Level 1 – Foundation Training Course (Part 2)

Lesson 4

In this lesson, you will learn the basics on how to use the recorder and how to get information out of different applications. I will show you some tips and tricks that you can use in real case scenarios.


UiPath allows you to record and replay actions. The outcome is a Recording Sequence which you can later modify. It’s a great way to start your automation process because it creates a new workflow that you can build upon. There are 4 types of recording: Basic, Desktop, Web, and Citrix.

Basic and Desktop are used for recording most desktop applications. Both of them are almost similar, with minor differences:

  • the Basic recording generates full selectors for each activity and no container. It’s suitable for single activities. The resulted automation is slower than the one generated by a Desktop recording because of the missing containers.
  • the Desktop recording generates a more complex workflow. It’s suitable for all types of desktop applications and multiple actions. Generates an AttachWindow container with the selector of the top-level window in which activities are enclosed, and partial selectors for each activity.

Web is for recording in web applications and browsers, generates containers and uses Simulate Type/Click input method by default

Citrix is for recording virtual machines, VNC, and Citrix environments.

It can record LEFT CLICKS on buttons, checkboxes and other clickable elements, and also TYPING into editable fields. On the other side, it doesn’t record keyboard shortcuts, modifier keys (CTRL and ALT key), right click, mouse hover, getting the text out of the apps, find elements and images, and copy to clipboard. All of these can be achieved using Manual Recording(stop the recording by pressing ESC and select the action you would like to perform from the following options: Start App, Click, Type, Copy, Element, Text or Image)

Recording controller (Wizards and Actions)


Keyboard shortcuts that you can use while recording:

Pause recording for 3 secondsF2
Stop recordingESC
Save and exit recordingESC+ESC
Exit the recordingRight click
Lesson 4 Quiz Passed Total score 100%

Lesson 5

You will learn about the various INPUT and OUTPUT, the technology behind it and when to use which one. Also, you will learn about a new technique called data scraping. It will help you extract information out of tabular and structured data.

UI interactions

All Ui interactions in UiPath fall in one of the 2 categories: Input or Output. Input is when the Robot is performing an action like clicking or typing. Output is when the action get data OUT of the application.

Input actionsOutput actions
ClickGetting text
Text typingFind element or image
Keyboard shortcutsClipboard operations
Right clickEtc
Mouse hover

Input methods

DefaultWindow MessageSimulate Type/Click
Works in BackgroundNoYesYes
Speed50% 50%100%
Supports KeysYes YesNo
Automatic Empty FieldNoNoYes

The Default method works all the time, but it has some downsides. It’s not as fast as the other methods and the application you interact with has to be active all the time.

First, I recommend you start with the Default method and depending on the requirements you can change it to one of the other two. Bear in mind that Simulate Type is the fastest but it doesn’t support Keys and clears the field before typing. And Windows Message works very well with background applications.

Output – Screen Scraping

I suggest you use Screen Scraping when the Robot has to find and retrieve bigger blocks of text. Also, when the information is behind complex UI, and it’s difficult to get or you need to find a specific UI property of an element (font, style, color, position…)

SpeedVery FastFastNot very Fast
Works in BackgroundYesNoNo
Gets Text PositionNoYesYes
Gets Hidden TextYesNoNo
Works with CITRIXNoNoYes

The FullText method it’s the best for the most situations. It captures all the text, even the hidden one.

The Native method can extract specific UI properties like font, style, and color. Also, it can capture the screen coordinates of any word or character. When it comes to text scraping, this method only captures the information from the editable fields.

The OCR method is not 100% accurate but it works when the others don’t. I usually use the method when scraping data from CITRIX or from any virtual environments. The OCR engine converts the images or the printed text into machine-encoded text. There are two available engines. Microsoft OCR is suitable for large images like scanned documents. Google OCR works better with low-resolution images like interface elements.

About Valeriu B

RPA Developer (UiPath certified)