diff options
| -rw-r--r-- | tests/test_auth.py | 8 | ||||
| -rw-r--r-- | tests/test_cli.py | 1 | ||||
| -rw-r--r-- | tests/test_forms.py | 1 | ||||
| -rw-r--r-- | webmentions_ssg/auth.py | 3 | ||||
| -rw-r--r-- | webmentions_ssg/forms.py | 4 | ||||
| -rw-r--r-- | webmentions_ssg/validation.py | 12 |
6 files changed, 25 insertions, 4 deletions
diff --git a/tests/test_auth.py b/tests/test_auth.py index 008382a..c302e8f 100644 --- a/tests/test_auth.py +++ b/tests/test_auth.py @@ -8,10 +8,16 @@ import pytest from flask import Flask from webmentions_ssg import db -from webmentions_ssg.auth import AUTH, is_safe_next_url, load_user +from webmentions_ssg.auth import AUTH, is_safe_next_url, load_user, validate_password from webmentions_ssg.models import User +@pytest.mark.parametrize("character", ["\x00", "\x1f", "\x7f", "\x80", "\x9f"]) +def test_validate_password_rejects_unicode_controls(character: str) -> None: + with pytest.raises(ValueError, match="control characters"): + validate_password(f"valid passphrase{character}") + + def test_user_loader_is_registered(app: Flask) -> None: assert AUTH._user_callback is load_user diff --git a/tests/test_cli.py b/tests/test_cli.py index e6214f2..ac673b4 100644 --- a/tests/test_cli.py +++ b/tests/test_cli.py @@ -608,6 +608,7 @@ def test_users_cli_create_rolls_back_on_duplicate_race(app: Flask) -> None: [ ("short", "at least 8 characters"), ("good passphrase with\tab", "control characters"), + ("good passphrase with\x80c1", "control characters"), ], ) @pytest.mark.parametrize("operation", ["create", "set-password"]) diff --git a/tests/test_forms.py b/tests/test_forms.py index 20e655c..a37a6d0 100644 --- a/tests/test_forms.py +++ b/tests/test_forms.py @@ -222,6 +222,7 @@ def test_endpoint_form_applies_configurable_url_byte_limit_to_both_fields( pytest.param("https://example.com/a b", id="ascii-space"), pytest.param("https://example.com/a\u00a0b", id="unicode-whitespace"), pytest.param("https://example.com/\x7fb", id="del"), + pytest.param("https://example.com/a\x80b", id="c1-control"), ], ) def test_endpoint_rejects_invalid_httpx_urls_and_whitespace( diff --git a/webmentions_ssg/auth.py b/webmentions_ssg/auth.py index 209bdd6..1707917 100644 --- a/webmentions_ssg/auth.py +++ b/webmentions_ssg/auth.py @@ -15,6 +15,7 @@ from flask_login import LoginManager from .extensions import db from .models import User +from .validation import is_control_character AUTH = LoginManager() @@ -38,7 +39,7 @@ def validate_password(password: str) -> None: raise ValueError( f"Password must be at least {MIN_PASSWORD_LENGTH} characters long." ) - if any(ord(char) < 32 or ord(char) == 127 for char in password): + if any(is_control_character(char) for char in password): raise ValueError("Password must not contain control characters.") diff --git a/webmentions_ssg/forms.py b/webmentions_ssg/forms.py index 0d83327..005bd20 100644 --- a/webmentions_ssg/forms.py +++ b/webmentions_ssg/forms.py @@ -5,7 +5,6 @@ """Webmention forms and URL validation.""" import ipaddress -import unicodedata from urllib.parse import urlsplit import httpx @@ -16,6 +15,7 @@ from wtforms.form import BaseForm from wtforms.validators import InputRequired, StopValidation from .net import is_public_address +from .validation import is_control_character class URLMaxBytes: @@ -57,7 +57,7 @@ class HTTPURL: raise StopValidation(field.gettext("Invalid URL.")) if any( - character.isspace() or unicodedata.category(character) == "Cc" + character.isspace() or is_control_character(character) for character in value ): raise StopValidation(field.gettext("Invalid URL.")) diff --git a/webmentions_ssg/validation.py b/webmentions_ssg/validation.py new file mode 100644 index 0000000..e39d4d5 --- /dev/null +++ b/webmentions_ssg/validation.py @@ -0,0 +1,12 @@ +# SPDX-FileCopyrightText: 2026 Dennis Fink <me+coding@dennisfink.me> +# +# SPDX-License-Identifier: BSD-3-Clause + +"""Shared character validation helpers.""" + +import unicodedata + + +def is_control_character(character: str) -> bool: + """Identify Unicode control characters, including C0, DEL, and C1.""" + return unicodedata.category(character) == "Cc" |
