hash-runner

hash-runner executes a command when a change is detected in specified files. It calculates the SHA256 hash of the files and compares them to the previous hash values stored in a file. If the hashes differ, the specified command is executed.

This tool is not an active file watcher that constantly checks for changes.

Use-case

It is designed to be used in conjunction with tools like turbo, where the watch mode may trigger unnecessary builds even when caching is used.

For example, consider the following turbo configuration:

{
  "$schema": "https://turbo.build/schema.json",
  "tasks": {
    "dev": {
      "dependsOn": ["@internal/some-package#build"],
      "cache": false,
      "persistent": true
    }
  }
}

When running turbo [watch] dev, it will trigger a build even when no changes are detected in @internal/some-package.

By using hash-runner, the build command will still run on @internal/some-package, but it won't actually execute the underlying build command unless changes are detected.

Installation

Node.js 20 or higher is required.

To install hash-runner, use npm:

npm install hash-runner --save-dev

Usage

hash-runner [--config <config-file>] [--force] [--silent]

CLI options:

• -c / --config <config-file>: Specify a custom configuration file.
• -f / --force: Force the creation of a new hash file and execute.
• -s / --silent: Suppress log output.

Configuration

hash-runner uses lilconfig to read configuration.

lilconfig will check the current directory for the following:

• a hash-runner property in package.json
• a .hash-runnerrc file in JSON or YAML format
• a .hash-runnerrc.json, .hash-runnerrc.js, .hash-runnerrc.ts, .hash-runnerrc.mjs, or .hash-runnerrc.cjs file
• a hash-runnerrc, hash-runnerrc.json, hash-runnerrc.js, hash-runnerrc.ts, hash-runnerrc.mjs, or hash-runnerrc.cjs file inside a .config subdirectory
• a hash-runner.config.js, hash-runner.config.ts, hash-runner.config.mjs, or hash-runner.config.cjs file

Configuration options

Option	Type	Required	Description
inputs	object	Yes	Contains input file configuration
inputs.includes	string[]	Yes	Array of glob patterns specifying files to include in hash calculation
inputs.excludes	string[]	No	Array of glob patterns specifying files to exclude from hash calculation
outputs	object	No	Contains output file configuration for tracking build artifacts
outputs.includes	string[]	Yes*	Array of glob patterns specifying output files to track (*required if outputs is defined)
outputs.excludes	string[]	No	Array of glob patterns specifying output files to exclude from tracking
execOnChange	string	Yes	Command to execute when changes are detected
hashFile	string	Yes	Path to file where hashes are stored (recommend adding to .gitignore)
parallelizeComparisonsChunkSize	number	No	Number of files per chunk for parallelizing hash comparison (default: 100)

Notes:

• The hash file and configuration file are automatically excluded from hash calculations
• If outputs are configured but missing or changed, the cache is considered stale and the command executes
• hash-runner exits with the status code of the executed command

Examples

Run tsc when changes are detected in files in the src directory:

package.json:

{
  "scripts": {
    "build": "hash-runner",
    "build:files": "tsc"
  }
}

Example configuration file (.hash-runnerrc.json):

{
  "inputs": {
    "includes": ["src/**/*.ts"],
    "excludes": ["dist/**"]
  },
  "outputs": {
    "includes": ["dist/**/*.js"],
    "excludes": ["dist/**/*.map"]
  },
  "execOnChange": "npm run build:files",
  "hashFile": ".hashes.json"
}

hash-runner.config.js:

module.exports = {
  inputs: {
    includes: ['src/**/*.ts'],
    excludes: ['dist/**']
  },
  outputs: {
    includes: ['dist/**/*.js'],
    excludes: ['dist/**/*.map']
  },
  execOnChange: 'npm run build:files',
  hashFile: '.hashes.json'
};

Basic Configuration (Inputs Only)

If you only need to track input files (similar to v3 behavior):

{
  "inputs": {
    "includes": ["src/**/*.ts"]
  },
  "execOnChange": "npm run build:files",
  "hashFile": ".hashes.json"
}

npm run build will only run tsc when changes are detected in files in the src directory.

Migration from v3 to v4

If you're upgrading from hash-runner v3, you'll need to manually update your configuration to the v4 format. For detailed migration instructions, see MIGRATING.md.

CI Mode

When running in a Continuous Integration (CI) environment, hash-runner will bypass hash checks and execute the specified command directly. This is controlled by the CI environment variable. To enable CI mode, set the CI environment variable to true:

{
  "scripts": {
    "build": "hash-runner",
    "build:ci": "CI=true hash-runner",
    "build:files": "tsc"
  }
}

This will bypass hash checks and execute the specified command directly.

API

In addition to the CLI, hash-runner can also be used programmatically:

import { HashRunner } from 'hash-runner';

const runner = new HashRunner('/path/to/config.json', { force: true });
await runner.run();

Constructor and Parameters

HashRunner(configPath?: string, options: HashRunnerOptions = {})

Parameters:

• configPath (optional): A string representing the path to the configuration file. If not specified, hash-runner will attempt to load the configuration from the current directory.
• options (optional): An object containing options to configure the behavior of hash-runner.

HashRunnerOptions

• force? (boolean): Force the creation of a new hash file and execute the command regardless of detected changes.
• silent? (boolean): Suppress log output when set to true.

Troubleshooting

This library uses debug to log messages. To enable debug messages, set the DEBUG environment variable to hash-runner.

DEBUG=hash-runner hash-runner