Generate an easily queryable JSON dataset about Brave public releases
This script stitches together data about public
Brave releases from Github and the
package.json data from brave-browser to produce
a central dataset about Brave versions.
Dependencies
Clone and install dependencies
git clone https://github.com/marshall/brave-versions
cd brave-versions
nvm use
yarn installRun brave-versions
yarn run brave-versionsBy default generated json is written to brave-versions.json
Configuration is currently done through command line switches and environment variables / dotenv.
Command line
$ brave-versions --help
Usage: brave-versions [options]
Options:
-V, --version output the version number
--brave-browser <dir> existing brave-browser git repo (default: "/Users/marshall/.brave-versions/brave-browser")
--cache-dir <dir> cache in dir (default: "/Users/marshall/.brave-versions")
--cache-github-releases enable cached github releases (default: false)
--no-git-pull skip git pull in brave-browser (default: git pull to update)
-o, --output <file> path to output json manifest (default: brave-versions.json)
-h, --help display help for command
Environment Variables
| Name | Description |
|---|---|
BRAVE_VERSIONS_DIR |
path where data is cached, by default this is $HOME/.brave-versions |
GITHUB_TOKEN |
optional authorization token for Github API (use to raise the Github API rate limit) |
The manifest generated by brave-versions is a JSON file with the following general structure:
{
"<git tag>": {
"tag": "<git tag>",
"name": "<version name>",
"channel": "stable|beta|nightly|dev",
"commit": "<git sha>",
"published": "<release publish date/time>",
"dependencies": {
"chrome": "<chrome version>",
"widevine": "<widevine version>"
},
"github": {
"release_id": "<github release id>",
"assets": [
{
"id": "<github asset id>",
"name": "<filename>",
"download_url": "<github download url>"
},
//...
]
}
},
//...
}