Compare commits

...

11 Commits

Author SHA1 Message Date
dgtlmoon
d07cb2a7a8 Merge branch 'master' into url-validation-improvements-2 2025-10-28 12:51:08 +01:00
dgtlmoon
f9cd4da2bb safer 2025-10-28 11:46:00 +01:00
dgtlmoon
4e699cc13b File:// type doesnt go through universal URL validation 2025-10-28 11:44:11 +01:00
dgtlmoon
964302cf3c blah 2025-10-28 11:15:58 +01:00
dgtlmoon
00b31cabd5 Fix test 2025-10-28 10:58:44 +01:00
dgtlmoon
ac13d8cbde refactor and handle spaces etc 2025-10-28 10:38:51 +01:00
dgtlmoon
a037cf7b9a Adding deeper test 2025-10-28 10:27:09 +01:00
dgtlmoon
8a7ea79fb3 Extra XSS checking 2025-10-28 10:23:27 +01:00
dgtlmoon
ec0d7cff21 Merge branch 'master' into url-validation-improvements-2 2025-10-28 10:14:58 +01:00
dgtlmoon
d48c82052a Unify safe URL checking to the one function 2025-10-28 09:59:13 +01:00
dgtlmoon
552e98519b Merge commit from fork 2025-10-28 09:26:17 +01:00
11 changed files with 151 additions and 78 deletions

View File

@@ -1,10 +1,9 @@
import os
from changedetectionio.strtobool import strtobool
from flask_restful import abort, Resource
from flask import request
import validators
from functools import wraps
from . import auth, validate_openapi_request
from ..validate_url import is_safe_valid_url
def default_content_type(content_type='text/plain'):
@@ -50,14 +49,13 @@ class Import(Resource):
urls = request.get_data().decode('utf8').splitlines()
added = []
allow_simplehost = not strtobool(os.getenv('BLOCK_SIMPLEHOSTS', 'False'))
for url in urls:
url = url.strip()
if not len(url):
continue
# If hosts that only contain alphanumerics are allowed ("localhost" for example)
if not validators.url(url, simple_host=allow_simplehost):
if not is_safe_valid_url(url):
return f"Invalid or unsupported URL - {url}", 400
if dedupe and self.datastore.url_exists(url):

View File

@@ -1,14 +1,12 @@
import os
from changedetectionio.strtobool import strtobool
from changedetectionio.html_tools import is_safe_url
from changedetectionio.validate_url import is_safe_valid_url
from flask_expects_json import expects_json
from changedetectionio import queuedWatchMetaData
from changedetectionio import worker_handler
from flask_restful import abort, Resource
from flask import request, make_response, send_from_directory
import validators
from . import auth
import copy
@@ -124,7 +122,7 @@ class Watch(Resource):
return validation_error, 400
# XSS etc protection
if request.json.get('url') and not is_safe_url(request.json.get('url')):
if request.json.get('url') and not is_safe_valid_url(request.json.get('url')):
return "Invalid URL", 400
watch.update(request.json)
@@ -232,9 +230,7 @@ class CreateWatch(Resource):
json_data = request.get_json()
url = json_data['url'].strip()
# If hosts that only contain alphanumerics are allowed ("localhost" for example)
allow_simplehost = not strtobool(os.getenv('BLOCK_SIMPLEHOSTS', 'False'))
if not validators.url(url, simple_host=allow_simplehost):
if not is_safe_valid_url(url):
return "Invalid or unsupported URL", 400
if json_data.get('proxy'):

View File

@@ -133,10 +133,10 @@ def get_socketio_path():
# Socket.IO will be available at {prefix}/socket.io/
return prefix
@app.template_global('is_safe_url')
def _is_safe_url(test_url):
from .html_tools import is_safe_url
return is_safe_url(test_url)
@app.template_global('is_safe_valid_url')
def _is_safe_valid_url(test_url):
from .validate_url import is_safe_valid_url
return is_safe_valid_url(test_url)
@app.template_filter('format_number_locale')
@@ -387,7 +387,7 @@ def changedetection_app(config=None, datastore_o=None):
# We would sometimes get login loop errors on sites hosted in sub-paths
# note for the future:
# if not is_safe_url(next):
# if not is_safe_valid_url(next):
# return flask.abort(400)
return redirect(url_for('watchlist.index'))

View File

@@ -28,11 +28,8 @@ from wtforms.utils import unset_value
from wtforms.validators import ValidationError
from validators.url import url as url_validator
from changedetectionio.widgets import TernaryNoneBooleanField
# default
# each select <option data-enabled="enabled-0-0"
from changedetectionio.blueprint.browser_steps.browser_steps import browser_step_ui_config
@@ -541,19 +538,10 @@ class validateURL(object):
def validate_url(test_url):
# If hosts that only contain alphanumerics are allowed ("localhost" for example)
try:
url_validator(test_url, simple_host=allow_simplehost)
except validators.ValidationError:
#@todo check for xss
message = f"'{test_url}' is not a valid URL."
from changedetectionio.validate_url import is_safe_valid_url
if not is_safe_valid_url(test_url):
# This should be wtforms.validators.
raise ValidationError(message)
from changedetectionio.html_tools import is_safe_url
if not is_safe_url(test_url):
# This should be wtforms.validators.
raise ValidationError('Watch protocol is not permitted by SAFE_PROTOCOL_REGEX or incorrect URL format')
raise ValidationError('Watch protocol is not permitted or invalid URL format')
class ValidateSinglePythonRegexString(object):

View File

@@ -15,8 +15,6 @@ TITLE_RE = re.compile(r"<title[^>]*>(.*?)</title>", re.I | re.S)
META_CS = re.compile(r'<meta[^>]+charset=["\']?\s*([a-z0-9_\-:+.]+)', re.I)
META_CT = re.compile(r'<meta[^>]+http-equiv=["\']?content-type["\']?[^>]*content=["\'][^>]*charset=([a-z0-9_\-:+.]+)', re.I)
SAFE_PROTOCOL_REGEX='^(http|https|ftp|file):'
# 'price' , 'lowPrice', 'highPrice' are usually under here
# All of those may or may not appear on different websites - I didnt find a way todo case-insensitive searching here
LD_JSON_PRODUCT_OFFER_SELECTORS = ["json:$..offers", "json:$..Offers"]
@@ -25,22 +23,6 @@ class JSONNotFound(ValueError):
def __init__(self, msg):
ValueError.__init__(self, msg)
def is_safe_url(test_url):
import os
# See https://github.com/dgtlmoon/changedetection.io/issues/1358
# Remove 'source:' prefix so we dont get 'source:javascript:' etc
# 'source:' is a valid way to tell us to return the source
r = re.compile(re.escape('source:'), re.IGNORECASE)
test_url = r.sub('', test_url)
pattern = re.compile(os.getenv('SAFE_PROTOCOL_REGEX', SAFE_PROTOCOL_REGEX), re.IGNORECASE)
if not pattern.match(test_url.strip()):
return False
return True
# Doesn't look like python supports forward slash auto enclosure in re.findall
# So convert it to inline flag "(?i)foobar" type configuration
@lru_cache(maxsize=100)

View File

@@ -1,5 +1,6 @@
from blinker import signal
from changedetectionio.html_tools import is_safe_url
from changedetectionio.validate_url import is_safe_valid_url
from changedetectionio.strtobool import strtobool
from changedetectionio.jinja2_custom import render as jinja_render
from . import watch_base
@@ -12,9 +13,6 @@ from .. import jinja2_custom as safe_jinja
from ..diff import ADDED_PLACEMARKER_OPEN
from ..html_tools import TRANSLATE_WHITESPACE_TABLE
# Allowable protocols, protects against javascript: etc
# file:// is further checked by ALLOW_FILE_URI
SAFE_PROTOCOL_REGEX='^(http|https|ftp|file):'
FAVICON_RESAVE_THRESHOLD_SECONDS=86400
@@ -63,7 +61,7 @@ class model(watch_base):
def link(self):
url = self.get('url', '')
if not is_safe_url(url):
if not is_safe_valid_url(url):
return 'DISABLED'
ready_url = url
@@ -84,7 +82,7 @@ class model(watch_base):
ready_url=ready_url.replace('source:', '')
# Also double check it after any Jinja2 formatting just incase
if not is_safe_url(ready_url):
if not is_safe_valid_url(ready_url):
return 'DISABLED'
return ready_url

View File

@@ -1,5 +1,6 @@
from changedetectionio.strtobool import strtobool
from changedetectionio.html_tools import is_safe_url
from changedetectionio.validate_url import is_safe_valid_url
from flask import (
flash
@@ -341,8 +342,10 @@ class ChangeDetectionStore:
logger.error(f"Error fetching metadata for shared watch link {url} {str(e)}")
flash("Error fetching metadata for {}".format(url), 'error')
return False
if not is_safe_url(url):
flash('Watch protocol is not permitted by SAFE_PROTOCOL_REGEX', 'error')
if not is_safe_valid_url(url):
flash('Watch protocol is not permitted or invalid URL format', 'error')
return None
if tag and type(tag) == str:

View File

@@ -53,7 +53,7 @@
<a class="pure-menu-heading" href="{{url_for('watchlist.index')}}">
<strong>Change</strong>Detection.io</a>
{% endif %}
{% if current_diff_url and is_safe_url(current_diff_url) %}
{% if current_diff_url and is_safe_valid_url(current_diff_url) %}
<a class="current-diff-url" href="{{ current_diff_url }}">
<span style="max-width: 30%; overflow: hidden">{{ current_diff_url }}</span></a>
{% else %}

View File

@@ -64,29 +64,21 @@ def test_jinja2_time_offset_in_url_query(client, live_server, measure_memory_usa
# Should not have template error
assert b'Invalid template' not in res.data
# https://techtonics.medium.com/secure-templating-with-jinja2-understanding-ssti-and-jinja2-sandbox-environment-b956edd60456
def test_jinja2_security_url_query(client, live_server, measure_memory_usage):
# Add our URL to the import page
test_url = url_for('test_return_query', _external=True)
# because url_for() will URL-encode the var, but we dont here
full_url = "{}?{}".format(test_url,
"date={{ ''.__class__.__mro__[1].__subclasses__()}}", )
full_url = test_url + "?date={{ ''.__class__.__mro__[1].__subclasses__()}}"
res = client.post(
url_for("ui.ui_views.form_quick_watch_add"),
data={"url": full_url, "tags": "test"},
follow_redirects=True
)
assert b"Watch added" in res.data
wait_for_all_checks(client)
assert b"Watch added" not in res.data
# It should report nothing found (no new 'has-unread-changes' class)
res = client.get(url_for("watchlist.index"))
assert b'is invalid and cannot be used' in res.data
# Some of the spewed output from the subclasses
assert b'dict_values' not in res.data
def test_timezone(mocker):
"""Verify that timezone is parsed."""

View File

@@ -25,7 +25,7 @@ def set_original_response():
return None
def test_bad_access(client, live_server, measure_memory_usage):
res = client.post(
url_for("imports.import_page"),
data={"urls": 'https://localhost'},
@@ -48,7 +48,7 @@ def test_bad_access(client, live_server, measure_memory_usage):
follow_redirects=True
)
assert b'Watch protocol is not permitted by SAFE_PROTOCOL_REGEX' in res.data
assert b'Watch protocol is not permitted or invalid URL format' in res.data
res = client.post(
url_for("ui.ui_views.form_quick_watch_add"),
@@ -56,7 +56,7 @@ def test_bad_access(client, live_server, measure_memory_usage):
follow_redirects=True
)
assert b'Watch protocol is not permitted by SAFE_PROTOCOL_REGEX' in res.data
assert b'Watch protocol is not permitted or invalid URL format' in res.data
res = client.post(
url_for("ui.ui_views.form_quick_watch_add"),
@@ -64,7 +64,7 @@ def test_bad_access(client, live_server, measure_memory_usage):
follow_redirects=True
)
assert b'Watch protocol is not permitted by SAFE_PROTOCOL_REGEX' in res.data
assert b'Watch protocol is not permitted or invalid URL format' in res.data
res = client.post(
@@ -73,8 +73,15 @@ def test_bad_access(client, live_server, measure_memory_usage):
follow_redirects=True
)
assert b'Watch protocol is not permitted by SAFE_PROTOCOL_REGEX' in res.data
assert b'Watch protocol is not permitted or invalid URL format' in res.data
res = client.post(
url_for("ui.ui_views.form_quick_watch_add"),
data={"url": 'https://i-wanna-xss-you.com?hereis=<script>alert(1)</script>', "tags": ''},
follow_redirects=True
)
assert b'Watch protocol is not permitted or invalid URL format' in res.data
def _runner_test_various_file_slash(client, file_uri):
@@ -111,8 +118,8 @@ def test_file_slash_access(client, live_server, measure_memory_usage):
test_file_path = os.path.abspath(__file__)
_runner_test_various_file_slash(client, file_uri=f"file://{test_file_path}")
_runner_test_various_file_slash(client, file_uri=f"file:/{test_file_path}")
_runner_test_various_file_slash(client, file_uri=f"file:{test_file_path}") # CVE-2024-56509
# _runner_test_various_file_slash(client, file_uri=f"file:/{test_file_path}")
# _runner_test_various_file_slash(client, file_uri=f"file:{test_file_path}") # CVE-2024-56509
def test_xss(client, live_server, measure_memory_usage):

View File

@@ -0,0 +1,109 @@
from functools import lru_cache
from loguru import logger
from urllib.parse import urlparse, urlunparse, parse_qsl, urlencode
def normalize_url_encoding(url):
"""
Safely encode a URL's query parameters, regardless of whether they're already encoded.
Why this is necessary:
URLs can arrive in various states - some with already encoded query parameters (%20 for spaces),
some with unencoded parameters (literal spaces), or a mix of both. The validators.url() function
requires proper encoding, but simply encoding an already-encoded URL would double-encode it
(e.g., %20 would become %2520).
This function solves the problem by:
1. Parsing the URL to extract query parameters
2. parse_qsl() automatically decodes parameters if they're encoded
3. urlencode() re-encodes them properly
4. Returns a consistently encoded URL that will pass validation
Example:
- Input: "http://example.com/test?time=2025-10-28 09:19" (space not encoded)
- Output: "http://example.com/test?time=2025-10-28+09%3A19" (properly encoded)
- Input: "http://example.com/test?time=2025-10-28%2009:19" (already encoded)
- Output: "http://example.com/test?time=2025-10-28+09%3A19" (properly encoded)
Returns a properly encoded URL string.
"""
try:
# Parse the URL into components (scheme, netloc, path, params, query, fragment)
parsed = urlparse(url)
# Parse query string - this automatically decodes it if encoded
# parse_qsl handles both encoded and unencoded query strings gracefully
query_params = parse_qsl(parsed.query, keep_blank_values=True)
# Re-encode the query string properly using standard URL encoding
encoded_query = urlencode(query_params, safe='')
# Reconstruct the URL with properly encoded query string
normalized = urlunparse((
parsed.scheme,
parsed.netloc,
parsed.path,
parsed.params,
encoded_query, # Use the re-encoded query
parsed.fragment
))
return normalized
except Exception as e:
# If parsing fails for any reason, return original URL
logger.debug(f"URL normalization failed for '{url}': {e}")
return url
@lru_cache(maxsize=10000)
def is_safe_valid_url(test_url):
from changedetectionio import strtobool
from changedetectionio.jinja2_custom import render as jinja_render
import os
import re
import validators
allow_file_access = strtobool(os.getenv('ALLOW_FILE_URI', 'false'))
safe_protocol_regex = '^(http|https|ftp|file):' if allow_file_access else '^(http|https|ftp):'
# See https://github.com/dgtlmoon/changedetection.io/issues/1358
# Remove 'source:' prefix so we dont get 'source:javascript:' etc
# 'source:' is a valid way to tell us to return the source
r = re.compile('^source:', re.IGNORECASE)
test_url = r.sub('', test_url)
# Check the actual rendered URL in case of any Jinja markup
try:
test_url = jinja_render(test_url)
except Exception as e:
logger.error(f'URL "{test_url}" is not correct Jinja2? {str(e)}')
return False
# Check query parameters and fragment
if re.search(r'[<>]', test_url):
logger.warning(f'URL "{test_url}" contains suspicious characters')
return False
# Normalize URL encoding - handle both encoded and unencoded query parameters
test_url = normalize_url_encoding(test_url)
# Be sure the protocol is safe (no file, etcetc)
pattern = re.compile(os.getenv('SAFE_PROTOCOL_REGEX', safe_protocol_regex), re.IGNORECASE)
if not pattern.match(test_url.strip()):
logger.warning(f'URL "{test_url}" is not safe, aborting.')
return False
# If hosts that only contain alphanumerics are allowed ("localhost" for example)
allow_simplehost = not strtobool(os.getenv('BLOCK_SIMPLEHOSTS', 'False'))
try:
if not test_url.strip().lower().startswith('file:') and not validators.url(test_url, simple_host=allow_simplehost):
logger.warning(f'URL "{test_url}" failed validation, aborting.')
return False
except validators.ValidationError:
logger.warning(f'URL f"{test_url}" failed validation, aborting.')
return False
return True