GPAL Fluent: High-level fluent C# API

//The handler is given two lists, everything the selector found and the subset that matched its filters, and matched is the one to read when the selector carries filters. Selected stays a plain bool because a row is either the chosen one or it is not; Checked is the nullable one because a box has a third state. Returning CallIfStatus.NotHandled reads the elements and leaves the unit of work to carry on as it would have. A page element also knows how it was found. Xpath and Css are the paths that located it, FullXPath is the absolute one from the root, and Xpaths is every path GPAL worked out for it, each with the number of elements it matches on that page. That count is the useful part: a path matching one element is a path worth keeping in a selector, and a path matching forty is a path that will find the wrong thing tomorrow. Reading them in a handler is how a workflow turns an element it found once into a selector it can rely on.

// the same handler shape, and the same property names, on either side

static CallIfStatus OnRowFound(IBrowser browser, List<IGPALElement> found, List<IGPALElement> matched, Selector selector, bool matchedAll)

{

foreach (IGPALElement element in matched)

if (true == element.Selected)

GPAL.PublishSimpleEvent(GPALEventType.NOTICE, $"[{element.Text}] is the selected one");

return CallIfStatus.NotHandled;

}

// a three state box: ticked, clear, or neither

static CallIfStatus OnBoxFound(IApplication application, List<IGPALAutomationElement> found, List<IGPALAutomationElement> matched, Selector selector, bool matchedAll)

{

foreach (IGPALAutomationElement element in matched)

GPAL.PublishSimpleEvent(GPALEventType.NOTICE, null == element.Checked

? $"[{element.Name}] is partly ticked"

: $"[{element.Name}] is [{element.Checked}]");

return CallIfStatus.NotHandled;

}