UFT Smart Identification Guide

Mastering Object Identification in UFT - Smart Identification

Smart identification in UFT is a powerful mechanism that ensures test reliability even when application objects change during testing cycles. This advanced feature provides a safety net when standard object identification methods fail, allowing automated tests to continue executing successfully despite UI modifications. In today's rapidly evolving development environments, where applications frequently update between test creation and execution, understanding and implementing smart identification effectively has become essential for maintaining stable test automation.

Understanding Object Identification in UFT

Object identification forms the foundation of how UFT interacts with applications during test execution. When you record a test, UFT captures properties of each object you interact with, creating a unique description that allows it to locate the same object during playback. This description typically includes properties like the object's type, name, and other distinguishing attributes. The standard identification process follows a hierarchy, first attempting to match objects using the learned description. If this fails, UFT may resort to ordinal identifiers like location or index. However, when these methods also fail, the Smart Identification mechanism activates as a sophisticated alternative approach.

The standard object identification relies on a set of properties that define an object uniquely in the application under test. These properties are stored in the Object Repository, which serves as a centralized location for storing object descriptions. During test execution, UFT scans the application, comparing objects in the UI against those stored in the repository. This process works well when applications remain static, but real-world applications often change, requiring more flexible identification mechanisms.

• UFT learns object properties during recording
• Objects can change between test creation and execution
• Smart identification provides a flexible alternative when standard methods fail
• The Object Repository stores object descriptions for use during playback
• Standard identification follows a hierarchy of properties and ordinal identifiers

The Need for Smart Identification

In dynamic development environments, applications frequently undergo changes that can break existing tests. Objects might be repositioned, renamed, or have their properties modified during development cycles. When these changes occur, UFT's standard object identification may fail, causing tests to halt with identification errors. This is where Smart Identification becomes crucial—it provides a safety net that allows tests to continue functioning despite minor application modifications.

Without Smart Identification, testers would need to constantly update their object repositories to match application changes, creating significant maintenance overhead. By implementing this fallback mechanism, UFT can identify objects even when some properties have changed, making tests more resilient to application evolution.

• Maintains test stability when applications change
• Reduces maintenance overhead by adapting to UI modifications
• Improves test reliability in evolving development environments

Smart identification serves as an essential component of a robust test automation strategy, particularly in agile development environments where applications evolve rapidly between sprints and releases.

How Smart Identification Works

When standard object identification fails, UFT activates Smart Identification if it's enabled. This mechanism operates by temporarily ignoring the learned object description and creating a new, more flexible set of properties to identify the object. Smart Identification uses two main components: the Base Filter Properties and the Optional Filter Properties. The base properties are essential for identifying the object, while optional properties provide additional discrimination if needed.

The Smart Identification process follows a systematic approach:

1. It discards the original object description that failed to match

2. It creates a new description using the Smart Identification properties

3. It applies a set of filters to narrow down potential matches

4. It scores potential matches based on how well they fit the description

5. It selects the object with the highest score as the match

This scoring mechanism evaluates each potential candidate, assigning points based on how many properties match. The object with the highest score becomes the identified object, allowing the test to proceed even when exact matches aren't found.

The algorithm considers both the presence of matching properties and their assigned weights or priorities. Properties that are more stable and reliable are typically given higher weights, increasing their influence on the final identification decision. This weighted approach ensures that the most critical identification factors have the greatest impact on determining the correct object.

' Example of a more advanced Smart Identification implementation
' This code demonstrates creating a custom Smart Identification mechanism

Class CustomSmartIdentification
    Private objRepository
    Private smartProperties
    Private scoreThreshold
    
    Private Sub Class_Initialize()
        ' Initialize default values
        Set objRepository = CreateObject("UFT.ObjectRepository")
        Set smartProperties = CreateObject("Scripting.Dictionary")
        scoreThreshold = 80
    End Sub
    
    Public Sub AddSmartProperty(PropertyName, Weight)
        ' Add a property to the Smart Identification set
        smartProperties.Add PropertyName, Weight
    End Sub
    
    Public Function FindObject(ObjectClass, ParentObject)
        ' Custom Smart Identification implementation
        Dim candidates, candidate, maxScore, bestMatch
        Dim score, propName, weight
        
        ' Get potential candidates
        Set candidates = GetPotentialCandidates(ObjectClass, ParentObject)
        maxScore = 0
        Set bestMatch = Nothing
        
        ' Evaluate each candidate
        For Each candidate In candidates
            score = 0
            ' Calculate score based on Smart Properties
            For Each propName In smartProperties.Keys
                weight = smartProperties(propName)
                If candidate.GetROProperty(propName) = ParentObject.GetTOProperty(propName) Then
                    score = score + weight
                End If
            Next
            
            ' Check if this candidate is the best match so far
            If score > maxScore And score >= scoreThreshold Then
                maxScore = score
                Set bestMatch = candidate
            End If
        Next
        
        Set FindObject = bestMatch
    End Function
End Class

Configuring Smart Identification Settings

Proper configuration of Smart Identification is essential for maximizing its effectiveness. You can access Smart Identification settings through the Object Identification dialog in UFT. These settings allow you to define which properties should be used for Smart Identification, configure the scoring mechanism, and set thresholds for acceptable matches.

To configure Smart Identification:

• Navigate to Tools > Options > GUI Testing tab > Object Identification
• Select the object class you want to configure
• Click the "Smart Identification" button
• Define the Base Filter Properties and Optional Filter Properties
• Set the ordinal identifier if needed
• Adjust the Smart Identification timeout if necessary

' Example of configuring Smart Identification programmatically
' This code demonstrates how to modify Smart Identification settings for a specific object class

Dim objDesc
Set objDesc = Description.Create()
objDesc("micclass").Value = "WebButton"
objDesc("html tag").Value = "INPUT"

' Configure Smart Identification properties
objDesc("smart_base_filter").Value = "name;html tag"
objDesc("smart_optional_filter").Value = "id;type"
objDesc("smart_score_threshold").Value = 80

' Apply the configuration to the object repository
Repository.Add "MyPage", "LoginButton", objDesc

When configuring Smart Identification, it's important to choose properties that are most likely to remain stable in your application. Properties like CSS selectors or XPath expressions often provide more reliable identification than dynamic properties that might change frequently.

The Base Filter Properties should include properties that are essential for identifying the object type, while Optional Filter Properties can include additional distinguishing characteristics. The scoring threshold determines the minimum score a candidate object must achieve to be considered a valid match.

' Example of handling Smart Identification in test code
' This code demonstrates how to check if Smart Identification was used and take appropriate action

Set objBrowser = Browser("micclass:=Browser")
Set objPage = objBrowser.Page("micclass:=Page")
Set objButton = objPage.WebButton("name:=Login", "html tag:=INPUT")

' Attempt to identify the button
If objButton.Exist(2) Then
    ' Button found - proceed with interaction
    objButton.Click
Else
    ' Button not found with standard properties
    ' Check if Smart Identification was used
    If objButton.GetROProperty("smart_identification_used") Then
        ' Log that Smart Identification was used
        Reporter.ReportEvent micWarning, "Object Identification", "Smart Identification was used for Login button"
        
        ' Optional: Take corrective action
        ' For example, update the object repository or use an alternative identification method
    Else
        ' Object not found at all - handle the error
        Reporter.ReportEvent micFail, "Object Identification", "Login button could not be identified"
        ExitTest
    End If
End If

Best Practices for Smart Identification

Implementing Smart Identification effectively requires careful consideration of several factors. First, it's essential to use Smart Identification sparingly—as a fallback mechanism rather than the primary identification method. Over-reliance on Smart Identification can mask underlying issues in your test design or application changes that should be addressed directly.

• Use specific properties whenever possible
• Regularly review Smart Identification matches
• Combine with other identification methods for robust tests
• Document Smart Identification configurations for team consistency

Another best practice is to implement consistent Smart Identification settings across your test suite. This ensures that objects are identified consistently across different tests and environments. Additionally, regularly reviewing Smart Identification matches helps identify when application changes require updates to your object repository rather than relying on the fallback mechanism.

When configuring Smart Identification:

• Focus on properties that are least likely to change
• Assign higher priority to more stable properties
• Test your configuration with different application versions
• Avoid including too many properties, as this might reduce flexibility

A well-balanced approach typically includes 3-5 carefully selected properties that uniquely identify the object while remaining stable. This balance between specificity and flexibility is key to effective Smart Identification implementation.

Common Challenges and Solutions

Despite its benefits, Smart Identification can present several challenges. One common issue is false positives, where Smart Identification incorrectly identifies an object because it matches with a similar but incorrect element. To mitigate this, ensure your Smart Identification properties are specific enough to uniquely identify the intended object while still being flexible enough to handle minor changes.

Another challenge is performance impact. Smart Identification requires additional processing to evaluate potential matches, which can slow down test execution. To address this, limit the scope of Smart Identification to objects that are most likely to change and ensure your Smart Identification properties are optimized for quick matching.

When troubleshooting Smart Identification issues:

• Review the test results to understand why standard identification failed
• Verify that the Smart Identification properties are appropriate for the object
• Check if the scoring threshold needs adjustment
• Consider if alternative identification methods might be more suitable

Over-reliance on Smart Identification might mask underlying issues in the application under test. It's important to use Smart Identification as a temporary solution while investigating why standard identification failed and addressing the root cause when possible.

Practical Examples and Use Cases

Smart identification proves particularly valuable in several common testing scenarios. Consider a web application where button labels change based on user language settings. With smart identification configured to use stable properties like HTML tag, name, or position rather than the label text, your tests can continue functioning across different language versions without modification.

Another practical use case is testing applications that undergo frequent UI updates. Instead of updating every test object after each release, you can configure smart identification to focus on properties that remain consistent across updates. This approach significantly reduces maintenance overhead while maintaining test coverage.

For instance, imagine testing a login form where the "Submit" button occasionally changes its text to "Login" or "Sign In." By configuring smart identification to prioritize the button's ID or position over its text property, your tests can successfully interact with the button regardless of the label variation.

In enterprise applications where components are reused across different modules with slight variations, Smart Identification can help maintain a single test object definition that works across all instances. This approach reduces redundancy and simplifies test maintenance.

Benefits and Limitations of Smart Identification

Smart identification offers several significant advantages in automated testing. It improves test resilience by allowing tests to continue execution even when objects change slightly. This reduces test maintenance efforts and extends the lifespan of automated tests. Additionally, smart identification can handle scenarios where objects are dynamically generated or have varying properties during different test runs.

However, smart identification is not without its limitations. It cannot handle drastic changes to object properties or structural modifications. Over-reliance on smart identification might mask underlying issues in the application under test. Furthermore, poorly configured smart identification might lead to false positives, where the test identifies the wrong object but continues execution, potentially causing test failures later in the execution flow.

• Benefits:
• Reduces test maintenance
• Handles minor UI changes
• Improves test resilience
• Extends test lifespan across application versions
• Provides a safety net for dynamic applications

• Limitations:
• Cannot handle major structural changes
• May mask application issues
• Risk of false positives
• Potential performance impact
• Requires careful configuration to be effective

Advanced Tips for Optimizing Smart Identification

To maximize the effectiveness of smart identification, consider these advanced optimization strategies. First, analyze your application thoroughly to identify which properties are most stable across different versions and scenarios. This analysis should guide your smart identification configuration decisions.

Second, balance between specificity and flexibility when selecting properties for smart identification. Too many properties might make the identification too rigid, while too few might lead to ambiguity. A well-balanced approach typically includes 3-5 carefully selected properties that uniquely identify the object while remaining stable.

Finally, regularly review and refine your smart identification configurations as your application evolves. What works in one version might not be effective in the next, so continuous optimization is key to maintaining reliable test execution.

Consider implementing a hybrid approach that combines Smart Identification with other object identification techniques. For example, you might use descriptive programming for critical objects while relying on Smart Identification for less stable elements. This balanced strategy ensures reliability where it matters most while maintaining flexibility for components that frequently change.

Conclusion

Smart identification in UFT is an essential feature for creating resilient automated tests that can adapt to inevitable changes in applications. By understanding how it works, configuring it properly, and implementing best practices, you can significantly reduce test maintenance while ensuring reliable test execution. When combined with a solid understanding of object identification principles and other testing techniques, smart identification becomes a powerful tool in your testing arsenal, enabling your automated tests to withstand the dynamic nature of modern applications.

As development methodologies continue to evolve with more frequent releases and shorter development cycles, the ability to create tests that can adapt to change without constant maintenance becomes increasingly valuable. Smart identification provides this adaptability, allowing test automation to keep pace with rapid application development while maintaining reliability and consistency.

By implementing smart identification strategically and thoughtfully, organizations can achieve a balance between test stability and flexibility, ensuring their automation efforts deliver consistent value throughout the application lifecycle.

Frequently Asked Questions

• What is Smart Identification in UFT?
  Smart Identification in UFT is a powerful mechanism that ensures test reliability when application objects change. It activates as a fallback when standard object identification methods fail, allowing tests to continue executing successfully despite UI modifications.
• How does Smart Identification work in UFT?
  When standard identification fails, Smart Identification creates a new, more flexible set of properties to identify the object. It uses Base Filter Properties and Optional Filter Properties, scores potential matches, and selects the object with the highest score as the correct match.
• How do I configure Smart Identification in UFT?
  Navigate to Tools > Options > GUI Testing tab > Object Identification, select the object class, click 'Smart Identification', define Base and Optional Filter Properties, set the scoring threshold, and adjust the timeout if needed. Focus on properties that are most likely to remain stable in your application.
• What are the best practices for using Smart Identification?
  Use Smart Identification sparingly as a fallback mechanism, not as the primary method. Implement consistent settings across your test suite, focus on stable properties, regularly review matches, and combine with other identification methods for robust tests.
• What are the benefits and limitations of Smart Identification?
  Benefits include reduced test maintenance, handling minor UI changes, improved test resilience, and extending test lifespan. Limitations include inability to handle major structural changes, potential for false positives, risk of masking application issues, and possible performance impact.