davext 0.12.1

base module for the DAVE debugger extension

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 related data directly from your process's memory. No need to recompile your code and/or instrument it with nasty macros or fancy libs, just start your debugger !

The DAVE GUI

The DAVE ascii formatters

2 channels 256 samples, min -9.9998E-01, max 1.0000E+00 {
  dSparkline[0] = "[0⎻⎺‾⎺⎻x⎼_⎽⎼x⎺‾⎺⎻x⎽_⎽—x⎺‾⎺⎻x⎽_⎽⎼x⎺‾⎺⎻x⎽_⎽—x⎺‾⎺⎻x⎽_⎼x⎺‾⎺—x_⎽—x⎺‾⎺⎻x⎽_⎽⎼x⎺‾⎺⎻x⎽_⎽—x⎺‾⎺—x_⎼x⎺‾⎺⎻x⎽_⎽⎼x⎺‾⎺x⎽_⎽⎼x⎺‾⎺—x_⎽⎼x⎺‾⎻x⎽_⎽—x‾⎺⎻x⎽_⎼x⎺‾⎺—x_⎽—x‾⎺⎻x_⎽⎼x‾⎺⎻x_⎽⎼x‾⎺⎻x_⎽—x‾⎺—x_]"
  dSparkline[1] = "[0⎻⎺‾⎺⎻x⎼_⎽⎼x⎺‾⎺⎻x⎽_⎽—x⎺‾⎺⎻x⎽_⎽⎼x⎺‾⎺⎻x⎽_⎽—x⎺‾⎺⎻x⎽_⎼x⎺‾⎺—x_⎽—x⎺‾⎺⎻x⎽_⎽⎼x⎺‾⎺⎻x⎽_⎽—x⎺‾⎺—x_⎼x⎺‾⎺⎻x⎽_⎽⎼x⎺‾⎺x⎽_⎽⎼x⎺‾⎺—x_⎽⎼x⎺‾⎻x⎽_⎽—x‾⎺⎻x⎽_⎼x⎺‾⎺—x_⎽—x‾⎺⎻x_⎽⎼x‾⎺⎻x_⎽⎼x‾⎺⎻x_⎽—x‾⎺—x_]"
  [...]
}

DAVE supports plotting for different types of audio "entities" that might live in your program's memory :

• audio sample containers (amplitude or spectral)
• IIR filters

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 IIR filter/coefficients are :

• juce::dsp::IIR::Filter
• juce::dsp::IIR:Coefficients
• juce::dsp::StateVariableTPTFilter
• juce::dsp::StateVariableFilter::Filter
• juce::dsp::StateVariableFilter::Parameters

You can also add supports for your own entity classes

Current support

Operating System	Status
Linux	✅
MacOS	✅
Windows	❌

Debugger platform	GDB	LLDB
CLI	✅	✅
VSCode	✅ (*)	❓ need testing
CLion	✅ (*)	✅ (*)
XCode	N/A	✅ (*)
Visual Studio	❌	❌

(*) - DAVE is not able to react when you go up/down in the stack using the IDE ui.

---

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 a working venv module

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.

You don't trust the script ? Check the Manual Installation Guide

Python : Python support is limited, deprecated 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

---

Licensing

This project is licensed under GPLv3.0. The goal is for you to be able to use it to debug any project, open or proprietary, but to prevent anyone from making a proprietary project out of DAVE's code.

If you have any question or request about licensing, don't hesitate to ask !

Wikipedia: Software under the GPL may be run for all purposes, including commercial purposes and even as a tool for creating proprietary software, such as when using GPL-licensed compilers.

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 24.04

When starting lldb on ubuntu 24.04 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/python3.12/
sudo ln -s /usr/lib/llvm-${VERSION}/lib/python3.12/dist-packages /usr/lib/local/lib/python3.12/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

PYTHON_VERSION="3.12"
LLDB_VERSION="14"
touch .venv/lib64/python${PYTHON_VERSION}/site-packages/lldb.pth
echo "/usr/lib/llvm-${LLDB_VERSION}/lib/python${PYTHON_VERSION}/site-packages/" > .venv/lib64/python${PYTHON_VERSION}/site-packages/lldb.pth