One CSRF token is assigned to every session ( i.e. every time you log in). So before you wish to get some data entered by user and send that as ajax call to some function which is protected by csrf_protect decorator, try to find the functions that are being called before you are getting this data from user. E.g. some template must be being rendered on which your user is entering data. That template is being rendered by some function. In this function you can get csrf token as follows: csrf = request.COOKIES['csrftoken'] Now pass this csrf value in context dictionary against which template in question is being rendered. Now in that template write this line: Now in your javascript function, before making ajax request, write this: var csrf = $('#csrf').val() this will pick value of token passed to template and store it in variable csrf. Now while making ajax call, in your post data, pass this value as well : "csrfmiddlewaretoken": csrf

这将工作，即使你没有实现django表单。

实际上，这里的逻辑是:您需要从请求中获得的令牌。 因此，您只需要在登录后立即找出被调用的函数。一旦你有了这个令牌，要么进行另一个ajax调用来获取它，要么将它传递给一些可以通过ajax访问的模板。

如果你的表单在不使用JS的情况下可以在Django中正确地发布，你应该可以使用ajax逐步增强它，而不需要任何修改或混乱地传递csrf令牌。只需序列化整个表单，它将自动拾取所有表单字段，包括隐藏的csrf字段:

$('#myForm').submit(function(){
    var action = $(this).attr('action');
    var that = $(this);
    $.ajax({
        url: action,
        type: 'POST',
        data: that.serialize()
        ,success: function(data){
            console.log('Success!');
        }
    });
    return false;
});

我已经用Django 1.3+和jQuery 1.5+进行了测试。显然，这适用于任何HTML表单，而不仅仅是Django应用程序。

One CSRF token is assigned to every session ( i.e. every time you log in). So before you wish to get some data entered by user and send that as ajax call to some function which is protected by csrf_protect decorator, try to find the functions that are being called before you are getting this data from user. E.g. some template must be being rendered on which your user is entering data. That template is being rendered by some function. In this function you can get csrf token as follows: csrf = request.COOKIES['csrftoken'] Now pass this csrf value in context dictionary against which template in question is being rendered. Now in that template write this line: Now in your javascript function, before making ajax request, write this: var csrf = $('#csrf').val() this will pick value of token passed to template and store it in variable csrf. Now while making ajax call, in your post data, pass this value as well : "csrfmiddlewaretoken": csrf

这将工作，即使你没有实现django表单。

实际上，这里的逻辑是:您需要从请求中获得的令牌。 因此，您只需要在登录后立即找出被调用的函数。一旦你有了这个令牌，要么进行另一个ajax调用来获取它，要么将它传递给一些可以通过ajax访问的模板。

这个问题是因为django希望cookie中的值作为表单数据的一部分传递回来。前一个答案中的代码是让javascript寻找cookie值并将其放入表单数据中。从技术角度来看，这是一种可爱的方式，但它看起来有点啰嗦。

在过去，我通过让javascript将令牌值放入post数据中更简单地做到了这一点。

如果在模板中使用{% csrf_token %}，则会发出一个包含该值的隐藏表单字段。但是，如果你使用{{csrf_token}}，你只会得到令牌的裸值，所以你可以在javascript中使用这个....

csrf_token = "{{ csrf_token }}";

然后，可以在散列中包含所需的键名，然后将其作为数据提交给ajax调用。

更新2022

在CSRF攻击中，无辜的终端用户被攻击者欺骗，提交了一个他们不打算提交的web请求

选项1

from django.views.decorators.csrf import csrf_exempt
from django.http.response import JsonResponse


@csrf_exempt
def commentDeletePost(request):
    if request.is_ajax() and request.method == 'POST':
        try:
            comment = Comment.objects.get(pk=request.POST['pk'])
            if comment.author != request.user:
                return JsonResponse({'e': 'Forbidden'}, status=403) 
            comment.delete()
            return JsonResponse({}, status=200)
        execpt Comment.DoesNotExist:
            return JsonResponse({'e': 'Not Found'}, status=404)

选项2

<div id="csrf">
    {% csrf_token %}
</div>
<script type="text/javascript">
    window.crud = {
        commentDelete: function(
            pk, 
            success,
            error, 
        ){
            $.ajax({
                headers: {'X-CSRFToken': document.getElementById('csrf').querySelector('input').value},
                type: "POST", 
                url: "{% url 'comment-delete-post' %}",
                data: {
                    pk: pk,
                }, 
                success: success, 
                error: error,
            })
        }, 
    }
</script>

two options have its own advantage. First option will discard csrf token, which will not protecte your site from csrf attacks, but it will allow user to send more than one request with same Ajax function. the second option will restrict user to send one Ajax request only since csrf token can only be used once, but it is more secure. I personally prefer option 1, since Ajax functions such as like, star, unlike requires more than one Ajax call, and it is not a risky function to allow user call more than once.

{% csrf_token %}在<form></form>内放置html模板

翻译过来就是:

<input type='hidden' name='csrfmiddlewaretoken' value='Sdgrw2HfynbFgPcZ5sjaoAI5zsMZ4wZR' />

所以为什么不像这样在你的JS中grep它:

token = $("#change_password-form").find('input[name=csrfmiddlewaretoken]').val()

然后传递它，例如做一些POST，比如:

$.post( "/panel/change_password/", {foo: bar, csrfmiddlewaretoken: token}, function(data){
    console.log(data);
});