)

(I see what you mean about the speed-up)

Cheers,

Graham

-----BEGIN PGP SIGNATURE-----

iEYEARECAAYFAkk2lzgACgkQOsmLt1NhivzHqACg2uQMOjcO4X/KoIrwF6m8RRQV
4t8AoKzbZQ3gjLdOfIVbFXdG3wj2f73SiQCVAgUBSTaXOFnrWVZ7aXD1AQKMNwP/
cBdboMUwGBLaffVjh0cDE3eP2c4WxKKx7xZrwg8wfo4lNeM69zngLbNEBTXNeiTy
udUgTx7kF7bnzwr46xuxO+7NABlLrKc9jwPtKeGAvM79qUR78+jz0WaZZBey7VPk
yWim4DnBHYmY13KN6UPCGe31R0kQKs1LuNfiHUlghLY=
=hggj
-----END PGP SIGNATURE-----

[percious]

Any help nailing down these last few bugs is appreciated from the
community. I just do not have the desire to track down MySQL based
errors.

The errors saying the database is already under control might be
happening if you are running tests on different versions of python
sumultaneously.

cheers.
-chris