.gdbinit modifications for using a Virtual/Conda environment with GDB Python scripts

.gdbinit

# All of your normal .gdbinit commands, functions, and setup tasks

# Update GDB's Python paths with the `sys.path` values of the local Python installation,
#  whether that is brew'ed Python, a virtualenv, or another system python.

# Convert GDB to interpret in Python
python
import os,subprocess,sys
# Execute a Python using the user's shell and pull out the sys.path (for site-packages)
paths = subprocess.check_output('python -c "import os,sys;print(os.linesep.join(sys.path).strip())"',shell=True).decode("utf-8").split()
# Extend GDB's Python's search path
sys.path.extend(paths)
end

This was created for users to get around GDB not loading the activated virtual environments Python modules

If you are looking for an equivalent .lldbinit version, check out the other Gist:
https://gist.github.com/tyhoff/7a286945ef75947ad49a347dbc8708ca

Hi @tyhoff ! thanks for providing this workaround.

Just a quick note: if you use conda, you can also simply install GDB through conda (from the conda-forge channel). The conda-forge GDB is built with the Python from conda-forge and so will automatically look for packages in the conda environment, so you would not need the above workaround. LLDB is also available from conda-forge, but I did not succeed in using it to debug on macOS.