loader-utils

What is loader-utils?

The loader-utils package provides utility functions for use with webpack loaders. It offers a variety of helper methods to make writing loaders easier and more standardized. These utilities include parsing query strings, hashing, getting options from loaders, and more.

What are loader-utils's main functionalities?

getOptions

This function is used to parse the options passed to a loader. It simplifies the process of handling loader options within a webpack configuration.

const loaderUtils = require('loader-utils');
const options = loaderUtils.getOptions(this);

stringifyRequest

This method helps to convert a module request into a string that can be used inside of a loader. It's useful for ensuring that the request is correctly formatted and can be resolved by webpack.

const loaderUtils = require('loader-utils');
const stringifiedRequest = loaderUtils.stringifyRequest(this, require.resolve('./file.js'));

getHashDigest

This utility generates a hash digest of the given content. It's commonly used for cache busting or to generate unique identifiers for file contents.

const loaderUtils = require('loader-utils');
const hashDigest = loaderUtils.getHashDigest(Buffer.from('some content'), 'sha512', 'hex', 7);

interpolateName

This function allows for the creation of a custom filename based on placeholders and content. It's often used in file-loader and url-loader to generate names based on the file content.

const loaderUtils = require('loader-utils');
const interpolatedName = loaderUtils.interpolateName(this, '[name]_[hash].[ext]', { content: source });

Other packages similar to loader-utils

schema-utils

Schema-utils validates options for webpack loaders and plugins against a JSON Schema. It is similar to loader-utils in that it helps with managing loader options, but it focuses on validation rather than utility functions.

file-loader

File-loader resolves import/require() on a file into a url and emits the file into the output directory. It's similar to loader-utils' interpolateName feature but is a standalone loader that focuses on file handling.

url-loader

Url-loader works like file-loader but can return a Data URL if the file is smaller than a byte limit. It shares some functionality with loader-utils, particularly in generating filenames and paths.

README

loader-utils

Methods

getOptions

Recommended way to retrieve the options of a loader invocation:

// inside your loader
const options = loaderUtils.getOptions(this);

1. If this.query is a string:
   • Tries to parse the query string and returns a new object
   • Throws if it's not a valid query string
2. If this.query is object-like, it just returns this.query
3. In any other case, it just returns null

Please note: The returned options object is read-only. It may be re-used across multiple invocations. If you pass it on to another library, make sure to make a deep copy of it:

const options = Object.assign(
	{},
	defaultOptions,
	loaderUtils.getOptions(this) // it is safe to pass null to Object.assign()
);
// don't forget nested objects or arrays
options.obj = Object.assign({}, options.obj); 
options.arr = options.arr.slice();
someLibrary(options);

clone is a good library to make a deep copy of the options.

Options as query strings

If the loader options have been passed as loader query string (loader?some&params), the string is parsed by using parseQuery.

parseQuery

Parses a passed string (e.g. loaderContext.resourceQuery) as a query string, and returns an object.

const params = loaderUtils.parseQuery(this.resourceQuery); // resource: `file?param1=foo`
if (params.param1 === "foo") {
	// do something
}

The string is parsed like this:

                             -> Error
?                            -> {}
?flag                        -> { flag: true }
?+flag                       -> { flag: true }
?-flag                       -> { flag: false }
?xyz=test                    -> { xyz: "test" }
?xyz=1                       -> { xyz: "1" } // numbers are NOT parsed
?xyz[]=a                     -> { xyz: ["a"] }
?flag1&flag2                 -> { flag1: true, flag2: true }
?+flag1,-flag2               -> { flag1: true, flag2: false }
?xyz[]=a,xyz[]=b             -> { xyz: ["a", "b"] }
?a%2C%26b=c%2C%26d           -> { "a,&b": "c,&d" }
?{data:{a:1},isJSON5:true}   -> { data: { a: 1 }, isJSON5: true }

stringifyRequest

Turns a request into a string that can be used inside require() or import while avoiding absolute paths. Use it instead of JSON.stringify(...) if you're generating code inside a loader.

Why is this necessary? Since webpack calculates the hash before module paths are translated into module ids, we must avoid absolute paths to ensure consistent hashes across different compilations.

This function:

• resolves absolute requests into relative requests if the request and the module are on the same hard drive
• replaces \ with / if the request and the module are on the same hard drive
• won't change the path at all if the request and the module are on different hard drives
• applies JSON.stringify to the result

loaderUtils.stringifyRequest(this, "./test.js");
// "\"./test.js\""

loaderUtils.stringifyRequest(this, ".\\test.js");
// "\"./test.js\""

loaderUtils.stringifyRequest(this, "test");
// "\"test\""

loaderUtils.stringifyRequest(this, "test/lib/index.js");
// "\"test/lib/index.js\""

loaderUtils.stringifyRequest(this, "otherLoader?andConfig!test?someConfig");
// "\"otherLoader?andConfig!test?someConfig\""

loaderUtils.stringifyRequest(this, require.resolve("test"));
// "\"../node_modules/some-loader/lib/test.js\""

loaderUtils.stringifyRequest(this, "C:\\module\\test.js");
// "\"../../test.js\"" (on Windows, in case the module and the request are on the same drive)

loaderUtils.stringifyRequest(this, "C:\\module\\test.js");
// "\"C:\\module\\test.js\"" (on Windows, in case the module and the request are on different drives)

loaderUtils.stringifyRequest(this, "\\\\network-drive\\test.js");
// "\"\\\\network-drive\\\\test.js\"" (on Windows, in case the module and the request are on different drives)

urlToRequest

Converts some resource URL to a webpack module request.

i Before call urlToRequest you need call isUrlRequest to ensure it is requestable url

const url = "path/to/module.js";

if (loaderUtils.isUrlRequest(url)) {
  // Logic for requestable url
  const request = loaderUtils.urlToRequest(url);
} else {
  // Logic for not requestable url
}

Simple example:

const url = "path/to/module.js";
const request = loaderUtils.urlToRequest(url); // "./path/to/module.js"

Module URLs

Any URL containing a ~ will be interpreted as a module request. Anything after the ~ will be considered the request path.

const url = "~path/to/module.js";
const request = loaderUtils.urlToRequest(url); // "path/to/module.js"

Root-relative URLs

URLs that are root-relative (start with /) can be resolved relative to some arbitrary path by using the root parameter:

const url = "/path/to/module.js";
const root = "./root";
const request = loaderUtils.urlToRequest(url, root); // "./root/path/to/module.js"

To convert a root-relative URL into a module URL, specify a root value that starts with ~:

const url = "/path/to/module.js";
const root = "~";
const request = loaderUtils.urlToRequest(url, root); // "path/to/module.js"

interpolateName

Interpolates a filename template using multiple placeholders and/or a regular expression. The template and regular expression are set as query params called name and regExp on the current loader's context.

const interpolatedName = loaderUtils.interpolateName(loaderContext, name, options);

The following tokens are replaced in the name parameter:

• [ext] the extension of the resource
• [name] the basename of the resource
• [path] the path of the resource relative to the context query parameter or option.
• [folder] the folder of the resource is in.
• [emoji] a random emoji representation of options.content
• [emoji:<length>] same as above, but with a customizable number of emojis
• [contenthash] the hash of options.content (Buffer) (by default it's the hex digest of the md5 hash)
• [<hashType>:contenthash:<digestType>:<length>] optionally one can configure
  • other hashTypes, i. e. sha1, md5, sha256, sha512
  • other digestTypes, i. e. hex, base26, base32, base36, base49, base52, base58, base62, base64
  • and length the length in chars
• [hash] the hash of options.content (Buffer) (by default it's the hex digest of the md5 hash)
• [<hashType>:hash:<digestType>:<length>] optionally one can configure
  • other hashTypes, i. e. sha1, md5, sha256, sha512
  • other digestTypes, i. e. hex, base26, base32, base36, base49, base52, base58, base62, base64
  • and length the length in chars
• [N] the N-th match obtained from matching the current file name against options.regExp

In loader context [hash] and [contenthash] are the same, but we recommend using [contenthash] for avoid misleading.

Examples

// loaderContext.resourcePath = "/app/js/javascript.js"
loaderUtils.interpolateName(loaderContext, "js/[hash].script.[ext]", { content: ... });
// => js/9473fdd0d880a43c21b7778d34872157.script.js

// loaderContext.resourcePath = "/app/js/javascript.js"
loaderUtils.interpolateName(loaderContext, "js/[contenthash].script.[ext]", { content: ... });
// => js/9473fdd0d880a43c21b7778d34872157.script.js

// loaderContext.resourcePath = "/app/page.html"
loaderUtils.interpolateName(loaderContext, "html-[hash:6].html", { content: ... });
// => html-9473fd.html

// loaderContext.resourcePath = "/app/flash.txt"
loaderUtils.interpolateName(loaderContext, "[hash]", { content: ... });
// => c31e9820c001c9c4a86bce33ce43b679

// loaderContext.resourcePath = "/app/img/image.gif"
loaderUtils.interpolateName(loaderContext, "[emoji]", { content: ... });
// => 👍

// loaderContext.resourcePath = "/app/img/image.gif"
loaderUtils.interpolateName(loaderContext, "[emoji:4]", { content: ... });
// => 🙍🏢📤🐝

// loaderContext.resourcePath = "/app/img/image.png"
loaderUtils.interpolateName(loaderContext, "[sha512:hash:base64:7].[ext]", { content: ... });
// => 2BKDTjl.png
// use sha512 hash instead of md5 and with only 7 chars of base64

// loaderContext.resourcePath = "/app/img/myself.png"
// loaderContext.query.name =
loaderUtils.interpolateName(loaderContext, "picture.png");
// => picture.png

// loaderContext.resourcePath = "/app/dir/file.png"
loaderUtils.interpolateName(loaderContext, "[path][name].[ext]?[hash]", { content: ... });
// => /app/dir/file.png?9473fdd0d880a43c21b7778d34872157

// loaderContext.resourcePath = "/app/js/page-home.js"
loaderUtils.interpolateName(loaderContext, "script-[1].[ext]", { regExp: "page-(.*)\\.js", content: ... });
// => script-home.js

getHashDigest

const digestString = loaderUtils.getHashDigest(buffer, hashType, digestType, maxLength);

• buffer the content that should be hashed
• hashType one of sha1, md5, sha256, sha512 or any other node.js supported hash type
• digestType one of hex, base26, base32, base36, base49, base52, base58, base62, base64
• maxLength the maximum length in chars

License

MIT (http://www.opensource.org/licenses/mit-license.php)

Changelog

1.2.3 (2018-12-27)

Bug Fixes

• interpolateName: don't interpolated hashType without hash or contenthash (#140) (3528fd9)

<a name="1.2.2"></a>