CharmFL

A Fault Localization tool for Python, which can be integrated in to the latest PyCharm IDE. After running the python unittests in your project is constructs a spectrum and the suspiciousness scores for each program element (e.g. statement, method). CharmFL also provides context for localizing faults. You are able to see different visualizations of your code and program spectra.

CharmFL uses pyan3 to construct the static call-graphs for your python project. From which we define "far" contexts for each program statement. For example, in the following program code, the far contexts are commented next to the statement:

def addToCart(product):
    if (product not in cart.keys()): 
        cart[str(product)] = 1 #Far context: getProductCount
    else:
        cart[str(product)] = cart[(str(product))] + 2  #Far context: getProductCount

def getProductCount(product):
    addToCart("apple") #Far context: addToCart
    if (product not in cart.keys()):
        return 0 #Far context: addToCart
    else:
        return cart[str(product)] #Far context: addToCart

Therefore, for each method the caller and callee methods are considered as far context. This is aggregated down to statement level, i.e. for each statement in the method the far context is considered as the method's far context.

Steps to set up

First, download the compressed ZIP from the releases

Then:

• Drag and drop into your PyCharm IDE

Or

• Go to File/Settings
• Click on Plugins
• Click on the Gear Icon
• Select "Install Plugin from Disk..."
• Select /CharmFL-1.9.1.zip

Preconditions

⚠️ This step is crucial for CharmFL to work!

Create a .coveragerc file in your root directory. This is important for CharmFL to run fault localization properly.

After creating the file, you should insert the following:

[run]
dynamic_context = test_function

You can download an example .coveragerc file.

⚠️ This step is crucial for creating Static Call Graphs!

Pyan use Graphviz for creating the static call graph html file.

In order to have it run properly please install Graphviz following their website: https://graphviz.org/

Main features

• Run Fault Localization on 3 different granularities (statement, method, class)
• Change metrics (Tarantula, Ochiai)
• Change tie strategy for ranking
• Generate the static call-graph for your python project
  • CharmFL highlights the far context for the selected method
• Heatmap visualization
• Interact with the Fault Localization algorithm