Compare commits

...

14 Commits

Author SHA1 Message Date
dgtlmoon
16bf6bafd1 Oops 2026-02-17 02:24:32 +01:00
dgtlmoon
449b2c2dc3 Merge branch 'master' into fixing-signals-update 2026-02-17 02:19:29 +01:00
dgtlmoon
a53f2a784d Pluggy plugin hook for before and after a watch is processed (#3888) 2026-02-17 01:58:41 +01:00
dgtlmoon
dd2c8dfd2b Fixing socket updates for status updates. 2026-02-17 01:57:24 +01:00
dgtlmoon
7558ca5fda 0.53.3 2026-02-16 20:41:07 +01:00
dgtlmoon
383c3b427f API - Adding automated test for API with NGINX sub-path, Skip validation errors about server path (allows use on sub-paths/reverse proxy etc) (#3886) 2026-02-16 20:32:35 +01:00
dgtlmoon
b01ba5d8a1 UI - Use version from code in version tab
Some checks failed
Build and push containers / metadata (push) Has been cancelled
Build and push containers / build-push-containers (push) Has been cancelled
Publish Python 🐍distribution 📦 to PyPI and TestPyPI / Build distribution 📦 (push) Has been cancelled
Publish Python 🐍distribution 📦 to PyPI and TestPyPI / Test the built package works basically. (push) Has been cancelled
Publish Python 🐍distribution 📦 to PyPI and TestPyPI / Publish Python 🐍 distribution 📦 to PyPI (push) Has been cancelled
ChangeDetection.io App Test / lint-code (push) Has been cancelled
ChangeDetection.io App Test / test-application-3-10 (push) Has been cancelled
ChangeDetection.io App Test / test-application-3-11 (push) Has been cancelled
ChangeDetection.io App Test / test-application-3-12 (push) Has been cancelled
ChangeDetection.io App Test / test-application-3-13 (push) Has been cancelled
2026-02-16 19:41:27 +01:00
dgtlmoon
86e5184cef 0.53.2 2026-02-16 18:52:31 +01:00
dgtlmoon
1dbf1f5db5 UI - Watch overview - Restock price, validate number before output (#3883) 2026-02-16 18:50:37 +01:00
dgtlmoon
c5bd7da647 Security - Adding small test and fixing overzealous filename cleaner (#3884) 2026-02-16 18:31:25 +01:00
dgtlmoon
549e167746 Datastore - On fresh installs, also scan for existing watch.json watches in subdirectories 2026-02-16 15:56:46 +01:00
dgtlmoon
9d38b45173 Security CVE-2026-25527 - Unauthenticated static path traversal in resources 2026-02-16 15:48:03 +01:00
dgtlmoon
3558e9ee10 Browser Steps - Minor code cleanup 2026-02-16 13:22:54 +01:00
dgtlmoon
4b94de7e0c UI - Browser Steps - First step was missing Clear / Remove / Pic buttons 2026-02-16 13:20:34 +01:00
24 changed files with 568 additions and 72 deletions

33
.github/nginx-reverse-proxy-test.conf vendored Normal file
View File

@@ -0,0 +1,33 @@
server {
listen 80;
server_name localhost;
# Test basic reverse proxy to changedetection.io
location / {
proxy_pass http://changedet-app:5000;
proxy_set_header Host $host;
proxy_set_header X-Real-IP $remote_addr;
proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
proxy_set_header X-Forwarded-Proto $scheme;
# WebSocket support
proxy_http_version 1.1;
proxy_set_header Upgrade $http_upgrade;
proxy_set_header Connection "upgrade";
}
# Test subpath deployment with X-Forwarded-Prefix
location /changedet-sub/ {
proxy_pass http://changedet-app:5000/;
proxy_set_header X-Forwarded-Prefix /changedet-sub;
proxy_set_header Host $host;
proxy_set_header X-Real-IP $remote_addr;
proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
proxy_set_header X-Forwarded-Proto $scheme;
# WebSocket support
proxy_http_version 1.1;
proxy_set_header Upgrade $http_upgrade;
proxy_set_header Connection "upgrade";
}
}

View File

@@ -324,6 +324,175 @@ jobs:
run: |
docker run --rm --network changedet-network test-changedetectionio bash -c 'cd changedetectionio;pytest tests/smtp/test_notification_smtp.py'
nginx-reverse-proxy:
runs-on: ubuntu-latest
needs: build
timeout-minutes: 10
env:
PYTHON_VERSION: ${{ inputs.python-version }}
steps:
- uses: actions/checkout@v6
- name: Download Docker image artifact
uses: actions/download-artifact@v7
with:
name: test-changedetectionio-${{ env.PYTHON_VERSION }}
path: /tmp
- name: Load Docker image
run: |
docker load -i /tmp/test-changedetectionio.tar
- name: Spin up services
run: |
docker network create changedet-network
# Start changedetection.io container with X-Forwarded headers support
docker run --name changedet-app --hostname changedet-app --network changedet-network \
-e USE_X_SETTINGS=true \
-d test-changedetectionio
sleep 3
- name: Start nginx reverse proxy
run: |
# Start nginx with our test configuration
docker run --name nginx-proxy --network changedet-network -d -p 8080:80 --rm \
-v ${{ github.workspace }}/.github/nginx-reverse-proxy-test.conf:/etc/nginx/conf.d/default.conf:ro \
nginx:alpine
sleep 2
- name: Test reverse proxy - root path
run: |
echo "=== Testing nginx reverse proxy at root path ==="
curl --retry-connrefused --retry 6 -s http://localhost:8080/ > /tmp/nginx-test-root.html
# Check for changedetection.io UI elements
if grep -q "checkbox-uuid" /tmp/nginx-test-root.html; then
echo "✓ Found checkbox-uuid in response"
else
echo "ERROR: checkbox-uuid not found in response"
cat /tmp/nginx-test-root.html
exit 1
fi
# Check for watchlist content
if grep -q -i "watch" /tmp/nginx-test-root.html; then
echo "✓ Found watch/watchlist content in response"
else
echo "ERROR: watchlist content not found"
cat /tmp/nginx-test-root.html
exit 1
fi
echo "✓ Root path reverse proxy working correctly"
- name: Test reverse proxy - subpath with X-Forwarded-Prefix
run: |
echo "=== Testing nginx reverse proxy at subpath /changedet-sub/ ==="
curl --retry-connrefused --retry 6 -s http://localhost:8080/changedet-sub/ > /tmp/nginx-test-subpath.html
# Check for changedetection.io UI elements
if grep -q "checkbox-uuid" /tmp/nginx-test-subpath.html; then
echo "✓ Found checkbox-uuid in subpath response"
else
echo "ERROR: checkbox-uuid not found in subpath response"
cat /tmp/nginx-test-subpath.html
exit 1
fi
echo "✓ Subpath reverse proxy working correctly"
- name: Test API through reverse proxy subpath
run: |
echo "=== Testing API endpoints through nginx subpath /changedet-sub/ ==="
# Extract API key from the changedetection.io datastore
API_KEY=$(docker exec changedet-app cat /datastore/changedetection.json | grep -o '"api_access_token": *"[^"]*"' | cut -d'"' -f4)
if [ -z "$API_KEY" ]; then
echo "ERROR: Could not extract API key from datastore"
docker exec changedet-app cat /datastore/changedetection.json
exit 1
fi
echo "✓ Extracted API key: ${API_KEY:0:8}..."
# Create a watch via API through nginx proxy subpath
echo "Creating watch via POST to /changedet-sub/api/v1/watch"
RESPONSE=$(curl -s -w "\n%{http_code}" -X POST "http://localhost:8080/changedet-sub/api/v1/watch" \
-H "x-api-key: ${API_KEY}" \
-H "Content-Type: application/json" \
-d '{
"url": "https://example.com/test-nginx-proxy",
"tag": "nginx-test"
}')
HTTP_CODE=$(echo "$RESPONSE" | tail -n1)
BODY=$(echo "$RESPONSE" | head -n-1)
if [ "$HTTP_CODE" != "201" ]; then
echo "ERROR: Expected HTTP 201, got $HTTP_CODE"
echo "Response: $BODY"
exit 1
fi
echo "✓ Watch created successfully (HTTP 201)"
# Extract the watch UUID from response
WATCH_UUID=$(echo "$BODY" | grep -o '"uuid": *"[^"]*"' | cut -d'"' -f4)
echo "✓ Watch UUID: $WATCH_UUID"
# Update the watch via PUT through nginx proxy subpath
echo "Updating watch via PUT to /changedet-sub/api/v1/watch/${WATCH_UUID}"
RESPONSE=$(curl -s -w "\n%{http_code}" -X PUT "http://localhost:8080/changedet-sub/api/v1/watch/${WATCH_UUID}" \
-H "x-api-key: ${API_KEY}" \
-H "Content-Type: application/json" \
-d '{
"paused": true
}')
HTTP_CODE=$(echo "$RESPONSE" | tail -n1)
BODY=$(echo "$RESPONSE" | head -n-1)
if [ "$HTTP_CODE" != "200" ]; then
echo "ERROR: Expected HTTP 200, got $HTTP_CODE"
echo "Response: $BODY"
exit 1
fi
if echo "$BODY" | grep -q 'OK'; then
echo "✓ Watch updated successfully (HTTP 200, response: OK)"
else
echo "ERROR: Expected response 'OK', got: $BODY"
echo "Response: $BODY"
exit 1
fi
# Verify the watch is paused via GET
echo "Verifying watch is paused via GET"
RESPONSE=$(curl -s "http://localhost:8080/changedet-sub/api/v1/watch/${WATCH_UUID}" \
-H "x-api-key: ${API_KEY}")
if echo "$RESPONSE" | grep -q '"paused": *true'; then
echo "✓ Watch is paused as expected"
else
echo "ERROR: Watch paused state not confirmed"
echo "Response: $RESPONSE"
exit 1
fi
echo "✓ API tests through nginx subpath completed successfully"
- name: Cleanup nginx test
if: always()
run: |
docker logs nginx-proxy || true
docker logs changedet-app || true
docker stop nginx-proxy changedet-app || true
docker rm nginx-proxy changedet-app || true
# Proxy tests
proxy-tests:
runs-on: ubuntu-latest

View File

@@ -2,7 +2,7 @@
# Read more https://github.com/dgtlmoon/changedetection.io/wiki
# Semver means never use .01, or 00. Should be .1.
__version__ = '0.53.1'
__version__ = '0.53.3'
from changedetectionio.strtobool import strtobool
from json.decoder import JSONDecodeError
@@ -610,7 +610,7 @@ def main():
@app.context_processor
def inject_template_globals():
return dict(right_sticky="v{}".format(datastore.data['version_tag']),
return dict(right_sticky="v"+__version__,
new_version_available=app.config['NEW_VERSION_AVAILABLE'],
has_password=datastore.data['settings']['application']['password'] != False,
socket_io_enabled=datastore.data['settings']['application'].get('ui', {}).get('socket_io_enabled', True),

View File

@@ -103,6 +103,7 @@ def validate_openapi_request(operation_id):
if request.method.upper() != 'GET':
# Lazy import - only loaded when actually validating a request
from openapi_core.contrib.flask import FlaskOpenAPIRequest
from openapi_core.templating.paths.exceptions import ServerNotFound, PathNotFound, PathError
spec = get_openapi_spec()
openapi_request = FlaskOpenAPIRequest(request)
@@ -110,6 +111,16 @@ def validate_openapi_request(operation_id):
if result.errors:
error_details = []
for error in result.errors:
# Skip path/server validation errors for reverse proxy compatibility
# Flask routing already validates that endpoints exist (returns 404 if not).
# OpenAPI validation here is primarily for request body schema validation.
# When behind nginx/reverse proxy, URLs may have path prefixes that don't
# match the OpenAPI server definitions, causing false positives.
if isinstance(error, PathError):
logger.debug(f"API Call - Skipping path/server validation (delegated to Flask): {error}")
continue
error_str = str(error)
# Extract detailed schema errors from __cause__
if hasattr(error, '__cause__') and hasattr(error.__cause__, 'schema_errors'):
for schema_error in error.__cause__.schema_errors:
@@ -117,9 +128,12 @@ def validate_openapi_request(operation_id):
msg = schema_error.message if hasattr(schema_error, 'message') else str(schema_error)
error_details.append(f"{field}: {msg}")
else:
error_details.append(str(error))
error_details.append(error_str)
# Only raise if we have actual validation errors (not path/server issues)
if error_details:
logger.error(f"API Call - Validation failed: {'; '.join(error_details)}")
raise BadRequest(f"Validation failed: {'; '.join(error_details)}")
raise BadRequest(f"Validation failed: {'; '.join(error_details)}")
except BadRequest:
# Re-raise BadRequest exceptions (validation failures)
raise

View File

@@ -379,6 +379,4 @@ def construct_blueprint(datastore: ChangeDetectionStore):
return browser_steps_blueprint
return browser_steps_blueprint

View File

@@ -194,9 +194,9 @@ def construct_blueprint(datastore: ChangeDetectionStore, update_q, worker_pool,
tag_limit = request.args.get('tag')
now = int(time.time())
# Mark watches as viewed in background thread to avoid blocking
def mark_viewed_background():
"""Background thread to mark watches as viewed - discarded after completion."""
# Mark watches as viewed - use background thread only for large watch counts
def mark_viewed_impl():
"""Mark watches as viewed - can run synchronously or in background thread."""
marked_count = 0
try:
for watch_uuid, watch in datastore.data['watching'].items():
@@ -209,15 +209,21 @@ def construct_blueprint(datastore: ChangeDetectionStore, update_q, worker_pool,
datastore.set_last_viewed(watch_uuid, now)
marked_count += 1
logger.info(f"Background marking complete: {marked_count} watches marked as viewed")
logger.info(f"Marking complete: {marked_count} watches marked as viewed")
except Exception as e:
logger.error(f"Error in background mark as viewed: {e}")
logger.error(f"Error marking as viewed: {e}")
# Start background thread and return immediately
thread = threading.Thread(target=mark_viewed_background, daemon=True)
thread.start()
# For small watch counts (< 10), run synchronously to avoid race conditions in tests
# For larger counts, use background thread to avoid blocking the UI
watch_count = len(datastore.data['watching'])
if watch_count < 10:
# Run synchronously for small watch counts
mark_viewed_impl()
else:
# Start background thread for large watch counts
thread = threading.Thread(target=mark_viewed_impl, daemon=True)
thread.start()
flash(gettext("Marking watches as viewed in background..."))
return redirect(url_for('watchlist.index', tag=tag_limit))
@ui_blueprint.route("/delete", methods=['GET'])

View File

@@ -304,12 +304,13 @@ html[data-darkmode="true"] .watch-tag-list.tag-{{ class_name }} {
</span>
{%- endif -%}
{%- if watch.get('restock') and watch['restock']['price'] != None -%}
{%- if watch['restock']['price'] != None -%}
{%- if watch.get('restock') and watch['restock'].get('price') -%}
{%- if watch['restock']['price'] is number -%}
<span class="restock-label price" title="{{ _('Price') }}">
{{ watch['restock']['price']|format_number_locale if watch['restock'].get('price') else '' }} {{ watch['restock'].get('currency','') }}
</span>
{%- endif -%}
{%- else -%} <!-- watch['restock']['price']' is not a number, cant output it -->
{%- endif -%}
{%- elif not watch.has_restock_info -%}
<span class="restock-label error">{{ _('No information') }}</span>
{%- endif -%}

View File

@@ -712,8 +712,14 @@ def changedetection_app(config=None, datastore_o=None):
def static_content(group, filename):
from flask import make_response
import re
group = re.sub(r'[^\w.-]+', '', group.lower())
filename = re.sub(r'[^\w.-]+', '', filename.lower())
# Strict sanitization: only allow a-z, 0-9, and underscore (blocks .. and other traversal)
group = re.sub(r'[^a-z0-9_-]+', '', group.lower())
filename = filename
# Additional safety: reject if sanitization resulted in empty strings
if not group or not filename:
abort(404)
if group == 'screenshot':
# Could be sensitive, follow password requirements

View File

@@ -129,6 +129,51 @@ class ChangeDetectionSpec:
"""
pass
@hookspec
def update_handler_alter(update_handler, watch, datastore):
"""Modify or wrap the update_handler before it processes a watch.
This hook is called after the update_handler (perform_site_check instance) is created
but before it calls call_browser() and run_changedetection(). Plugins can use this to:
- Wrap the handler to add logging/metrics
- Modify handler configuration
- Add custom preprocessing logic
Args:
update_handler: The perform_site_check instance that will process the watch
watch: The watch dict being processed
datastore: The application datastore
Returns:
object or None: Return a modified/wrapped handler, or None to keep the original.
If multiple plugins return handlers, they are chained in registration order.
"""
pass
@hookspec
def update_finalize(update_handler, watch, datastore, processing_exception):
"""Called after watch processing completes (success or failure).
This hook is called in the finally block after all processing is complete,
allowing plugins to perform cleanup, update metrics, or log final status.
The plugin can access update_handler.last_logging_insert_id if it was stored
during update_handler_alter, and use processing_exception to determine if
the processing succeeded or failed.
Args:
update_handler: The perform_site_check instance (may be None if creation failed)
watch: The watch dict that was processed (may be None if not loaded)
datastore: The application datastore
processing_exception: The exception from the main processing block, or None if successful.
This does NOT include cleanup exceptions - only exceptions from
the actual watch processing (fetch, diff, etc).
Returns:
None: This hook doesn't return a value
"""
pass
# Set up Plugin Manager
plugin_manager = pluggy.PluginManager(PLUGIN_NAMESPACE)
@@ -499,4 +544,66 @@ def get_plugin_template_paths():
template_paths.append(templates_dir)
logger.debug(f"Added plugin template path: {templates_dir}")
return template_paths
return template_paths
def apply_update_handler_alter(update_handler, watch, datastore):
"""Apply update_handler_alter hooks from all plugins.
Allows plugins to wrap or modify the update_handler before it processes a watch.
Multiple plugins can chain modifications - each plugin receives the result from
the previous plugin.
Args:
update_handler: The perform_site_check instance to potentially modify
watch: The watch dict being processed
datastore: The application datastore
Returns:
object: The (potentially modified/wrapped) update_handler
"""
# Get all plugins that implement the update_handler_alter hook
results = plugin_manager.hook.update_handler_alter(
update_handler=update_handler,
watch=watch,
datastore=datastore
)
# Chain results - each plugin gets the result from the previous one
current_handler = update_handler
if results:
for result in results:
if result is not None:
logger.debug(f"Plugin modified update_handler for watch {watch.get('uuid')}")
current_handler = result
return current_handler
def apply_update_finalize(update_handler, watch, datastore, processing_exception):
"""Apply update_finalize hooks from all plugins.
Called in the finally block after watch processing completes, allowing plugins
to perform cleanup, update metrics, or log final status.
Args:
update_handler: The perform_site_check instance (may be None)
watch: The watch dict that was processed (may be None)
datastore: The application datastore
processing_exception: The exception from processing, or None if successful
Returns:
None
"""
try:
# Call all plugins that implement the update_finalize hook
plugin_manager.hook.update_finalize(
update_handler=update_handler,
watch=watch,
datastore=datastore,
processing_exception=processing_exception
)
except Exception as e:
# Don't let plugin errors crash the worker
logger.error(f"Error in update_finalize hook: {e}")
logger.exception(f"update_finalize hook exception details:")

View File

@@ -360,16 +360,13 @@ $(document).ready(function () {
// Add the extra buttons to the steps
$('ul#browser_steps li').each(function (i) {
var s = '<div class="control">' + '<a data-step-index=' + i + ' class="pure-button button-secondary button-green button-xsmall apply" >Apply</a>&nbsp;';
if (i > 0) {
// The first step never gets these (Goto-site)
s += `<a data-step-index="${i}" class="pure-button button-secondary button-xsmall clear" >Clear</a>&nbsp;` +
`<a data-step-index="${i}" class="pure-button button-secondary button-red button-xsmall remove" >Remove</a>`;
s += `<a data-step-index="${i}" class="pure-button button-secondary button-xsmall clear" >Clear</a>&nbsp;` +
`<a data-step-index="${i}" class="pure-button button-secondary button-red button-xsmall remove" >Remove</a>`;
// if a screenshot is available
if (browser_steps_available_screenshots.includes(i.toString())) {
var d = (browser_steps_last_error_step === i+1) ? 'before' : 'after';
s += `&nbsp;<a data-step-index="${i}" class="pure-button button-secondary button-xsmall show-screenshot" title="Show screenshot from last run" data-type="${d}">Pic</a>&nbsp;`;
}
// if a screenshot is available
if (browser_steps_available_screenshots.includes(i.toString())) {
var d = (browser_steps_last_error_step === i+1) ? 'before' : 'after';
s += `&nbsp;<a data-step-index="${i}" class="pure-button button-secondary button-xsmall show-screenshot" title="Show screenshot from last run" data-type="${d}">Pic</a>&nbsp;`;
}
s += '</div>';
$(this).append(s)

View File

@@ -235,6 +235,8 @@ class ChangeDetectionStore(DatastoreUpdatesMixin, FileSavingDataStore):
# No datastore yet - check if this is a fresh install or legacy migration
self.init_fresh_install(include_default_watches=include_default_watches,
version_tag=version_tag)
# Maybe they copied a bunch of watch subdirs across too
self._load_state()
def init_fresh_install(self, include_default_watches, version_tag):
# Generate app_guid FIRST (required for all operations)

View File

@@ -331,6 +331,7 @@ def prepare_test_function(live_server, datastore_path):
# Cleanup: Clear watches and queue after test
try:
from changedetectionio.flask_app import update_q
from pathlib import Path
# Clear the queue to prevent leakage to next test
while not update_q.empty():
@@ -340,6 +341,18 @@ def prepare_test_function(live_server, datastore_path):
break
datastore.data['watching'] = {}
# Delete any old watch metadata JSON files
base_path = Path(datastore.datastore_path).resolve()
max_depth = 2
for file in base_path.rglob("*.json"):
# Calculate depth relative to base path
depth = len(file.relative_to(base_path).parts) - 1
if depth <= max_depth and file.is_file():
file.unlink()
except Exception as e:
logger.warning(f"Error during datastore cleanup: {e}")

View File

@@ -9,7 +9,7 @@ by testing various scenarios that should trigger validation errors.
import time
import json
from flask import url_for
from .util import live_server_setup, wait_for_all_checks
from .util import live_server_setup, wait_for_all_checks, delete_all_watches
def test_openapi_validation_invalid_content_type_on_create_watch(client, live_server, measure_memory_usage, datastore_path):
@@ -27,6 +27,7 @@ def test_openapi_validation_invalid_content_type_on_create_watch(client, live_se
# Should get 400 error due to OpenAPI validation failure
assert res.status_code == 400, f"Expected 400 but got {res.status_code}"
assert b"Validation failed" in res.data, "Should contain validation error message"
delete_all_watches(client)
def test_openapi_validation_missing_required_field_create_watch(client, live_server, measure_memory_usage, datastore_path):
@@ -44,6 +45,7 @@ def test_openapi_validation_missing_required_field_create_watch(client, live_ser
# Should get 400 error due to missing required field
assert res.status_code == 400, f"Expected 400 but got {res.status_code}"
assert b"Validation failed" in res.data, "Should contain validation error message"
delete_all_watches(client)
def test_openapi_validation_invalid_field_in_request_body(client, live_server, measure_memory_usage, datastore_path):
@@ -83,6 +85,7 @@ def test_openapi_validation_invalid_field_in_request_body(client, live_server, m
# Backend validation now returns "Unknown field(s):" message
assert b"Unknown field" in res.data, \
"Should contain validation error about unknown fields"
delete_all_watches(client)
def test_openapi_validation_import_wrong_content_type(client, live_server, measure_memory_usage, datastore_path):
@@ -100,6 +103,7 @@ def test_openapi_validation_import_wrong_content_type(client, live_server, measu
# Should get 400 error due to content-type mismatch
assert res.status_code == 400, f"Expected 400 but got {res.status_code}"
assert b"Validation failed" in res.data, "Should contain validation error message"
delete_all_watches(client)
def test_openapi_validation_import_correct_content_type_succeeds(client, live_server, measure_memory_usage, datastore_path):
@@ -117,6 +121,7 @@ def test_openapi_validation_import_correct_content_type_succeeds(client, live_se
# Should succeed
assert res.status_code == 200, f"Expected 200 but got {res.status_code}"
assert len(res.json) == 2, "Should import 2 URLs"
delete_all_watches(client)
def test_openapi_validation_get_requests_bypass_validation(client, live_server, measure_memory_usage, datastore_path):
@@ -141,6 +146,7 @@ def test_openapi_validation_get_requests_bypass_validation(client, live_server,
# Should return JSON with watch list (empty in this case)
assert isinstance(res.json, dict), "Should return JSON dictionary for watch list"
delete_all_watches(client)
def test_openapi_validation_create_tag_missing_required_title(client, live_server, measure_memory_usage, datastore_path):
@@ -158,10 +164,13 @@ def test_openapi_validation_create_tag_missing_required_title(client, live_serve
# Should get 400 error due to missing required field
assert res.status_code == 400, f"Expected 400 but got {res.status_code}"
assert b"Validation failed" in res.data, "Should contain validation error message"
delete_all_watches(client)
def test_openapi_validation_watch_update_allows_partial_updates(client, live_server, measure_memory_usage, datastore_path):
"""Test that watch updates allow partial updates without requiring all fields (positive test)."""
#xxx
api_key = live_server.app.config['DATASTORE'].data['settings']['application'].get('api_access_token')
# First create a valid watch
@@ -198,4 +207,5 @@ def test_openapi_validation_watch_update_allows_partial_updates(client, live_ser
)
assert res.status_code == 200
assert res.json.get('title') == 'Updated Title Only', "Title should be updated"
assert res.json.get('url') == 'https://example.com', "URL should remain unchanged"
assert res.json.get('url') == 'https://example.com', "URL should remain unchanged"
delete_all_watches(client)

View File

@@ -6,8 +6,6 @@ from flask import url_for
from .util import set_original_response, set_modified_response, live_server_setup, wait_for_all_checks, extract_rss_token_from_UI, \
extract_UUID_from_client, delete_all_watches
sleep_time_for_fetch_thread = 3
# Basic test to check inscriptus is not adding return line chars, basically works etc
def test_inscriptus():

View File

@@ -6,10 +6,6 @@ from urllib.request import urlopen
from .util import set_original_response, set_modified_response, live_server_setup, wait_for_all_checks
import os
sleep_time_for_fetch_thread = 3
def test_check_extract_text_from_diff(client, live_server, measure_memory_usage, datastore_path):
import time
with open(os.path.join(datastore_path, "endpoint-content.txt"), "w") as f:

View File

@@ -41,7 +41,6 @@ def set_modified_ignore_response(datastore_path):
def test_render_anchor_tag_content_true(client, live_server, measure_memory_usage, datastore_path):
"""Testing that the link changes are detected when
render_anchor_tag_content setting is set to true"""
sleep_time_for_fetch_thread = 3
# Give the endpoint time to spin up
time.sleep(1)

View File

@@ -100,7 +100,6 @@ def test_normal_page_check_works_with_ignore_status_code(client, live_server, me
# Tests the whole stack works with staus codes ignored
def test_403_page_check_works_with_ignore_status_code(client, live_server, measure_memory_usage, datastore_path):
sleep_time_for_fetch_thread = 3
set_original_response(datastore_path=datastore_path)
@@ -112,8 +111,7 @@ def test_403_page_check_works_with_ignore_status_code(client, live_server, measu
uuid = client.application.config.get('DATASTORE').add_watch(url=test_url)
client.get(url_for("ui.form_watch_checknow"), follow_redirects=True)
# Give the thread time to pick it up
time.sleep(sleep_time_for_fetch_thread)
wait_for_all_checks(client)
# Goto the edit page, check our ignore option
# Add our URL to the import page

View File

@@ -2,10 +2,9 @@
import time
from flask import url_for
from . util import live_server_setup
import os
from .util import live_server_setup, delete_all_watches, wait_for_all_checks
# Should be the same as set_original_ignore_response(datastore_path=datastore_path) but with a little more whitespacing
@@ -50,10 +49,7 @@ def set_original_ignore_response(datastore_path):
# If there was only a change in the whitespacing, then we shouldnt have a change detected
def test_check_ignore_whitespace(client, live_server, measure_memory_usage, datastore_path):
sleep_time_for_fetch_thread = 3
# Give the endpoint time to spin up
time.sleep(1)
set_original_ignore_response(datastore_path=datastore_path)
@@ -74,17 +70,17 @@ def test_check_ignore_whitespace(client, live_server, measure_memory_usage, data
uuid = client.application.config.get('DATASTORE').add_watch(url=test_url)
client.get(url_for("ui.form_watch_checknow"), follow_redirects=True)
time.sleep(sleep_time_for_fetch_thread)
wait_for_all_checks(client)
# Trigger a check
client.get(url_for("ui.form_watch_checknow"), follow_redirects=True)
set_original_ignore_response_but_with_whitespace(datastore_path)
time.sleep(sleep_time_for_fetch_thread)
wait_for_all_checks(client)
# Trigger a check
client.get(url_for("ui.form_watch_checknow"), follow_redirects=True)
# Give the thread time to pick it up
time.sleep(sleep_time_for_fetch_thread)
wait_for_all_checks(client)
# It should report nothing found (no new 'has-unread-changes' class)
res = client.get(url_for("watchlist.index"))

View File

@@ -24,6 +24,30 @@ def set_original_response(datastore_path):
f.write(test_return_data)
return None
def test_favicon(client, live_server, measure_memory_usage, datastore_path):
# Attempt to fetch it, make sure that works
SVG_BASE64 = 'PHN2ZyB4bWxucz0iaHR0cDovL3d3dy53My5vcmcvMjAwMC9zdmciIHZpZXdCb3g9IjAgMCAxIDEiLz4='
uuid = client.application.config.get('DATASTORE').add_watch(url='https://localhost')
live_server.app.config['DATASTORE'].data['watching'][uuid].bump_favicon(url="favicon-set-type.svg",
favicon_base_64=SVG_BASE64
)
res = client.get(url_for('static_content', group='favicon', filename=uuid))
assert res.status_code == 200
assert len(res.data) > 10
res = client.get(url_for('static_content', group='..', filename='__init__.py'))
assert res.status_code != 200
res = client.get(url_for('static_content', group='.', filename='../__init__.py'))
assert res.status_code != 200
# Traverse by filename protection
res = client.get(url_for('static_content', group='js', filename='../styles/styles.css'))
assert res.status_code != 200
def test_bad_access(client, live_server, measure_memory_usage, datastore_path):
res = client.post(
@@ -478,3 +502,80 @@ def test_logout_with_redirect(client, live_server, measure_memory_usage, datasto
# Cleanup
del client.application.config['DATASTORE'].data['settings']['application']['password']
def test_static_directory_traversal(client, live_server, measure_memory_usage, datastore_path):
"""
Test that the static file serving route properly blocks directory traversal attempts.
This tests the fix for GHSA-9jj8-v89v-xjvw (CVE pending).
The vulnerability was in /static/<group>/<filename> where the sanitization regex
allowed dots, enabling "../" traversal to read application source files.
The fix changed the regex from r'[^\w.-]+' to r'[^a-z0-9_]+' which blocks dots.
"""
# Test 1: Direct .. traversal attempt (URL-encoded)
res = client.get(
"/static/%2e%2e/flask_app.py",
follow_redirects=False
)
# Should be blocked (404 or 403)
assert res.status_code in [404, 403], f"Expected 404/403, got {res.status_code}"
# Should NOT contain application source code
assert b"def static_content" not in res.data
assert b"changedetection_app" not in res.data
# Test 2: Direct .. traversal attempt (unencoded)
res = client.get(
"/static/../flask_app.py",
follow_redirects=False
)
assert res.status_code in [404, 403], f"Expected 404/403, got {res.status_code}"
assert b"def static_content" not in res.data
# Test 3: Multiple dots traversal
res = client.get(
"/static/..../flask_app.py",
follow_redirects=False
)
assert res.status_code in [404, 403], f"Expected 404/403, got {res.status_code}"
assert b"def static_content" not in res.data
# Test 4: Try to access other application files
for filename in ["__init__.py", "datastore.py", "store.py"]:
res = client.get(
f"/static/%2e%2e/{filename}",
follow_redirects=False
)
assert res.status_code in [404, 403], f"File {filename} should be blocked"
# Should not contain Python code indicators
assert b"import" not in res.data or b"# Test" in res.data # Allow "1 Imported" etc
# Test 5: Verify legitimate static files still work
# Note: We can't test actual files without knowing what exists,
# but we can verify the sanitization doesn't break valid groups
res = client.get(
"/static/images/test.png", # Will 404 if file doesn't exist, but won't traverse
follow_redirects=False
)
# Should get 404 (file not found) not 403 (blocked)
# This confirms the group name "images" is valid
assert res.status_code == 404
# Test 6: Ensure hyphens and dots are blocked in group names
res = client.get(
"/static/../../../etc/passwd",
follow_redirects=False
)
assert res.status_code in [404, 403]
assert b"root:" not in res.data
# Test 7: Test that underscores still work (they're allowed)
res = client.get(
"/static/visual_selector_data/test.json",
follow_redirects=False
)
# visual_selector_data is a real group, but requires auth
# Should get 403 (not authenticated) or 404 (file not found), not a path traversal
assert res.status_code in [403, 404]

View File

@@ -6,9 +6,6 @@ from urllib.request import urlopen
from .util import set_original_response, set_modified_response, live_server_setup, delete_all_watches
import re
sleep_time_for_fetch_thread = 3
def test_share_watch(client, live_server, measure_memory_usage, datastore_path):
set_original_response(datastore_path=datastore_path)

View File

@@ -6,7 +6,6 @@ from urllib.request import urlopen
from .util import set_original_response, set_modified_response, live_server_setup, wait_for_all_checks
from ..diff import ADDED_STYLE
sleep_time_for_fetch_thread = 3
def test_check_basic_change_detection_functionality_source(client, live_server, measure_memory_usage, datastore_path):
set_original_response(datastore_path=datastore_path)
@@ -72,7 +71,10 @@ def test_check_ignore_elements(client, live_server, measure_memory_usage, datast
follow_redirects=True
)
time.sleep(sleep_time_for_fetch_thread)
client.get(url_for("ui.form_watch_checknow"), follow_redirects=True)
wait_for_all_checks(client)
res = client.get(
url_for("ui.ui_preview.preview_page", uuid="first"),

View File

@@ -2,7 +2,8 @@
import time
from flask import url_for
from . util import live_server_setup, delete_all_watches
from .util import live_server_setup, delete_all_watches, wait_for_all_checks
import os
@@ -25,9 +26,6 @@ def set_original_ignore_response(datastore_path):
def test_trigger_regex_functionality_with_filter(client, live_server, measure_memory_usage, datastore_path):
# live_server_setup(live_server) # Setup on conftest per function
sleep_time_for_fetch_thread = 3
set_original_ignore_response(datastore_path=datastore_path)
# Give the endpoint time to spin up
@@ -38,8 +36,7 @@ def test_trigger_regex_functionality_with_filter(client, live_server, measure_me
uuid = client.application.config.get('DATASTORE').add_watch(url=test_url)
client.get(url_for("ui.form_watch_checknow"), follow_redirects=True)
# it needs time to save the original version
time.sleep(sleep_time_for_fetch_thread)
wait_for_all_checks(client)
### test regex with filter
res = client.post(
@@ -52,8 +49,9 @@ def test_trigger_regex_functionality_with_filter(client, live_server, measure_me
follow_redirects=True
)
# Give the thread time to pick it up
time.sleep(sleep_time_for_fetch_thread)
client.get(url_for("ui.form_watch_checknow"), follow_redirects=True)
wait_for_all_checks(client)
client.get(url_for("ui.ui_diff.diff_history_page", uuid="first"))
@@ -62,7 +60,8 @@ def test_trigger_regex_functionality_with_filter(client, live_server, measure_me
f.write("<html>some new noise with cool stuff2 ok</html>")
client.get(url_for("ui.form_watch_checknow"), follow_redirects=True)
time.sleep(sleep_time_for_fetch_thread)
wait_for_all_checks(client)
# It should report nothing found (nothing should match the regex and filter)
res = client.get(url_for("watchlist.index"))
@@ -73,7 +72,8 @@ def test_trigger_regex_functionality_with_filter(client, live_server, measure_me
f.write("<html>some new noise with <span id=in-here>cool stuff6</span> ok</html>")
client.get(url_for("ui.form_watch_checknow"), follow_redirects=True)
time.sleep(sleep_time_for_fetch_thread)
wait_for_all_checks(client)
res = client.get(url_for("watchlist.index"))
assert b'has-unread-changes' in res.data

View File

@@ -160,6 +160,7 @@ def extract_UUID_from_client(client):
return uuid.strip()
def delete_all_watches(client=None):
wait_for_all_checks(client)
uuids = list(client.application.config.get('DATASTORE').data['watching'])
for uuid in uuids:
@@ -180,6 +181,23 @@ def delete_all_watches(client=None):
time.sleep(0.2)
# Delete any old watch metadata
from pathlib import Path
base_path = Path(
client.application.config.get('DATASTORE').datastore_path
).resolve()
max_depth = 2
for file in base_path.rglob("*.json"):
# Calculate depth relative to base path
depth = len(file.relative_to(base_path).parts) - 1
if depth <= max_depth and file.is_file():
file.unlink()
def wait_for_all_checks(client=None):
"""
Waits until the queue is empty and workers are idle.

View File

@@ -4,11 +4,10 @@ import changedetectionio.content_fetchers.exceptions as content_fetchers_excepti
from changedetectionio.processors.text_json_diff.processor import FilterNotFoundInResponse
from changedetectionio import html_tools
from changedetectionio import worker_pool
from changedetectionio.flask_app import watch_check_update
from changedetectionio.queuedWatchMetaData import PrioritizedItem
from changedetectionio.pluggy_interface import apply_update_handler_alter, apply_update_finalize
import asyncio
import importlib
import os
import sys
import time
@@ -56,6 +55,7 @@ async def async_update_worker(worker_id, q, notification_q, app, datastore, exec
while not app.config.exit.is_set():
update_handler = None
watch = None
processing_exception = None # Reset at start of each iteration to prevent state bleeding
try:
# Efficient blocking via run_in_executor (no polling overhead!)
@@ -119,7 +119,7 @@ async def async_update_worker(worker_id, q, notification_q, app, datastore, exec
# to prevent race condition with wait_for_all_checks()
fetch_start_time = round(time.time())
try:
if uuid in list(datastore.data['watching'].keys()) and datastore.data['watching'][uuid].get('url'):
changed_detected = False
@@ -136,6 +136,8 @@ async def async_update_worker(worker_id, q, notification_q, app, datastore, exec
logger.info(f"Worker {worker_id} processing watch UUID {uuid} Priority {queued_item_data.priority} URL {watch['url']}")
try:
# Retrieve signal by name to ensure thread-safe access across worker threads
watch_check_update = signal('watch_check_update')
watch_check_update.send(watch_uuid=uuid)
# Processor is what we are using for detecting the "Change"
@@ -154,6 +156,9 @@ async def async_update_worker(worker_id, q, notification_q, app, datastore, exec
update_handler = processor_module.perform_site_check(datastore=datastore,
watch_uuid=uuid)
# Allow plugins to modify/wrap the update_handler
update_handler = apply_update_handler_alter(update_handler, watch, datastore)
update_signal = signal('watch_small_status_comment')
update_signal.send(watch_uuid=uuid, status="Fetching page..")
@@ -498,6 +503,8 @@ async def async_update_worker(worker_id, q, notification_q, app, datastore, exec
gc.collect()
except Exception as e:
# Store the processing exception for plugin finalization hook
processing_exception = e
logger.error(f"Worker {worker_id} unexpected error processing {uuid}: {e}")
logger.exception(f"Worker {worker_id} full exception details:")
@@ -509,6 +516,11 @@ async def async_update_worker(worker_id, q, notification_q, app, datastore, exec
finally:
# Always cleanup - this runs whether there was an exception or not
if uuid:
# Capture references for plugin finalize hook BEFORE cleanup
# (cleanup may delete these variables, but plugins need the original references)
finalize_handler = update_handler # Capture now, before cleanup deletes it
finalize_watch = watch # Capture now, before any modifications
# Call quit() as backup (Puppeteer/Playwright have internal cleanup, but this acts as safety net)
try:
if update_handler and hasattr(update_handler, 'fetcher') and update_handler.fetcher:
@@ -518,11 +530,9 @@ async def async_update_worker(worker_id, q, notification_q, app, datastore, exec
logger.exception(f"Worker {worker_id} full exception details:")
try:
# Release UUID from processing (thread-safe)
worker_pool.release_uuid_from_processing(uuid, worker_id=worker_id)
# Send completion signal
# Send completion signal - retrieve by name to ensure thread-safe access
if watch:
watch_check_update = signal('watch_check_update')
watch_check_update.send(watch_uuid=watch['uuid'])
# Clean up all memory references BEFORE garbage collection
@@ -547,6 +557,31 @@ async def async_update_worker(worker_id, q, notification_q, app, datastore, exec
logger.error(f"Worker {worker_id} error during cleanup: {cleanup_error}")
logger.exception(f"Worker {worker_id} full exception details:")
# Call plugin finalization hook after all cleanup is done
# Use captured references from before cleanup
try:
apply_update_finalize(
update_handler=finalize_handler,
watch=finalize_watch,
datastore=datastore,
processing_exception=processing_exception
)
except Exception as finalize_error:
logger.error(f"Worker {worker_id} error in finalize hook: {finalize_error}")
logger.exception(f"Worker {worker_id} full exception details:")
finally:
# Clean up captured references to allow immediate garbage collection
del finalize_handler
del finalize_watch
# Release UUID from processing AFTER all cleanup and hooks complete (thread-safe)
# This ensures wait_for_all_checks() waits for finalize hooks to complete
try:
worker_pool.release_uuid_from_processing(uuid, worker_id=worker_id)
except Exception as release_error:
logger.error(f"Worker {worker_id} error releasing UUID: {release_error}")
logger.exception(f"Worker {worker_id} full exception details:")
del(uuid)
# Brief pause before continuing to avoid tight error loops (only on error)