gump-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Adam R. B. Jack" <aj...@trysybase.com>
Subject Re: cvs commit: gump/python/gump/test sync.py pyunit.py
Date Thu, 04 Mar 2004 23:40:40 GMT
> this is my first attempt at doing something in Python for gumpy.

Awesome, I can't wait for us to be free of the bonds of 'rsync'. (I don't
know how we'll get a feel for this, when thrown into the deep end w/ the
CVS/SVN data we have, but I guess we wait a little then just plunge.)

> Since this code is not called from any where yet, it should not break
> gumpy, the worst thing which could
> break are the tests.

Great, we'll see how those go for a while then migrate.

First, it seems you really figure this stuff out -- including my home grown
pyunit stuff, cool.

Second, in checking these out I noticed a couple of minor things:

- In Sync.execute if you change the except below to makedirs to a finally,
we ought get an exception when (if ) makedir fails.

- It see that epurate is called unconditionally, could there be an optional
switch on Sync to do that, or not? If we are to replace cp as well as rsynch
we need 'copy on top of w/o clean-up'.

A ponderance:

- If we made the functions at the bottom of the module into methods on the
class, and if we made Sync inherit from Annotatable (& call
Annotatable.__init__(self) in __init__, we could log information to the
object, and then (if we wanted to know what was done, perhaps for debugging)
we could copy those annotations off (perhaps transfer them to the Module
object for listing in the forrest page.)

> By the way, I have run the tests under Windows and Boa Constructor with
> Python 2.2.

They run for me using Activestate Python 2.3. :-) [I know I ought work on
2.2, but I was hoping we could move. I'll test this on Linux on 2.2 sometime
soon.]

> I have seen that the logging sometimes work and sometimes not.

Gosh, interesting. I've never see that. I also never (overtly) noticed the
logging config making much difference, I assumed that writting to stderr was
the default. Could it be that in some cases your output is buffered and that
output not sync'ed? [No pun intended. ;-)] I'm clutching at straws, I've no
idea...

> Under cygwin I can run my test OK under Python 2.3.3 cygwin version.

This is just to be thorough, I take it, to ensure you cope in case folks
do -- but they oughtn't (right) for other reason (launching Java/Ant) that
you've stated before. Right?

regards

Adam


---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@gump.apache.org
For additional commands, e-mail: general-help@gump.apache.org


Mime
View raw message