software-update-checker

A command line interface to check if local software installs have updates.

Tired of opening countless programs to check for updates? So was I, which is why I created this project, which acts as a centralized spot to not only get the versions of your locally installed software, but to get their latest versions, and compare that against your local version to see if you have an update available.

How does it work? You configure a software by specifying a command to execute to produce a version string (e.g. git --version), and a URL containing the latest version of that software (e.g. https://github.com/git-for-windows/git/releases/latest), specifying a RegEx Group to extract the exact version strings.

---

Usage

The command line interface (CLI) tool can be used in both an interactive (with prompts and selections) and non-interactive (with runtime flags) way, depending on usage needs. For ideas on how to configure your software, see the example configurations.

Interactive

Simply run the executable to have a list of options displayed. Select an option and follow the additional prompts. Targets (found with the --help flag) can be added with the --interactive flag to open to a specific part of the tool. The interactive mode will persist until a user specifies the Exit option or closes the command session (with ctrl + c).

Non-Interactive

To see a list of available targets and flags, run the executable with the --help flag. All targets have their own --help flags to list their specific flags.

Example Configurations

Reference the following files for what software configurations should look like based on your platform:

• Linux
• MacOS
• Windows

---

Development

Want to modify software-update-checker yourself? Follow these steps below:

Requirements

• Node.js
• Powershell for e2e tests

Install Dependencies

To pull down all dependencies required by this project, run

npm install

Debug

To run the project quickly for development, run

npm start

Package

To create executables for the project that can be run standalone (without Node.js), run

npm run dist

Executables will be created in the dist directory

Lint

This project is configured to be linted with ESLint to make sure code is syntactially correct, and also uses Prettier to make sure code is formatted consistently. Jenkinsfiles are linted with npm-groovy-lint.

To check code for proper syntax and formatting, run

npm run lint

To attempt to fix lint issues automatically, run

npm run lint:fix

Test

To test the project for regressions, run

npm test

This will lint the project, then run unit, functional, and e2e tests.

Unit Tests

Unit tests are for stateless logic tests on small, contained parts of the code base (ideally not reliant on/mock anything outside their scope), and can be run with

npm run test:unit

Note: To run a specific test, execute

npm run test:unit -- -t 'test name'

With spaces separating describe blocks and test names

Functional Tests

Functional tests are for portions of the code that cannot easily be mocked or rely on elements outside their scope, such as executables and interacting with a live website. They can be run with

npm run test:func

Note: To run a specific test, execute

npm run test:func -- -t 'test name'

With spaces separating describe blocks and test names

End to End Tests

End to End (E2E) Tests run against the final, distributable executable and are meant to test overarching scenarios of the tool to simulate as close as possible the real interactions and use/edge cases.

They required code to be packaged, and can be run with

npm run test:e2e

Note: To run a specific test, execute

npm run test:e2e -- -t 'test name'

With spaces separating describe blocks and test names

E2E test execution can be configured with the following environment variables:

Name	Required	Default	Description	Example(s)
E2E_EXEC_MIN_QUIET_PERIOD	yes	250	The minimum amount of time (in milliseconds) that the E2E interactive execution should wait before sending prompt responses.	300
E2E_LOCAL_IP_ADDRESS	yes	127.0.0.1	The IP address of the machine running the E2E tests	::1

Code Coverage

By default, code coverage will be generated for unit and functional tests in the coverage directory in unit and func sub-directories respectively. Their individual coverages can be merged into an overall coverage report with

npm run coverage:merge

There is no code coverage for e2e tests as those do not run against source code, but binaries.

To view code coverage in the browser, run

npm run coverage:view

Append :unit or :func to view coverage for unit or functional tests only, respectively.