I have a form that askes for a phone number. I need to make sure that only digits [0-9] get saved in the database.
In the Django documentation it says:
What happens when you save?
3) Prepare the data for the database. Each field is asked to provide its current value in a data type that can be written to the database.
How does this happen? Or more specifically, how can I make sure this is cleaned? I know that I can just override the models save method, but it seems like there is a better way and I'm just not sure how to do it.
I guess I could write a custom field for it, but that seems like overkill here.
Also, I realize that I can put the validation on the form, but it really feels like stripping out the characters belongs on the model.
Your question specifically about point 3 is a little different from "cleaning" in the way django uses the term.
3) Prepare the data for the database. Each field is asked to provide its current value in a data type that can be written to the database.
Point 3 is about converting the python object values to one suitable for a database. Specifically, this is done in Field.get_prep_value and Field.get_db_prep_value
https://docs.djangoproject.com/en/dev/howto/custom-model-fields/#django.db.models.Field.get_prep_value
It's the opposite of to_python which takes a DB value and converts it to a python object.
As for ensuring only digits 0-9 get stored, that would be done in a Fields clean method (subclass IntegerField), form clean method, form clean_FIELDNAME method, or model clean.
You can add a custom Form Cleaning method to your objects model - take a look at this article https://docs.djangoproject.com/en/dev/ref/forms/validation/#form-field-default-cleaning
Look at "Cleaning a specific field attribute"
use django model form + custom form field cleaning
Below is a quick example of what you might be looking for, where MyModel is the model containing the phone number field, which I named it tel here.
import re
class MyForm(ModelForm):
class Meta:
model = MyModel
def clean_tel(self):
tel = self.cleaned_data.get('tel', '') # this is from user input
# use regular expression to check if tel contains only digits; you might wanna enhance the regular expression to restrict the tel number to have certain number of digits.
result = re.match(r'\d+', tel)
if result:
return tel # tel is clean so return it
else:
raise ValidationError("Phone number contains invalid character.")
Related
In my app, I have a document number which consists of several fields of Document model like:
{{doc_code}}{{doc_num}}-{{doc_year}}
doc_num is an integer in the model, but for the user, it is a five digits string, where empty spaces are filled by zero, like 00024, or 00573.
doc_year is a date field in the model, but in full document number, it is the two last digits of the year.
So for users, the document number is for example - TR123.00043-22.
I want to implement searching on the documents list page.
One approach is to autogenerate the full_number field from doc_code, doc_num and doc_year fields in the save method of Document model and filter on this full_number.
Anothe is to use Concat function before using of filter on query.
First by concatinate full_code field
docs = Document.annotate(full_code=Concat('doc_code', 'doc_num', Value('-'), 'doc_year', output_field=CharField()))
and than filter by full_code field
docs = docs.filter(full_code__icontain=keyword)
But how to pass doc_num as five digits string and doc_year as two last digits of year to Concat function?
Or what could be a better solution for this task?
Concat will only take field names and string values, so you don't really have many options there that I know of.
As you note, you can set an extra field on save. That's probably the best approach if you are going to be using it in multiple places.
The save function would look something ike
def save(self, *args, **kwargs):
super().save()
self.full_code = str(self.doc_code) + f"{doc_num:05d}") + '-' + time.strftime("%y", doc_year))
self.save()
doc_num requires python>= 3.6, other methods for earlier pythons can be seen here
doc_year assumes it is a datetime type. If it is just a four digit int then something like str(doc_year)[-2:] should work instead.
Alternately, if you are only ever going to use it rarely you could loop through your recordset adding an additional field
docs=Document.objects.all() #or whatever filter is appropriate
for doc in docs:
doc.full_code = f"{doc.doc_code}{doc.doc_num}-{time.strftime("%y", doc_year)}
#or f"{doc.doc_code}{doc.doc_num}-{str(doc_year)[-2:]} if doc_year not datetime
and then convert it to a list so you don't make another DB call and lose your new field, and filter it via list comprehension.
filtered_docs = [x for x in list(docs) if search_term in x.full_code]
pass filtered_docs to your template and away you go.
So, I came across this when I was working on a project.
I had mistakenly placed a "," after a field in one of my models and Django did all the migrations while ignoring that particular field. It took me a while to realize that a little "," after the field is responsible for my field not being reflected in the database.
However, I understand that there shouldn't be a coma but I was kind of expecting Django to give me an error or at least a warning.
Something like maybe:
"Invalid syntax in models.py near FieldName"
EDIT:
"one or more model fields are stored as tuple/s are you sure you want to do so?"
But it ignores that particular field and keeps on migrating. My question is why does Django let that happen? Is this the expected behaviour and shouldn't Django notify for such things? or why this is being passed silently.
Here is an example to have a look at.
class person(models.Model):
name = models.CharField(max_length=10)
surname = models.CharField(max_length=10),
age = models.PositiveIntegerField()
Now, if you create migrations and apply them Django will simply ignore the surname field here and apply the migrations without any errors, why is it so?
It is not invalid syntax. By adding a trailing comma, you wrap the field in a singleton tuple. So the type of person.surname is tuple.
For example if you write:
>>> a = 1,
>>> a
(1,)
>>> type(a)
<class 'tuple'>
A model can, besides the model fields contain all sorts of things: constants, subclasses, methods, etc.
One could do an exhaustive search in all the fields, etc. to check if a tuple wraps a model field, but that could take considerable time, it might result in evaluating lazy attributes, and it might even get stuck in an infinite loop.
It might however be something that can be added to flake8-django [GitHub].
I have to store data, a part of them is predefined but the user can chose to custom it.
What is the best way to store these data in the database?
2 fields, 1 will be an integer field for predefined option and the second will be a string for the custom user input
1 string field, which will contains a json like {predefined: 2, custom: ''}
1 string field which will contains custom string or predefined option id (converted during the request process)
1 string field which will contains the fulltext option even if it is a predefined (some of these predefined options can be long text)
I tried the 1) but double the number of fields for each "custom ready" data doesn't seem to be perfect...
Any idea ?
Considering you might need the following (it's not very clear from your question):
a form where there is an input field for the customizable part of the string
an easy way to refer to the complete string
a way to administer/manage/validate the non-customizable string
=> use two fields:
class TheModel(Model):
# if you have a certain constant number of choices, use ChoiceField
# otherwise use a ForeingKey and create a different model for those
non_customizable_prefix = ChoiceField(null=False, blank=False, ...)
# unique? validators? max/min length? null/blank?
customizable_part = CharField(...)
#property
def complete_string(self):
return '{}{}'.format(self. non_customizable_prefix, self. customizable_part)
This model will provide you with two separate input fields in Django forms or the Django admin, offering easy ways to make the non_customizable_prefix read only or only modifiable with certain privileges.
is there any possibility to define a Field which gets incremented, but starts with a letter? So the values should be like: S123, S124, S125.
I' d like to use the id field for it, but some reason i' d like to have it as a CharField.
As a workaround i could use:
id = models.CharField(max_length = 32, primary_key = True)
and redefine the save method, so i always precalculate its value, but this won' t be that robust than a "real" solution, and also my solution would too slow with the calculation.
Is there a proper solution for my problem?
Django: 1.9.2
Python: 3.4.2
.
I don't think you should manually define a primary key. Django usually uses relational database to build an app, which means it would rely on some key field to join other tables to do the lookup when it needs to. Having primary keys like S123 makes it extra hard to maintain because you need to store the same thing as a reference in other tables.
What I would suggest is storing the letter part and the digits separately. You could use the default id field django created as the digit part and create your own field to store the letter part. Then you would use a property method to return the value you want to have. Roughly:
class Foo(models.Model):
letter = models.CharField(max_length=1)
#property
def symbol(self):
return '%s%s' % (self.letter, self.id)
Then you could do:
foo = Foo.objects.create(letter='S')
print foo.symbol # this would print S1, S2, etc.
In case you don't know, here's an explanation of #property in python.
I have a ModelForm, in which I'm having a CharField, which is declared as unique in the Model.
But I have 2 problems:
If I fill in the form with a field having the same name I don't get an error message.
I'd like this field not to contain white spaces.
Is it possible to do that using a ModelForm?
You can do something close to this:
class MyModelForm(forms.ModelForm):
# your field definitions go here
def clean_myuniquefield(self):
# strip all spaces
data = str(self.cleaned_data['myuniquefield']).replace(' ', '')
model = self._meta.model
# check if entry already exists
try:
obj = model.objects.get(myuniquefield=data)
except model.DoesNotExist:
return data
raise forms.ValidationError("Value already exists!")
To get rid of spaces, make a clean_fieldname function to strip the spaces.
http://docs.djangoproject.com/en/dev/ref/forms/validation/#ref-forms-validation
As for uniqueness, also note about the meta-field unique_together. I don't know if you need it, but I didn't know about it until I dug around.
If you really need to do uniqueness checking before trying to add and failing, you can also do that in the clean_* function. However, it might be better to assume that the database will take care of it and fail in a standard way, and just set up your error messages properly. That way, if you change constraints later, it will flow through more easily. And if others have to maintain your code, it will be more standard.
Hope this helps.