feat: Improved WebSelect Handling: Added Combobox Support, Enhanced Element Detection, and Smarter Option Matching

src/kleinanzeigen_bot/__init__.py

@@ -1059,9 +1059,13 @@ async def __set_special_attributes(self, ad_cfg:Ad) -> None:
             try:
                 # finding element by name cause id are composed sometimes eg. autos.marke_s+autos.model_s for Modell by cars
                 special_attr_elem = await self.web_find(By.XPATH, f"//*[contains(@name, '{special_attribute_key}')]")
-            except TimeoutError as ex:
-                LOG.debug("Attribute field '%s' could not be found.", special_attribute_key)
-                raise TimeoutError(f"Failed to set special attribute [{special_attribute_key}] (not found)") from ex
+            except TimeoutError:
+                # Trying to find element by ID instead cause sometimes there is NO name attribute...
+                try:
+                    special_attr_elem = await self.web_find(By.ID, special_attribute_key)
+                except TimeoutError as ex:
+                    LOG.debug("Attribute field '%s' could not be found.", special_attribute_key)
+                    raise TimeoutError(f"Failed to set special attribute by either ID or Name [{special_attribute_key}] (not found)") from ex
 
             try:
                 elem_id:str = str(special_attr_elem.attrs.id)
@@ -1071,6 +1075,9 @@ async def __set_special_attributes(self, ad_cfg:Ad) -> None:
                 elif special_attr_elem.attrs.type == "checkbox":
                     LOG.debug("Attribute field '%s' seems to be a checkbox...", special_attribute_key)
                     await self.web_click(By.ID, elem_id)
+                elif special_attr_elem.attrs.type == "text" and special_attr_elem.attrs.get("role") == "combobox":
+                    LOG.debug(_("Attribute field '%s' seems to be a Combobox (i.e. text input with filtering dropdown)..."), special_attribute_key)
+                    await self.web_select_combobox(By.ID, elem_id, special_attribute_value_str)
                 else:
                     LOG.debug("Attribute field '%s' seems to be a text input...", special_attribute_key)
                     await self.web_input(By.ID, elem_id, special_attribute_value_str)

src/kleinanzeigen_bot/resources/translations.de.yaml

@@ -125,6 +125,7 @@ kleinanzeigen_bot/__init__.py:
     "Attribute field '%s' is not of kind radio button.": "Attributfeld '%s' ist kein Radiobutton."
     "Attribute field '%s' seems to be a checkbox...": "Attributfeld '%s' scheint eine Checkbox zu sein..."
     "Attribute field '%s' seems to be a text input...": "Attributfeld '%s' scheint ein Texteingabefeld zu sein..."
+    "Attribute field '%s' seems to be a Combobox (i.e. text input with filtering dropdown)...": "Attributfeld '%s' scheint eine Combobox zu sein (d.h. Texteingabefeld mit Dropdown-Filter)..."
 
   download_ads:
     "Scanning your ad overview...": "Scanne Anzeigenübersicht..."
@@ -401,8 +402,14 @@ kleinanzeigen_bot/utils/web_scraping_mixin.py:
   web_find:
     "Unsupported selector type: %s": "Nicht unterstützter Selektor-Typ: %s"
 
+  web_select_combobox:
+    "Combobox input field does not have 'aria-controls' attribute to locate dropdown options.": "Das Eingabefeld der Combobox hat kein 'aria-controls'-Attribut, um Dropdown-Optionen zu finden."
+    "Cannot locate combobox dropdown options.": "Kann Combobox-Dropdown-Optionen nicht finden."
+    "No <li> options found in combobox dropdown with id '%s'.": "Keine <li>-Optionen im Combobox-Dropdown mit der ID '%s' gefunden."
+
   web_find_all:
     "Unsupported selector type: %s": "Nicht unterstützter Selektor-Typ: %s"
+
   close_browser_session:
     "Closing Browser session...": "Schließe Browser-Sitzung..."
 

src/kleinanzeigen_bot/utils/web_scraping_mixin.py

@@ -848,22 +848,101 @@ async def web_select(self, selector_type:By, selector_value:str, selected_value:
             timeout_error_message = f"No clickable HTML element with selector: {selector_type}='{selector_value}' found"
         )
         elem = await self.web_find(selector_type, selector_value)
+
+        js_value = json.dumps(selected_value)  # safe escaping for JS
         await elem.apply(f"""
             function (element) {{
-              for(let i=0; i < element.options.length; i++)
-                {{
-                  if(element.options[i].value == "{selected_value}") {{
-                    element.selectedIndex = i;
-                    element.dispatchEvent(new Event('change', {{ bubbles: true }}));
-                    break;
+                const wanted = String({js_value});
+
+                // 1) Try by value
+                for (let i = 0; i < element.options.length; i++) {{
+                    if (element.options[i].value === wanted) {{
+                        element.selectedIndex = i;
+                        element.dispatchEvent(new Event('change', {{ bubbles: true }}));
+                        return;
+                    }}
+                }}
+
+                // 2) Fallback by displayed text (trimmed)
+                const needle = wanted.trim();
+                for (let i = 0; i < element.options.length; i++) {{
+                    const opt = element.options[i];
+                    const shown = (opt.label ?? opt.text ?? opt.textContent ?? '').trim();
+                    if (shown === needle) {{
+                        element.selectedIndex = i;
+                        element.dispatchEvent(new Event('change', {{ bubbles: true }}));
+                        return;
+                    }}
                 }}
-              }}
-              throw new Error("Option with value {selected_value} not found.");
+
+                throw new Error("Option not found by value or displayed text: " + wanted);
             }}
         """)
         await self.web_sleep()
         return elem
 
+    async def web_select_combobox(self, selector_type:By, selector_value:str, selected_value:Any, timeout:int | float = 5) -> Element:
+        """
+        Selects an <li> Option of a <input/> HTML Combobox element by visible text.
+
+        :param timeout: timeout in seconds
+        :raises TimeoutError: if element could not be found within time
+        :raises UnexpectedTagNameException: if element is not a <select> element
+        """
+        input_field = await self.web_find(selector_type, selector_value, timeout = timeout)
+        await input_field.clear_input()
+        await input_field.send_keys(str(selected_value))
+        await self.web_sleep()
+
+        # From the Inputfield, get the attribute "aria-controls" which POINTS to the Dropdown ul #id:
+        dropdown_id = input_field.attrs.get("aria-controls")
+        if not dropdown_id:
+            LOG.error(_("Combobox input field does not have 'aria-controls' attribute to locate dropdown options."))
+            raise TimeoutError(_("Cannot locate combobox dropdown options."))
+
+        dropdown_elem = await self.web_find(By.ID, dropdown_id)
+        js_value = json.dumps(selected_value)  # safe escaping for JS
+
+        # This selects the correct <li> by visible text inside the dropdown. It includes normalization, i.e. trimming
+        # leading/trailing spaces and collapsing multiple spaces to single spaces for matching. It is done case-insensitive.
+        ok = await dropdown_elem.apply(f"""
+        function (element) {{
+          const selected = String({js_value});
+          const normalize = s => (s ?? '').replace(/\\s+/g, ' ').trim().toLowerCase();
+          // Normalize whitespace and convert to lowercase for comparison
+
+          // Get all <li> elements inside the dropdown
+          const items = element.querySelectorAll(':scope > li[role="option"], :scope > li');
+
+          for (const li of items) {{
+            // The visible label is typically inside the last <span>
+            const labelEl = li.querySelector(':scope > span:last-of-type');
+            const label = normalize(labelEl ? labelEl.textContent : li.textContent);
+
+            // Compare normalized lowercase values
+            if (label === normalize(selected)) {{
+              // Scroll to make sure the element is visible
+              try {{
+                li.scrollIntoView({{block: 'nearest'}});
+              }} catch (e) {{}}
+
+              // Click the matched element
+              li.click();
+              return true;
+            }}
+          }}
+
+          // Return false if no matching item was found
+          return false;
+        }}
+        """)
+        if not ok:
+            LOG.error(_("No <li> options found in combobox dropdown with id '%s'."), dropdown_id)
+            raise TimeoutError(_("Cannot locate combobox dropdown options."))
+
+        await self.web_sleep()
+        return dropdown_elem
+
     async def _validate_chrome_version_configuration(self) -> None:
         """
         Validate Chrome version configuration for Chrome 136+ security requirements.

tests/unit/test_web_scraping_mixin.py

@@ -156,6 +156,62 @@ async def test_web_input_clear_failure(self, web_scraper:WebScrapingMixin, mock_
         with pytest.raises(Exception, match = "Cannot clear input"):
             await web_scraper.web_input(By.ID, "test-id", "test text")
 
+    @pytest.mark.asyncio
+    async def test_web_select_combobox_missing_dropdown_options(self, web_scraper: WebScrapingMixin) -> None:
+        """Test combobox selection when aria-controls attribute is missing."""
+        input_field = AsyncMock(spec=Element)
+        input_field.attrs = {}
+        input_field.clear_input = AsyncMock()
+        input_field.send_keys = AsyncMock()
+        web_scraper.web_find = AsyncMock(return_value=input_field)
+
+        with pytest.raises(TimeoutError, match="Cannot locate combobox dropdown options."):
+            await web_scraper.web_select_combobox(By.ID, "combo-id", "Option", timeout=0.1)
+
+        input_field.clear_input.assert_awaited_once()
+        input_field.send_keys.assert_awaited_once_with("Option")
+
+    @pytest.mark.asyncio
+    async def test_web_select_combobox_selects_matching_option(self, web_scraper: WebScrapingMixin) -> None:
+        """Test combobox selection matches a visible <li> option."""
+        input_field = AsyncMock(spec=Element)
+        input_field.attrs = {"aria-controls": "dropdown-id"}
+        input_field.clear_input = AsyncMock()
+        input_field.send_keys = AsyncMock()
+
+        dropdown_elem = AsyncMock(spec=Element)
+        dropdown_elem.apply = AsyncMock(return_value=True)
+
+        web_scraper.web_find = AsyncMock(side_effect=[input_field, dropdown_elem])
+        web_scraper.web_sleep = AsyncMock()
+
+        result = await web_scraper.web_select_combobox(By.ID, "combo-id", "Visible Label")
+
+        assert result is dropdown_elem
+        input_field.clear_input.assert_awaited_once()
+        input_field.send_keys.assert_awaited_once_with("Visible Label")
+        dropdown_elem.apply.assert_awaited_once()
+        assert web_scraper.web_sleep.await_count == 2
+
+    @pytest.mark.asyncio
+    async def test_web_select_combobox_no_matching_option_raises(self, web_scraper: WebScrapingMixin) -> None:
+        """Test combobox selection raises when no <li> matches the entered text."""
+        input_field = AsyncMock(spec=Element)
+        input_field.attrs = {"aria-controls": "dropdown-id"}
+        input_field.clear_input = AsyncMock()
+        input_field.send_keys = AsyncMock()
+
+        dropdown_elem = AsyncMock(spec=Element)
+        dropdown_elem.apply = AsyncMock(return_value=False)
+
+        web_scraper.web_find = AsyncMock(side_effect=[input_field, dropdown_elem])
+        web_scraper.web_sleep = AsyncMock()
+
+        with pytest.raises(TimeoutError, match="Cannot locate combobox dropdown options."):
+            await web_scraper.web_select_combobox(By.ID, "combo-id", "Missing Label")
+
+        dropdown_elem.apply.assert_awaited_once()
+
     @pytest.mark.asyncio
     async def test_web_open_timeout(self, web_scraper:WebScrapingMixin, mock_browser:AsyncMock) -> None:
         """Test page load timeout in web_open."""