wtforms表单类子类和字段顺序

时间:2011-05-01 12:11:11

标签: python subclassing wtforms

我有一个UserForm类:

class UserForm(Form):
    first_name = TextField(u'First name', [validators.Required()])
    last_name = TextField(u'Last name', [validators.Required()])
    middle_name = TextField(u'Middle name', [validators.Required()])
    username = TextField(u'Username', [validators.Required()])
    password = TextField(u'Password', [validators.Required()], widget=PasswordInput())
    email = TextField(u'Email', [validators.Optional(), validators.Email()])

并希望在UpdateUserForm中将密码字段设为可选:

class UpdateUserForm(UserForm):
    password = TextField(u'Password', [validators.Optional()], widget=PasswordInput())

但密码字段位于电子邮件字段之后,而不是之前。

如何在子类化时保留字段顺序?

此外,当我尝试更改密码字段验证器时,它不起作用 - 密码仍然是必需的:/为什么?

class UpdateUserForm(UserForm):
    def __init__(self, **kwargs):
        self.password.validators = [validators.Optional()]
        super(UpdateUserForm, self).__init__(**kwargs)

class UpdateUserForm(UserForm):
    def __init__(self, **kwargs):
        self.password = TextField(u'Password', [validators.Optional()], widget=PasswordInput())
        super(UpdateUserForm, self).__init__(**kwargs)

一些想法......

class UpdateUserForm(UserForm):
    def __init__(self, formdata=None, obj=None, prefix='', **kwargs):
        self._unbound_fields[4][1] = TextField(u'Password', [validators.Optional()], widget=PasswordInput())
        UserForm.__init__(self, formdata=None, obj=None, prefix='', **kwargs)

最后,我需要的是:

class UpdateUserForm(UserForm):
    def __init__(self, formdata=None, obj=None, prefix='', **kwargs):
        UserForm.__init__(self, formdata, obj, prefix, **kwargs)
        self['password'].validators = [validators.Optional()]
        self['password'].flags.required = False

6 个答案:

答案 0 :(得分:11)

关于在迭代表单对象时重写字段的第一个问题,这就是我所做的:

class BaseForm(Form):
    def __iter__(self):
        field_order = getattr(self, 'field_order', None)
        if field_order:
            temp_fields = []
            for name in field_order:
                if name == '*':
                    temp_fields.extend([f for f in self._unbound_fields if f[0] not in field_order])
                else:
                    temp_fields.append([f for f in self._unbound_fields if f[0] == name][0])
            self._unbound_fields = temp_fields
        return super(BaseForm, self).__iter__()

class BaseUserForm(BaseForm):
    password = PasswordField('Password', [Required()])
    full_name = TextField('Full name', [Required()])

class NewUserForm(BaseUserForm):
    username = Textfield('Username', [Required()])
    field_order = ('username', '*')

这样,当您渲染NewUserForm时(可能来自按字段迭代表单呈现字段的模板),您将看到usernamepasswordfull_name。通常情况下,您最后会看到username

答案 1 :(得分:4)

我通过在__order类上定义一个额外的Form属性并覆盖__iter__方法来解决这个问题,以便根据定义首先对返回的迭代器数据进行排序。它可能效率不高,但表单上没有那么多字段,可能会导致任何问题。它也适用于子类表单中的字段。

class MyForm(Form):
    field3 = TextField()
    field1 = TextField()
    field2 = TextField()

    __order = ('field1', 'field2', 'field3')

    def __iter__(self):
        fields = list(super(MyForm, self).__iter__())
        get_field = lambda field_id: next((fld for fld in fields
                                           if fld.id == field_id))
        return (get_field(field_id) for field_id in self.__order)

答案 2 :(得分:3)

这就是我完成你想要做的事情:

class UserForm(wtforms.Form):                                                   
    def __init__(self, *args, **kwargs):                                        
        super(UserForm,self).__init__(*args, **kwargs)                          

        if kwargs.get('update', None):                                          
            self['passwd'].validators.append(wtforms.validators.Optional())
            self['passwd'].flags.required = False     
        else:                                                                   
            self['passwd'].validators.append(wtforms.validators.Required()) 

    passwd = UnicodeField(                                                      
        u'Password',                                                            
        [                                                                       
            wtforms.validators.length(max=50),                                  
            wtforms.validators.EqualTo(                                         
                'confirm',                                                      
                message='Passwords must match'                                  
                )                                                               
            ],                                                                  
        widget = wtforms.widgets.PasswordInput()                                
        )                                                                       

    confirm = wtforms.PasswordField(u'Password Verify')

然后,当我实例化UserForm时,我在编辑时传递update = True。这似乎对我有用。

答案 3 :(得分:2)

这是因为字段排序由UnboundField.creation_counter类定义,该类使用Field类在代码中出现的顺序。

>>> x1 = UserForm()
>>> x2 = UpdateUserForm()
>>> [(f[0], f[1].creation_counter) for f in x1._unbound_fields]
[('first_name', 22), ('last_name', 23), ('middle_name', 24), ('username', 25), ('password', 26), ('email', 27)]
>>> [(f[0], f[1].creation_counter) for f in x2._unbound_fields]
[('first_name', 22), ('last_name', 23), ('middle_name', 24), ('username', 25), ('email', 27), ('password', 28)]
>>> 

由于这很难解决(因为使用这种方法,wtforms会变得很神奇),处理此问题的最佳方法是按所需顺序定义字段。

class BaseForm(Form):
    first_name = TextField(u'First name', [validators.Required()])
    last_name = TextField(u'Last name', [validators.Required()])
    middle_name = TextField(u'Middle name', [validators.Required()])
    username = TextField(u'Username', [validators.Required()])

class UserForm(BaseForm):
    password = TextField(u'Password', [validators.Required()], widget=PasswordInput())
    email = TextField(u'Email', [validators.Optional(), validators.Email()])

class UpdateUserForm(BaseForm):
    password = TextField(u'Password', [validators.Optional()], widget=PasswordInput())
    email = TextField(u'Email', [validators.Optional(), validators.Email()])

但如果你是完美主义者或需要遵守DRY principle

class BaseForm(Form):
    first_name = TextField(u'First name', [validators.Required()])
    last_name = TextField(u'Last name', [validators.Required()])
    middle_name = TextField(u'Middle name', [validators.Required()])
    username = TextField(u'Username', [validators.Required()])

class UserForm(BaseForm):
    password = TextField(u'Password', [validators.Required()], widget=PasswordInput())

class UpdateUserForm(BaseForm):
    password = TextField(u'Password', [validators.Optional()], widget=PasswordInput())

BaseForm.email = TextField(u'Email', [validators.Optional(), validators.Email()])

答案 4 :(得分:2)

要强制对表单字段进行排序,您可以使用以下方法:

from collections import OrderedDict

def order_fields(fields, order):
    return OrderedDict((k,fields[k]) for k in order)

并在表单构造函数中调用它,如下所示:

class FancyForm(Form, ParentClass1, ParentClass2...):
    x = TextField()
    y = TextField()
    z = TextField()

    _order = 'x y z'.split()


    def __init__(self, *args, **kwargs):
        super(FancyForm, self).__init__(*args, **kwargs)
        self._fields = order_fields(self._fields, 
                                    self._order + ParentClass1._order + ParentClass2._order)

答案 5 :(得分:0)

我已将两个答案合并到以下代码段中:

def __iter__(self):
    ordered_fields = collections.OrderedDict()

    for name in getattr(self, 'field_order', []):
        ordered_fields[name] = self._fields.pop(name)

    ordered_fields.update(self._fields)

    self._fields = ordered_fields

    return super(BaseForm, self).__iter__()

它在BaseForm上的 iter 我的每个表单都是其中的一个。基本上,field_order中定义的所有内容都按此顺序排列,其余字段按原样呈现。