Skip to content

Object Locator


Object locators are created during Recording/Learning and used during Playback to identify learned objects and simulated objects. There are many types of locators.

In most libraries (Web, Desktop, Mobile) only one locator is generated when you record or learn an object.

Legacy Generic library supports 4 different locators added each time you learn or record object:

  • Location: This locator uses the object's index relative to encapsulating objects for identification. The location is stored as a period separated list of indexes. For instance, 1.2.3 would be "the third object in the second object in the first object."  The name, class, and role of the object are also stored.
  • LocationPath: This locator remembers name, class, and role property information for the object and all of its encapsulating objects.
  • LocationRect: This locator stores screen coordinates.
  • Ordinal: This locator creates an array of object name/object class combinations.  Each object is assigned an index in  the array.


The locator for the selected object is shown in the Properties view:

properties dialog

How Locators are Stored

The locator for each object is specified in saved_script_objects array in Objects.js of a test. Locator information is stored in locations array of each object.

"Simulated": {
    "locations": [
            "locator_name": "Location",
            "location": {
                "location": "",
                "window_name": "param:window_name",
                "window_class": "param:window_class"
            "locator_name": "LocationPath",
            "location": {
                // ...
            "locator_name": "LocationRect",
            "location": {
                "window_name": "param:window_name",
                "window_class": "param:window_class",
                "rect": {
                    "object_name": "",
                    "object_class": "param:object_class",
                    "object_role": "param:object_role",
                    "x": 573,
                    "y": 59,
                    "w": 220,
                    "h": 66
    "window_class": "MSPaintApp",
    "object_text": "",
    "object_role": "ROLE_SYSTEM_GROUPING",
    "object_class": "NetUIHWND",
    "version": 0,
    "object_type": "SeSSimulated",
    "object_flavor": "Simulated",
    "object_name": "Simulated",
    "ignore_object_name": true,
    "window_name": "Untitled - Paint"

Locator Parameters

If a piece of information in the locator matches a piece of object info (object_name, object_class, object_role, object_text) then it is stored in the locator as param:object_info.  For example:

"object_name": "param:object_name",
"object_class": "param:object_class",
"object_role": "param:object_role"

Over-riding Locator Parameters (RVL)

See also RVL.SetLocatorOpts and this KB Article for more detailed explanation and example of overriding locator parameters in RVL.

Over-riding Locator Parameters (JS)

You can over-ride the information used to locate your object at runtime.  Normally, to refer to an object, you use the SeS function:


To override locator parameters, specify the new value in the function call. In the following example, we over-ride the object_name parameter for object Username:

SeS('Username', {object_name:"regex:.*"})

You may want to change a parameter value for every locator/object in the program.  For instance, perhaps the url of the webpage has changed.  Use the global variable g_locatorparams as in the following example:

function Test()
    // Here we use direct parameter overriding
    SeS('Obj1', {url:"http://newaddr/"}).DoAction();
    SeS('Obj2', {url:"http://newaddr/"}).DoAction();

    // And this is equivalent to above
    g_locatorparams["url"] = "http://newaddr/";

    // ...

See Also