If you’re installing into a fresh KNIME 4.0, please get the update site URL here.
This release provides no new features, rather it ensures compatibility to KNIME 4.0, and thus involves a breaking change. Due to changes in KNIME 4.0 we had to remove the jBrowser plugin – we feel that this is not a big thing, as (1) it was always an optional installation and thus not widely used anyways, and (2) there are now other (better) options for running a headless, bundled browser (e.g. Chrome, Chromium, or Firefox).
Beside that, we removed the “Execute Selenium script” node – the related “Selenium Builder” project for creating the JSON scripts is not maintained since a long while, and the node caused confusions, as there are other, incompatible Selenium JSON script formats which could not be processed by this node anyways.
For more information, please check the change log.
Let us know if you have feedback, comments, or issues!
In case you’re updating an existing KNIME installation where the jBrowserDriver has been installed, you need to uninstall jBrowserDriver before, else wise you’ll see a “Cannot complete the install because of a conflicting dependency”.
For that, go to About KNIME → Installation Details → Installed Software and there uninstall “Selenium Nodes for KNIME: jBrowserDriver” by selecting the entry and clicking “Uninstall…”
The Find Elements node now has a Select button which allows you to grab page elements more easily than before. Say “goodbye” to writing queries by hand or copying selectors from your browser. Instead, simply point at the desired element in the browser window, click on it , and a matching CSS selector will be created for you. This feature works in all Chrome-based browsers (Chrome, Chromium and Opera).
The Find Elements node will show you all matched elements for a given query in realtime — both in the browser window and in the DOM outline. Modify your query and you will immediately see whether you hit the desired elements. This feature is available for all Chrome-based browsers — as stated above.
What Else?
Chris Terrell from EverydayVBA made a three minute video which showcases the new functionality. Watch here and also check out his other videos!
There are several more changes, e.g. a new Reset on Load node. Please check our changelog for all the details.
Update 4.2 adds a new predefined snippet to disable SSL-related errors (see here). Beside that, we’ve updated the bundled Chromium browser to version 80.0.3987.117.
We released Selenium Nodes 4.3.1 a few days ago. Notable changes include:
A brand-new Create PDF node which allows you to write PDF files using Chromium-based headless browsers. How does it work for you? We’re looking forward to your feedback!
Start WebDriver: Input port for the factory is now optional – you can configure the WebDriver directly within the node configuration, the WebDriver Factory is no longer needed.
Several new Capability snippets for the Firefox browser (kudos @armingrudd)
Updated driver binaries and bundled Chromium
Reset on Load: Fix when used in metanode or wrapped metanode or component or however these things are called now (thanks to @Rich_ard)
The pdf creator works very well. Features It is great to scan a web page with all its links and texts (not in the form of an image or image format) and save it as pdf.
also the web driver features have been improved operationally
Since a few minutes, there’s Selenium Nodes 4.4 on the update site. The changes:
Navigate: Add “Flow Variable” to URL input.
Send Keys: Add “Flow Variable” to text input.
Wait: Add “Flow Variable” to time input.
Find Elements: Add “Flow Variable” to query and result column input.
Take Full Page Screenshot: New node for taking “full page” screenhots (i.e. longer than the visible window bounds) using Chrome. Currently labeled as “beta” — your feedback is welcome!
And there’s one thing which will make it much easier for you to build Selenium Workflows: Many nodes now have an integrated “Find Elements” functionality, which will make your workflows much cleaner, easier to operate and faster to execute. Small change, big impact!
When opening existing workflows, you’ll see a notification that some nodes have been updated. Don’t hesitate to click the “Migrate Workflow …” button to update to the latest node versions (a backup copy will be automatically retained and the nodes are of course backwards-compatible).
(Info) Requires at least KNIME 4.1 (please make sure you’re using an update site URL corresponding to your KNIME version)
(Add) Extract Attribute, Clear, Click, Extract CSS Property, Drag and Drop, Extract Inner/OuterHTML, Send File Path, Send Keys, Submit, Table Extractor: New node versions which allow to pick WebElements directly within the dialog without having to put a “Find Elements” input node first (old versions still work and become “deprecated”)
(Change) Extract Inner/OuterHTML renamed to Extract Text and add “textContent” option
(Change) Use NodePit Licensing plugin for license key management
(Change) Execute JavaScript: Support boolean and long flow variables which finally made it to KNIME 4.1
(Change) Execute JavaScript: Support collection flow variables and input cells
(Change) Execute JavaScript: Support JSON input and output
(Change) Add migration logic to replace deprecated nodes
There was an issue in the 4.7 release which could lead to a conflicting dependency and thus prevented installation on some environments. We just released version 4.7.1 (code name consequently “Aschermittwoch” ) which fixes this problem.
Note: It’s important that you also update Palladian to the newest version which we just rolled out, and which is v2.4 (an announcement and overview over the changes will follow). Find the update site URLs here:
Selenium Node v4.8.1 fixes issues introduced by KNIME 4.4 and updates the included drivers and Chromium version.
We have labelled HtmlUnit as “deprecated”, i.e. we plan to remove it with the next major 5.0 release. Why? HtmlUnit is not state of the art for current websites (very limited JS support), it’s unclear if the Selenium team will maintain it in the future, and thus we strongly recommend to switch to headless Chrome or Firefox instead. If still you rely on HtmlUnit for any reason, please let us know!
This one is labelled as “major” release, as we have upgraded the underlying Selenium library - this means that we removed some old, no-longer state-of-the-art functionality, which was previously labelled as “deprecated”. More important, these changes allow us to make use of new functionalities now and coming up in future releases. Notable, there is a new couple of nodes for Network Dumps:
This allows you to “record” all network requests which happen within your workflow execution including the URL and the request and response payload. You can use it e.g. to verify outgoing requests, to extract and analyze API responses, to grab all downloaded images, and probably much more!
Update: We’ve received a report about a “NoSuchMethodError” error when starting a browser after the upgrade - if this should happen to you, please follow the steps outlined here to get it resolved.
What You See Is What You Select
Here’s Selenium Nodes 4.7 (Codename “Alaaf”)
) which fixes this problem.