This very simple middleware will set the 'user' attribute of a mod_python request.. this way the username is logged into the apache log ... useful when making webalizer statistics to see the activity of users .. (http://sct.sphene.net)
Useage: `{% load setting %}` `{% setting DEBUG %}` or... `{% setting MEDIA_ROOT %}` You get the gist.
It's cheap, hackish, and dirty, but it works. \\o/ Lines 19 & 20 are what you should edit.
Sometimes you'll have a list of ORM objects that aren't in a QuerySet, but you still want to sort them by date. For instance, you have a list of shows with a date_time attribute. Maybe you used a list comprehension to filter them...who knows. Regardless, you have a list (not a QuerySet) of Django objects.
In this type of model you are allowed to define a model with a generic type. For instance, a location can be an address, GPS coordinates, an intersection and many others types. Using a many to many field, models can have multiple locations without worrying about the type of location referencing. New locations types can be added without changing the references in other models. This code is also used in Django's built in ContentTypes app.
This class simplies the Feed class of django. The differences are: 1. Don't need define title and description template file 2. default feed generator class is Atom1Feed 3. According feed_url, EasyFeed can auto get the domain field, and you can also specify the domain parameter.(feed_url should be a full domain path, so you can use [Get the full request path](http://www.djangosnippets.org/snippets/41/) to get the full path of the feed url.) 4. There is a helper function render_feed() to return a response value. example --------- Feed class: class BookCommentsFeed(EasyFeed): def __init__(self, feed_url, book_id): super(BookCommentsFeed, self).__init__(feed_url) self.book_id = book_id self.book = Book.objects.get(id=int(book_id)) def link(self): return '/book/%s' % self.book_id def items(self): return self.book.comment_set.all().order_by('-createtime')[:15] def title(self): return 'Comments of: ' + self.book.title def description(self): return self.book.description def item_link(self, item): return '/book/%s/%s' % (self.book_id, item.chapter.num) def item_description(self, item): return item.content def item_title(self, item): return '#%d Comment for: ' % item.id + item.chapter.title def item_pubdate(self, item): return item.createtime def item_author_name(self, item): return item.username And the view code is: from feeds import * from utils.easyfeed import render_feed from utils.common import get_full_path def bookcomments(request, book_id): return render_feed(BookCommentsFeed(get_full_path(request), book_id))
Enables cookie based authentication with apache. I needed user authentication for some static files, but couldn't use the method described [here](http://www.djangoproject.com/documentation/apache_auth/) as this prompts the user for his credentials, making him log in twice. There is some overhead in the code, because it runs all request middleware components (only session and auth would be needed). All arguments described in the link above are supported. I use it like this in the apache config: <Location "/protected/location"> PythonPath "['/path/to/proj/'] + sys.path" PythonOption DJANGO_SETTINGS_MODULE myproj.settings PythonOption DjangoPermissionName '<permission.codename>' PythonAccessHandler my_proj.modpython #this should point to accesshandler SetHandler None </Location>
This middleware will add a log of the SQL queries executed at the bottom of every page. You can (should) use BeautifulSoup to place this in a specific location. Note: If you serve non-html content, it would be wise to do a mimetype check.
This will perform a regular expression search/replace on a string in your template. `{% load replace %}` `{{ mystring|replace:"/l(u+)pin/m\1gen" }}` If: `mystring = 'lupin, luuuuuupin, and luuuuuuuuuuuuupin are lè pwn'` then it will return: `mugen, muuuuuugen, and muuuuuuuuuuuuugen are lè pwn` The argument is in the following format: [delim char]regexp search[delim char]regexp replace
Allows you to override form fields created by form_for_model or form_for_instance, without having to redefine whole fields for example in form.base_fields.
Use this code to generate downloadable [vCard][] objects. See the [VObject docs][1] for more details on the API. [1]: http://vobject.skyhouseconsulting.com/ [vcard]: http://en.wikipedia.org/wiki/VCard
[Snippet #2](http://www.djangosnippets.org/snippets/2/) demonstrated some cool tricks possible with manager methods. This example shows how to assign and use a custom manager method. In this snippet the `belongs_to_user` method returns an Account queryset containing only those accounts associated with the specified user. The method is useful because it hides the implementation of User in the Account model. Line 17 associates the custom manager with the Account model.
A patch (against django svn trunk [4649](http://code.djangoproject.com/browser/django/trunk/?rev=4649)) that allows users to log in with Basic HTTP Authentication i.s.o. login forms using some simple middleware (entire patch is ~50 lines). I was unaware of http://code.djangoproject.com/wiki/GenericAuthorization so I'm not sure about its usefulness in the long run. You can enable it by including 'django.contrib.auth.middleware.BasicAuthenticationMiddleware' in your MIDDLEWARE_CLASSES and then adding the following lines in your settings.py: BASIC_WWW_AUTHENTICATION = True WWW_AUTHENTICATION_REALM = "djangolures.com" Updated: See also http://code.djangoproject.com/ticket/3609 (patch now availble here as well).
Using newforms you can create forms from existing models easily and automatically with either `form_for_model(class)` or `form_for_instance(instance)`. Usually the automagically generated form fields are sufficient; however, sometimes you need to restrict selections in a choice field. You can also set the default selection when instantiating the form. In this example, if `acct` is not contained in the 'account' field choices, the selection defaults to the first entry. This example is probably not good practice when using `form_for_instance` because the existing value 'selection' of the choice field is lost and must be reset manually (see above).
I use this script to keep track of the cache usage for various sites. It presently only supports memcached because that's all that I use, but leave a comment with patches for other systems and I'll add them.