MouseRobot provides a good support for automation of repetitive tasks when working with web-interfaces. The program is able to distinguish and find all the major elements on the web pages: text boxes, table cells, links, captions, headers, pictures, radio buttons, lists, etc.
Although MouseRobot may be used with any browser, the best combination of reliability and speed of performance is achieved by using Microsoft Internet Explorer and Mozilla Firefox. This is due to the fact that these browsers have developed interfaces to interact with external programs. When a page is displayed in these browsers, MouseRobot is able to distinguish its individual elements, and automatically scroll the contents if needed.
Supported pages and content
MouseRobot works well with both static and dynamic content. However, MouseRobot is not able to distinguish inscriptions, objects and controls within the image, Flash-elements, Java-applets and elements of Silverlight.
How MouseRobot sees the page contents
For a successful automation, it is important to know how MouseRobot sees and interprets the contents of the page and what the guiding principles of the program are when looking for the object on page.
During execution, MouseRobot searches for the object that is like the desired one. In this case, the similarity is determined by the properties, size and position of the object.
As an example, let us consider the search page of google.com. The figure below presents it the way we see it, plus its allocated objects that MouseRobot is able to distinguish:
Each object on a page is characterized by type (TEXT, GRAPHIC, PUSHBUTTON, etc.), by a title or name, internet address, to which it refers, as well as the size and its position on the page.
Since MouseRobot cannot read the inscriptions on the pictures, as well as it is not able to perceive the design, then it has much less information than we do to search for items.
The program will "see" the same page approximately like this:
|•||Note that the intervals between the inscriptions, which we usually do not even notice, are also the elements of the text page and, consequently, MouseRobot will have to verify them while searching. For complex and large web pages the number of these elements can be calculated in thousands, thus the program may require considerable time (several seconds) for their analysis.|
|•||In addition, if desired object does not have any of the characteristics that distinguish it from other similar objects on the page, MouseRobot can make a mistake and finds another object instead of required one or can't make decision and stops the execution.|
For example, text entry fields on web-forms often have the same size and properties, and differ only in position on the page.
As you can see, the fields 1, 2, 3, and A, B, C are very similar to each other, and it would have been very difficult to distinguish between them if there is no labels with the explanation on their left.
However, these labels on the web pages often do not relate to the text fields, and therefore Mouse Robot in most cases will not be able to rely on them when searching for a specific field. As a result, this web-form takes the following form:
|•||It is better not to rely on the fields themselves, but on easily identifiable text labels, and then move on to the fields pressing the Tab. Many windows support the movement between the elements by pressing Tab and rapid movement to a particular item by pressing Alt + letter (the desired letter is automatically highlighted if you press Alt).|
For example, to move to the drop-down list "Pages per sheet" is proposed to press Alt + h, and to move to the "Scale to paper size" - Alt + z.
The same technique (to find and set the input focus on a static text label, and then to switch to the input fields and navigate between them, pressing Tab or Shift + Tab) allows you to effectively automate the data entry not only on web pages, but also in regular Windows applications.
|•||Another important feature that should be considered when automating web pages is the time of page loading. If regular Windows-based applications open a specific window in approximately the same time then, the speed of opening web pages is affected by the connection status and server load. Thus, before working with a Web page, make sure that it is available, for example find any field on the page by step "Search object on screen" and increase the maximum performance time this step "Step timeout" up to 10-15 seconds (depending on the page size and speed of the internet) for reliable execution.|