This solves the problem with **choices** described [here](http://www.b-list.org/weblog/2007/nov/02/handle-choices-right-way/) Define your choices like this (in models.py): statuses = MyChoices(BIDDING_STARTED=10, BIDDING_ENDED=20) And then: status = models.IntegerField( default=statuses.BIDDING_STARTED, choices=statuses.get_choices() )
I like the per-site caching offered by Django (it's simple) and I like the cache to be dependent on the chosen language. Unfortunately with the current `CacheMiddleware` you can either have the one or the other but not both at the same time. `VaryOnLangCacheMiddleware` is a small wrapper around `CacheMiddleware`. It looks at the `request.LANGUAGE_CODE` (set by `LocaleMiddleware`) and appends it to your `CACHE_MIDDLEWARE_KEY_PREFIX`. So "mysiteprefix" becomes "mysiteprefix_en" or "mysiteprefix_de-AT" depending on the user's chosen language. Site-wide, so no messing with per-view decorators and stuff. To use this, make sure `VaryOnLangCacheMiddleware` comes below `LocaleMiddleware` in your settings.py. If you haven't set your `CACHE_MIDDLEWARE_KEY_PREFIX`, it's works, too. **Update:** Replaced `super()` calls with `CacheMiddleware.xxx(self, ...)`.
Somebody mentioned in #django the other day that they have multiple databases with the same schema... Well *lucky me* so do I!! This is one way to work with this issue. I've also included migrate_table_structure just in case the schema doesn't exist in the new database yet. As per the multiple-db-support branch, write all of your databases into the OTHER_DATABASES in settings.py. You can now copy models from your various models.py files and use them in different databases at the same time. This can also be used to migrate databases from one dialect to the other without having to translate your data into an interim format (e.g. csv, XML). You can just do: qs = MyModel.objects.filter(**filters) NewModel = duplicate_model_and_rels(MyModel, 'new_db') #Assuming that the schema is already in new_db: for mod in qs: new = NewModel() new.__dict__ = mod.__dict__ new.save() I tried this using some hacks with SQLAlchemy, and the above approach is a huge amount quicker! I've used this to copy some stuff from an oracle db, into a sqlite db so i could carry on working later and transferred about 20,000 in 5 mins or so. GOTCHAS ======== This only works against my copy of multi-db as I've made a couple of changes. My copy is substantially the same as my patch attached to ticket 4747 though, so it might work to a point (especially the data migration aspect). If it doesn't work hit me up and I'll send you my patch against trunk. I'm not too crazy about the code in copy_field, it works fine, but looks ugly... If anyone knows of a better way to achieve the same, please let me know.
Here is a Django view that turns code like this: @variables { $varcolor: #333; } body { color: $varcolor; padding: 20px; } .space { padding: 10px; } .small { font-size: 10px; } #banana(.space, .small) { margin-bottom: 10px; } And turns it into something like this: body { color: #333; padding: 20px; } #banana { padding: 10px; font-size: 10px; margin-bottom: 10px; } .small { font-size: 10px; } .space { padding: 10px; } Notice the variables declaration at the top. The other feature is *extending* - here #banana extends .space and .small. The url.py entry might look something like this: (r'^css/(?P<css>(\w|-)+)\.css$','csspp.csspp'), Here referencing csspp.py in your path (root directory of your site probably). The code also looks for a CSS_DIR setting in your settings file. You will probably want to point straight to your media/css/ directory. **Known problems** * There is now way of extending selectors that are already extending something else. In the example code there is now way to extend #banana since it is already extending .small and .space.
Middleware for communicating with Flash Player via Flashticle and Django. Setup a view at /gateway/math/multiply like so: def multiply(request, m1, m2): return m1 * m2 Then in your Flex/Flash app you call "math.multiply" on a NetConnection pointing to http://domain.com/gateway/ Does not yet support authentication.
Make sure we don't create an infinite loop with a self-referencing foreign key. Many times I have needed category models that reference themselves, providing a flexible way to make children categories, grandchildren categories, etc. If you chain the slugs (or even ids) there's a chance you could end up with an infinite loop. Assumes a required, unique slug field ('slug') and an optional self-referencing foreign key ('parent'). All_data doesn't give you the object's ID, so we will find it via the unique slug. If either is not present we pass -- if there's no parent chosen it's not a problem, and if there is no slug present the submission will fail on that validation instead. It is worth noting that if the user changes the slug field on submission AND picks a bad parent it will not be caught. Infinite loop cases: 1. A references A 2. A tries to reference B, which is currently referencing A
If you ever have a need to display something like: "last update 5 days ago" "user logged in 2 mins ago" you can use this script to determine how long ago a timestamp is versus now().
Useful for transferring datetime values to JS: briefer and overcoming the 1 vs. 0 based month enumeration difference between python and JS.
s = Scaffold(SomeForm) print s.as_dl()
A CompressedTextField to transparently save data gzipped in the database and uncompress at retrieval. Full description at my blog: [arnebrodowski.de/...Field-for-Django.html](http://www.arnebrodowski.de/blog/435-Implementing-a-CompressedTextField-for-Django.html)
! Note - no longer needed Save this script in the same directory as manage.py and run it through the command line. It picks up project Command class instances. Something that will hopefully be fixed in the Django SVN version soon. Heres an example of a command: #utils/management/commands/sqlallall.py from django.core.management import call_command from django.core.management.base import BaseCommand from django.db import models class Command(BaseCommand): help = "Returns sqlall for all installed apps." def handle(self, *args, **options): """ Returns sqlall for all installed apps. """ for app in models.get_apps(): call_command("sqlall", app.__name__.split(".")[-2])
This allows for urls in the form of `/grandparent-slug/parent-slug/self-slug/` where the number of parent slugs could be 0 to many. You'll need to make sure that it is your last urlpattern because it is basically a catch-all that would supersede any other urlpatterns. Assumes your page model has these two fields: * `slug = models.SlugField(prepopulate_from=("title",), unique=True)` * `parent = models.ForeignKey("self", blank=True, null=True)`
A revised version of [zeeg's Sphinx Search ORM](http://www.djangosnippets.org/snippets/231/), using my Sphinx client and adding support for Sphinx's excerpt generator. It's still missing support for search modes/order_by/filter/exclude, but it should be easy and I will add the relevant methods soon as I need them. Usage is the same as zeeg's class, except that you can pass a field name (or tuple for related objects) to its constructor, that will be used for excerpts: class MyModel(models.Model): search = SphinxSearch(excerpts_field='description') MyModel.search.query('query') MyModel.search.query('query').count() Returns an ordered list of the objects in your database.
I hate when my unittest hits database. Especially when each test case needs different dataset. So I wrote this db mock, that's local to specific test and uses sqlite3 in-memory db. Usage (nosetests): class TestMainNoData(DbMock): 'testing main function with no meaningful data' def test_no_posts(self): 'there are no posts' assert models.Post.objects.count() == 0, 'should be no feeds'
This is based on [Snippet 161](/snippets/161/) It marks duplicated SQL queries. To avoid duplicates read: [Caching and Queryset](http://www.djangoproject.com/documentation/db-api/#caching-and-querysets) Sept. 07: Updated for current trunk: 'response' behaves like 'response.header' 22. October '07: Log into directory.