davext 0.8.0

No project description provided

Base python module for the DAVE project

Important: It is not recommended to install DAVE directly from pypi, you should rather use the install script as show in the "Project Setup" section.

Debugger Audio Visualization Extension

DAVE is a GDB & LLDB extension which provide commands to visualize audio data directly from your buffers/classes. No need to recompile your code and/or instrument it with nasty macros or fancy libs, just start your debugger !

The DAVE GUI

Currently supported audio containers (both in float and double) are :

1D (mono) containers:

• std::array
• std::vector
• std::span
• Microsoft's gsl::span
• C array
• pointer
• choc::buffer::MonoView
• choc::buffer::MonoVBuffer

2D (multichannel) containers:

• Any nesting of C/C++ standards 1D containers
• juce::AudioBuffer
• juce::dsp::AudioBlock
• choc::buffer::InterleavedView
• choc::buffer::InterleavedBuffer
• choc::buffer::ChannelArrayView
• choc::buffer::ChannelArrayBuffer

Currently supported OS are Linux and MacOS. Both GNU and LLVM stdlib implementation are supported.

---

To get started first follow the Project Setup guide below, then you can read the User Guide to get familiar with DAVE.

If you want to develop/experiment with dave, follow the Development Guide

Project setup

Requirements

DAVE requires a python >= 3.10 installation with working venv and tkinter modules

Installation

The simplest way to install dave and its bindings is to use the install script :

# via the install script
## using curl
bash -c "$(curl -fsSL https://raw.githubusercontent.com/maxmarsc/dave/refs/heads/main/dave/assets/dave_install.sh)"

## using wget
bash -c "$(wget https://raw.githubusercontent.com/maxmarsc/dave/refs/heads/main/dave/assets/dave_install.sh -O -)"

It will install the python dave modules, the debuggers bindings, and the dave cli tool to help manage your dave installation.

MacOS : On MacOS distribution you might need to install tkinter using homebrew/macports. Python : Python support is limited and requires a different installation method, see the section below

---

After binding, starts your debugger, you should see the following message :

[dave] Successfully loaded

And the dave debugger commands should be available :

• dave show
• dave delete
• dave freeze
• dave concat

See the User Guide on how to use these.

Update

If you want/need to update dave, you can use the dave cli tool :

# Update dave
dave update

Uninstallation

If you just want to remove the dave bindings run

dave unbind

If you want to completely remove dave from your system run

dave uninstall

Python support

Python support is limited because both python debuggers I have investigated (pdb and debugpy) does not provide an API complete enough to provide full DAVE support. Most importantly Python debugger does not provide automatic update on breakpoints.

To install DAVE for python, install the davext pypi package in your environment.

Only 1D and 2D numpy tensors are supported, and you need to manually import dave from the debugger CLI, like this:

import dave.debuggers.pdb as pydave

Then you can use the following commands :

• pydave.show analog to dave show
• pydave.concat analog to dave concat
• pydave.freeze analog to dave freeze
• pydave.delete analog to dave delete
• pydave.update which forces the update of the containers data

---

Milestones

• Freeze
• Concatenate
• Save to disc
• Add tooltip hover for action buttons
• out-of-scope detection
• react to step over/up/down
• Mid/Side
• Interleaved
• Easy installation
• improve logging system
• document commands
• add license
• llvm libc/libstd (need to be tested)
• JUCE
• GSL
• Add samplerate support
• Add a way for the user to add custom container support
• Add command to help diagnostic type for custom container support
• Add argument parsing to show (view, layout, settings...)
• Add versionning selection to dave_install
• Add FISH shell support
• command aliases
• minimize call to render functions
• CHOC
• Eigen
• Document the code
• find testing strategy
• Restart dave process
• Improve GUI proportions
• container deletion (GUI + delete command)
• LLDB

Troubleshooting

Set logging level

When running into an issue, please activate the debug log level, by setting the env variable DAVE_LOGLEVEL to debug before starting the debugger.

LLDB on Ubuntu 22.04

When starting lldb on ubuntu 22.04 you might get this error :

Traceback (most recent call last):
  File "<string>", line 1, in <module>
ModuleNotFoundError: No module named 'lldb'

This is a known bug. In the meantime you can fix these using a symbolic link :

sudo mkdir -p /usr/lib/local/lib/python3.10/
sudo ln -s /usr/lib/llvm-${VERSION}/lib/python3.10/dist-packages /usr/lib/local/lib/python3.10/dist-packages

LLDB init

To always load the .lldbinit file in the current working directory, add the following command to ~/.lldbinit:

settings set target.load-cwd-lldbinit true

LLDB python module linting

for venv

touch venv/lib64/python3.10/site-packages/lldb.pth
echo "/usr/lib/llvm-${VERSION}/lib/python3.10/dist-packages/" > venv/lib64/python3.10/site-packages/lldb.pth