fix(schemas): use EmailStr for email fields in user create schemas

[armorbreak001]

Problem

UserCreateSchema and SimpleUserCreateSchema use email: str, which accepts any string — including invalid email addresses. This means garbage data passes Pydantic validation and either causes a DB-level error or gets stored as-is, with no user-facing 422 response explaining the format requirement.

The underlying Django model uses EmailField(unique=True), so the serializer should enforce the same validation at the API boundary.

Fix

Change email: str → email: pydantic.EmailStr in both schemas:

• server/apps/model_fk/serializers.py (UserCreateSchema)
• server/apps/model_simple/serializers.py (SimpleUserCreateSchema)

This matches the existing pattern already used in the project's tests and examples (pydantic.EmailStr).

Closes #945

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 100.00%. Comparing base (d210a70) to head (1b7d5b7).

Additional details and impacted files

@@            Coverage Diff            @@
##            master      #946   +/-   ##
=========================================
  Coverage   100.00%   100.00%           
=========================================
  Files          369       369           
  Lines        15539     15539           
  Branches       503       503           
=========================================
  Hits         15539     15539           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[sobolevn]

@Stranger6667 do you have an experience integrating schemathesis and pydantic.EmailStr validator? What is the best way to solve

| - Response violates schema
  | 
  |     "w@ydifdjaglsttf.닷넷" is not a "email"
  | 
  |     Schema at /components/schemas/SimpleUserSchema/properties/email:
  | 
  |         {
  |             "format": "email",
  |             "type": "string",
  |             "title": "Email"
  |         }
  | 
  |     Value:
  | 
  |         "w@ydifdjaglsttf.\ub2f7\ub137"
  | 
  | [200] OK:
  | 
  |     `[{"email":"0@a.com","customer_service_uid":"e3e70682-c209-1cac-a29f-6fbed82c07cd","id":1,"created_at":"2026-04-21T08:49:15.540062Z"},{"email":"qaT@v.cooking","customer_service_uid":"cfdd7215-6621-50e9-b7e2-67dc84f117b8","id":2,"created_at":"2026-04-21T08:49:15.644746Z"},{"email":"tB0@w.aramco","customer_service_uid":"e149bd09-0df5-1245-84b0-6badfa7576c5","id":3,"created_at":"2026-04-21T08:49:15.712245Z"},{"email":"_O9Ap%@a.com","customer_service_uid":"e3e70682-c209-5cac-a29f-6fbed82c07cd","id":4,"created_at // Output truncated...`
  | 
  | Reproduce with:
  | 
  |     curl -X GET --insecure http://localhost/api/model-simple/user-detailed/
  | 
  |  (1 sub-exception)
  +-+---------------- 1 ----------------

?

[Stranger6667]

Probably to change format: email to format: idn-email at least for the output side. I assume Schemathesis generated an ascii value originally that matches format: email (w@ydifdjaglsttf.xn--3eb46j), then it got normalized somewhere on the way and becomes invalid for format: email. So, an alternative would be to disable this normalization step