As we’re preparing for our 2021 FTS release here is a sneak-peak into the capabilities our Product & Engineering teams are preparing. As always, we count on our great Community to try out the newest features in our products and we welcome the feedback.
Check out the new Table Extraction experience! Besides a new UI, it comes with a lot of improvements requested by you, over time:
- In-place column settings with sorting and formatting capabilities
- ExtractDataTable with DateTime & Number columns support
- Extraction limits and data preview at hand
This long-requested feature has been in our mental backlog for a while, and we believe we finally found a solution for it. So from now on, when you need an easy way to verify that a specific action (like a click or type) was actually executed successfully – that it had the intended outcome, you got it!
Just add a verification from the context menu of any UI activity, and the verification will be done either automatically (for Type actions), or by indicating an element on the screen that is expected to appear, disappear, or change. If the verification fails, maybe because the webpage wasn’t yet fully loaded, or any other reason, the action is retried until successful.
More options are available under Properties > Input > Verify execution.
This feature addresses the need of targeting elements regardless of their specific content, such as trying to click on the first element in a search list, or just trying to click on a dynamic element (e.g. a link that can have different texts). Technically it was always possible, but it required some advanced knowledge and some fiddling. Now, it’s a single click from the hover-menu that appears over any targets that contain text. Under the hood, it does multiple things, like changing the attributes to make them dynamic, removing text references, disabling incompatible targeting methods, and so on.
In some situations, especially responsive website, it’s useful to have the window a very specific size, so it looks & behaves exactly as expected. The previously existing methods of setting a window to a specific size involved a bit of trial&error, and this features aims to correct that through a simple setting: Resize window to design-time size . When enabled, the applications will always be resized to the width/height they had when they were indicated in Studio(X). It’s available as an option for all Use App/Browser containers.
Now you can record multiple apps on the fly, without the need to click on “Change application” button, for a more fluid, task-oriented experience.
In the selection screen there always was the option to define image-only targets using the F3 shortcut key, but that functionality was lacking in the Recorder.
From now on, image targets can be indicated on the fly, by simply doing a box-selection around the element you want to target. Works both in the selection screen and the recorder.
Check:text was a “hidden gem”of the Fuzzy selector, that could allow much easier targeting of elements that contained text. In a way, it’s similar to the older UIA Click Text activity, but now part of the Modern experience, under Fuzzy selector, in the selection screen. With support for wildcards, variables, the whole shabang.
Check app state can be used to verify the state of an application, but previously it was limited to the current application. Now, it can be used stand-alone, outside of an Use App/Browser container, in order to check if another application is open, for example.
Some apps & websites use hidden elements for various parts of their UI, and sometimes this can confuse our targeting system, because robots can see everything Well, now you can enforce the Visibility flag, which will ensure that any hidden elements are completely ignored, both at design-time and runtime.
Nobody like recurring alerts, so now various popups shown in UIAutomation have the option to be hidden for the duration of the session. Simply reloading a project brings them back.
A few options like the Pause (F2) timer, auto-apply flag, and recorder Expert mode are persisted per machine, as they are more related to how a user works than a specific project.
This new activity will simply output the URL of the current tab, so it can be used in automations.
Very often, you need to launch not only executables but also open files via native applications like .doc via MS Word. Thus, we have improved the StartProcess activity for you to be able to do that.
Now we can steer the SAP Multiple labeled table (ALV LIST) via UI automation.
We have enhanced the UI automation driver to generate proper selectors for elements rendered under ShadowDOM. Note the new attribute autogenerated inside the selector when targeting such elements:
<html app='chrome.exe' title='VirusTotal' /> <webctrl tag='HOME-VIEW' /> <webctrl shadowhostid='home-view' tag='VT-UI-MAIN-UPLOAD-FORM' /> <webctrl id='infoIcon' **shadowhostid='uploadForm'** tag='VT-UI-BUTTON' />
You can now use our Australia based Document Understanding public endpoints as well, in case server regions are of concern to you. Do check out the documentation here [About Licensing], the Document Understanding Endpoints section, for more information.
Until now, Data Manager would only allow you to define up to 40 fields for data extraction. This restriction has been lifted and you can now define up to 300 fields! Check out the new Data Manager experience available in the Cloud platform as preview.
The UiPath Document OCR machine learning model is now capable of recognizing checkboxes as well! Give it a try by using it in the Digitize Document scope activity. As expected, this enhancement to our OCR engine will bring immediate benefits when used with the Form Extractor, Intelligent Form Extractor, and Machine Learning Extractor
The Validation Station got a huge facelift when using and viewing the Text version of a document:
You can now see color coded words, just like in the Original document view;
You can now select and highlight disparate words from the Text version;
You can now choose to either select cursive text (like until now), or tokens in a given area, switching between interaction modes:
In the Validation Station, you can now see crops of the document corresponding to the values you are reviewing, right under the reported value – so you don’t need to cross check left- and right-side panels anymore: