CraTer-Tool

1. Introduction

CraTer-Tool is a light-weight prototype of CraTer, it can quickly analysis a Java crash and predict whether the fault reside inside or outside of the stack trace.

The key ideas of the prototype are: 1) build a classification model based on the extracted features from historical crashes; 2) use the built model to predict the fault position of new-submited crash.

2. Usage example

Please download the whole project since the crater-tool.jar requires the training data in the files/ folder. For a quick start, we provide a usage example in the example/ folder.

2.1 INPUT:

1. The path of the faulty project is in the example/test_proj/.

2. The stack trace of crash is in the example/trace_iobe.txt which record the stack trace of the IndexOutOfBoundsException crash,

Exception in thread "main" java.lang.IndexOutOfBoundsException: Index: 1, Size: 0
	at java.util.ArrayList.rangeCheck(ArrayList.java:659)
	at java.util.ArrayList.get(ArrayList.java:435)
	at IOBE.find_elem_by_idx(IOBE.java:27)
	at IOBE.trigger_null_pointer_exp(IOBE.java:12)
	at IOBE.main(IOBE.java:5)

Then we use the following command to predict the fault position of a crash.

$ java -jar crater-tool.jar -projPath example/test_proj/ -projStackTrace example/trace_iobe.txt

2.2 OUTPUT

After executing the above command in the terminal, we can get the following output, including the crash informaton, extrcted features, and prediction results.

According to the prediction results, crater-tool.jar guess that the fault may reside inside of the stack trace.

>>>>> Crash Information: 
Crashed Java project: example/test_proj/
   Crash stack trace: example/trace_iobe.txt
      Exception Type: IndexOutOfBoundsException
         Trace lines: 5
   Number of Classes: 1
   Number of Methods: 3
  polymorphic or not: false
      Crash Position: IOBE, find_elem_by_idx, 27
       Crash Trigger: IOBE, main, 5
------------------------------------------------

>>>>> Extracted Features (all 89 features): 
  13.000,   5.000,   1.000,   3.000,   1.000,   4.000,  16.000,   4.000,   4.000,   2.000,
   2.000,   3.000,   0.000,   5.000,   1.000,   1.000,   1.000,   3.000,   2.000,   0.000,
   0.000,   0.000,   0.000,   0.000,   0.000,   0.000,   0.000,   0.000,   0.000,   0.000,
   1.000,   1.000,   0.000,   0.000,   0.667,   0.000,   0.000,   0.000,   0.000,   0.000,
   0.000,   0.000,   0.000,   0.000,   0.000,   0.000,   0.333,   0.333,   0.000,   0.000,
   3.000,   0.000,   5.000,   1.000,   1.000,   1.000,   3.000,   1.000,   0.000,   0.000,
   0.000,   0.000,   0.000,   0.000,   0.000,   0.000,   0.000,   0.000,   0.000,   1.000,
   0.000,   0.000,   0.000,   0.333,   0.000,   0.000,   0.000,   0.000,   0.000,   0.000,
   0.000,   0.000,   0.000,   0.000,   0.000,   0.333,   0.000,   0.000,   0.000,
------------------------------------------------

>>>>> Prediction Results:
Classifier    : weka.classifiers.meta.FilteredClassifier
Possibility   : INSIDE - 39.97%, OUTSIDE - 60.03%.
Recommandation: The root-cause-line of the given crash may reside OUTSIDE of the stack trace. Try to check 
the code through the method invocations, 

	at IOBE.find_elem_by_idx(IOBE.java:27)
	at IOBE.trigger_null_pointer_exp(IOBE.java:12)
	at IOBE.main(IOBE.java:5)

------------------------------------------------

3. Parameter tuning

• If you wanna change the training set of crater-tool.jar, you can find and replace the files/training_set.csv.
• If you wanna change the classifier used in crater-tool.jar, you can find and modify the code in Line 77 of the 'cstar.yongfeng.launcher.Entry.java'