RSS Fixes and improvements - Ability to set "RSS Color HTML Format" in Settings, detect and filter content with bad content that could break RSS (#3055)

pull/3062/head
dgtlmoon 2025-03-26 12:08:15 +01:00 zatwierdzone przez GitHub
rodzic b53e1985ac
commit f26ea55e9c
Nie znaleziono w bazie danych klucza dla tego podpisu
ID klucza GPG: B5690EEEBB952194
8 zmienionych plików z 235 dodań i 105 usunięć

Wyświetl plik

@ -1,102 +1 @@
import time
import datetime
import pytz
from flask import Blueprint, make_response, request, url_for
from loguru import logger
from feedgen.feed import FeedGenerator
from changedetectionio.store import ChangeDetectionStore
from changedetectionio.safe_jinja import render as jinja_render
def construct_blueprint(datastore: ChangeDetectionStore):
rss_blueprint = Blueprint('rss', __name__)
# Import the login decorator if needed
# from changedetectionio.auth_decorator import login_optionally_required
@rss_blueprint.route("", methods=['GET'])
def feed():
now = time.time()
# Always requires token set
app_rss_token = datastore.data['settings']['application'].get('rss_access_token')
rss_url_token = request.args.get('token')
if rss_url_token != app_rss_token:
return "Access denied, bad token", 403
from changedetectionio import diff
limit_tag = request.args.get('tag', '').lower().strip()
# Be sure limit_tag is a uuid
for uuid, tag in datastore.data['settings']['application'].get('tags', {}).items():
if limit_tag == tag.get('title', '').lower().strip():
limit_tag = uuid
# Sort by last_changed and add the uuid which is usually the key..
sorted_watches = []
# @todo needs a .itemsWithTag() or something - then we can use that in Jinaj2 and throw this away
for uuid, watch in datastore.data['watching'].items():
# @todo tag notification_muted skip also (improve Watch model)
if datastore.data['settings']['application'].get('rss_hide_muted_watches') and watch.get('notification_muted'):
continue
if limit_tag and not limit_tag in watch['tags']:
continue
watch['uuid'] = uuid
sorted_watches.append(watch)
sorted_watches.sort(key=lambda x: x.last_changed, reverse=False)
fg = FeedGenerator()
fg.title('changedetection.io')
fg.description('Feed description')
fg.link(href='https://changedetection.io')
for watch in sorted_watches:
dates = list(watch.history.keys())
# Re #521 - Don't bother processing this one if theres less than 2 snapshots, means we never had a change detected.
if len(dates) < 2:
continue
if not watch.viewed:
# Re #239 - GUID needs to be individual for each event
# @todo In the future make this a configurable link back (see work on BASE_URL https://github.com/dgtlmoon/changedetection.io/pull/228)
guid = "{}/{}".format(watch['uuid'], watch.last_changed)
fe = fg.add_entry()
# Include a link to the diff page, they will have to login here to see if password protection is enabled.
# Description is the page you watch, link takes you to the diff JS UI page
# Dict val base_url will get overriden with the env var if it is set.
ext_base_url = datastore.data['settings']['application'].get('active_base_url')
# Because we are called via whatever web server, flask should figure out the right path (
diff_link = {'href': url_for('ui.ui_views.diff_history_page', uuid=watch['uuid'], _external=True)}
fe.link(link=diff_link)
# @todo watch should be a getter - watch.get('title') (internally if URL else..)
watch_title = watch.get('title') if watch.get('title') else watch.get('url')
fe.title(title=watch_title)
html_diff = diff.render_diff(previous_version_file_contents=watch.get_history_snapshot(dates[-2]),
newest_version_file_contents=watch.get_history_snapshot(dates[-1]),
include_equal=False,
line_feed_sep="<br>")
# @todo Make this configurable and also consider html-colored markup
# @todo User could decide if <link> goes to the diff page, or to the watch link
rss_template = "<html><body>\n<h4><a href=\"{{watch_url}}\">{{watch_title}}</a></h4>\n<p>{{html_diff}}</p>\n</body></html>\n"
content = jinja_render(template_str=rss_template, watch_title=watch_title, html_diff=html_diff, watch_url=watch.link)
fe.content(content=content, type='CDATA')
fe.guid(guid, permalink=False)
dt = datetime.datetime.fromtimestamp(int(watch.newest_history_key))
dt = dt.replace(tzinfo=pytz.UTC)
fe.pubDate(dt)
response = make_response(fg.rss_str())
response.headers.set('Content-Type', 'application/rss+xml;charset=utf-8')
logger.trace(f"RSS generated in {time.time() - now:.3f}s")
return response
return rss_blueprint
RSS_FORMAT_TYPES = [('plaintext', 'Plain text'), ('html', 'HTML Color')]

Wyświetl plik

@ -0,0 +1,147 @@
from changedetectionio.safe_jinja import render as jinja_render
from changedetectionio.store import ChangeDetectionStore
from feedgen.feed import FeedGenerator
from flask import Blueprint, make_response, request, url_for, redirect
from loguru import logger
import datetime
import pytz
import re
import time
BAD_CHARS_REGEX=r'[\x00-\x08\x0B\x0C\x0E-\x1F]'
# Anything that is not text/UTF-8 should be stripped before it breaks feedgen (such as binary data etc)
def scan_invalid_chars_in_rss(content):
for match in re.finditer(BAD_CHARS_REGEX, content):
i = match.start()
bad_char = content[i]
hex_value = f"0x{ord(bad_char):02x}"
# Grab context
start = max(0, i - 20)
end = min(len(content), i + 21)
context = content[start:end].replace('\n', '\\n').replace('\r', '\\r')
logger.warning(f"Invalid char {hex_value} at pos {i}: ...{context}...")
# First match is enough
return True
return False
def clean_entry_content(content):
cleaned = re.sub(BAD_CHARS_REGEX, '', content)
return cleaned
def construct_blueprint(datastore: ChangeDetectionStore):
rss_blueprint = Blueprint('rss', __name__)
# Some RSS reader situations ended up with rss/ (forward slash after RSS) due
# to some earlier blueprint rerouting work, it should goto feed.
@rss_blueprint.route("/", methods=['GET'])
def extraslash():
return redirect(url_for('rss.feed'))
# Import the login decorator if needed
# from changedetectionio.auth_decorator import login_optionally_required
@rss_blueprint.route("", methods=['GET'])
def feed():
now = time.time()
# Always requires token set
app_rss_token = datastore.data['settings']['application'].get('rss_access_token')
rss_url_token = request.args.get('token')
if rss_url_token != app_rss_token:
return "Access denied, bad token", 403
from changedetectionio import diff
limit_tag = request.args.get('tag', '').lower().strip()
# Be sure limit_tag is a uuid
for uuid, tag in datastore.data['settings']['application'].get('tags', {}).items():
if limit_tag == tag.get('title', '').lower().strip():
limit_tag = uuid
# Sort by last_changed and add the uuid which is usually the key..
sorted_watches = []
# @todo needs a .itemsWithTag() or something - then we can use that in Jinaj2 and throw this away
for uuid, watch in datastore.data['watching'].items():
# @todo tag notification_muted skip also (improve Watch model)
if datastore.data['settings']['application'].get('rss_hide_muted_watches') and watch.get('notification_muted'):
continue
if limit_tag and not limit_tag in watch['tags']:
continue
watch['uuid'] = uuid
sorted_watches.append(watch)
sorted_watches.sort(key=lambda x: x.last_changed, reverse=False)
fg = FeedGenerator()
fg.title('changedetection.io')
fg.description('Feed description')
fg.link(href='https://changedetection.io')
html_colour_enable = False
if datastore.data['settings']['application'].get('rss_content_format') == 'html':
html_colour_enable = True
for watch in sorted_watches:
dates = list(watch.history.keys())
# Re #521 - Don't bother processing this one if theres less than 2 snapshots, means we never had a change detected.
if len(dates) < 2:
continue
if not watch.viewed:
# Re #239 - GUID needs to be individual for each event
# @todo In the future make this a configurable link back (see work on BASE_URL https://github.com/dgtlmoon/changedetection.io/pull/228)
guid = "{}/{}".format(watch['uuid'], watch.last_changed)
fe = fg.add_entry()
# Include a link to the diff page, they will have to login here to see if password protection is enabled.
# Description is the page you watch, link takes you to the diff JS UI page
# Dict val base_url will get overriden with the env var if it is set.
ext_base_url = datastore.data['settings']['application'].get('active_base_url')
# @todo fix
# Because we are called via whatever web server, flask should figure out the right path (
diff_link = {'href': url_for('ui.ui_views.diff_history_page', uuid=watch['uuid'], _external=True)}
fe.link(link=diff_link)
# @todo watch should be a getter - watch.get('title') (internally if URL else..)
watch_title = watch.get('title') if watch.get('title') else watch.get('url')
fe.title(title=watch_title)
try:
html_diff = diff.render_diff(previous_version_file_contents=watch.get_history_snapshot(dates[-2]),
newest_version_file_contents=watch.get_history_snapshot(dates[-1]),
include_equal=False,
line_feed_sep="<br>",
html_colour=html_colour_enable
)
except FileNotFoundError as e:
html_diff = f"History snapshot file for watch {watch.get('uuid')}@{watch.last_changed} - '{watch.get('title')} not found."
# @todo Make this configurable and also consider html-colored markup
# @todo User could decide if <link> goes to the diff page, or to the watch link
rss_template = "<html><body>\n<h4><a href=\"{{watch_url}}\">{{watch_title}}</a></h4>\n<p>{{html_diff}}</p>\n</body></html>\n"
content = jinja_render(template_str=rss_template, watch_title=watch_title, html_diff=html_diff, watch_url=watch.link)
# Out of range chars could also break feedgen
if scan_invalid_chars_in_rss(content):
content = clean_entry_content(content)
fe.content(content=content, type='CDATA')
fe.guid(guid, permalink=False)
dt = datetime.datetime.fromtimestamp(int(watch.newest_history_key))
dt = dt.replace(tzinfo=pytz.UTC)
fe.pubDate(dt)
response = make_response(fg.rss_str())
response.headers.set('Content-Type', 'application/rss+xml;charset=utf-8')
logger.trace(f"RSS generated in {time.time() - now:.3f}s")
return response
return rss_blueprint

Wyświetl plik

@ -78,7 +78,10 @@
{{ render_field(form.application.form.pager_size) }}
<span class="pure-form-message-inline">Number of items per page in the watch overview list, 0 to disable.</span>
</div>
<div class="pure-control-group">
{{ render_field(form.application.form.rss_content_format) }}
<span class="pure-form-message-inline">Love RSS? Does your reader support HTML? Set it here</span>
</div>
<div class="pure-control-group">
{{ render_checkbox_field(form.application.form.extract_title_as_title) }}
<span class="pure-form-message-inline">Note: This will automatically apply to all existing watches.</span>

Wyświetl plik

@ -429,7 +429,7 @@ def changedetection_app(config=None, datastore_o=None):
import changedetectionio.conditions.blueprint as conditions
app.register_blueprint(conditions.construct_blueprint(datastore), url_prefix='/conditions')
import changedetectionio.blueprint.rss as rss
import changedetectionio.blueprint.rss.blueprint as rss
app.register_blueprint(rss.construct_blueprint(datastore), url_prefix='/rss')
# watchlist UI buttons etc

Wyświetl plik

@ -3,6 +3,7 @@ import re
from loguru import logger
from wtforms.widgets.core import TimeInput
from changedetectionio.blueprint.rss import RSS_FORMAT_TYPES
from changedetectionio.conditions.form import ConditionFormRow
from changedetectionio.strtobool import strtobool
@ -739,6 +740,9 @@ class globalSettingsApplicationForm(commonSettingsForm):
render_kw={"style": "width: 5em;"},
validators=[validators.NumberRange(min=0,
message="Should be atleast zero (disabled)")])
rss_content_format = SelectField('RSS Content format', choices=RSS_FORMAT_TYPES)
removepassword_button = SubmitField('Remove password', render_kw={"class": "pure-button pure-button-primary"})
render_anchor_tag_content = BooleanField('Render anchor tag content', default=False)
shared_diff_access = BooleanField('Allow access to view diff page when password is enabled', default=False, validators=[validators.Optional()])

Wyświetl plik

@ -1,4 +1,7 @@
from os import getenv
from changedetectionio.blueprint.rss import RSS_FORMAT_TYPES
from changedetectionio.notification import (
default_notification_body,
default_notification_format,
@ -9,6 +12,8 @@ from changedetectionio.notification import (
_FILTER_FAILURE_THRESHOLD_ATTEMPTS_DEFAULT = 6
DEFAULT_SETTINGS_HEADERS_USERAGENT='Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/87.0.4280.66 Safari/537.36'
class model(dict):
base_config = {
'note': "Hello! If you change this file manually, please be sure to restart your changedetection.io instance!",
@ -48,6 +53,7 @@ class model(dict):
'password': False,
'render_anchor_tag_content': False,
'rss_access_token': None,
'rss_content_format': RSS_FORMAT_TYPES[0][0],
'rss_hide_muted_watches': True,
'schema_version' : 0,
'shared_diff_access': False,

Wyświetl plik

@ -49,6 +49,22 @@ def set_original_cdata_xml():
f.write(test_return_data)
def set_html_content(content):
test_return_data = f"""<html>
<body>
Some initial text<br>
<p>{content}</p>
<br>
So let's see what happens. <br>
</body>
</html>
"""
# Write as UTF-8 encoded bytes
with open("test-datastore/endpoint-content.txt", "wb") as f:
f.write(test_return_data.encode('utf-8'))
def test_setup(client, live_server, measure_memory_usage):
live_server_setup(live_server)
@ -164,3 +180,58 @@ def test_rss_xpath_filtering(client, live_server, measure_memory_usage):
assert b'Some other description' not in res.data # Should NOT be selected by the xpath
res = client.get(url_for("ui.form_delete", uuid="all"), follow_redirects=True)
def test_rss_bad_chars_breaking(client, live_server):
"""This should absolutely trigger the RSS builder to go into worst state mode
- source: prefix means no html conversion (which kinda filters out the bad stuff)
- Binary data
- Very long so that the saving is performed by Brotli (and decoded back to bytes)
Otherwise feedgen should support regular unicode
"""
#live_server_setup(live_server)
with open("test-datastore/endpoint-content.txt", "w") as f:
ten_kb_string = "A" * 10_000
f.write(ten_kb_string)
test_url = url_for('test_endpoint', _external=True)
res = client.post(
url_for("imports.import_page"),
data={"urls": "source:"+test_url},
follow_redirects=True
)
assert b"1 Imported" in res.data
wait_for_all_checks(client)
# Set the bad content
with open("test-datastore/endpoint-content.txt", "w") as f:
jpeg_bytes = "\xff\xd8\xff\xe0\x00\x10XXXXXXXX\x00\x01\x02\x00\x00\x01\x00\x01\x00\x00" # JPEG header
jpeg_bytes += "A" * 10_000
f.write(jpeg_bytes)
res = client.get(url_for("ui.form_watch_checknow"), follow_redirects=True)
assert b'Queued 1 watch for rechecking.' in res.data
wait_for_all_checks(client)
rss_token = extract_rss_token_from_UI(client)
uuid = next(iter(live_server.app.config['DATASTORE'].data['watching']))
assert live_server.app.config['DATASTORE'].data['watching'][uuid].history_n == 2
# Check RSS feed is still working
res = client.get(
url_for("rss.feed", uuid=uuid, token=rss_token),
follow_redirects=False # Important! leave this off! it should not redirect
)
assert res.status_code == 200
#assert live_server.app.config['DATASTORE'].data['watching'][uuid].history_n == 2
#assert live_server.app.config['DATASTORE'].data['watching'][uuid].history_n == 2

Wyświetl plik

@ -173,7 +173,7 @@ def live_server_setup(live_server):
return resp
# Tried using a global var here but didn't seem to work, so reading from a file instead.
with open("test-datastore/endpoint-content.txt", "r") as f:
with open("test-datastore/endpoint-content.txt", "rb") as f:
resp = make_response(f.read(), status_code)
if uppercase_headers:
resp.headers['CONTENT-TYPE'] = ctype if ctype else 'text/html'