• Stars
    star
    6,459
  • Rank 6,148 (Top 0.2 %)
  • Language
    C++
  • License
    Apache License 2.0
  • Created almost 8 years ago
  • Updated 3 months ago

Reviews

There are no reviews yet. Be the first to send feedback to the community and the maintainers!

Repository Details

Draco is a library for compressing and decompressing 3D geometric meshes and point clouds. It is intended to improve the storage and transmission of 3D graphics.

draco-ci

News

Attention GStatic users: the Draco team strongly recommends using the versioned URLs for accessing Draco GStatic content. If you are using the URLs that include the v1/decoders substring within the URL, edge caching and GStatic propagation delays can result in transient errors that can be difficult to diagnose when new Draco releases are launched. To avoid the issue pin your sites to a versioned release.

Version 1.5.6 release:

  • Using the versioned www.gstatic.com WASM and Javascript decoders continues to be recommended. To use v1.5.6, use this URL:
  • The CMake flag DRACO_DEBUG_MSVC_WARNINGS has been replaced with DRACO_DEBUG_COMPILER_WARNINGS, and the behavior has changed. It is now a boolean flag defined in draco_options.cmake.
  • Bug fixes.
  • Security fixes.

Version 1.5.5 release:

Version 1.5.4 release:

Version 1.5.3 release:

Version 1.5.2 release

  • This is the same as v1.5.1 with the following two bug fixes:
    • Fixes DRACO_TRANSCODER_SUPPORTED enabled builds.
    • ABI version updated.

Version 1.5.1 release

  • Adds assertion enabled Emscripten builds to the release, and a subset of the assertion enabled builds to GStatic. See the file listing below.
  • Custom paths to third party dependencies are now supported. See BUILDING.md for more information.
  • The CMake configuration file draco-config.cmake is now tested and known to work for using Draco in Linux, MacOS, and Windows CMake projects. See the install_test subdirectory of src/draco/tools for more information.
  • Bug fixes.

Version 1.5.0 release

  • Adds the draco_transcoder tool. See the section below on the glTF transcoding tool, and BUILDING.md for build and dependency information.
  • Some changes to configuration variables have been made for this release:
    • The DRACO_GLTF flag has been renamed to DRACO_GLTF_BITSTREAM to help increase understanding of its purpose, which is to limit Draco features to those included in the Draco glTF specification.
    • Variables exported in CMake via draco-config.cmake and find-draco.cmake (formerly FindDraco.cmake) have been renamed. It's unlikely that this impacts any existing projects as the aforementioned files were not formed correctly. See PR775 for full details of the changes.
  • A CMake version file has been added.
  • The CMake install target now uses absolute paths direct from CMake instead of building them using CMAKE_INSTALL_PREFIX. This was done to make Draco easier to use for downstream packagers and should have little to no impact on users picking up Draco from source.
  • Certain MSVC warnings have had their levels changed via compiler flag to reduce the amount of noise output by the MSVC compilers. Set MSVC warning level to 4, or define DRACO_DEBUG_MSVC_WARNINGS at CMake configuration time to restore previous behavior.
  • Bug fixes.

Version 1.4.3 release

Version 1.4.1 release

Version 1.4.0 release

  • WASM and JavaScript decoders are hosted from a static URL.
  • Changed npm modules to use WASM, which increased performance by ~200%.
  • Updated Emscripten to 2.0.
    • This causes the Draco codec modules to return a promise instead of the module directly.
    • Please see the example code on how to handle the promise.
  • Changed NORMAL quantization default to 8.
  • Added new array API to decoder and deprecated DecoderBuffer.
    • See PR #513 for more information.
  • Changed WASM/JavaScript behavior of catching exceptions.
    • See issue #629 for more information.
  • Code cleanup.
  • Emscripten builds now disable NODEJS_CATCH_EXIT and NODEJS_CATCH_REJECTION.
    • Authors of a CLI tool might want to add their own error handlers.
  • Added Maya plugin builds.
  • Unity plugin builds updated.
  • Bug fixes.

Version 1.3.6 release

  • WASM and JavaScript decoders are now hosted from a static URL
  • Changed web examples to pull Draco decoders from static URL
  • Added new API to Draco WASM decoder, which increased performance by ~15%
  • Decreased Draco WASM decoder size by ~20%
  • Added support for generic and multiple attributes to Draco Unity plug-ins
  • Added new API to Draco Unity, which increased decoder performance by ~15%
  • Changed quantization defaults:
    • POSITION: 11
    • NORMAL: 7
    • TEX_COORD: 10
    • COLOR: 8
    • GENERIC: 8
  • Code cleanup
  • Bug fixes

Version 1.3.5 release

  • Added option to build Draco for Universal Scene Description
  • Code cleanup
  • Bug fixes

Version 1.3.4 release

  • Released Draco Animation code
  • Fixes for Unity
  • Various file location and name changes

Version 1.3.3 release

  • Added ExpertEncoder to the Javascript API
    • Allows developers to set quantization options per attribute id
  • Bug fixes

Version 1.3.2 release

  • Bug fixes

Version 1.3.1 release

  • Fix issue with multiple attributes when skipping an attribute transform

Version 1.3.0 release

  • Improved kD-tree based point cloud encoding
    • Now applicable to point clouds with any number of attributes
    • Support for all integer attribute types and quantized floating point types
  • Improved mesh compression up to 10% (on average ~2%)
    • For meshes, the 1.3.0 bitstream is fully compatible with 1.2.x decoders
  • Improved Javascript API
    • Added support for all signed and unsigned integer types
    • Added support for point clouds to our Javascript encoder API
  • Added support for integer properties to the PLY decoder
  • Bug fixes

Previous releases

https://github.com/google/draco/releases

Description

Draco is a library for compressing and decompressing 3D geometric meshes and point clouds. It is intended to improve the storage and transmission of 3D graphics.

Draco was designed and built for compression efficiency and speed. The code supports compressing points, connectivity information, texture coordinates, color information, normals, and any other generic attributes associated with geometry. With Draco, applications using 3D graphics can be significantly smaller without compromising visual fidelity. For users, this means apps can now be downloaded faster, 3D graphics in the browser can load quicker, and VR and AR scenes can now be transmitted with a fraction of the bandwidth and rendered quickly.

Draco is released as C++ source code that can be used to compress 3D graphics as well as C++ and Javascript decoders for the encoded data.

Contents

Building

See BUILDING for building instructions.

Usage

Unity

For the best information about using Unity with Draco please visit https://github.com/atteneder/DracoUnity

For a simple example of using Unity with Draco see README in the unity folder.

WASM and JavaScript Decoders

It is recommended to always pull your Draco WASM and JavaScript decoders from:

https://www.gstatic.com/draco/v1/decoders/

Users will benefit from having the Draco decoder in cache as more sites start using the static URL.

Command Line Applications

The default target created from the build files will be the draco_encoder and draco_decoder command line applications. Additionally, draco_transcoder is generated when CMake is run with the DRACO_TRANSCODER_SUPPORTED variable set to ON (see BUILDING for more details). For all applications, if you run them without any arguments or -h, the applications will output usage and options.

Encoding Tool

draco_encoder will read OBJ, STL or PLY files as input, and output Draco-encoded files. We have included Stanford's Bunny mesh for testing. The basic command line looks like this:

./draco_encoder -i testdata/bun_zipper.ply -o out.drc

A value of 0 for the quantization parameter will not perform any quantization on the specified attribute. Any value other than 0 will quantize the input values for the specified attribute to that number of bits. For example:

./draco_encoder -i testdata/bun_zipper.ply -o out.drc -qp 14

will quantize the positions to 14 bits (default is 11 for the position coordinates).

In general, the more you quantize your attributes the better compression rate you will get. It is up to your project to decide how much deviation it will tolerate. In general, most projects can set quantization values of about 11 without any noticeable difference in quality.

The compression level (-cl) parameter turns on/off different compression features.

./draco_encoder -i testdata/bun_zipper.ply -o out.drc -cl 8

In general, the highest setting, 10, will have the most compression but worst decompression speed. 0 will have the least compression, but best decompression speed. The default setting is 7.

Encoding Point Clouds

You can encode point cloud data with draco_encoder by specifying the -point_cloud parameter. If you specify the -point_cloud parameter with a mesh input file, draco_encoder will ignore the connectivity data and encode the positions from the mesh file.

./draco_encoder -point_cloud -i testdata/bun_zipper.ply -o out.drc

This command line will encode the mesh input as a point cloud, even though the input might not produce compression that is representative of other point clouds. Specifically, one can expect much better compression rates for larger and denser point clouds.

Decoding Tool

draco_decoder will read Draco files as input, and output OBJ, STL or PLY files. The basic command line looks like this:

./draco_decoder -i in.drc -o out.obj

glTF Transcoding Tool

draco_transcoder can be used to add Draco compression to glTF assets. The basic command line looks like this:

./draco_transcoder -i in.glb -o out.glb

This command line will add geometry compression to all meshes in the in.glb file. Quantization values for different glTF attributes can be specified similarly to the draco_encoder tool. For example -qp can be used to define quantization of the position attribute:

./draco_transcoder -i in.glb -o out.glb -qp 12

C++ Decoder API

If you'd like to add decoding to your applications you will need to include the draco_dec library. In order to use the Draco decoder you need to initialize a DecoderBuffer with the compressed data. Then call DecodeMeshFromBuffer() to return a decoded mesh object or call DecodePointCloudFromBuffer() to return a decoded PointCloud object. For example:

draco::DecoderBuffer buffer;
buffer.Init(data.data(), data.size());

const draco::EncodedGeometryType geom_type =
    draco::GetEncodedGeometryType(&buffer);
if (geom_type == draco::TRIANGULAR_MESH) {
  unique_ptr<draco::Mesh> mesh = draco::DecodeMeshFromBuffer(&buffer);
} else if (geom_type == draco::POINT_CLOUD) {
  unique_ptr<draco::PointCloud> pc = draco::DecodePointCloudFromBuffer(&buffer);
}

Please see src/draco/mesh/mesh.h for the full Mesh class interface and src/draco/point_cloud/point_cloud.h for the full PointCloud class interface.

Javascript Encoder API

The Javascript encoder is located in javascript/draco_encoder.js. The encoder API can be used to compress mesh and point cloud. In order to use the encoder, you need to first create an instance of DracoEncoderModule. Then use this instance to create MeshBuilder and Encoder objects. MeshBuilder is used to construct a mesh from geometry data that could be later compressed by Encoder. First create a mesh object using new encoderModule.Mesh() . Then, use AddFacesToMesh() to add indices to the mesh and use AddFloatAttributeToMesh() to add attribute data to the mesh, e.g. position, normal, color and texture coordinates. After a mesh is constructed, you could then use EncodeMeshToDracoBuffer() to compress the mesh. For example:

const mesh = {
  indices : new Uint32Array(indices),
  vertices : new Float32Array(vertices),
  normals : new Float32Array(normals)
};

const encoderModule = DracoEncoderModule();
const encoder = new encoderModule.Encoder();
const meshBuilder = new encoderModule.MeshBuilder();
const dracoMesh = new encoderModule.Mesh();

const numFaces = mesh.indices.length / 3;
const numPoints = mesh.vertices.length;
meshBuilder.AddFacesToMesh(dracoMesh, numFaces, mesh.indices);

meshBuilder.AddFloatAttributeToMesh(dracoMesh, encoderModule.POSITION,
  numPoints, 3, mesh.vertices);
if (mesh.hasOwnProperty('normals')) {
  meshBuilder.AddFloatAttributeToMesh(
    dracoMesh, encoderModule.NORMAL, numPoints, 3, mesh.normals);
}
if (mesh.hasOwnProperty('colors')) {
  meshBuilder.AddFloatAttributeToMesh(
    dracoMesh, encoderModule.COLOR, numPoints, 3, mesh.colors);
}
if (mesh.hasOwnProperty('texcoords')) {
  meshBuilder.AddFloatAttributeToMesh(
    dracoMesh, encoderModule.TEX_COORD, numPoints, 3, mesh.texcoords);
}

if (method === "edgebreaker") {
  encoder.SetEncodingMethod(encoderModule.MESH_EDGEBREAKER_ENCODING);
} else if (method === "sequential") {
  encoder.SetEncodingMethod(encoderModule.MESH_SEQUENTIAL_ENCODING);
}

const encodedData = new encoderModule.DracoInt8Array();
// Use default encoding setting.
const encodedLen = encoder.EncodeMeshToDracoBuffer(dracoMesh,
                                                   encodedData);
encoderModule.destroy(dracoMesh);
encoderModule.destroy(encoder);
encoderModule.destroy(meshBuilder);

Please see src/draco/javascript/emscripten/draco_web_encoder.idl for the full API.

Javascript Decoder API

The Javascript decoder is located in javascript/draco_decoder.js. The Javascript decoder can decode mesh and point cloud. In order to use the decoder, you must first create an instance of DracoDecoderModule. The instance is then used to create DecoderBuffer and Decoder objects. Set the encoded data in the DecoderBuffer. Then call GetEncodedGeometryType() to identify the type of geometry, e.g. mesh or point cloud. Then call either DecodeBufferToMesh() or DecodeBufferToPointCloud(), which will return a Mesh object or a point cloud. For example:

// Create the Draco decoder.
const decoderModule = DracoDecoderModule();
const buffer = new decoderModule.DecoderBuffer();
buffer.Init(byteArray, byteArray.length);

// Create a buffer to hold the encoded data.
const decoder = new decoderModule.Decoder();
const geometryType = decoder.GetEncodedGeometryType(buffer);

// Decode the encoded geometry.
let outputGeometry;
let status;
if (geometryType == decoderModule.TRIANGULAR_MESH) {
  outputGeometry = new decoderModule.Mesh();
  status = decoder.DecodeBufferToMesh(buffer, outputGeometry);
} else {
  outputGeometry = new decoderModule.PointCloud();
  status = decoder.DecodeBufferToPointCloud(buffer, outputGeometry);
}

// You must explicitly delete objects created from the DracoDecoderModule
// or Decoder.
decoderModule.destroy(outputGeometry);
decoderModule.destroy(decoder);
decoderModule.destroy(buffer);

Please see src/draco/javascript/emscripten/draco_web_decoder.idl for the full API.

Javascript Decoder Performance

The Javascript decoder is built with dynamic memory. This will let the decoder work with all of the compressed data. But this option is not the fastest. Pre-allocating the memory sees about a 2x decoder speed improvement. If you know all of your project's memory requirements, you can turn on static memory by changing CMakeLists.txt accordingly.

Metadata API

Starting from v1.0, Draco provides metadata functionality for encoding data other than geometry. It could be used to encode any custom data along with the geometry. For example, we can enable metadata functionality to encode the name of attributes, name of sub-objects and customized information. For one mesh and point cloud, it can have one top-level geometry metadata class. The top-level metadata then can have hierarchical metadata. Other than that, the top-level metadata can have metadata for each attribute which is called attribute metadata. The attribute metadata should be initialized with the correspondent attribute id within the mesh. The metadata API is provided both in C++ and Javascript. For example, to add metadata in C++:

draco::PointCloud pc;
// Add metadata for the geometry.
std::unique_ptr<draco::GeometryMetadata> metadata =
  std::unique_ptr<draco::GeometryMetadata>(new draco::GeometryMetadata());
metadata->AddEntryString("description", "This is an example.");
pc.AddMetadata(std::move(metadata));

// Add metadata for attributes.
draco::GeometryAttribute pos_att;
pos_att.Init(draco::GeometryAttribute::POSITION, nullptr, 3,
             draco::DT_FLOAT32, false, 12, 0);
const uint32_t pos_att_id = pc.AddAttribute(pos_att, false, 0);

std::unique_ptr<draco::AttributeMetadata> pos_metadata =
    std::unique_ptr<draco::AttributeMetadata>(
        new draco::AttributeMetadata(pos_att_id));
pos_metadata->AddEntryString("name", "position");

// Directly add attribute metadata to geometry.
// You can do this without explicitly add |GeometryMetadata| to mesh.
pc.AddAttributeMetadata(pos_att_id, std::move(pos_metadata));

To read metadata from a geometry in C++:

// Get metadata for the geometry.
const draco::GeometryMetadata *pc_metadata = pc.GetMetadata();

// Request metadata for a specific attribute.
const draco::AttributeMetadata *requested_pos_metadata =
  pc.GetAttributeMetadataByStringEntry("name", "position");

Please see src/draco/metadata and src/draco/point_cloud for the full API.

NPM Package

Draco NPM NodeJS package is located in javascript/npm/draco3d. Please see the doc in the folder for detailed usage.

three.js Renderer Example

Here's an example of a geometric compressed with Draco loaded via a Javascript decoder using the three.js renderer.

Please see the javascript/example/README.md file for more information.

GStatic Javascript Builds

Prebuilt versions of the Emscripten-built Draco javascript decoders are hosted on www.gstatic.com in version labeled directories:

https://www.gstatic.com/draco/versioned/decoders/VERSION/*

As of the v1.4.3 release the files available are:

Beginning with the v1.5.1 release assertion enabled builds of the following files are available:

Support

For questions/comments please email [email protected]

If you have found an error in this library, please file an issue at https://github.com/google/draco/issues

Patches are encouraged, and may be submitted by forking this project and submitting a pull request through GitHub. See CONTRIBUTING for more detail.

License

Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at

http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.

References

Bunny model from Stanford's graphic department https://graphics.stanford.edu/data/3Dscanrep/

More Repositories

1

material-design-icons

Material Design icons by Google (Material Symbols)
50,560
star
2

guava

Google core libraries for Java
Java
48,313
star
3

zx

A tool for writing better scripts
JavaScript
42,760
star
4

styleguide

Style guides for Google-originated open-source projects
HTML
37,420
star
5

leveldb

LevelDB is a fast key-value storage library written at Google that provides an ordered mapping from string keys to string values.
C++
36,205
star
6

googletest

GoogleTest - Google Testing and Mocking Framework
C++
34,040
star
7

material-design-lite

Material Design Components in HTML/CSS/JS
HTML
32,281
star
8

comprehensive-rust

This is the Rust course used by the Android team at Google. It provides you the material to quickly teach Rust.
Rust
27,842
star
9

python-fire

Python Fire is a library for automatically generating command line interfaces (CLIs) from absolutely any Python object.
Python
26,842
star
10

mediapipe

Cross-platform, customizable ML solutions for live and streaming media.
C++
25,626
star
11

gson

A Java serialization/deserialization library to convert Java Objects into JSON and back
Java
23,317
star
12

flatbuffers

FlatBuffers: Memory Efficient Serialization Library
C++
23,037
star
13

iosched

The Google I/O Android App
Kotlin
21,772
star
14

ExoPlayer

This project is deprecated and stale. The latest ExoPlayer code is available in https://github.com/androidx/media
Java
21,710
star
15

eng-practices

Google's Engineering Practices documentation
19,942
star
16

web-starter-kit

Web Starter Kit - a workflow for multi-device websites
HTML
18,422
star
17

flexbox-layout

Flexbox for Android
Kotlin
18,230
star
18

fonts

Font files available from Google Fonts, and a public issue tracker for all things Google Fonts
HTML
18,222
star
19

filament

Filament is a real-time physically based rendering engine for Android, iOS, Windows, Linux, macOS, and WebGL2
C++
17,554
star
20

cadvisor

Analyzes resource usage and performance characteristics of running containers.
Go
17,078
star
21

gvisor

Application Kernel for Containers
Go
15,733
star
22

libphonenumber

Google's common Java, C++ and JavaScript library for parsing, formatting, and validating international phone numbers.
C++
15,728
star
23

WebFundamentals

Former git repo for WebFundamentals on developers.google.com
JavaScript
13,851
star
24

yapf

A formatter for Python files
Python
13,755
star
25

brotli

Brotli compression format
TypeScript
13,363
star
26

tink

Tink is a multi-language, cross-platform, open source library that provides cryptographic APIs that are secure, easy to use correctly, and hard(er) to misuse.
Java
13,318
star
27

deepdream

13,212
star
28

wire

Compile-time Dependency Injection for Go
Go
12,919
star
29

guetzli

Perceptual JPEG encoder
C++
12,917
star
30

guice

Guice (pronounced 'juice') is a lightweight dependency injection framework for Java 11 and above, brought to you by Google.
Java
12,458
star
31

blockly

The web-based visual programming editor.
TypeScript
12,392
star
32

sanitizers

AddressSanitizer, ThreadSanitizer, MemorySanitizer
C
11,410
star
33

or-tools

Google's Operations Research tools:
C++
11,144
star
34

dopamine

Dopamine is a research framework for fast prototyping of reinforcement learning algorithms.
Jupyter Notebook
10,529
star
35

grumpy

Grumpy is a Python to Go source code transcompiler and runtime.
Go
10,464
star
36

oss-fuzz

OSS-Fuzz - continuous fuzzing for open source software.
Shell
10,389
star
37

auto

A collection of source code generators for Java.
Java
10,234
star
38

go-github

Go library for accessing the GitHub v3 API
Go
10,206
star
39

go-cloud

The Go Cloud Development Kit (Go CDK): A library and tools for open cloud development in Go.
Go
9,546
star
40

sentencepiece

Unsupervised text tokenizer for Neural Network-based text generation.
C++
8,657
star
41

tsunami-security-scanner

Tsunami is a general purpose network security scanner with an extensible plugin system for detecting high severity vulnerabilities with high confidence.
Java
8,232
star
42

re2

RE2 is a fast, safe, thread-friendly alternative to backtracking regular expression engines like those used in PCRE, Perl, and Python. It is a C++ library.
C++
8,190
star
43

traceur-compiler

Traceur is a JavaScript.next-to-JavaScript-of-today compiler
JavaScript
8,173
star
44

trax

Trax — Deep Learning with Clear Code and Speed
Python
8,051
star
45

pprof

pprof is a tool for visualization and analysis of profiling data
Go
7,875
star
46

skia

Skia is a complete 2D graphic library for drawing Text, Geometries, and Images.
C++
7,874
star
47

benchmark

A microbenchmark support library
C++
7,812
star
48

magika

Detect file content types with deep learning
Rust
7,680
star
49

android-classyshark

Android and Java bytecode viewer
Java
7,492
star
50

accompanist

A collection of extension libraries for Jetpack Compose
Kotlin
7,442
star
51

closure-compiler

A JavaScript checker and optimizer.
Java
7,394
star
52

agera

Reactive Programming for Android
Java
7,227
star
53

latexify_py

A library to generate LaTeX expression from Python code.
Python
7,160
star
54

diff-match-patch

Diff Match Patch is a high-performance library in multiple languages that manipulates plain text.
Python
7,132
star
55

flutter-desktop-embedding

Experimental plugins for Flutter for Desktop
C++
7,102
star
56

glog

C++ implementation of the Google logging module
C++
7,017
star
57

jsonnet

Jsonnet - The data templating language
Jsonnet
6,938
star
58

model-viewer

Easily display interactive 3D models on the web and in AR!
TypeScript
6,858
star
59

lovefield

Lovefield is a relational database for web apps. Written in JavaScript, works cross-browser. Provides SQL-like APIs that are fast, safe, and easy to use.
JavaScript
6,847
star
60

error-prone

Catch common Java mistakes as compile-time errors
Java
6,818
star
61

gops

A tool to list and diagnose Go processes currently running on your system
Go
6,375
star
62

gopacket

Provides packet processing capabilities for Go
Go
6,289
star
63

automl

Google Brain AutoML
Jupyter Notebook
6,230
star
64

osv-scanner

Vulnerability scanner written in Go which uses the data provided by https://osv.dev
Go
6,222
star
65

flax

Flax is a neural network library for JAX that is designed for flexibility.
Jupyter Notebook
6,085
star
66

grafika

Grafika test app
Java
6,071
star
67

snappy

A fast compressor/decompressor
C++
6,068
star
68

physical-web

The Physical Web: walk up and use anything
Java
6,017
star
69

j2objc

A Java to iOS Objective-C translation tool and runtime.
Java
5,990
star
70

gemma.cpp

lightweight, standalone C++ inference engine for Google's Gemma models.
C++
5,961
star
71

ios-webkit-debug-proxy

A DevTools proxy (Chrome Remote Debugging Protocol) for iOS devices (Safari Remote Web Inspector).
C
5,918
star
72

seesaw

Seesaw v2 is a Linux Virtual Server (LVS) based load balancing platform.
Go
5,634
star
73

EarlGrey

🍵 iOS UI Automation Test Framework
Objective-C
5,616
star
74

seq2seq

A general-purpose encoder-decoder framework for Tensorflow
Python
5,577
star
75

google-java-format

Reformats Java source code to comply with Google Java Style.
Java
5,538
star
76

mesop

Rapidly build AI apps in Python
Python
5,401
star
77

wireit

Wireit upgrades your npm/pnpm/yarn scripts to make them smarter and more efficient.
TypeScript
5,385
star
78

syzkaller

syzkaller is an unsupervised coverage-guided kernel fuzzer
Go
5,350
star
79

uuid

Go package for UUIDs based on RFC 4122 and DCE 1.1: Authentication and Security Services.
Go
5,284
star
80

clusterfuzz

Scalable fuzzing infrastructure.
Python
5,283
star
81

battery-historian

Battery Historian is a tool to analyze battery consumers using Android "bugreport" files.
Go
5,249
star
82

gemma_pytorch

The official PyTorch implementation of Google's Gemma models
Python
5,242
star
83

bbr

5,156
star
84

gumbo-parser

An HTML5 parsing library in pure C99
HTML
5,141
star
85

git-appraise

Distributed code review system for Git repos
Go
5,122
star
86

google-authenticator

Open source version of Google Authenticator (except the Android app)
Java
5,077
star
87

gts

☂️ TypeScript style guide, formatter, and linter.
TypeScript
5,071
star
88

closure-library

Google's common JavaScript library
JavaScript
4,881
star
89

grr

GRR Rapid Response: remote live forensics for incident response
Python
4,757
star
90

cameraview

[DEPRECATED] Easily integrate Camera features into your Android app
Java
4,734
star
91

pytype

A static type analyzer for Python code
Python
4,731
star
92

liquidfun

2D physics engine for games
C++
4,559
star
93

clasp

🔗 Command Line Apps Script Projects
TypeScript
4,525
star
94

google-ctf

Google CTF
Python
4,477
star
95

gxui

An experimental Go cross platform UI library.
Go
4,450
star
96

santa

A binary authorization and monitoring system for macOS
Objective-C++
4,402
star
97

bloaty

Bloaty: a size profiler for binaries
C++
4,386
star
98

tcmalloc

C++
4,339
star
99

ko

Build and deploy Go applications on Kubernetes
Go
4,329
star
100

orbit

C/C++ Performance Profiler
C++
4,190
star