Initial Communication

Before the official start of the first week, I had already engaged in discussions with my mentor regarding the scope and overall codebase of noWorkFlow.

This week was all about getting familiarized with noWorkFlow and its dependencies, which I’ll be using for my project in GSoC 2024. I started by reading some documentation and tutorials about AST (Abstract Syntax Tree) to gaining a better understanding of their structures and functionalities.

AST is a data structure that represent a code in the form of tree. In the context of my project, AST plays a pivotal role in enabling the comparison of provenance data from different trials. By traversing and inspecting the AST of source code, I can extract essential information and identify similarities or discrepancies between trials.

Learnings

My mentors has given some guidance for the documentation of AST, especially from Green Tree Snakes. I also watch some Python Conference that specifically talks about AST.

I began practicing writing basic code with AST to read and analyze source code with a different types of Python constructs using ast.NodeVisitor. The output initially aimed to generate a graph representing the code’s structure, but for now, it will be a table that compares different node types across multiple trials.

Difficulty

I encountered several challenges during the setup of my development environment, many of which arose due to outdated documentation of noWorkFlow. These obstacles prompted me to search online for solutions to troubleshoot effectively.

The difficulties I faced included:

• Incompatibility with my current Python versions ✔️
• Discrepancies in driver paths conflicting with environment paths ✔️
• Ongoing integration issues with IPython and noWorkFlow ❌

Currently, the solution for the development environment involves prioritizing Python development initially, with the integration of IPython/Jupyter postponed to later stages. Next week, I plan to start learning SQLAlchemy, another key tool in noWorkFlow.

Thanks to my mentor, João Felipe Pimentel, for helping me get started and guiding me.

Looking forward to sharing more progress with you next week. See you then!