Gauge updates for July

What’s new in release 0.9.0 and more

Zabil Cheriya Maliackal

Published in

Gauge

3 min readJul 21, 2017

Cleaner Command Line Interface

Before 0.9.0, Gauge used a flat list of options for running commands. For example,

# Initializing a project
$ gauge --init java# Installing a plugin
$ gauge --install# Running a spec with tags 
$ gauge specs --tags="smoke"# Ignores that don't work together
$ gauge --install java --tags="smoke"

As Gauge is a test runner, plug-in/template manager and refactoring tool rolled into one executable, we've removed clutter by grouping these features.

With sub-commands it’s easier to discover features and use the right options (flags). For example,

# Install a plug-in
$ gauge install ruby# Use templates to create a project
$ gauge init java# Running tests
$ gauge run specs # Context specific options
$ gauge run specs --tags smoke# Show option as invalid.
$ gauge install java --tags smoke
Error: unknown flag: --tags

Here’s a list of all the sub commands.

> gauge
Usage:
  gauge <command> [flags] [args]Examples:
  gauge run specs/
  gauge run --parallel specs/Commands:
  config      Change global configurations
  daemon      Run as a daemon
  docs        Generate documentation using specified plugin
  format      Formats the specified spec files
  help        Help about any command
  init        Initialize project structure in the current directory
  install     Download and install plugin(s)
  refactor    Refactor steps
  run         Run specs
  telemetry   Configure options for sending anonymous usage stats
  uninstall   Uninstalls a plugin
  update      Updates a plugin
  validate    Check for validation and parse errors
  version     Print Gauge and plugin versions

There’s help for each command,

> gauge help init
Initialize project structure in the current directory.Usage:
  gauge init [flags] <template>Examples:
  gauge init javaFlags:
  -h, --help        Help for init
  -t, --templates   Lists all available templates

Plus, man pages offline and online!

Telemetry

This release of gauge uses telemetry to collect anonymous usage data. More about why, here.

If you don’t want to help improve gauge, easily opt out by running,

> gauge telemetry off

Language Server Protocol

Gauge is betting on LSP to make it Bring Your Own Editor platform.

Release 0.9.0 implements a language server for Gauge that can be wired up to editors to auto-suggest steps, concepts, etc. while authoring scenarios.

Here’s a plug-in using the Gauge language server to auto-suggest in Visual Studio Code,

Easier builds

Build Gauge, from the master branch, with a single command

> go get github.com/getgauge/gauge

There’s no build script, configuration or need to clone the git repository. It just works.

The gaugebinary compiles to $GOPATH/bin. Use it run on bleeding edge to test new features and fixes.

These changes make Gauge single binary (from 0.9.0). Drop it anywhere in $PATH and run it like you normally do.

There’s more!

For a full list of changes visit our release page.

Thanks

A big thanks to Mikhail Dolinin and Rabih Halage for their valuable pull requests!