Merge branch 'master' into selenium-proxy-fix

Co-authored-by: silversub <silversub@gmail.com>

changedetectionio/blueprint/browser_steps/__init__.py

@@ -168,9 +168,7 @@ def construct_blueprint(datastore: ChangeDetectionStore):
             step_optional_value = request.form.get('optional_value')
             is_last_step = strtobool(request.form.get('is_last_step'))
 
-            # @todo try.. accept.. nice errors not popups..
             try:
-
                 browsersteps_sessions[browsersteps_session_id]['browserstepper'].call_action(action_name=step_operation,
                                          selector=step_selector,
                                          optional_value=step_optional_value)

changedetectionio/blueprint/browser_steps/browser_steps.py

@@ -61,23 +61,6 @@ class steppable_browser_interface():
 
     def __init__(self, start_url):
         self.start_url = start_url
-        
-    def safe_page_operation(self, operation_fn, default_return=None):
-        """Safely execute a page operation with error handling"""
-        if self.page is None:
-            logger.warning("Attempted operation on None page object")
-            return default_return
-            
-        try:
-            return operation_fn()
-        except Exception as e:
-            logger.debug(f"Page operation failed: {str(e)}")
-            # Try to reclaim memory if possible
-            try:
-                self.page.request_gc()
-            except:
-                pass
-            return default_return
 
     # Convert and perform "Click Button" for example
     def call_action(self, action_name, selector=None, optional_value=None):
@@ -109,20 +92,11 @@ class steppable_browser_interface():
         if optional_value and ('{%' in optional_value or '{{' in optional_value):
             optional_value = jinja_render(template_str=optional_value)
 
-        try:
-            action_handler(selector, optional_value)
-            # Safely wait for timeout
-            def wait_timeout():
-                self.page.wait_for_timeout(1.5 * 1000)
-            self.safe_page_operation(wait_timeout)
-            logger.debug(f"Call action done in {time.time()-now:.2f}s")
-        except Exception as e:
-            logger.error(f"Error executing action '{call_action_name}': {str(e)}")
-            # Request garbage collection to free up resources after error
-            try:
-                self.page.request_gc()
-            except:
-                pass
+
+        action_handler(selector, optional_value)
+        # Safely wait for timeout
+        self.page.wait_for_timeout(1.5 * 1000)
+        logger.debug(f"Call action done in {time.time()-now:.2f}s")
 
     def action_goto_url(self, selector=None, value=None):
         if not value:
@@ -130,11 +104,7 @@ class steppable_browser_interface():
             return None
             
         now = time.time()
-        
-        def goto_operation():
-            return self.page.goto(value, timeout=0, wait_until='load')
-            
-        response = self.safe_page_operation(goto_operation)
+        response = self.page.goto(value, timeout=0, wait_until='load')
         logger.debug(f"Time to goto URL {time.time()-now:.2f}s")
         return response
 
@@ -147,53 +117,40 @@ class steppable_browser_interface():
         if not value or not len(value.strip()):
             return
             
-        def click_operation():
-            elem = self.page.get_by_text(value)
-            if elem.count():
-                elem.first.click(delay=randint(200, 500), timeout=self.action_timeout)
-                
-        self.safe_page_operation(click_operation)
+        elem = self.page.get_by_text(value)
+        if elem.count():
+            elem.first.click(delay=randint(200, 500), timeout=self.action_timeout)
+
 
     def action_click_element_containing_text_if_exists(self, selector=None, value=''):
         logger.debug("Clicking element containing text if exists")
         if not value or not len(value.strip()):
             return
             
-        def click_if_exists_operation():
-            elem = self.page.get_by_text(value)
-            logger.debug(f"Clicking element containing text - {elem.count()} elements found")
-            if elem.count():
-                elem.first.click(delay=randint(200, 500), timeout=self.action_timeout)
+        elem = self.page.get_by_text(value)
+        logger.debug(f"Clicking element containing text - {elem.count()} elements found")
+        if elem.count():
+            elem.first.click(delay=randint(200, 500), timeout=self.action_timeout)
                 
-        self.safe_page_operation(click_if_exists_operation)
 
     def action_enter_text_in_field(self, selector, value):
         if not selector or not len(selector.strip()):
             return
 
-        def fill_operation():
-            self.page.fill(selector, value, timeout=self.action_timeout)
-            
-        self.safe_page_operation(fill_operation)
+        self.page.fill(selector, value, timeout=self.action_timeout)
 
     def action_execute_js(self, selector, value):
         if not value:
             return None
             
-        def evaluate_operation():
-            return self.page.evaluate(value)
-            
-        return self.safe_page_operation(evaluate_operation)
+        return self.page.evaluate(value)
 
     def action_click_element(self, selector, value):
         logger.debug("Clicking element")
         if not selector or not len(selector.strip()):
             return
 
-        def click_operation():
-            self.page.click(selector=selector, timeout=self.action_timeout + 20 * 1000, delay=randint(200, 500))
-            
-        self.safe_page_operation(click_operation)
+        self.page.click(selector=selector, timeout=self.action_timeout + 20 * 1000, delay=randint(200, 500))
 
     def action_click_element_if_exists(self, selector, value):
         import playwright._impl._errors as _api_types
@@ -201,16 +158,14 @@ class steppable_browser_interface():
         if not selector or not len(selector.strip()):
             return
             
-        def click_if_exists_operation():
-            try:
-                self.page.click(selector, timeout=self.action_timeout, delay=randint(200, 500))
-            except _api_types.TimeoutError:
-                return
-            except _api_types.Error:
-                # Element was there, but page redrew and now its long long gone
-                return
+        try:
+            self.page.click(selector, timeout=self.action_timeout, delay=randint(200, 500))
+        except _api_types.TimeoutError:
+            return
+        except _api_types.Error:
+            # Element was there, but page redrew and now its long long gone
+            return
                 
-        self.safe_page_operation(click_if_exists_operation)
 
     def action_click_x_y(self, selector, value):
         if not value or not re.match(r'^\s?\d+\s?,\s?\d+\s?$', value):
@@ -222,10 +177,8 @@ class steppable_browser_interface():
             x = int(float(x.strip()))
             y = int(float(y.strip()))
             
-            def click_xy_operation():
-                self.page.mouse.click(x=x, y=y, delay=randint(200, 500))
+            self.page.mouse.click(x=x, y=y, delay=randint(200, 500))
                 
-            self.safe_page_operation(click_xy_operation)
         except Exception as e:
             logger.error(f"Error parsing x,y coordinates: {str(e)}")
 
@@ -233,27 +186,17 @@ class steppable_browser_interface():
         if not selector or not len(selector.strip()):
             return
 
-        def select_operation():
-            self.page.select_option(selector, label=value, timeout=self.action_timeout)
-
-        self.safe_page_operation(select_operation)
+        self.page.select_option(selector, label=value, timeout=self.action_timeout)
 
     def action_scroll_down(self, selector, value):
-        def scroll_operation():
-            # Some sites this doesnt work on for some reason
-            self.page.mouse.wheel(0, 600)
-            self.page.wait_for_timeout(1000)
-            
-        self.safe_page_operation(scroll_operation)
+        # Some sites this doesnt work on for some reason
+        self.page.mouse.wheel(0, 600)
+        self.page.wait_for_timeout(1000)
 
     def action_wait_for_seconds(self, selector, value):
         try:
             seconds = float(value.strip()) if value else 1.0
-            
-            def wait_operation():
-                self.page.wait_for_timeout(seconds * 1000)
-                
-            self.safe_page_operation(wait_operation)
+            self.page.wait_for_timeout(seconds * 1000)
         except (ValueError, TypeError) as e:
             logger.error(f"Invalid value for wait_for_seconds: {str(e)}")
 
@@ -263,14 +206,11 @@ class steppable_browser_interface():
             
         import json
         v = json.dumps(value)
-        
-        def wait_for_text_operation():
-            self.page.wait_for_function(
-                f'document.querySelector("body").innerText.includes({v});', 
-                timeout=30000
-            )
+        self.page.wait_for_function(
+            f'document.querySelector("body").innerText.includes({v});',
+            timeout=30000
+        )
             
-        self.safe_page_operation(wait_for_text_operation)
 
     def action_wait_for_text_in_element(self, selector, value):
         if not selector or not value:
@@ -280,82 +220,60 @@ class steppable_browser_interface():
         s = json.dumps(selector)
         v = json.dumps(value)
         
-        def wait_for_text_in_element_operation():
-            self.page.wait_for_function(
-                f'document.querySelector({s}).innerText.includes({v});', 
-                timeout=30000
-            )
-            
-        self.safe_page_operation(wait_for_text_in_element_operation)
+        self.page.wait_for_function(
+            f'document.querySelector({s}).innerText.includes({v});',
+            timeout=30000
+        )
 
     # @todo - in the future make some popout interface to capture what needs to be set
     # https://playwright.dev/python/docs/api/class-keyboard
     def action_press_enter(self, selector, value):
-        def press_operation():
-            self.page.keyboard.press("Enter", delay=randint(200, 500))
+        self.page.keyboard.press("Enter", delay=randint(200, 500))
             
-        self.safe_page_operation(press_operation)
 
     def action_press_page_up(self, selector, value):
-        def press_operation():
-            self.page.keyboard.press("PageUp", delay=randint(200, 500))
-            
-        self.safe_page_operation(press_operation)
+        self.page.keyboard.press("PageUp", delay=randint(200, 500))
 
     def action_press_page_down(self, selector, value):
-        def press_operation():
-            self.page.keyboard.press("PageDown", delay=randint(200, 500))
-            
-        self.safe_page_operation(press_operation)
+        self.page.keyboard.press("PageDown", delay=randint(200, 500))
 
     def action_check_checkbox(self, selector, value):
         if not selector:
             return
-            
-        def check_operation():
-            self.page.locator(selector).check(timeout=self.action_timeout)
-            
-        self.safe_page_operation(check_operation)
+
+        self.page.locator(selector).check(timeout=self.action_timeout)
 
     def action_uncheck_checkbox(self, selector, value):
         if not selector:
             return
             
-        def uncheck_operation():
-            self.page.locator(selector).uncheck(timeout=self.action_timeout)
+        self.page.locator(selector).uncheck(timeout=self.action_timeout)
             
-        self.safe_page_operation(uncheck_operation)
 
     def action_remove_elements(self, selector, value):
         """Removes all elements matching the given selector from the DOM."""
         if not selector:
             return
             
-        def remove_operation():
-            self.page.locator(selector).evaluate_all("els => els.forEach(el => el.remove())")
-            
-        self.safe_page_operation(remove_operation)
+        self.page.locator(selector).evaluate_all("els => els.forEach(el => el.remove())")
 
     def action_make_all_child_elements_visible(self, selector, value):
         """Recursively makes all child elements inside the given selector fully visible."""
         if not selector:
             return
             
-        def make_visible_operation():
-            self.page.locator(selector).locator("*").evaluate_all("""
-                els => els.forEach(el => {
-                    el.style.display = 'block';   // Forces it to be displayed
-                    el.style.visibility = 'visible';   // Ensures it's not hidden
-                    el.style.opacity = '1';   // Fully opaque
-                    el.style.position = 'relative';   // Avoids 'absolute' hiding
-                    el.style.height = 'auto';   // Expands collapsed elements
-                    el.style.width = 'auto';   // Ensures full visibility
-                    el.removeAttribute('hidden');   // Removes hidden attribute
-                    el.classList.remove('hidden', 'd-none');  // Removes common CSS hidden classes
-                })
-            """)
-            
-        self.safe_page_operation(make_visible_operation)
+        self.page.locator(selector).locator("*").evaluate_all("""
+            els => els.forEach(el => {
+                el.style.display = 'block';   // Forces it to be displayed
+                el.style.visibility = 'visible';   // Ensures it's not hidden
+                el.style.opacity = '1';   // Fully opaque
+                el.style.position = 'relative';   // Avoids 'absolute' hiding
+                el.style.height = 'auto';   // Expands collapsed elements
+                el.style.width = 'auto';   // Ensures full visibility
+                el.removeAttribute('hidden');   // Removes hidden attribute
+                el.classList.remove('hidden', 'd-none');  // Removes common CSS hidden classes
+            })
+        """)
 
 # Responsible for maintaining a live 'context' with the chrome CDP
 # @todo - how long do contexts live for anyway?

changedetectionio/conditions/__init__.py

@@ -5,7 +5,7 @@ from json_logic.builtins import BUILTINS
 from .exceptions import EmptyConditionRuleRowNotUsable
 from .pluggy_interface import plugin_manager  # Import the pluggy plugin manager
 from . import default_plugin
-
+from loguru import logger
 # List of all supported JSON Logic operators
 operator_choices = [
     (None, "Choose one - Operator"),
@@ -113,12 +113,14 @@ def execute_ruleset_against_all_plugins(current_watch_uuid: str, application_dat
                             application_datastruct=application_datastruct,
                             ephemeral_data=ephemeral_data
                         )
-                        
+                        logger.debug(f"Trying plugin {plugin}....")
+
                         # Set a timeout of 10 seconds
                         try:
                             new_execute_data = future.result(timeout=10)
                             if new_execute_data and isinstance(new_execute_data, dict):
                                 EXECUTE_DATA.update(new_execute_data)
+
                         except concurrent.futures.TimeoutError:
                             # The plugin took too long, abort processing for this watch
                             raise Exception(f"Plugin {plugin.__class__.__name__} took more than 10 seconds to run.")

changedetectionio/conditions/plugins/levenshtein_plugin.py

@@ -9,15 +9,20 @@ def levenshtein_ratio_recent_history(watch, incoming_text=None):
     try:
         from Levenshtein import ratio, distance
         k = list(watch.history.keys())
-        if len(k) >= 2:
-            # When called from ui_edit_stats_extras, we don't have incoming_text
-            if incoming_text is None:
-                a = watch.get_history_snapshot(timestamp=k[-1])  # Latest snapshot
-                b = watch.get_history_snapshot(timestamp=k[-2])  # Previous snapshot
-            else:
-                a = watch.get_history_snapshot(timestamp=k[-2]) # Second newest, incoming_text will be "newest"
-                b = incoming_text
-            
+        a = None
+        b = None
+
+        # When called from ui_edit_stats_extras, we don't have incoming_text
+        if incoming_text is None:
+            a = watch.get_history_snapshot(timestamp=k[-1])  # Latest snapshot
+            b = watch.get_history_snapshot(timestamp=k[-2])  # Previous snapshot
+
+        # Needs atleast one snapshot
+        elif len(k) >= 1: # Should be atleast one snapshot to compare against
+            a = watch.get_history_snapshot(timestamp=k[-1]) # Latest saved snapshot
+            b = incoming_text if incoming_text else k[-2]
+
+        if a and b:
             distance_value = distance(a, b)
             ratio_value = ratio(a, b)
             return {
@@ -53,7 +58,7 @@ def add_data(current_watch_uuid, application_datastruct, ephemeral_data):
     # ephemeral_data['text'] will be the current text after filters, they may have edited filters but not saved them yet etc
 
     if watch and 'text' in ephemeral_data:
-        lev_data = levenshtein_ratio_recent_history(watch, ephemeral_data['text'])
+        lev_data = levenshtein_ratio_recent_history(watch, ephemeral_data.get('text',''))
         if isinstance(lev_data, dict):
             res['levenshtein_ratio'] = lev_data.get('ratio', 0)
             res['levenshtein_similarity'] = lev_data.get('percent_similar', 0)

changedetectionio/content_fetchers/playwright.py

@@ -194,7 +194,6 @@ class fetcher(Fetcher):
             browsersteps_interface.page = self.page
 
             response = browsersteps_interface.action_goto_url(value=url)
-            self.headers = response.all_headers()
 
             if response is None:
                 context.close()
@@ -202,6 +201,8 @@ class fetcher(Fetcher):
                 logger.debug("Content Fetcher > Response object from the browser communication was none")
                 raise EmptyReply(url=url, status_code=None)
 
+            self.headers = response.all_headers()
+
             try:
                 if self.webdriver_js_execute_code is not None and len(self.webdriver_js_execute_code):
                     browsersteps_interface.action_execute_js(value=self.webdriver_js_execute_code, selector=None)

changedetectionio/content_fetchers/puppeteer.py

@@ -147,7 +147,7 @@ class fetcher(Fetcher):
                          is_binary,
                          empty_pages_are_a_change
                          ):
-
+        import re
         self.delete_browser_steps_screenshots()
         extra_wait = int(os.getenv("WEBDRIVER_DELAY_BEFORE_CONTENT_READY", 5)) + self.render_extract_delay
 
@@ -172,6 +172,17 @@ class fetcher(Fetcher):
         # headless - ask a new page
         self.page = (pages := await browser.pages) and len(pages) or await browser.newPage()
 
+        if '--window-size' in self.browser_connection_url:
+            # Be sure the viewport is always the window-size, this is often not the same thing
+            match = re.search(r'--window-size=(\d+),(\d+)', self.browser_connection_url)
+            if match:
+                logger.debug(f"Setting viewport to same as --window-size in browser connection URL {int(match.group(1))},{int(match.group(2))}")
+                await self.page.setViewport({
+                    "width": int(match.group(1)),
+                    "height": int(match.group(2))
+                })
+                logger.debug(f"Puppeteer viewport size {self.page.viewport}")
+
         try:
             from pyppeteerstealth import inject_evasions_into_page
         except ImportError:
@@ -218,7 +229,6 @@ class fetcher(Fetcher):
 
         response = await self.page.goto(url, waitUntil="load")
 
-
         if response is None:
             await self.page.close()
             await browser.close()

changedetectionio/content_fetchers/requests.py

@@ -28,6 +28,7 @@ class fetcher(Fetcher):
 
         import chardet
         import requests
+        from requests.exceptions import ProxyError, ConnectionError, RequestException
 
         if self.browser_steps_get_valid_steps():
             raise BrowserStepsInUnsupportedFetcher(url=url)
@@ -52,14 +53,19 @@ class fetcher(Fetcher):
         if strtobool(os.getenv('ALLOW_FILE_URI', 'false')) and url.startswith('file://'):
             from requests_file import FileAdapter
             session.mount('file://', FileAdapter())
-
-        r = session.request(method=request_method,
-                            data=request_body.encode('utf-8') if type(request_body) is str else request_body,
-                            url=url,
-                            headers=request_headers,
-                            timeout=timeout,
-                            proxies=proxies,
-                            verify=False)
+        try:
+            r = session.request(method=request_method,
+                                data=request_body.encode('utf-8') if type(request_body) is str else request_body,
+                                url=url,
+                                headers=request_headers,
+                                timeout=timeout,
+                                proxies=proxies,
+                                verify=False)
+        except Exception as e:
+            msg = str(e)
+            if proxies and 'SOCKSHTTPSConnectionPool' in msg:
+                msg = f"Proxy connection failed? {msg}"
+            raise Exception(msg) from e
 
         # If the response did not tell us what encoding format to expect, Then use chardet to override what `requests` thinks.
         # For example - some sites don't tell us it's utf-8, but return utf-8 content

changedetectionio/content_fetchers/res/stock-not-in-stock.js

@@ -51,6 +51,7 @@ async () => {
             'niet op voorraad',
             'no disponible',
             'no featured offers available',
+            'no longer available',
             'no longer in stock',
             'no tickets available',
             'non disponibile',
@@ -125,6 +126,20 @@ async () => {
         // so it's good to filter to just the 'above the fold' elements
         // and it should be atleast 100px from the top to ignore items in the toolbar, sometimes menu items like "Coming soon" exist
 
+        function elementIsInEyeBallRange(element) {
+            // outside the 'fold' or some weird text in the heading area
+            // .getBoundingClientRect() was causing a crash in chrome 119, can only be run on contentVisibility != hidden
+            // Note: theres also an automated test that places the 'out of stock' text fairly low down
+            // Skip text that could be in the header area
+            if (element.getBoundingClientRect().bottom + window.scrollY <= 300 ) {
+                return false;
+            }
+            // Skip text that could be much further down (like a list of "you may like" products that have 'sold out' in there
+            if (element.getBoundingClientRect().bottom + window.scrollY >= 1300 ) {
+                return false;
+            }
+            return true;
+        }
 
 // @todo - if it's SVG or IMG, go into image diff mode
 
@@ -161,9 +176,7 @@ async () => {
         for (let i = elementsToScan.length - 1; i >= 0; i--) {
             const element = elementsToScan[i];
 
-            // outside the 'fold' or some weird text in the heading area
-            // .getBoundingClientRect() was causing a crash in chrome 119, can only be run on contentVisibility != hidden
-            if (element.getBoundingClientRect().top + window.scrollY >= vh || element.getBoundingClientRect().top + window.scrollY <= 100) {
+            if (!elementIsInEyeBallRange(element)) {
                 continue
             }
 
@@ -177,11 +190,11 @@ async () => {
             } catch (e) {
                 console.warn('stock-not-in-stock.js scraper - handling element for gettext failed', e);
             }
-
             if (elementText.length) {
                 // try which ones could mean its in stock
                 if (negateOutOfStockRegex.test(elementText) && !elementText.includes('(0 products)')) {
                     console.log(`Negating/overriding 'Out of Stock' back to "Possibly in stock" found "${elementText}"`)
+                    element.style.border = "2px solid green"; // highlight the element that was detected as in stock
                     return 'Possibly in stock';
                 }
             }
@@ -190,10 +203,8 @@ async () => {
         // OTHER STUFF THAT COULD BE THAT IT'S OUT OF STOCK
         for (let i = elementsToScan.length - 1; i >= 0; i--) {
             const element = elementsToScan[i];
-            // outside the 'fold' or some weird text in the heading area
-            // .getBoundingClientRect() was causing a crash in chrome 119, can only be run on contentVisibility != hidden
-            // Note: theres also an automated test that places the 'out of stock' text fairly low down
-            if (element.getBoundingClientRect().top + window.scrollY >= vh + 250 || element.getBoundingClientRect().top + window.scrollY <= 100) {
+
+            if (!elementIsInEyeBallRange(element)) {
                 continue
             }
             elementText = "";
@@ -208,6 +219,7 @@ async () => {
                 for (const outOfStockText of outOfStockTexts) {
                     if (elementText.includes(outOfStockText)) {
                         console.log(`Selected 'Out of Stock' - found text "${outOfStockText}" - "${elementText}" - offset top ${element.getBoundingClientRect().top}, page height is ${vh}`)
+                        element.style.border = "2px solid red"; // highlight the element that was detected as out of stock
                         return outOfStockText; // item is out of stock
                     }
                 }

changedetectionio/content_fetchers/res/xpath_element_scraper.js

@@ -202,7 +202,6 @@ async (options) => {
         // Foreach filter, go and find it on the page and add it to the results so we can visualise it again
         for (const f of include_filters) {
             bbox = false;
-            q = false;
 
             if (!f.length) {
                 console.log("xpath_element_scraper: Empty filter, skipping");
@@ -255,7 +254,7 @@ async (options) => {
                             console.log("xpath_element_scraper: Got filter by ownerElement element, scroll from top was " + scroll_y)
                         } catch (e) {
                             console.log(e)
-                            console.log("xpath_element_scraper: error looking up q.ownerElement")
+                            console.log("xpath_element_scraper: error looking up node.ownerElement")
                         }
                     }
 

changedetectionio/content_fetchers/webdriver_selenium.py

@@ -76,8 +76,7 @@ class fetcher(Fetcher):
         for opt in CHROME_OPTIONS:
             options.add_argument(opt)
 
-        if self.proxy:
-            options.proxy = self.proxy
+        options.add_argument(f"--proxy-server={self.proxy}")
 
         self.driver = webdriver.Remote(
             command_executor=self.browser_connection_url,

changedetectionio/run_proxy_tests.sh

@@ -82,3 +82,26 @@ done
 
 
 docker kill squid-one squid-two squid-custom
+
+# Test that the UI is returning the correct error message when a proxy is not available
+
+# Requests
+docker run --network changedet-network \
+  test-changedetectionio \
+  bash -c 'cd changedetectionio && pytest tests/proxy_list/test_proxy_noconnect.py'
+
+# Playwright
+docker run --network changedet-network \
+  test-changedetectionio \
+  bash -c 'cd changedetectionio && PLAYWRIGHT_DRIVER_URL=ws://sockpuppetbrowser:3000 pytest tests/proxy_list/test_proxy_noconnect.py'
+
+# Puppeteer fast
+docker run --network changedet-network \
+  test-changedetectionio \
+  bash -c 'cd changedetectionio && FAST_PUPPETEER_CHROME_FETCHER=1 PLAYWRIGHT_DRIVER_URL=ws://sockpuppetbrowser:3000 pytest tests/proxy_list/test_proxy_noconnect.py'
+
+# Selenium - todo - fix proxies
+docker run --network changedet-network \
+  -e "WEBDRIVER_URL=http://selenium:4444/wd/hub" \
+  test-changedetectionio \
+  bash -c 'cd changedetectionio && pytest tests/proxy_list/test_proxy_noconnect.py'

changedetectionio/tests/proxy_list/test_proxy_noconnect.py

@@ -0,0 +1,53 @@
+#!/usr/bin/env python3
+
+from flask import url_for
+from ..util import live_server_setup, wait_for_all_checks
+import os
+from ... import strtobool
+
+
+# Just to be sure the UI outputs the right error message on proxy connection failed
+def test_proxy_noconnect_custom(client, live_server, measure_memory_usage):
+    live_server_setup(live_server)
+
+    # Goto settings, add our custom one
+    res = client.post(
+        url_for("settings.settings_page"),
+        data={
+            "requests-time_between_check-minutes": 180,
+            "application-ignore_whitespace": "y",
+            "application-fetch_backend": 'html_webdriver' if os.getenv('PLAYWRIGHT_DRIVER_URL') else 'html_requests',
+            "requests-extra_proxies-0-proxy_name": "custom-test-proxy",
+            # test:awesome is set in tests/proxy_list/squid-passwords.txt
+            "requests-extra_proxies-0-proxy_url": "http://THISPROXYDOESNTEXIST:3128",
+        },
+        follow_redirects=True
+    )
+
+    assert b"Settings updated." in res.data
+
+    res = client.post(
+        url_for("imports.import_page"),
+        # Because a URL wont show in squid/proxy logs due it being SSLed
+        # Use plain HTTP or a specific domain-name here
+        data={"urls": "https://changedetection.io/CHANGELOG.txt"},
+        follow_redirects=True
+    )
+
+    assert b"1 Imported" in res.data
+    wait_for_all_checks(client)
+
+    res = client.get(url_for("watchlist.index"))
+    assert b'Page.goto: net::ERR_PROXY_CONNECTION_FAILED' in res.data
+
+    # Requests
+    check_string = b'Proxy connection failed?'
+
+    if os.getenv('PLAYWRIGHT_DRIVER_URL') or strtobool(os.getenv('FAST_PUPPETEER_CHROME_FETCHER', 'False')):
+        check_string = b'ERR_PROXY_CONNECTION_FAILED'
+
+    if os.getenv("WEBDRIVER_URL"):
+        check_string = b'ERR_PROXY_CONNECTION_FAILED'
+
+    assert check_string in res.data
+

changedetectionio/tests/restock/test_restock.py

@@ -14,6 +14,8 @@ from changedetectionio.notification import (
 def set_original_response():
     test_return_data = """<html>
        <body>
+       <section id=header style="padding: 50px; height: 350px">This is the header which should be ignored always - <span>add to cart</span></section>
+       <!-- stock-not-in-stock.js will ignore text in the first 300px, see elementIsInEyeBallRange(), sometimes "add to cart" and other junk is here -->
      Some initial text<br>
      <p>Which is across multiple lines</p>
      <br>
@@ -52,8 +54,6 @@ def test_restock_detection(client, live_server, measure_memory_usage):
 
     set_original_response()
     #assert os.getenv('PLAYWRIGHT_DRIVER_URL'), "Needs PLAYWRIGHT_DRIVER_URL set for this test"
-
-    time.sleep(1)
     live_server_setup(live_server)
     #####################
     notification_url = url_for('test_notification_endpoint', _external=True).replace('http://localhost', 'http://changedet').replace('http', 'json')
@@ -84,7 +84,8 @@ def test_restock_detection(client, live_server, measure_memory_usage):
     # Is it correctly show as NOT in stock?
     wait_for_all_checks(client)
     res = client.get(url_for("watchlist.index"))
-    assert b'not-in-stock' in res.data
+    assert b'processor-restock_diff' in res.data # Should have saved in restock mode
+    assert b'not-in-stock' in res.data # should be out of stock
 
     # Is it correctly shown as in stock
     set_back_in_stock_response()

changedetectionio/tests/test_conditions.py

@@ -196,7 +196,11 @@ def test_condition_validate_rule_row(client, live_server):
     )
     assert res.status_code == 200
     assert b'false' in res.data
-
+    # cleanup for the next
+    client.get(
+        url_for("ui.form_delete", uuid="all"),
+        follow_redirects=True
+    )
 
 
 
@@ -235,4 +239,107 @@ def test_wordcount_conditions_plugin(client, live_server, measure_memory_usage):
     )
 
     # Assert the word count is counted correctly
-    assert b'<td>13</td>' in res.data
+    assert b'<td>13</td>' in res.data
+
+    # cleanup for the next
+    client.get(
+        url_for("ui.form_delete", uuid="all"),
+        follow_redirects=True
+    )
+
+# If there was only a change in the whitespacing, then we shouldnt have a change detected
+def test_lev_conditions_plugin(client, live_server, measure_memory_usage):
+    #live_server_setup(live_server)
+
+    with open("test-datastore/endpoint-content.txt", "w") as f:
+        f.write("""<html>
+       <body>
+     Some initial text<br>
+     <p>Which is across multiple lines</p>
+     <br>
+     So let's see what happens.  <br>
+     </body>
+     </html>
+    """)
+
+    # Add our URL to the import page
+    test_url = url_for('test_endpoint', _external=True)
+    res = client.post(
+        url_for("ui.ui_views.form_quick_watch_add"),
+        data={"url": test_url, "tags": '', 'edit_and_watch_submit_button': 'Edit > Watch'},
+        follow_redirects=True
+    )
+    assert b"Watch added in Paused state, saving will unpause" in res.data
+
+    uuid = next(iter(live_server.app.config['DATASTORE'].data['watching']))
+    # Give the thread time to pick it up
+    wait_for_all_checks(client)
+    res = client.post(
+        url_for("ui.ui_edit.edit_page", uuid=uuid, unpause_on_save=1),
+        data={
+            "url": test_url,
+            "fetch_backend": "html_requests",
+            "conditions_match_logic": "ALL",  # ALL = AND logic
+            "conditions-0-field": "levenshtein_ratio",
+            "conditions-0-operator": "<",
+            "conditions-0-value": "0.8" # needs to be more of a diff to trigger a change
+        },
+        follow_redirects=True
+    )
+
+    assert b"unpaused" in res.data
+
+    wait_for_all_checks(client)
+    res = client.get(url_for("watchlist.index"))
+    assert b'unviewed' not in res.data
+
+    # Check the content saved initially, even tho a condition was set - this is the first snapshot so shouldnt be affected by conditions
+    res = client.get(
+        url_for("ui.ui_views.preview_page", uuid=uuid),
+        follow_redirects=True
+    )
+    assert b'Which is across multiple lines' in res.data
+
+
+    ############### Now change it a LITTLE bit...
+    with open("test-datastore/endpoint-content.txt", "w") as f:
+        f.write("""<html>
+       <body>
+     Some initial text<br>
+     <p>Which is across multiple lines</p>
+     <br>
+     So let's see what happenxxxxxxxxx.  <br>
+     </body>
+     </html>
+    """)
+
+    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)
+
+    res = client.get(url_for("watchlist.index"))
+    assert b'unviewed' not in res.data #because this will be like 0.90 not 0.8 threshold
+
+    ############### Now change it a MORE THAN 50%
+    test_return_data = """<html>
+       <body>
+     Some sxxxx<br>
+     <p>Which is across a lines</p>
+     <br>
+     ok.  <br>
+     </body>
+     </html>
+    """
+
+    with open("test-datastore/endpoint-content.txt", "w") as f:
+        f.write(test_return_data)
+    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)
+    res = client.get(url_for("watchlist.index"))
+    assert b'unviewed' in res.data
+    # cleanup for the next
+    client.get(
+        url_for("ui.form_delete", uuid="all"),
+        follow_redirects=True
+    )

changedetectionio/tests/test_ignore_text.py

@@ -32,13 +32,14 @@ def test_strip_text_func():
     stripped_content = html_tools.strip_ignore_text(test_content, ignore)
     assert stripped_content == "Some initial text\n\nWhich is across multiple lines\n\n\n\nSo let's see what happens."
 
-def set_original_ignore_response():
-    test_return_data = """<html>
+def set_original_ignore_response(ver_stamp="123"):
+    test_return_data = f"""<html>
        <body>
      Some initial text<br>
      <p>Which is across multiple lines</p>
      <br>
      So let's see what happens.  <br>
+     <link href="https://www.somesite/wp-content/themes/cooltheme/style2.css?v={ver_stamp}" rel="stylesheet"/>
      </body>
      </html>
 
@@ -48,13 +49,14 @@ def set_original_ignore_response():
         f.write(test_return_data)
 
 
-def set_modified_original_ignore_response():
-    test_return_data = """<html>
+def set_modified_original_ignore_response(ver_stamp="123"):
+    test_return_data = f"""<html>
        <body>
      Some NEW nice initial text<br>
      <p>Which is across multiple lines</p>
      <br>
      So let's see what happens.  <br>
+     <link href="https://www.somesite/wp-content/themes/cooltheme/style2.css?v={ver_stamp}" rel="stylesheet"/>
      <p>new ignore stuff</p>
      <p>blah</p>
      </body>
@@ -67,14 +69,15 @@ def set_modified_original_ignore_response():
 
 
 # Is the same but includes ZZZZZ, 'ZZZZZ' is the last line in ignore_text
-def set_modified_ignore_response():
-    test_return_data = """<html>
+def set_modified_ignore_response(ver_stamp="123"):
+    test_return_data = f"""<html>
        <body>
      Some initial text<br>
      <p>Which is across multiple lines</p>
      <P>ZZZZz</P>
      <br>
      So let's see what happens.  <br>
+     <link href="https://www.somesite/wp-content/themes/cooltheme/style2.css?v={ver_stamp}" rel="stylesheet"/>
      </body>
      </html>
 
@@ -165,9 +168,9 @@ def test_check_ignore_text_functionality(client, live_server, measure_memory_usa
     assert b'Deleted' in res.data
 
 # When adding some ignore text, it should not trigger a change, even if something else on that line changes
-def test_check_global_ignore_text_functionality(client, live_server, measure_memory_usage):
-    #live_server_setup(live_server)
-    ignore_text = "XXXXX\r\nYYYYY\r\nZZZZZ"
+def _run_test_global_ignore(client, as_source=False, extra_ignore=""):
+    ignore_text = "XXXXX\r\nYYYYY\r\nZZZZZ\r\n"+extra_ignore
+
     set_original_ignore_response()
 
     # Goto the settings page, add our ignore text
@@ -186,6 +189,10 @@ def test_check_global_ignore_text_functionality(client, live_server, measure_mem
 
     # Add our URL to the import page
     test_url = url_for('test_endpoint', _external=True)
+    if as_source:
+        # Switch to source mode so we can test that too!
+        test_url = "source:"+test_url
+
     res = client.post(
         url_for("imports.import_page"),
         data={"urls": test_url},
@@ -203,12 +210,15 @@ def test_check_global_ignore_text_functionality(client, live_server, measure_mem
         follow_redirects=True
     )
     assert b"Updated watch." in res.data
-
+    wait_for_all_checks(client)
     # Check it saved
     res = client.get(
         url_for("settings.settings_page"),
     )
-    assert bytes(ignore_text.encode('utf-8')) in res.data
+
+    for i in ignore_text.splitlines():
+        assert bytes(i.encode('utf-8')) in res.data
+
 
     # Trigger a check
     client.get(url_for("ui.form_watch_checknow"), follow_redirects=True)
@@ -221,7 +231,8 @@ def test_check_global_ignore_text_functionality(client, live_server, measure_mem
 
     # Make a change which includes the ignore text, it should be ignored and no 'change' triggered
     # It adds text with "ZZZZzzzz" and "ZZZZ" is in the ignore list
-    set_modified_ignore_response()
+    # And tweaks the ver_stamp which should be picked up by global regex ignore
+    set_modified_ignore_response(ver_stamp=time.time())
 
     # Trigger a check
     client.get(url_for("ui.form_watch_checknow"), follow_redirects=True)
@@ -243,3 +254,11 @@ def test_check_global_ignore_text_functionality(client, live_server, measure_mem
 
     res = client.get(url_for("ui.form_delete", uuid="all"), follow_redirects=True)
     assert b'Deleted' in res.data
+
+def test_check_global_ignore_text_functionality(client, live_server):
+    #live_server_setup(live_server)
+    _run_test_global_ignore(client, as_source=False)
+
+def test_check_global_ignore_text_functionality_as_source(client, live_server):
+    #live_server_setup(live_server)
+    _run_test_global_ignore(client, as_source=True, extra_ignore='/\?v=\d/')

docker-compose.yml

@@ -72,7 +72,7 @@ services:
   
       # Comment out ports: when using behind a reverse proxy , enable networks: etc.
       ports:
-        - 5000:5000
+        - 127.0.0.1:5000:5000
       restart: unless-stopped
 
      # Used for fetching pages via WebDriver+Chrome where you need Javascript support.
@@ -82,7 +82,7 @@ services:
      # If WEBDRIVER or PLAYWRIGHT are enabled, changedetection container depends on that
      # and must wait before starting (substitute "browser-chrome" with "playwright-chrome" if last one is used)
 #      depends_on:
-#          sockpuppetbrowser:
+#          browser-sockpuppet-chrome:
 #              condition: service_started
 
 

requirements.txt

@@ -53,7 +53,8 @@ lxml >=4.8.0,<6,!=5.2.0,!=5.2.1
 # XPath 2.0-3.1 support - 4.2.0 broke something?
 elementpath==4.1.5
 
-selenium~=4.14.0
+selenium==4.31.0
+
 
 # https://github.com/pallets/werkzeug/issues/2985
 # Maybe related to pytest?
@@ -90,6 +91,8 @@ extruct
 # For cleaning up unknown currency formats
 babel
 
+levenshtein
+
 # Needed for > 3.10, https://github.com/microsoft/playwright-python/issues/2096
 greenlet >= 3.0.3