Re #3115 fix multiprocessing start

Dockerfile

@@ -68,7 +68,7 @@ COPY changedetection.py /app/changedetection.py
 # Github Action test purpose(test-only.yml).
 # On production, it is effectively LOGGER_LEVEL=''.
 ARG LOGGER_LEVEL=''
-ENV LOGGER_LEVEL "$LOGGER_LEVEL"
+ENV LOGGER_LEVEL="$LOGGER_LEVEL"
 
 WORKDIR /app
 CMD ["python", "./changedetection.py", "-d", "/datastore"]

changedetection.py

@@ -3,4 +3,6 @@
 # Only exists for direct CLI usage
 
 import changedetectionio
-changedetectionio.main()
+
+if __name__ == '__main__':
+    changedetectionio.main()

changedetectionio/blueprint/tags/__init__.py

@@ -104,6 +104,9 @@ def construct_blueprint(datastore: ChangeDetectionStore):
             uuid = list(datastore.data['settings']['application']['tags'].keys()).pop()
 
         default = datastore.data['settings']['application']['tags'].get(uuid)
+        if not default:
+            flash("Tag not found", "error")
+            return redirect(url_for('watchlist.index'))
 
         form = group_restock_settings_form(
                                        formdata=request.form if request.method == 'POST' else None,

changedetectionio/blueprint/tags/templates/edit-tag.html

@@ -66,7 +66,7 @@
                     <div  class="pure-control-group inline-radio">
                       {{ render_checkbox_field(form.notification_muted) }}
                     </div>
-                    {% if is_html_webdriver %}
+                    {% if 1 %}
                     <div class="pure-control-group inline-radio">
                       {{ render_checkbox_field(form.notification_screenshot) }}
                         <span class="pure-form-message-inline">

changedetectionio/blueprint/ui/edit.py

@@ -213,9 +213,6 @@ def construct_blueprint(datastore: ChangeDetectionStore, update_q, queuedWatchMe
             if request.method == 'POST' and not form.validate():
                 flash("An error occurred, please see below.", "error")
 
-            visualselector_data_is_ready = datastore.visualselector_data_is_ready(uuid)
-
-
             # JQ is difficult to install on windows and must be manually added (outside requirements.txt)
             jq_support = True
             try:
@@ -225,11 +222,12 @@ def construct_blueprint(datastore: ChangeDetectionStore, update_q, queuedWatchMe
 
             watch = datastore.data['watching'].get(uuid)
 
+            # if system or watch is configured to need a chrome type browser
             system_uses_webdriver = datastore.data['settings']['application']['fetch_backend'] == 'html_webdriver'
-
-            watch_uses_webdriver = False
+            watch_needs_selenium_or_playwright = False
             if (watch.get('fetch_backend') == 'system' and system_uses_webdriver) or watch.get('fetch_backend') == 'html_webdriver' or watch.get('fetch_backend', '').startswith('extra_browser_'):
-                watch_uses_webdriver = True
+                watch_needs_selenium_or_playwright = True
+
 
             from zoneinfo import available_timezones
 
@@ -247,14 +245,17 @@ def construct_blueprint(datastore: ChangeDetectionStore, update_q, queuedWatchMe
                 'has_default_notification_urls': True if len(datastore.data['settings']['application']['notification_urls']) else False,
                 'has_extra_headers_file': len(datastore.get_all_headers_in_textfile_for_watch(uuid=uuid)) > 0,
                 'has_special_tag_options': _watch_has_tag_options_set(watch=watch),
-                'watch_uses_webdriver': watch_uses_webdriver,
                 'jq_support': jq_support,
                 'playwright_enabled': os.getenv('PLAYWRIGHT_DRIVER_URL', False),
                 'settings_application': datastore.data['settings']['application'],
+                'system_has_playwright_configured': os.getenv('PLAYWRIGHT_DRIVER_URL'),
+                'system_has_webdriver_configured': os.getenv('WEBDRIVER_URL'),
+                'visual_selector_data_ready': datastore.visualselector_data_is_ready(watch_uuid=uuid),
                 'timezone_default_config': datastore.data['settings']['application'].get('timezone'),
                 'using_global_webdriver_wait': not default['webdriver_delay'],
                 'uuid': uuid,
-                'watch': watch
+                'watch': watch,
+                'watch_needs_selenium_or_playwright': watch_needs_selenium_or_playwright,
             }
 
             included_content = None

changedetectionio/conditions/__init__.py

@@ -94,11 +94,11 @@ def execute_ruleset_against_all_plugins(current_watch_uuid: str, application_dat
     EXECUTE_DATA = {}
     result = True
     
-    ruleset_settings = application_datastruct['watching'].get(current_watch_uuid)
+    watch = application_datastruct['watching'].get(current_watch_uuid)
 
-    if ruleset_settings.get("conditions"):
-        logic_operator = "and" if ruleset_settings.get("conditions_match_logic", "ALL") == "ALL" else "or"
-        complete_rules = filter_complete_rules(ruleset_settings['conditions'])
+    if watch and watch.get("conditions"):
+        logic_operator = "and" if watch.get("conditions_match_logic", "ALL") == "ALL" else "or"
+        complete_rules = filter_complete_rules(watch['conditions'])
         if complete_rules:
             # Give all plugins a chance to update the data dict again (that we will test the conditions against)
             for plugin in plugin_manager.get_plugins():

changedetectionio/content_fetchers/playwright.py

@@ -26,9 +26,11 @@ def capture_full_page(page):
     step_size = SCREENSHOT_SIZE_STITCH_THRESHOLD # Size that won't cause GPU to overflow
     screenshot_chunks = []
     y = 0
-    
-    # If page height is larger than current viewport, use a larger viewport for better capturing
+
     if page_height > page.viewport_size['height']:
+        if page_height < step_size:
+            step_size = page_height # Incase page is bigger than default viewport but smaller than proposed step size
+        logger.debug(f"Setting bigger viewport to step through large page width W{page.viewport_size['width']}xH{step_size} because page_height > viewport_size")
         # Set viewport to a larger size to capture more content at once
         page.set_viewport_size({'width': page.viewport_size['width'], 'height': step_size})
 
@@ -59,7 +61,10 @@ def capture_full_page(page):
         p.join()
         logger.debug(
             f"Screenshot (chunked/stitched) - Page height: {page_height} Capture height: {SCREENSHOT_MAX_TOTAL_HEIGHT} - Stitched together in {time.time() - start:.2f}s")
-
+        # Explicit cleanup
+        del screenshot_chunks
+        del p
+        del parent_conn, child_conn
         screenshot_chunks = None
         return screenshot
 
@@ -286,12 +291,28 @@ class fetcher(Fetcher):
                     pass
                 
                 # Clean up resources properly
-                context.close()
-                context = None
+                try:
+                    self.page.request_gc()
+                except:
+                    pass
 
-                self.page.close()
+                try:
+                    self.page.close()
+                except:
+                    pass
                 self.page = None
 
-                browser.close()
-                borwser = None
+                try:
+                    context.close()
+                except:
+                    pass
+                context = None
+
+                try:
+                    browser.close()
+                except:
+                    pass
+                browser = None
+
+
 

changedetectionio/content_fetchers/puppeteer.py

@@ -46,9 +46,10 @@ async def capture_full_page(page):
     screenshot_chunks = []
     y = 0
     if page_height > page.viewport['height']:
+        if page_height < step_size:
+            step_size = page_height # Incase page is bigger than default viewport but smaller than proposed step size
         await page.setViewport({'width': page.viewport['width'], 'height': step_size})
 
-
     while y < min(page_height, SCREENSHOT_MAX_TOTAL_HEIGHT):
         await page.evaluate(f"window.scrollTo(0, {y})")
         screenshot_chunks.append(await page.screenshot(type_='jpeg',

changedetectionio/content_fetchers/webdriver_selenium.py

@@ -65,7 +65,17 @@ class fetcher(Fetcher):
         # request_body, request_method unused for now, until some magic in the future happens.
 
         options = ChromeOptions()
-        options.add_argument("--headless")
+
+        # Load Chrome options from env
+        CHROME_OPTIONS = [
+            line.strip()
+            for line in os.getenv("CHROME_OPTIONS", "").strip().splitlines()
+            if line.strip()
+        ]
+
+        for opt in CHROME_OPTIONS:
+            options.add_argument(opt)
+
         if self.proxy:
             options.proxy = self.proxy
 
@@ -80,7 +90,9 @@ class fetcher(Fetcher):
             self.quit()
             raise
 
-        self.driver.set_window_size(1280, 1024)
+        if not "--window-size" in os.getenv("CHROME_OPTIONS", ""):
+            self.driver.set_window_size(1280, 1024)
+
         self.driver.implicitly_wait(int(os.getenv("WEBDRIVER_DELAY_BEFORE_CONTENT_READY", 5)))
 
         if self.webdriver_js_execute_code is not None:
@@ -88,6 +100,7 @@ class fetcher(Fetcher):
             # Selenium doesn't automatically wait for actions as good as Playwright, so wait again
             self.driver.implicitly_wait(int(os.getenv("WEBDRIVER_DELAY_BEFORE_CONTENT_READY", 5)))
 
+
         # @todo - how to check this? is it possible?
         self.status_code = 200
         # @todo somehow we should try to get this working for WebDriver

changedetectionio/html_tools.py

@@ -435,7 +435,9 @@ def cdata_in_document_to_text(html_content: str, render_anchor_tag_content=False
 
     return re.sub(pattern, repl, html_content)
 
-def html_to_text(html_content: str, render_anchor_tag_content=False, is_rss=False) -> str:
+
+def html_to_text_sub_worker(conn, html_content: str, render_anchor_tag_content=False, is_rss=False):
+
     from inscriptis import get_text
     from inscriptis.model.config import ParserConfig
 
@@ -470,9 +472,19 @@ def html_to_text(html_content: str, render_anchor_tag_content=False, is_rss=Fals
         html_content = re.sub(r'</title>', r'</h1>', html_content)
 
     text_content = get_text(html_content, config=parser_config)
+    conn.send(text_content)
+    conn.close()
 
-    return text_content
+# NOTE!! ANYTHING LIBXML, HTML5LIB ETC WILL CAUSE SOME SMALL MEMORY LEAK IN THE LOCAL "LIB" IMPLEMENTATION OUTSIDE PYTHON
+def html_to_text(html_content: str, render_anchor_tag_content=False, is_rss=False):
+    from multiprocessing import Process, Pipe
 
+    parent_conn, child_conn = Pipe()
+    p = Process(target=html_to_text_sub_worker, args=(child_conn, html_content, render_anchor_tag_content, is_rss))
+    p.start()
+    text = parent_conn.recv()
+    p.join()
+    return text
 
 # Does LD+JSON exist with a @type=='product' and a .price set anywhere?
 def has_ldjson_product_info(content):

changedetectionio/templates/_helpers.html

@@ -98,15 +98,13 @@
 
 
 {% macro playwright_warning() %}
-    <p><strong>Error - Playwright support for Chrome based fetching is not enabled.</strong> Alternatively try our <a href="https://changedetection.io">very affordable subscription based service which has all this setup for you</a>.</p>
+    <p><strong>Error - This watch needs Chrome (with playwright/sockpuppetbrowser), but Chrome based fetching is not enabled.</strong> Alternatively try our <a href="https://changedetection.io">very affordable subscription based service which has all this setup for you</a>.</p>
     <p>You may need to <a href="https://github.com/dgtlmoon/changedetection.io/blob/09ebc6ec6338545bdd694dc6eee57f2e9d2b8075/docker-compose.yml#L31">Enable playwright environment variable</a> and uncomment the <strong>sockpuppetbrowser</strong> in the <a href="https://github.com/dgtlmoon/changedetection.io/blob/master/docker-compose.yml">docker-compose.yml</a> file.</p>
     <br>
-    <p>(Also Selenium/WebDriver can not extract full page screenshots reliably so Playwright is recommended here)</p>
-
 {% endmacro %}
 
-{% macro only_webdriver_type_watches_warning() %}
-    <p><strong>Sorry, this functionality only works with Playwright/Chrome enabled watches.<br>You need to <a href="#request">Set the fetch method to Playwright/Chrome mode and resave</a> and have the Playwright connection enabled.</strong></p><br>
+{% macro only_playwright_type_watches_warning() %}
+    <p><strong>Sorry, this functionality only works with Playwright/Chrome enabled watches.<br>You need to <a href="#request">Set the fetch method to Playwright/Chrome mode and resave</a> and have the SockpuppetBrowser/Playwright or Selenium enabled.</strong></p><br>
 {% endmacro %}
 
 {% macro render_time_schedule_form(form, available_timezones, timezone_default_config) %}

changedetectionio/templates/edit.html

@@ -1,6 +1,6 @@
 {% extends 'base.html' %}
 {% block content %}
-{% from '_helpers.html' import render_field, render_checkbox_field, render_button, render_time_schedule_form, playwright_warning, only_webdriver_type_watches_warning, render_conditions_fieldlist_of_formfields_as_table %}
+{% from '_helpers.html' import render_field, render_checkbox_field, render_button, render_time_schedule_form, playwright_warning, only_playwright_type_watches_warning, render_conditions_fieldlist_of_formfields_as_table %}
 {% from '_common_fields.html' import render_common_settings_form %}
 <script src="{{url_for('static_content', group='js', filename='tabs.js')}}" defer></script>
 <script src="{{url_for('static_content', group='js', filename='vis.js')}}" defer></script>
@@ -204,7 +204,9 @@ Math: {{ 1 + 1 }}") }}
             </div>
 
             <div class="tab-pane-inner" id="browser-steps">
-            {% if playwright_enabled and watch_uses_webdriver %}
+            {% if watch_needs_selenium_or_playwright %}
+                {# Only works with playwright #}
+                {% if system_has_playwright_configured %}
                 <img class="beta-logo" src="{{url_for('static_content', group='images', filename='beta-logo.png')}}" alt="New beta functionality">
                 <fieldset>
                     <div class="pure-control-group">
@@ -223,7 +225,6 @@ Math: {{ 1 + 1 }}") }}
                         <div class="flex-wrapper" >
 
                             <div id="browser-steps-ui" class="noselect">
-
                                 <div class="noselect"  id="browsersteps-selector-wrapper" style="width: 100%">
                                     <span class="loader" >
                                         <span id="browsersteps-click-start">
@@ -245,15 +246,16 @@ Math: {{ 1 + 1 }}") }}
                     </div>
                 </fieldset>
                 {% else %}
-                    <span class="pure-form-message-inline">
-                        {% if not watch_uses_webdriver %}
-                            {{ only_webdriver_type_watches_warning() }}
-                        {% endif %}
-                        {%  if not playwright_enabled %}
-                            {{ playwright_warning() }}
-                        {% endif %}
-                    </span>
+                    {# it's configured to use selenium or chrome but system says its not configured #}
+                    {{ playwright_warning() }}
+                    {% if system_has_webdriver_configured %}
+                        <strong>Selenium/Webdriver cant be used here because it wont fetch screenshots reliably.</strong>
+                    {% endif %}
                 {% endif %}
+            {% else %}
+                {# "This functionality needs chrome.." #}
+                {{ only_playwright_type_watches_warning() }}
+            {% endif %}
             </div>
 
 
@@ -262,7 +264,7 @@ Math: {{ 1 + 1 }}") }}
                     <div  class="pure-control-group inline-radio">
                       {{ render_checkbox_field(form.notification_muted) }}
                     </div>
-                    {% if watch_uses_webdriver %}
+                    {% if watch_needs_selenium_or_playwright %}
                     <div class="pure-control-group inline-radio">
                       {{ render_checkbox_field(form.notification_screenshot) }}
                         <span class="pure-form-message-inline">
@@ -379,7 +381,9 @@ Math: {{ 1 + 1 }}") }}
 
                 <fieldset>
                     <div class="pure-control-group">
-                        {% if playwright_enabled and watch_uses_webdriver %}
+                        {% if watch_needs_selenium_or_playwright %}
+                            {% if system_has_playwright_configured %}
+                                {%  if visual_selector_data_ready %}
                             <span class="pure-form-message-inline" id="visual-selector-heading">
                                 The Visual Selector tool lets you select the <i>text</i> elements that will be used for the change detection. It automatically fills-in the filters in the "CSS/JSONPath/JQ/XPath Filters" box of the <a href="#filters-and-triggers">Filters & Triggers</a> tab. Use <strong>Shift+Click</strong> to select multiple items.
                             </span>
@@ -396,14 +400,20 @@ Math: {{ 1 + 1 }}") }}
                                 <canvas id="selector-canvas"></canvas>
                             </div>
                             <div id="selector-current-xpath" style="overflow-x: hidden"><strong>Currently:</strong>&nbsp;<span class="text">Loading...</span></div>
+                            {% else %}
+                                <strong>Error, The Visual selector data is not ready, it needs to complete atleast one fetch, please queue the item and reload.</strong>
+                            {% endif %}
                         {% else %}
-                            {% if not watch_uses_webdriver %}
-                                {{ only_webdriver_type_watches_warning() }}
-                            {% endif %}
-                            {% if not playwright_enabled %}
-                                {{ playwright_warning() }}
-                            {% endif %}
+                            {# The watch needed chrome but system says that playwright is not ready #}
+                            {{ playwright_warning() }}
                         {% endif %}
+                            {% if system_has_webdriver_configured %}
+                                <strong>Selenium/Webdriver cant be used here because it wont fetch screenshots reliably.</strong>
+                            {% endif %}
+                    {% else %}
+                        {# "This functionality needs chrome.." #}
+                        {{ only_playwright_type_watches_warning() }}
+                    {% endif %}
                     </div>
                 </fieldset>
             </div>

docker-compose.yml

@@ -9,15 +9,20 @@ services:
 #        - ./proxies.json:/datastore/proxies.json
 
   #    environment:
-  #        Default listening port, can also be changed with the -p option
+  #        Default listening port, can also be changed with the -p option (not to be confused with ports: below)
   #      - PORT=5000
   #
   #        Log levels are in descending order. (TRACE is the most detailed one)
   #        Log output levels: TRACE, DEBUG(default), INFO, SUCCESS, WARNING, ERROR, CRITICAL
   #      - LOGGER_LEVEL=TRACE
   #
-  #       Alternative WebDriver/selenium URL, do not use "'s or 's!
-  #      - WEBDRIVER_URL=http://browser-chrome:4444/wd/hub
+  #
+  #       Uncomment below and the "sockpuppetbrowser" to use a real Chrome browser (It uses the "playwright" protocol)
+  #      - PLAYWRIGHT_DRIVER_URL=ws://browser-sockpuppet-chrome:3000
+  #
+  #
+  #       Alternative WebDriver/selenium URL, do not use "'s or 's! (old, deprecated, does not support screenshots very well)
+  #      - WEBDRIVER_URL=http://browser-selenium-chrome:4444/wd/hub
   #
   #       WebDriver proxy settings webdriver_proxyType, webdriver_ftpProxy, webdriver_noProxy,
   #                                webdriver_proxyAutoconfigUrl, webdriver_autodetect,
@@ -25,9 +30,6 @@ services:
   #
   #             https://selenium-python.readthedocs.io/api.html#module-selenium.webdriver.common.proxy
   #
-  #       Alternative target "Chrome" Playwright URL, do not use "'s or 's!
-  #       "Playwright" is a driver/librarythat allows changedetection to talk to a Chrome or similar browser.
-  #      - PLAYWRIGHT_DRIVER_URL=ws://sockpuppetbrowser:3000
   #
   #       Playwright proxy settings playwright_proxy_server, playwright_proxy_bypass, playwright_proxy_username, playwright_proxy_password
   #
@@ -43,7 +45,7 @@ services:
   #        Base URL of your changedetection.io install (Added to the notification alert)
   #      - BASE_URL=https://mysite.com
   #        Respect proxy_pass type settings, `proxy_set_header Host "localhost";` and `proxy_set_header X-Forwarded-Prefix /app;`
-  #        More here https://github.com/dgtlmoon/changedetection.io/wiki/Running-changedetection.io-behind-a-reverse-proxy-sub-directory
+  #        More here https://github.com/dgtlmoon/changedetection.io/wiki/Running-changedetection.io-behind-a-reverse-proxy
   #      - USE_X_SETTINGS=1
   #
   #        Hides the `Referer` header so that monitored websites can't see the changedetection.io hostname.
@@ -86,8 +88,8 @@ services:
 
      # Sockpuppetbrowser is basically chrome wrapped in an API for allowing fast fetching of web-pages.
      # RECOMMENDED FOR FETCHING PAGES WITH CHROME, be sure to enable the "PLAYWRIGHT_DRIVER_URL" env variable in the main changedetection container
-#    sockpuppetbrowser:
-#        hostname: sockpuppetbrowser
+#    browser-sockpuppet-chrome:
+#        hostname: browser-sockpuppet-chrome
 #        image: dgtlmoon/sockpuppetbrowser:latest
 #        cap_add:
 #            - SYS_ADMIN
@@ -102,14 +104,18 @@ services:
      # Used for fetching pages via Playwright+Chrome where you need Javascript support.
      # Note: Works well but is deprecated, does not fetch full page screenshots (doesnt work with Visual Selector)
      #       Does not report status codes (200, 404, 403) and other issues
-#    browser-chrome:
-#        hostname: browser-chrome
+#    browser-selenium-chrome:
+#        hostname: browser-selenium-chrome
 #        image: selenium/standalone-chrome:4
 #        environment:
 #            - VNC_NO_PASSWORD=1
 #            - SCREEN_WIDTH=1920
 #            - SCREEN_HEIGHT=1080
 #            - SCREEN_DEPTH=24
+#          CHROME_OPTIONS: |
+#            --window-size=1280,1024
+#            --headless
+#            --disable-gpu
 #        volumes:
 #            # Workaround to avoid the browser crashing inside a docker container
 #            # See https://github.com/SeleniumHQ/docker-selenium#quick-start

docs/api_v1/index.html

@@ -5,13 +5,13 @@
   <meta name="description" content="Manage your changedetection.io watches via API, requires the `x-api-key` header which is found in the settings UI.">
   <meta name="viewport" content="width=device-width, initial-scale=1.0">
   <meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
-  <link href="assets/bootstrap.min.css?v=1701595483622" rel="stylesheet" media="screen">
-  <link href="assets/prism.css?v=1701595483622" rel="stylesheet" />
-  <link href="assets/main.css?v=1701595483622" rel="stylesheet" media="screen, print">
-  <link href="assets/favicon.ico?v=1701595483622" rel="icon" type="image/x-icon">
-  <link href="assets/apple-touch-icon.png?v=1701595483622" rel="apple-touch-icon" sizes="180x180">
-  <link href="assets/favicon-32x32.png?v=1701595483622" rel="icon" type="image/png" sizes="32x32">
-  <link href="assets/favicon-16x16.png?v=1701595483622" rel="icon" type="image/png" sizes="16x16">
+  <link href="assets/bootstrap.min.css?v=1744573753999" rel="stylesheet" media="screen">
+  <link href="assets/prism.css?v=1744573753999" rel="stylesheet" />
+  <link href="assets/main.css?v=1744573753999" rel="stylesheet" media="screen, print">
+  <link href="assets/favicon.ico?v=1744573753999" rel="icon" type="image/x-icon">
+  <link href="assets/apple-touch-icon.png?v=1744573753999" rel="apple-touch-icon" sizes="180x180">
+  <link href="assets/favicon-32x32.png?v=1744573753999" rel="icon" type="image/png" sizes="32x32">
+  <link href="assets/favicon-16x16.png?v=1744573753999" rel="icon" type="image/png" sizes="16x16">
 </head>
 
 <body class="container-fluid">
@@ -928,6 +928,6 @@
   </div>
 </div>
 
-<script src="assets/main.bundle.js?v=1701595483622"></script>
+<script src="assets/main.bundle.js?v=1744573753999"></script>
 </body>
 </html>