Output or screen scraping is used in activities that are designed to receive text from a specified element or region. There are three options available.
| Method | Speed | Accuracy | Background Execution | Extract Text Position | Support for Citrix |
|---|---|---|---|---|---|
| ApiCall | 100 | 100% | Yes | No | No |
| Window Messages | 80 | 100% | Yes | No | No |
| OCR | 40 | 99% | No | Yes | Yes |
ApiCall is the default method. It is very accurate and works in most cases.
OS Windows uses a message-passing model. We simulate this model using Windowmessage to simulate mouse or keyboard events. In this activity, window messages simulate mouse input elements in the form of an internal message for the Windows system.
When using virtual environments, or if none of the above-mentioned methods work, it is possible to use OCR. This is not 100% accurate but will obtain results as long as there is visible text. It works based on the Tesseract engine.