refactor: improve language normalization and date handling utilities

- Introduced a new `normalize_lang` function to standardize language codes across the application, ensuring consistent handling of user language preferences.
- Refactored date handling utilities by adding `parse_utc_iso` and `parse_utc_iso_naive` functions for better parsing of ISO 8601 date strings, enhancing timezone awareness.
- Updated various modules to utilize the new language normalization and date parsing functions, improving code clarity and maintainability.
- Enhanced error handling in date validation to raise specific `DateRangeValidationError` exceptions, providing clearer feedback on validation issues.
- Improved test coverage for date range validation and language normalization functionalities, ensuring robustness and reliability.

duty_teller/api/telegram_auth.py

@@ -6,6 +6,8 @@ import json
 import time
 from urllib.parse import unquote
 
+from duty_teller.i18n.lang import normalize_lang
+
 # Telegram algorithm: https://core.telegram.org/bots/webapps#validating-data-received-via-the-mini-app
 # Data-check string: sorted key=value with URL-decoded values, then HMAC-SHA256(WebAppData, token) as secret.
 
@@ -31,14 +33,6 @@ def validate_init_data(
     return username
 
 
-def _normalize_lang(language_code: str | None) -> str:
-    """Normalize to 'ru' or 'en' for i18n."""
-    if not language_code or not isinstance(language_code, str):
-        return "en"
-    code = language_code.strip().lower()
-    return "ru" if code.startswith("ru") else "en"
-
-
 def validate_init_data_with_reason(
     init_data: str,
     bot_token: str,
@@ -107,7 +101,7 @@ def validate_init_data_with_reason(
         return (None, None, "user_invalid", "en")
     if not isinstance(user, dict):
         return (None, None, "user_invalid", "en")
-    lang = _normalize_lang(user.get("language_code"))
+    lang = normalize_lang(user.get("language_code"))
     raw_id = user.get("id")
     if raw_id is None:
         return (None, None, "no_user_id", lang)