From 8dc39d4a3d5de0f884f9c0d16e0158dc4b2f0378 Mon Sep 17 00:00:00 2001 From: dgtlmoon Date: Wed, 12 Nov 2025 17:38:18 +0100 Subject: [PATCH] RSS feeds for a single watches! --- changedetectionio/blueprint/rss/__init__.py | 18 ++- changedetectionio/blueprint/rss/blueprint.py | 139 +++++++++++++----- .../settings/templates/settings.html | 2 +- changedetectionio/blueprint/ui/edit.py | 7 +- .../blueprint/ui/templates/edit.html | 1 + changedetectionio/forms.py | 2 +- changedetectionio/model/App.py | 4 +- changedetectionio/notification/handler.py | 2 + changedetectionio/store.py | 55 ++++--- changedetectionio/templates/base.html | 9 +- changedetectionio/tests/test_backend.py | 3 +- changedetectionio/tests/test_rss.py | 73 +++++++++ 12 files changed, 249 insertions(+), 66 deletions(-) diff --git a/changedetectionio/blueprint/rss/__init__.py b/changedetectionio/blueprint/rss/__init__.py index d4e09194..adecd339 100644 --- a/changedetectionio/blueprint/rss/__init__.py +++ b/changedetectionio/blueprint/rss/__init__.py @@ -1 +1,17 @@ -RSS_FORMAT_TYPES = [('plaintext', 'Plain text'), ('html', 'HTML Color')] +from copy import deepcopy +from loguru import logger + +from changedetectionio.model import USE_SYSTEM_DEFAULT_NOTIFICATION_FORMAT_FOR_WATCH +from changedetectionio.notification import valid_notification_formats +RSS_CONTENT_FORMAT_DEFAULT = 'text' + +# Some stuff not related +RSS_FORMAT_TYPES = deepcopy(valid_notification_formats) +if RSS_FORMAT_TYPES.get('markdown'): + del RSS_FORMAT_TYPES['markdown'] + +if RSS_FORMAT_TYPES.get(USE_SYSTEM_DEFAULT_NOTIFICATION_FORMAT_FOR_WATCH): + del RSS_FORMAT_TYPES[USE_SYSTEM_DEFAULT_NOTIFICATION_FORMAT_FOR_WATCH] + +if not RSS_FORMAT_TYPES.get(RSS_CONTENT_FORMAT_DEFAULT): + logger.critical(f"RSS_CONTENT_FORMAT_DEFAULT not in the acceptable list {RSS_CONTENT_FORMAT_DEFAULT}") diff --git a/changedetectionio/blueprint/rss/blueprint.py b/changedetectionio/blueprint/rss/blueprint.py index 602aa3db..7a27dcc7 100644 --- a/changedetectionio/blueprint/rss/blueprint.py +++ b/changedetectionio/blueprint/rss/blueprint.py @@ -37,6 +37,51 @@ def clean_entry_content(content): def construct_blueprint(datastore: ChangeDetectionStore): rss_blueprint = Blueprint('rss', __name__) + # Helper function to generate GUID for RSS entries + def generate_watch_guid(watch): + """Generate a unique GUID for a watch RSS entry.""" + return f"{watch['uuid']}/{watch.last_changed}" + + # Helper function to generate diff content for a watch + def generate_watch_diff_content(watch, dates, rss_content_format): + """ + Generate HTML diff content for a watch given its history dates. + Returns the rendered HTML content ready for RSS/display. + """ + from changedetectionio import diff + + # Same logic as watch-overview.html + if datastore.data['settings']['application']['ui'].get('use_page_title_in_list') or watch.get('use_page_title_in_list'): + watch_label = watch.label + else: + watch_label = watch.get('url') + + try: + html_diff = diff.render_diff( + previous_version_file_contents=watch.get_history_snapshot(timestamp=dates[-2]), + newest_version_file_contents=watch.get_history_snapshot(timestamp=dates[-1]), + include_equal=False + ) + + requested_output_format = datastore.data['settings']['application'].get('rss_content_format') + url, html_diff, n_title = apply_service_tweaks(url='', n_body=html_diff, n_title=None, requested_output_format=requested_output_format) + + except FileNotFoundError as e: + html_diff = f"History snapshot file for watch {watch.get('uuid')}@{watch.last_changed} - '{watch.get('title')} not found." + + # @note: We use 
 because nearly all RSS readers render only HTML (Thunderbird for example cant do just plaintext)
+        rss_template = "
{{watch_label}} had a change.\n\n{{html_diff}}\n
"
+        if 'html' in rss_content_format:
+            rss_template = "\n
{{watch_label}}
\n
{{html_diff}}
\n\n"
+
+        content = jinja_render(template_str=rss_template, watch_label=watch_label, 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)
+
+        return content, watch_label
+
     # 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'])
@@ -51,6 +96,8 @@ def construct_blueprint(datastore: ChangeDetectionStore):
         # Always requires token set
         app_rss_token = datastore.data['settings']['application'].get('rss_access_token')
         rss_url_token = request.args.get('token')
+        rss_content_format = datastore.data['settings']['application'].get('rss_content_format')
+
         if rss_url_token != app_rss_token:
             return "Access denied, bad token", 403
 
@@ -81,10 +128,6 @@ def construct_blueprint(datastore: ChangeDetectionStore):
         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())
@@ -95,7 +138,7 @@ def construct_blueprint(datastore: ChangeDetectionStore):
             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)
+                guid = generate_watch_guid(watch)
                 fe = fg.add_entry()
 
                 # Include a link to the diff page, they will have to login here to see if password protection is enabled.
@@ -109,38 +152,9 @@ def construct_blueprint(datastore: ChangeDetectionStore):
 
                 fe.link(link=diff_link)
 
-                # Same logic as watch-overview.html
-                if datastore.data['settings']['application']['ui'].get('use_page_title_in_list') or watch.get('use_page_title_in_list'):
-                    watch_label = watch.label
-                else:
-                    watch_label = watch.get('url')
+                content, watch_label = generate_watch_diff_content(watch, dates, rss_content_format)
 
                 fe.title(title=watch_label)
-                try:
-
-                    html_diff = diff.render_diff(previous_version_file_contents=watch.get_history_snapshot(timestamp=dates[-2]),
-                                                 newest_version_file_contents=watch.get_history_snapshot(timestamp=dates[-1]),
-                                                 include_equal=False,
-                                                 line_feed_sep="
"
-                                                 )
-
-
-                    requested_output_format = 'htmlcolor' if html_colour_enable else 'html'
-                    html_diff = apply_service_tweaks(url='', n_body=html_diff, n_title=None, requested_output_format=requested_output_format)
-
-                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  goes to the diff page, or to the watch link
-                rss_template = "\n
{{watch_title}}
\n
{{html_diff}}
\n\n"
-
-                content = jinja_render(template_str=rss_template, watch_title=watch_label, 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))
@@ -152,4 +166,59 @@ def construct_blueprint(datastore: ChangeDetectionStore):
         logger.trace(f"RSS generated in {time.time() - now:.3f}s")
         return response
 
+    @rss_blueprint.route("/watch/", methods=['GET'])
+    def rss_single_watch(uuid):
+        """
+        Display the most recent change for a single watch as RSS feed.
+        Returns RSS XML with a single entry showing the diff between the last two snapshots.
+        """
+        # Always requires token set
+        app_rss_token = datastore.data['settings']['application'].get('rss_access_token')
+        rss_url_token = request.args.get('token')
+        rss_content_format = datastore.data['settings']['application'].get('rss_content_format')
+
+        if rss_url_token != app_rss_token:
+            return "Access denied, bad token", 403
+
+        # Get the watch by UUID
+        watch = datastore.data['watching'].get(uuid)
+        if not watch:
+            return f"Watch with UUID {uuid} not found", 404
+
+        # Check if watch has at least 2 history snapshots
+        dates = list(watch.history.keys())
+        if len(dates) < 2:
+            return f"Watch {uuid} does not have enough history snapshots to show changes (need at least 2)", 400
+
+        # Add uuid to watch for proper functioning
+        watch['uuid'] = uuid
+
+        # Generate the diff content using the shared helper function
+        content, watch_label = generate_watch_diff_content(watch, dates, rss_content_format)
+
+        # Create RSS feed with single entry
+        fg = FeedGenerator()
+        fg.title(f'changedetection.io - {watch.label}')
+        fg.description('Changes')
+        fg.link(href='https://changedetection.io')
+
+        # Add single entry for this watch
+        guid = generate_watch_guid(watch)
+        fe = fg.add_entry()
+
+        # Include a link to the diff page
+        diff_link = {'href': url_for('ui.ui_views.diff_history_page', uuid=watch['uuid'], _external=True)}
+        fe.link(link=diff_link)
+
+        fe.title(title=watch_label)
+        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')
+        return response
+
     return rss_blueprint
\ No newline at end of file
diff --git a/changedetectionio/blueprint/settings/templates/settings.html b/changedetectionio/blueprint/settings/templates/settings.html
index 9913e095..5d4b5536 100644
--- a/changedetectionio/blueprint/settings/templates/settings.html
+++ b/changedetectionio/blueprint/settings/templates/settings.html
@@ -86,7 +86,7 @@
                         
                         

                             {{ render_checkbox_field(form.application.form.rss_reader_mode) }}
-                            Transforms RSS/RDF feed watches into beautiful text only
+                            When watching RSS/Atom feeds, convert them into clean text for better change detection.
                         

                     
                 
diff --git a/changedetectionio/blueprint/ui/edit.py b/changedetectionio/blueprint/ui/edit.py
index f68bdbd1..d394a906 100644
--- a/changedetectionio/blueprint/ui/edit.py
+++ b/changedetectionio/blueprint/ui/edit.py
@@ -236,7 +236,7 @@ def construct_blueprint(datastore: ChangeDetectionStore, update_q, queuedWatchMe
 
             # Import the global plugin system
             from changedetectionio.pluggy_interface import collect_ui_edit_stats_extras
-            
+            app_rss_token = datastore.data['settings']['application'].get('rss_access_token'),
             template_args = {
                 'available_processors': processors.available_processors(),
                 'available_timezones': sorted(available_timezones()),
@@ -252,6 +252,11 @@ def construct_blueprint(datastore: ChangeDetectionStore, update_q, queuedWatchMe
                 'has_special_tag_options': _watch_has_tag_options_set(watch=watch),
                 'jq_support': jq_support,
                 'playwright_enabled': os.getenv('PLAYWRIGHT_DRIVER_URL', False),
+                'app_rss_token': app_rss_token,
+                'rss_uuid_feed' : {
+                    'label': watch.label,
+                    'url': url_for('rss.rss_single_watch', uuid=watch['uuid'], token=app_rss_token)
+                },
                 'settings_application': datastore.data['settings']['application'],
                 'system_has_playwright_configured': os.getenv('PLAYWRIGHT_DRIVER_URL'),
                 'system_has_webdriver_configured': os.getenv('WEBDRIVER_URL'),
diff --git a/changedetectionio/blueprint/ui/templates/edit.html b/changedetectionio/blueprint/ui/templates/edit.html
index f6e7f6a0..bb4c50a0 100644
--- a/changedetectionio/blueprint/ui/templates/edit.html
+++ b/changedetectionio/blueprint/ui/templates/edit.html
@@ -476,6 +476,7 @@ Math: {{ 1 + 1 }}") }}
                        class="pure-button button-error">Clear History{% endif %}
                     Clone & Edit
+                 RSS Feed for this watch
                 
             
         
diff --git a/changedetectionio/forms.py b/changedetectionio/forms.py
index ce441df9..51760efe 100644
--- a/changedetectionio/forms.py
+++ b/changedetectionio/forms.py
@@ -1000,7 +1000,7 @@ class globalSettingsApplicationForm(commonSettingsForm):
                               validators=[validators.NumberRange(min=0,
                                                                  message="Should be atleast zero (disabled)")])
 
-    rss_content_format = SelectField('RSS Content format', choices=RSS_FORMAT_TYPES)
+    rss_content_format = SelectField('RSS Content format', choices=list(RSS_FORMAT_TYPES.items()))
 
     removepassword_button = SubmitField('Remove password', render_kw={"class": "pure-button pure-button-primary"})
     render_anchor_tag_content = BooleanField('Render anchor tag content', default=False)
diff --git a/changedetectionio/model/App.py b/changedetectionio/model/App.py
index 0152aed6..cbf81370 100644
--- a/changedetectionio/model/App.py
+++ b/changedetectionio/model/App.py
@@ -1,7 +1,7 @@
 from os import getenv
 from copy import deepcopy
 
-from changedetectionio.blueprint.rss import RSS_FORMAT_TYPES
+from changedetectionio.blueprint.rss import RSS_FORMAT_TYPES, RSS_CONTENT_FORMAT_DEFAULT
 
 from changedetectionio.notification import (
     default_notification_body,
@@ -54,7 +54,7 @@ class model(dict):
                     'password': False,
                     'render_anchor_tag_content': False,
                     'rss_access_token': None,
-                    'rss_content_format': RSS_FORMAT_TYPES[0][0],
+                    'rss_content_format': RSS_CONTENT_FORMAT_DEFAULT,
                     'rss_hide_muted_watches': True,
                     'rss_reader_mode': False,
                     'scheduler_timezone_default': None,  # Default IANA timezone name
diff --git a/changedetectionio/notification/handler.py b/changedetectionio/notification/handler.py
index 1e80db8b..f3be4104 100644
--- a/changedetectionio/notification/handler.py
+++ b/changedetectionio/notification/handler.py
@@ -187,6 +187,8 @@ def replace_placemarkers_in_text(text, url, requested_output_format):
 
 def apply_service_tweaks(url, n_body, n_title, requested_output_format):
 
+    logger.debug(f"Applying markup in '{requested_output_format}' mode")
+
     # Re 323 - Limit discord length to their 2000 char limit total or it wont send.
     # Because different notifications may require different pre-processing, run each sequentially :(
     # 2000 bytes minus -
diff --git a/changedetectionio/store.py b/changedetectionio/store.py
index a05ec623..ba223ae3 100644
--- a/changedetectionio/store.py
+++ b/changedetectionio/store.py
@@ -6,6 +6,7 @@ from flask import (
     flash
 )
 
+from .blueprint.rss import RSS_CONTENT_FORMAT_DEFAULT
 from .html_tools import TRANSLATE_WHITESPACE_TABLE
 from .model import App, Watch, USE_SYSTEM_DEFAULT_NOTIFICATION_FORMAT_FOR_WATCH
 from copy import deepcopy, copy
@@ -775,6 +776,28 @@ class ChangeDetectionStore:
 
         return updates_available
 
+    def add_notification_url(self, notification_url):
+
+        logger.debug(f">>> Adding new notification_url - '{notification_url}'")
+
+        notification_urls = self.data['settings']['application'].get('notification_urls', [])
+
+        if notification_url in notification_urls:
+            return notification_url
+
+        with self.lock:
+            notification_urls = self.__data['settings']['application'].get('notification_urls', [])
+
+            if notification_url in notification_urls:
+                return notification_url
+
+            # Append and update the datastore
+            notification_urls.append(notification_url)
+            self.__data['settings']['application']['notification_urls'] = notification_urls
+            self.needs_write = True
+
+        return notification_url
+
     # Run all updates
     # IMPORTANT - Each update could be run even when they have a new install and the schema is correct
     #             So therefor - each `update_n` should be very careful about checking if it needs to actually run
@@ -1087,25 +1110,15 @@ class ChangeDetectionStore:
         formats['markdown'] = 'Markdown'
         re_run(formats)
 
-    def add_notification_url(self, notification_url):
-        
-        logger.debug(f">>> Adding new notification_url - '{notification_url}'")
-
-        notification_urls = self.data['settings']['application'].get('notification_urls', [])
-
-        if notification_url in notification_urls:
-            return notification_url
-
-        with self.lock:
-            notification_urls = self.__data['settings']['application'].get('notification_urls', [])
-
-            if notification_url in notification_urls:
-                return notification_url
-
-            # Append and update the datastore
-            notification_urls.append(notification_url)
-            self.__data['settings']['application']['notification_urls'] = notification_urls
-            self.needs_write = True
-
-        return notification_url
 
+    # RSS types should be inline with the same names as notification types
+    def update_24(self):
+        rss_format = self.data['settings']['application'].get('rss_content_format')
+        if not rss_format or 'text' in rss_format:
+            # might have been 'plaintext, 'plain text' or something
+            self.data['settings']['application']['rss_content_format'] = RSS_CONTENT_FORMAT_DEFAULT
+        elif 'html' in rss_format:
+            self.data['settings']['application']['rss_content_format'] = 'htmlcolor'
+        else:
+            # safe fallback to text
+            self.data['settings']['application']['rss_content_format'] = RSS_CONTENT_FORMAT_DEFAULT
diff --git a/changedetectionio/templates/base.html b/changedetectionio/templates/base.html
index 5b398d30..dd56ed27 100644
--- a/changedetectionio/templates/base.html
+++ b/changedetectionio/templates/base.html
@@ -8,8 +8,13 @@
     
     Change Detection{{extra_title}}
     {% if app_rss_token %}
-      
-    {% endif %}
+        
+
+        {% if rss_uuid_feed %}
+        
+
+        {%- endif -%}
+    {%- endif -%}
     
     
     {% if extra_stylesheets %}
diff --git a/changedetectionio/tests/test_backend.py b/changedetectionio/tests/test_backend.py
index e9e81fe5..0fa74094 100644
--- a/changedetectionio/tests/test_backend.py
+++ b/changedetectionio/tests/test_backend.py
@@ -77,10 +77,9 @@ def test_check_basic_change_detection_functionality(client, live_server, measure
     assert b'' not in content
+            assert 'body' not in content
+            assert '(changed) Which is across multiple lines\n'
+            assert 'modified head title had a change.' # Because it picked it up  as watch_title in default template
+        elif expected_type == 'html':
+            assert '<p>' in content
+            assert '<body>' in content
+            assert '<p>(changed) Which is across multiple lines<br>' in content
+            assert f'href="{url}">modified head title had a change.</a>'
+        elif expected_type == 'htmlcolor':
+            assert '<body>' in content
+            assert ' role="note" aria-label="Changed text" title="Changed text">Which is across multiple lines</span>' in content
+            assert f'href="{url}">modified head title had a change.</a>'
+        else:
+            raise Exception(f"Unknown type {expected_type}")
+
+
+    item = root.findall('.//item')[0].findtext('description')
+    check_formatting(expected_type=rss_content_format, content=item, url=test_url)
+
+    # Now the default one is over, lets try all the others
+    for k in list(RSS_FORMAT_TYPES.keys()):
+        res = client.post(
+            url_for("settings.settings_page"),
+            data={"application-rss_content_format": k},
+            follow_redirects=True
+        )
+        assert b'Settings updated' in res.data
+
+        res = client.get(
+            url_for('rss.rss_single_watch', uuid=uuid, token=app_rss_token),
+            follow_redirects=False
+        )
+        assert res.status_code == 200
+        root = ET.fromstring(res.data)
+        item = root.findall('.//item')[0].findtext('description')
+        check_formatting(expected_type=k, content=item, url=test_url)