On Nov 11, 11:38 am, oggie rob <oz.robhar...@gmail.com> wrote:
> Hi all,
> I'm working on a project which uses pyamf and produces the following
> exception during login:
>
> AttributeError: 'LazyModule' object has no attribute 'Manager'
I'm a bit late in responding here, but
Hi all,
I'm working on a project which uses pyamf and produces the following
exception during login:
AttributeError: 'LazyModule' object has no attribute 'Manager'
However debugging this has been difficult. The error shows up with the
following setup:
- django 1.0
- pyamf 0.3.1
- apache
The short story is all of these should be doable in the admin, and
most will also be quite simple.
The long story is below. Note, however, that some of my suggestions
are based on the old admin rather than newforms-admin, but from what I
have seen of newforms-admin, it should be fairly similar
If the purpose is to check for permissions/different display with
logged in vs. anonymous clients/etc, why not eliminate the chance of a
failure in the logout method and just create a new Client? Gives more
flexibility (by calling the client you want) and keeps things simple.
I use a method to
There are no restrictions at all on this. Remember that the view is
called via urls.py, so you define your different views & files there.
Just a point of interest, though - models don't always match up with
views. In fact, most of my views contain at least two or three models.
Rather than break
; wrote:
> Oggie,
> Thanks for the reply. So you're saying that it would be better to
> send my shopping cart data to every template through my all my
> view's? If that's the case then I assume there is no way to send
> request.session information using a custom template?
>
>
If you really need to use a tag, look at
http://www.djangoproject.com/documentation/templates_python/#passing-template-variables-to-the-tag
But I think you probably would find it easier to add an extra context
processor & avoid the tag altogether.
-rob
On Aug 3, 6:01 am, Greg <[EMAIL
On Aug 1, 11:22 pm, oggie rob <[EMAIL PROTECTED]> wrote:
> This is what middleware is made
> for.http://www.djangoproject.com/documentation/middleware/
>
> -rob
Sorry, I have recently been working with shopping carts and had
redirection on my brain. Fabien's suggestion is
This is what middleware is made for.
http://www.djangoproject.com/documentation/middleware/
-rob
On Aug 1, 9:51 pm, Greg <[EMAIL PROTECTED]> wrote:
> I've created a shopping cart. I want to put something on the top of
> every page that let's people know how many items they have in their
>
Yes. "filter" returns a set (hence the [] around "",
indicating a collection). If you are using a pk, where you can
guarantee a single returned value, use "get":
>>> p = Profile.objects.get(id=1)
>>> p.department
-rob
On Aug 1, 10:35 pm, james_027 <[EMAIL PROTECTED]> wrote:
> hi,
>
> i am
> However, if I add a new transaction, I want to see the Users listed in
> the User ChoiceField by username, rather than userid which it
> currently is. I have over 1000 users and so finding the correct user
> is a nightmare.
You need to use raw_id_admin instead of that edit_inline stuff. You
The admin interface contains a history feature, you might want to look
at how that works. Apart from a table to record the actual changes,
there is a "relationship" table that links the model that is changed
with the log record (i.e. the log isn't tied to a particular table
because of this).
When you use a ManyToMany field, a "private" table (not tied to a
model like other tables) is created in the database with something
like the following (created via the contrib.auth.User's
user_permissions field:
id, user_id, permission_id
If you wanted to create your own table that mirrors this
> My problems are this. First off, I was able to get the development
> admin working with my project fine, but when I try to serve it up
> using apache and mod_python, it's looking for the project under my
> Python25 directory (c:\dev\Python25), not in c:\dev\testproj.
I think you need to say
> Is it possible?
> Or should I remove auto_now and update the date field whenever
> updating the model?
If you can't update to trunk, this would probably be your easiest
solution. It should only be a few lines of code.
-rob
--~--~-~--~~~---~--~~
You received
Re: TortoiseSVN, take a look at http://tortoisesvn.net/node/87
Short story - unless you can get your admin to change your firewall,
you can't use svn since "code" doesn't respond to https, no matter
which platform you are using.
-rob
On Jul 22, 8:20 pm, james_027 <[EMAIL PROTECTED]> wrote:
>
With 0.91, you use the "magic" name for db-api operations, rather than
the model name. For example:
from django.models import my_model_module
my_model_module.somemodels.get_object(...)
I think sometime before 0.91 was finished, the directory structure
changed, so you might access your models
> PythonPath "['/home/rith'] + sys.path"
>
>
> This is the templates in settings.py:
> TEMPLATE_DIRS = (
> "/home/rith/soccerux/templates",
> )
>
> Any idea why it's not working with apache? Thank you.
Apache probably can't "read" any files/directories because your home
directory
> In short, a C extension module may cache data from one sub interpreter
> and then use it in the context of a different sub interpreter causing
> incorrect or errornous behaviour. The problem with Decimal support in
> pyscopg falls into this category. It is entirely possible that the OP
> may
ersion_handler
> PythonInterpreter version2
>
>
> Finally, you might want to have different manage.py stubs with an
> extra first line importing the right django module as django into the
> namespace to get going.
>
> Best regards...
>
> Niels
>
> On Jul 19, 12
Keith-Magee" <[EMAIL PROTECTED]>
wrote:
> On 7/19/07, oggie rob <[EMAIL PROTECTED]> wrote:
>
>
>
> > Please, if you've seen the same issue or have any helpful ideas to try
> > to stop the error, let me know.
>
> I think I've seen the same problem (or, a
I've run into a problem that appears to be caused by multiple django
versions running on the same server. I have a preforked mod_python
installation and the problem I'm seeing is this: after using the trunk-
based application, I try to use the 0.91-based application and get a
500 error. The root
Right! Meant to but hit the wrong list. Please ignore this thread...
On Jul 14, 9:42 pm, theju <[EMAIL PROTECTED]> wrote:
> Probably this post would be best answered on the Django-developers
> list.
>
> Cheers
> Thejaswi Puthraya
--~--~-~--~~~---~--~~
You
Hi all,
Playing around with the test client I found it frustrating that
"context" can either be a single context or a list, depending on how
the template is rendered. My feeling is they should always return
lists. Likewise for the "template" field. I can see that no matter
what there is a chance
The statements seem fine, just that your models don't reflect what you
are after.
1. if you set null=True, the database lines would read "NULL" instead
of "NOT NULL". blank=True refers to whether the value can be empty
(i.e. NULL is not equal to ''). However for most CharFields you will
want to
You need to add the Admin class to ContactPerson
-rob
On Jul 12, 6:59 am, Manuel Meyer <[EMAIL PROTECTED]> wrote:
> -BEGIN PGP SIGNED MESSAGE-
> Hash: SHA256
>
> Hey,
>
> from other projects, running on the same django instance, I know that
> there is a possibility to display a '+'
Since you haven't got an answer yet...
I'm not a project developer, so I can't say what their intentions are,
but having worked extensively using oldforms and recently using
newforms, I doubt very much that the Manipulator framework will hang
around for anything other than temporary backward
I could be a bit more pythonic:
>>> from django.db.models.query import QOr
>>> ql = [Q(birthdate__day=(startdate + timedelta(days=x)).day) &
>>> Q(birthdate__month=(startdate + timedelta(days=x)).month) for x in range(7)]
>>> Person.objects.filter(QOr(*ql))
-rob
On Jun 19, 2:51 pm, Tim Chase
t.
Thanks for the help,
-rob
On Jun 6, 5:49 pm, "Russell Keith-Magee" <[EMAIL PROTECTED]>
wrote:
> On 6/6/07, oggie rob <[EMAIL PROTECTED]> wrote:
>
>
>
> > Hi all,
> > I have some models and need to add the capability of versioning the
> > data. Basi
Hi all,
I have some models and need to add the capability of versioning the
data. Basically for each object that represents a page, I want to have
a "running" version and an "editing" version of the same page, so that
users can see the "approved" version yet I can work on a newer one.
(I've found
have with my approach ...
>
> patrick
>
> Am 03.06.2007 um 00:40 schrieb oggie rob:
>
>
>
> > The advantage is you get to organize your additions in the base
> > template (which is where you should strive to manage layout & L as
> > much as possible). Your solut
You might be limited to a relatively small number of Django apps (i.e.
separate entries in the httpd.conf file) on a single dedicated server,
since each would require its own PythonInterpreter, but depending on
traffic I would guess you could have more than that on there. Having
said that, if you
Its hard to tell what you mean by "really long once I start factoring
time and uniqueness". Do you mean really large data sets or really
wide lines in your python file?
If the former, you will probably need to be more specific about what
the long query contains. If the latter, look at chaining
As always, depends. Just think about how source control (i.e. code
ownership), re-usability and IDE integration/navigation would fit with
each approach. Contrary to Russ, I have my form classes near to my
view functions, but I won't bet my C++ Builder 4 CD that its the best
way :)
Probably not
> As i try to login into the admin, i alway get a message saying my
> browser does not accepr cookies!!
> But it isn't true!
>
> Any suggestion?
IE has a funny setting wrt media files & security settings - if you
use a separate server for media (e.g. media.company.com vs. www.company.com
as the
Check Unix permissions for the todo directory and/or settings.py file.
If apache can't access them, it won't run properly.
-rob
On May 4, 6:49 am, RollyF <[EMAIL PROTECTED]> wrote:
> Could someone point me to a solution somewhere? I have a directory
> call apps and inside is a django project,
On Apr 26, 12:16 pm, Laurie Harper <[EMAIL PROTECTED]> wrote:
> oggie rob wrote:
> >> (r'^%s(?P[^?#]*)' % (root_url), 'apps.cms.views.render'),
> > Note that '?' is a special character, ...
>
> Nope; it's special outside a character class ([...]), but in a char
That is abysmal performance, though. If your dev server performs okay
(i.e. it is not the way you have written your index view) you should
probably look for another provider.
-rob
On Apr 24, 6:44 am, Roboto <[EMAIL PROTECTED]> wrote:
> RESOLVED.
>
> I am on Shared Hosting with GrokThis.Net.
The last I heard, choices only worked for CharFields. Probably it is
trying to compare 600.0 with '600.0' and failing, so it resorts to the
unmatched "-", or failing to save the value properly in the first
place. Docs don't clarify, though - it might have changed recently.
If so, your best
>
> (r'^%s(?P[^?#]*)' % (root_url), 'apps.cms.views.render'),
>
Note that '?' is a special character, but more importantly your
expression probably matches both the initial request and the
redirected one. Check it out:
>>> r = re.compile('/about/(?P[^?#]*)')
>>> r.match('/about/?user1')
You're missing the __str__ method. Not sure if that is causing the
"Error" though.
-rob
On Apr 18, 10:39 pm, Michael Lake <[EMAIL PROTECTED]> wrote:
> Hi all
>
> In the Admin interface I'm getting the folloing error for a model:
> Error: variable not passed into template?>Data
>
> class
usr/lib/python2.4/site-packages']
>
> So this is the problem I guess - apache can't access django, but I'm
> not sure why... any ideas?
>
> oggie rob: Nah, I'm not discouraged, I really like there attidue...
> linux isn't my forte though it seems!
>
> --
> Ollie
>
>
What happens when you run:
sudo -u apache python
>>> import django
>>> django
BTW, I use openhosting and have a few django sites running so don't be
discouraged!
-rob
On Apr 18, 12:59 pm, Oliver Charles <[EMAIL PROTECTED]> wrote:
> Ok,
>
> I've done a ton of googling on this, and not come up
> page_data['topic_modification_date'] = datetime.datetime.date()
> page_data['topic_modification_date_time'] = datetime.datetime.time()
What James is saying is suggesting is that do_html2python only expects
strings or lists of strings, but you were trying to pass in a datetime
object. Give it
> The actual code is here -
http://www.satchmoproject.com/trac/browser/satchmo/trunk/satchmo/shop...
and the save function starts at line 150.
I had a look and was wondering about the
self.clean_data['other_field'] references within various clean_x
methods. I thought, perhaps cyclic references
You will need to provide more specific information to get an answer to
that question. Namely, you will need to provide:
1) How long "several seconds" is
2) What is your view (or is it admin view)?
3) What would satisfy "good" performance
4) What do you see when you don't use the web server &
Try using firebug (extension for Firefox). You can see how it is
instantiated in the admin site and where things are going wrong.
I suspect you either are missing the "vSelectMultipleField" (is that
right?) class, the multiple="multiple" attribute, or have not got all
of the JS included. As for
urllib.urlopen('http://127.0.0.1:8000/putProduct').read()
On Apr 9, 9:09 am, "johnny" <[EMAIL PROTECTED]> wrote:
> How do you send a request to localhost djanog dev server from Python
> Terminal?
> I have started mysql and python dev server locally on my machine.
> Now, I want to open
> python
On Apr 8, 1:10 am, "oliver" <[EMAIL PROTECTED]> wrote:
> Hi Rob,
> would you mind posting the code you wrote as an example? Would be very
> helpful.
>
> thank!
> oli
I posted my first djangosnippet to show it:
http://www.djangosnippets.org/snippets/162/
HTH,
-rob
For a list display, it can be very simple. Write a method within the
model that returns an "img" html tag, and use the "allow_tags"
decorator on that method (just like you would for short_description).
Then add that method to your list_display variable in Admin. I did
this for thumbnails.
I
Ahh... botched post.
Anyway, check the Page Info "Type" & post your view method if there
are still problems.
-rob
On Apr 2, 9:02 pm, "oggie rob" <[EMAIL PROTECTED]> wrote:
> Try looking at Firefox's "Page Info". That will tell you whether
> Djan
Try looking at Firefox's "Page Info". That will tell you whether
Django or Firefox is the issue. It should say Why don't you post some
of your view method. It might be
On Apr 2, 7:27 pm, queezy <[EMAIL PROTECTED]> wrote:
> Hi All!
>
> I used response = HttpResponse(pdfbytes,
> I was expecting to just be able to do:
> f = SomeClass.objects.filter(a=a, b=b, c=c)
>
> Is there some way to do this with the current code base?
I'm not sure of the exact syntax, but you can use a dictionary to
achieve this fairly easily. Something like
args = {}
if a == None:
> Anyway, back to the question: do think there is a way with standard django
> to access two databases (maybe sequentially, changing database connections)
> in one script? Or is the only way to use intermediate storage like a file?
The simplest way I can think of is to write your own SQL to
I can't tell if this is the cause, but you should remove the "pass"
right under "class Admin":
-rob
--~--~-~--~~~---~--~~
You received this message because you are subscribed to the Google Groups
"Django users" group.
To post to this group, send email to
You need to include django/contrib/admin/media/js/calendar.js & admin/
DateTimeShortcut.js, and put 'class="vDateField"' or "vTimeField" for
each widget. The html is dynamically generated by javascript when you
load the page (based on class name), which is why you can't find the
js code from the
> Does anybody have any experience with this and can give war-stories about
> good or bad ways to do this?
I just finished updating an aggregate-rich view and I think it is
better to use SQL than messing with filters for most complex things.
My basic rule of thumb is to use python functions to
> is there a way to save the request so i can send a response later
(when the are new items).
You can't send a much-delayed response to browser clients. the http
request will time out if it doesn't get a response in time.
I think you need to look to AJAX. Get a response immediately & keep
>
>
> We're talking about *named* virtual hosting + SSL. Named
> virtual hosting uses the "Host:" HTTP/1.1 request header
> to decide which site should receive the request. But to
> extract that from the SSL'd content we need the site's
> private key. The problem is we don't know which site
>
> 1. One virtual host per site (project1.mycompany.com,
>project2.mycompany.com, etc.)
>
>Pros: Each project has its own admin site. The sqlite database
>file for each project is independent of the others.
>
>Cons: For SSL, each virtual host needs its own IP address and its
>
I had this same problem using a shared system. In my case, I think it
was actually caused by subversion (run through mod_python with SSL).
In particular, I noticed a problem whenever I committed changes to
large binary files.
Unfortunately I never managed to narrow down the problem. However I
> and have forgotten what I did. Now, when I try to call up my admin
> pages, I get nothing. A blank. So my hunch is that apache isn't
> serving up my admin media files, but I can't figure out why. Any help
> is greatly appreciated.
I don't think it is not your media files. If you had no media
Essentially ak pointed out the error. It is expecting 8bit encoding and
the euro is apparently higher than 8 bits (i.e. 20AC is greater than
FF). Just keep plugging around (particularly from the shell & looking
at the db) and you'll probably find it.
-rob
On Jan 25, 10:12 pm, "Christian
> Has anybody seen such Tracebacks ?
Hi Dirk,
I think I have seen this, but only using IE on the development server
with admin-like pages. I think it might be issues with the media going
through the dev server (i.e. probably not an issue with mod_python). If
you were able to hack the way those
Since we're being helpful...
You might be interested in the clock change I submitted some time ago.
It shows the clock in 30 minute increments instead of just 4 areas. You
can either replace your django copy or use the patched file in your own
admin directory somewhere.
> foo = models.ForeignKey( bar, null=True, blank=True )
Give this one a go again. This is the form that should work. Perhaps
there was another problem that got in the way? BTW there is nothing
wrong with using a null field for your foreign key relationship in most
schemas, from a practical
> First i'd like to know if anyone has already done this kind of thing?
I have a pretty basic time tracking and invoicing feature in my
database. As people have pointed out, the biggest issue is printing
consistently. However, you can avoid this headache (yes, it is going to
be a headache!) if
My guess would be caching or database discrepencies. Have you tried
accessing the production database through the dev server? Also have you
tried to delete .pyc files?
-rob
--~--~-~--~~~---~--~~
You received this message because you are subscribed to the Google
> I've thought about an url like "/company1/admin" that will display me only
> the "periodical" and "article" that belongs to company1.
Just be sure this is what you want. There may be times where you wish
to completely separate each company's data - in which case, you can run
multiple
> I'd like to remove the underscore from it.
Underscores are not the norm in python class names, but they will still
work fine (I think?). In your case, add "verbose_name" and possibly
"verbose_name_plural" to your Meta/META class or use CapWords.
-rob
> Found out the issue was the fcgi needing to be killed after changes.
That's where the dev server is nice... in general its better to develop
with the dev server because you don't have to restart
-rob
--~--~-~--~~~---~--~~
You received this message because
> admin.logentry: 'user' has relation with uninstalled model User
>admin.logentry: 'content_type' has relation with uninstalled model
> ContentType
I would guess you need "UPDATE" statements for the content_types table.
You should be able to see what those are (and they will vary for
Yes, I've disabled it now, and all is good!
--~--~-~--~~~---~--~~
You received this message because you are subscribed to the Google Groups
"Django users" group.
To post to this group, send email to django-users@googlegroups.com
To unsubscribe from this group,
> I started to rev back mod_python and ran into a few dependency problems. I'll
> keep trying and see if that helps. Will let you know...
The REAL problem between my working (openhosting.com virtual) Fedora
system and this one was SELinux. Aarrgghh!
And of course I didn't have any problems in
Thanks for the replies everybody. This is good info and I need to know
more about apache in general.
Unfortunately it doesn't solve the problem! I am using the vanilla
Fedora rpm for httpd which builds in prefork.c and has generally
default module directives. The funny behaviour is that it runs
Hi all,
I have a problem with a new server that I am setting up.
Details are:
Fedora Core 5
apache httpd 2.2.0 (rpm)
postgresql 8.1.3 (rpm)
psycopg 1.1.21(source - and a real pain)
python 2.4 (rpm)
pretty weak hardware (800 mHz PIII, 256 M RAM)
So, this app runs fine on two other machines
Tone,
One other thing to consider: the choice of how to integrate developers
is important, but probably more so is deciding on a release schedule.
Your choice of tools and developer cooperation will be dictated by this
somewhat.
For example, if you decide you will make changes "as they are
> I ALWAYS get the didnt match response!
This is where print statements are a very simple diagnosis of the
problem.
Just print out m.password and request.POST['password'] so that you can
see them too. The problem will probably be very obvious.
-rob
Hi Matt,
All that the SESSION_COOKIE_AGE changes is the expire_date on a
"session" entry in one of your database tables. If you change
SESSION_COOKIE_AGE, but are still logged in, then that date is still
going to be about two weeks in the future.
Try logging out, and setting the COOKIE_AGE to
> I did a quick hack today and made it toggable. If you want to aply a
filter, show it. If not, leave it there and browse the listings.
This is nice - can you write up a bug?
-rob
--~--~-~--~~~---~--~~
You received this message because you are subscribed to
Why don't you post some code? Nothing is obvious from what you've said
-rob
--~--~-~--~~~---~--~~
You received this message because you are subscribed to the Google Groups
"Django users" group.
To post to this group, send email to django-users@googlegroups.com
You can quite easily do this with nested lists. Just have the following
structure for each item in the allyears list:
["2005", ["a","b","c"],]
then you can reference each item by number:
{% for yearitem in allyears %}
{{ yearitem.0 }}
{% for item in yearitem.1 %}
{{ item }}
{% endfor
> - latest development from trunk?
If you plan to deploy anytime in the next month or two (yes, this is a
real possibility with Django!), you should start with trunk but settle
on a revision before you get too far into the project. I wouldn't
recommend updating from trunk once you've deployed
> A front-end to a distributed tinder-box setup. Reporting build
results that are either pulled from other locations or posted to the
server.
You could contribute to trac: http://www.edgewall.com/trac/details.html
Although they might not be thrilled about another package to install
all the other
> This seems like a bug to me - certainly it is inconsistent behaviour between
> {% if %} (where you can do 'var1|lower' etc) and {% ifeqaul %}. I would file
> a bug report.
I don't think this is inconsistent. You normally put filters within
variables. I don't think you can put it within
> You could remove the "^" in the regex to say it matches any strings that
> prefixed the intended url pattern:
> r'^cefinban/recettes/(?P\d+)/$',
> 'cefinban.recettes.views.detail')
> to
> r'recettes/(?P\d+)/$', 'cefinban.recettes.views.detail')
But that doesn't really solve the problem, just
> But what about web apps where users normally enter, update, and delete the
> bulk of the application's data? Is it considered good practice to use the
> built-in admin content management functionality as general purpose input
> screens?
I can't say if this is "good practice" or not. I'm
Hi Luke,
You need "blank=True" instead of "null=True" for CharFields and
TextFields (they should be non-null even if nothing is typed in there).
-rob
> I've just upgraded from Django0.90 to Django0.91 my production server
(running a django app on a LAN).
[kicks tires a bit] Yep, thar's yer problem right there. Ya put tha
tires on backwards.
Seriously, you should plan a little better than that! First of all it
is known the 0.x is not backward
> Please take a look at my reply in that post. I think both topics are
awefully similar, should we simply "close" my post and discuss in here
instead?
Hi Roberto,
I'm not sure the goals you are asking for are the same. What I
mentioned applied to the model only, not to the admin interface.
You
> but in the user-management part, the password has to be entered in an
> md5-hashed form.
hi gabor,
I've been thinking about this as well and my solution is probably going
to be something like the following:
1) Create a custom form with only the username/password/confirm fields.
This will take
Patrik K says:
> just being curious: what´s the argument against defining the rows in the
model?
Hi Patrick,
I think it would be better use of display design if you kept the rows
out of the model. You can achieve as much as you are asking by using
css.
For example, if you do the following:
1)
Tom Tobin wrote:
> I don't know; I couldn't tell which of the dozen of your (accidentally, I
> assume) copy-and-pasted sections of your email to read. :-p
Then the answer is obvious - since the post is so obviously distinct
from DRY, Django is not the right solution!
Next?
PythonistL says:
> There is no command like
> new_data = manipulator.flatten_data()
Ahh, woops!
I've had my head in admin code so much lately I forgot this wasn't a
standard manipulator method.
(FYI, the method I was referring to is created by "magic", and has
probably already been removed from
> Id rather not go to:
> r'/albums/(?P\d+)/$' and keep the more readable urls
You can still use this format and get readable urls:
r'albums/(?P)/$'
The album_dir is simply a parameter that is passed to your view
function. If you set your regex up you can limit it/adjust it to allow
for
Hi Emanuele,
To force a user to make a choice wrt boolean fields, you have two
options: a BooleanField or (as you seem to be looking for) a
NullBooleanField, which actually represents three states (including
'unselected'). This will mirror what you are trying to achieve with the
CharField
I did a similar thing not long ago. I used '/places/edit/%s/' %
new_place.id (the difference being %s vs %i), but I don't think that
matters.
Also (and I'm not sure how much this matters), rjwittams suggested in a
recent change that the do_html2python() call should occur before the
"if errors:"
Something that I noticed in the last update I did was that the
raw_id_admin field changed (again!). Originally it looked like this:
verbose_field_name [ ] selected_object.__repr__()
I think this changed sometime in the last few revs so that the
selected_object.__repr__() disappeared. With my
Hi Emanuele,
A quick glance at the use of 'member_name' in the output showed a line
that contained this:
'member_name': 'fornitura.0.materiale_id'
Is that supposed to be in your model?
-rob
Babak,
It shouldn't matter. Take a look at django.core.meta.__init__
manipulator_methods[k[len(MANIPULATOR_FUNCTIONS_PREFIX):]] = v
and
_reassign_globals(manipulator_methods, new_mod,
new_mod.__dict__['AddManipulator'])
_reassign_globals(manipulator_methods, new_mod,
1 - 100 of 111 matches
Mail list logo