v1: Dataflow graph

[Shofiya2003]

Dataflow graph consists of two nodes: FileNode[nodeType, name, mode, version] and ProcessNode[nodeType, tid, exec_epoch]

Preview of the graph:
Command: bash -c "head ../flake.nix ; head ../flake.lock"

Command: ./PROBE record tests/c/createFile.exe

Additional Changes:

• Fixed the happens before graph to include edge of CloneOp node to the first occurrence of matching pthread_id in the kernel thread instead of all the occurrences

[charmoniumQ]

Remove mode from FileNode and remove TID from process node; the threads in a process can communicate with each other, so if one of them knows the contents of a file, they all do.

[Shofiya2003]

I will resolve the errors caused by pulling the latest changes, address the failing checks, and mark it for review by tomorrow.

[Shofiya2003]

I've created a makeshift function to generate a cmd for the ProcessNode since it's not being recorded during the provenance. However, it doesn't handle some edge cases well. Should we consider removing cmd from ProcessNode for now? I can create a PR to add cmd later when we start recording it.

[charmoniumQ]

I can put something in libprobe when the user does an ExecOp that has the arguments and environment that got executed.

[Shofiya2003]

I can put something in libprobe when the user does an ExecOp that has the arguments and environment that got executed.

that would help!

[charmoniumQ]

I still think we don't need any new __init__.py files. Let's resolve that over Discord.