This is a bit of a hack, but as far as I can see currently the only way to specify a validation error that is specific to a field in form.clean(). I am aware of clean_<fieldname>, but those are difficult to use when the validation process for a field involves other fields as well, because the necessary data might at that point not be yet available in form.cleaned_data.
Example (in project/application/models.py): register_custom_permissions_simple((("is_editor", "User is editor"),)) In a view: if not request.user.has_perm('application.is_editor'): return HttpResonseRedirect(LoginUrl)
A simple login form that does the actual authentification itself. **Usage:** if request.method == "POST": loginform = LoginForm(request.POST) if loginform.login(): return HttpResponseRedirect(redir_url) else: loginform = LoginForm()
This is an extendend version of the Rails Flash implementation by Sean Patrick Hogan that supports different message types. **Setting a flash message:** request.flash.error = 'Item could not be saved' request.flash['error'] = 'Item could not be saved' request.flash['foo'] = 'bar' **Displaying a flash in the view:** <!-- show the error message --> {% if flash.error %}An error occured:{{ flash.error }}{% endif %} <!-- just show the first message found --> {% if flash %}An error occured:{{ flash }}{% endif %} <!-- show all messages --> {% for msg in flash %}{{ msg.type }}: {{ msg.msg }}{% endfor %} Note that it still works with simple strings as well. Feel free to just use it like this: request.flash = "Message" And: {% if flash %}{{ flash }}{% endif %} However, be aware that once you did this, you destroyed the Flash() dict and thus lost the extended functionality. You can use request.flash.clear() to remove all messages.
I wanted to make the objects of a particular model approvable and store the timestamp of when that happened. In other frameworks/languages, I used to combined those in one "approved_at" field, which would be NULL if an object was currently unapproved. I tried different approaches to implement this in django, and this is the best I came up with so far. Basically, the code in __setattr__ makes sure that the field, once set, will not be updated again. Overriding setattr__() could also be a solution to determining if a field value has changed in save(), a question that seems come up from time to time in #django.
Can be used if a form field should not be editable, but the current value or the value that will be automatically used should still be visible to the user. __init__ takes two additional parameters: **value** is the actual value to be used when saving the form, while **display** determines what is shown to the user when rendering. If *display* is not specified, *value* itself will be used instead. If *display* is a *ModelChoiceField*, *value* is assumed to be a primary key of the model, and the widget will automatically try to retrieve and use the string representation of the corresponding item.
If you have many models that all share the same fields, this might be an option. Please note that order matters: Your model need to inherit from TimestampedModelBase first, and models.Model second. The fields are added directly to each model, e.g. while they will be duplicated on the database level, you only have to define them once in your python code. Not sure if there is a way to automate the call to TimestampedModelInit(). Tested with trunk rev. 5699. There is probably a slight chance that future revisions might break this.