StackCollector

This is a Pharo project which launches projects, collect the execution trace and then returns this trace. Note that it only supports launching of Python projects for now.

Installation

In a Pharo 13 image, launch the following code in a Playground

Metacello new
	baseline: 'StackCollector';
	repository: 'github://Divergence-Driven-Debugging/D3-StackCollector';
	onConflictUseIncoming;
	load

Setup python environment

The Python trace collector relies on a local Python virtual environment and the debugpy DAP adapter. From the root directory of your Pharo image, create a virtual environment named venv and install debugpy:

python3 -m venv venv
source venv/bin/activate
pip install -r pharo-local/iceberg/Divergence-Driven-Debugging/D3-StackCollector/requirements.txt

The trace collector expects the following structure:

<image-directory>/
├── venv/
│   ├── bin/
│   │   └── python3
├── Pharo.image
└── pharo-local/

Setup external tools

You also have to install external tools for other langages (Java, JavaScript, ...) and add their path to the .env file:

• java-debug adapter from Microsoft (JAVA_DEBUG in .env);
• java language serveur (JDTLS in .emv);
• vscode-js-debug (JS_DEBUG in .env).

Usage

Once you've install the requierements, you can use the trace collector by doing in your Pharo Playground:

result := TraceCollector new
	      collectWithoutAssignmentFor: ProjectLangage 
		  onProject: 'my/project/directory' 
		  withEntryPoint: '/my/project/directory/entrypoint'

Where:

• ProjectLangage is a subclass of SupportedLanguage corresponding to your project langage (ex: PythonLanguage, JSLanguage, JavaLanguage, ...)
• 'my/project/directory' is the path to your project directory
• '/my/project/directory/entrypoint' is the path to the entry point of your project (ex: a main.py file in a python project)

Then you will get an array corresponding to the execution trace in result variable.