showify

Arguably the most comprehensive cross-platform library for stringifying any JavaScript value into a human-readable format, handling nearly every scenario you might encounter.

Features

Stringify any JavaScript value in a format similar to util.inspect in Node.js. Special cases like Map, Set, array empty slots, wrapper objects for primitives, ArrayBuffer, DataView, WeakMap, WeakSet, and more are handled properly.
Support for maximum recursion depth in nested objects.
Circular references displayed with a reference pointer (e.g., <ref *1> { foo: { bar: [Circular *1] } }) instead of just [Circular].
Auto indentation and line breaking for long outputs, just like console.log/util.inspect in Node.js and Deno.
ANSI color support.
Highly customizable options for indentation, quote style, hidden properties, colors, and more.
Custom serializers for your own types, with a user-friendly API that works on a tree-like structure instead of raw strings, handling maximum recursion depth, indentation, and circular references automatically — without any extra effort.

Quickstart

Simply import the show function and use it to stringify any JavaScript value:

import { show } from "showify";

const value = {
  foo: "bar",
  "Hello\nworld": [-0, 2n, NaN],
  [Symbol("qux")]: { quux: "corge" },
  map: new Map([
    ["foo", "bar"],
    [{ bar: 42 }, "qux"],
  ]),
};
value.circular = value;

console.log(show(value, { indent: 2, trailingComma: "auto", colors: true }));
// <ref *1> {
//   foo: "bar",
//   "Hello\nworld": [-0, 2n, NaN],
//   map: Map(2) { "foo" => "bar", { bar: 42 } => "qux" },
//   circular: [Circular *1],
//   [Symbol(qux)]: { quux: "corge" },
// }

The show() function accepts an optional second argument for options. Some common options are listed here:

depth: Maximum recursion depth, defaults to Infinity.
indent: Number of spaces to indent, defaults to 0. To enable indentation and auto line breaking, set indent to a positive integer, e.g., 2.
breakLength: Maximum line length before breaking, defaults to 80. This option is ignored if indent is 0.
sorted: Whether to sort the keys of objects (including Maps and Sets) in the resulting string, defaults to false.
quoteStyle: Preferred quote style for strings, should be "single", "double", "backtick", or an array of them to try in order, defaults to ["double", "single", "backtick"].
trailComma: Whether to add a trailing comma to the last element of an array or object, should be "none", "always" or "auto" (add trailing comma only when the last item is on a separate line), defaults to "none".
colors: Enable ANSI colors, defaults to false.

showify supports many other options. For a complete list of options, see the available options section below.

Note that showify uses slightly different default options compared to util.inspect in Node.js. If you want to achieve the exactly same default output as util.inspect in Node.js, see the related FAQ.

Installation

To install showify via npm (or any other package manager you prefer):

npm install showify

If you prefer a lightweight version with a smaller bundle size, you may also consider @showify/lite, which removes ANSI color support and several other features that are not needed for most use cases.

Comparison

Here’s a comparison of the features of showify, util.inspect in Node.js, and other libraries:

✅ 1st-class, built-in, and ready to use with no additional configuration or code.
🟡 Supported, but has limitations or is not as complete as showify.
🔶 Supported and documented, but not out-of-the-box or requires extra user-code to implement.
❌ Not officially supported or documented.

Feature	showify	`util.inspect`	pretty-format	object-inspect
Bundle size		❓
Indentation	✅	🟡 uncustomizable	✅	✅
Auto line breaking	✅	✅	❌	❌
Circular references	✅	✅	✅	✅
Circular reference pointers	✅	✅	❌	❌
Maximum recursion depth	✅	✅	✅	✅
ANSI colors	✅	✅	🔶	❌
Sorted keys	✅	✅	✅	❌
Auto select proper quote style for strings	✅	✅	❌	❌
Custom quote style	✅	❌	❌	🟡 [1]
Auto line break for long multiline strings	✅	✅	❌	❌
Trailing comma	✅	❌	🟡 forced	❌
Hidden (non-enumerable) properties	✅ [2]	✅	❌	❌
Proper getters/setters inspection [3]	✅	✅	❌	❌
`Symbol.toStringTag` support	✅	✅	❌	🟡
Custom serializers	✅	✅	✅	✅
Promise inspection	🟡 [4]	✅	❌	❌
WeakMap/WeakSet inspection	🟡 [5]	✅	❌	❌
Proxy inspection	❌	✅	❌	❌

[1]: object-inspect only supports single or double quotes, while showify supports automatic quote style.

[2]: In addition to showHidden: boolean from Node.js, showify also provides an "exclude-meta" option, which behaves like true but removes common meta properties like the length property of arrays from the output.

[3]: Whether to treat getters/setters specially. Getters may throw errors, so it’s important to account for such cases.

[4]: Promise objects are displayed as Promise { <state unknown> }, since it’s impossible to determine the state of a Promise without awaiting it.

[5]: WeakMap and WeakSet objects are displayed as ${className} { <items unknown> } even when showHidden is "always", since it’s impossible to retrieve the values of a WeakMap or WeakSet without knowing the keys.

Aside from the features listed above, showify also supports many more special cases than other libraries, such as wrapper objects for primitives (e.g., new String("foo")), async/generator functions, ES6 classes, and more.

Documentation

Limitations

Module: Module detection may be misled by Symbol.toStringTag. If an object has a non-writable, non-enumerable, and non-configurable Symbol.toStringTag property set to "Module", it will be considered a module.
Promise: Promise objects are displayed as Promise { <state unknown> }, since it’s impossible to determine the state of a Promise without awaiting it.
Proxy: Proxy objects are treated as regular objects, since it’ impossible to identify whether an object is a Proxy in vanilla JavaScript.
WeakMap/WeakSet: WeakMap and WeakSet objects are displayed as ${className} { <items unknown> } even when showHidden is "always", since it’s impossible to retrieve the values of a WeakMap or WeakSet without knowing the keys.

Known differences from `util.inspect` in Node.js

Default options: showify uses slightly different default options compared to util.inspect in Node.js. For example, showify uses infinite depth, prefers double quotes, and does not break lines by default, while util.inspect uses a depth of 2, prefers single quotes, and breaks lines by default. See the related FAQ if you want to achieve the exact same default output as util.inspect in Node.js.
Custom serialization: util.inspect supports custom serialization via the special Symbol(nodejs.util.inspect.custom) property, while showify does not. Instead, showify offers similar functionality with a serializers option in the show function.
Break length: showify tries to break lines exactly at breakLength characters, while util.inspect may break lines at slightly different positions due to a different algorithm for calculating the break position.
Array break: When breaking arrays, showify always places one element per line, while util.inspect might place multiple elements on a single line, e.g., [\n 1, 2, 3, 4, 5,\n 6, 7, 8, 9, 10\n].
Circular reference pointer: showify always displays a reference pointer when the same object is referenced, while util.inspect does not always do so if the same object is referenced multiple times.
Symbol.toStringTag: util.inspect shows [Symbol(Symbol.toStringTag)]: "..." for some objects (like generators and generator functions) when showHidden is true, even if Symbol.toStringTag is not an own property of the object. showify does not show this in such cases.

A complete list of rules for stringifying JavaScript values

Common rules:

Maximum recursion depth: Values are stringified up to the specified depth (defaults to Infinity). When stringifying a value, if the current depth exceeds the specified depth, showify checks if further recursion is necessary. If recursion is not needed, it simply stringifies the value as is. If further recursion is needed, showify displays [${className}] and stops further recursion. A special case occurs when the value’s [[Prototype]] is null, in which case the value is displayed as [Object: null prototype].
Break length: Lines are broken at breakLength (defaults to 80) characters if possible. When stringifying a value, showify first tries the inline format. If the inline format doesn't fit within the breakLength, it switches to the multiline format. This process is repeated recursively for each child value.
Circular references: Objects with circular references are displayed with a reference pointer. For example, <ref *1> { foo: [ [Circular *1] ], bar: <ref *2> { inner: [Circular *2], obj: [Circular *1] } }.
Extra keys: Extra keys of any special object (e.g., wrapper objects for primitives, errors, promises, functions, arrays) are displayed. For example, [Function: foo] { bar: "baz" } or [1, 2, 3, foo: "bar"].
Class name: An object’s ${className} is defined as:
- If the object is a prototype of a class (by checking if value === value.constructor.prototype), Object.getPrototypeOf(value).constructor.name.
- Otherwise, value.constructor.name.
- If the name from the above two steps is empty, "Object".
Symbol.toStringTag: If an object has a Symbol.toStringTag property that is not already shown and is not equal to its ${className}, it is displayed as the following:
- For Dates and RegExps, it is displayed with brackets around after the class name, e.g., Date [MyTag] 2025-02-13T11:42:41.196Z or MyRegExp [MyTag] /(?:)/ { foo: 'bar' }.
- For Errors, it is displayed with brackets around after the error name, e.g., Error [MyTag]: error message\n at ... or [TypeError [MyTag]: error message].
- For ES6 classes, it is displayed with brackets around after the class name, e.g., [class MyClass [MyTag]] or [class MyClass [MyTag] extends MySuperClass].
- For other objects, the tag is displayed with brackets around after its original prefix (if it does not have a prefix, the prefix is defined as ${className}), e.g., Object [MyTag] { foo: "bar" }or[String: "foo"] [MyTag].

Primitive values:

Undefined: undefined is displayed as undefined. Gray if colors is true.
Null: null is displayed as null. Bold if colors is true.
Boolean: true and false are displayed as true and false, respectively. Yellow if colors is true.
Number: Numbers are displayed as their literal values, e.g., 42. Yellow if colors is true.
BigInt: BigInts are displayed as their literal values with n suffix, e.g., 42n. Yellow if colors is true.
String: Strings are displayed as their literal values with quotes based on the quoteStyle option, e.g., "foo". Green if colors is true.
Multi-line strings: Multi-line strings are broken into "${s1}\n" + "${s2}\n" + ... if its indentation plus the length of the string exceeds breakLength. Each part is colorized separately if colors is true, and +s are not colorized.
Long strings: If maxStringLength is set and the string is longer than maxStringLength, it is displayed as "${s}"... ${n} more character${n === 1 ? "" : "s"}, e.g., "foo bar"... 3 more characters. This also applies to multi-line strings, e.g., "foo\n" + "bar"... 1 more character".
Symbol: Symbols are displayed as their literal values, e.g., Symbol(foo). Green if colors is true.
Wrapper objects for primitives: Wrapped primitive values are displayed as [${Type}: ${value}], e.g., [String: "foo"]. Their colors are the same as their primitive values if colors is true. If a wrapped primitive’s ${className} is not equal to its ${Type}, it is displayed as [${Type} (${className}): ${value}], e.g., [String (MyString): "foo"].

Special objects:

Date: Date objects are displayed as date.toISOString(). If ${className} is not "Date", it is displayed as ${className} ${date.toISOString()}.
RegExp: RegExp objects are displayed as re.toString(). If ${className} is not "RegExp", it is displayed as ${className} ${re.toString()}.
Error: Error objects are displayed as error.stack if available and valid, or [${error.stack}] if error.stack is available but invalid, or [${className}: ${error.message}] if message is available, or [${className}] otherwise.
Promise: Promise objects are displayed as Promise { <state unknown> }.
ArrayBuffer: [Uint8Contents] and byteLength are displayed for ArrayBuffer objects. For example, ArrayBuffer { [Uint8Contents]: <2a 00 00 00 00 00 00 00>, byteLength: 8 }.
DataView: byteLength, byteOffset and buffer are displayed for DataView objects. For example, DataView { byteLength: 8, byteOffset: 0, buffer: ArrayBuffer { [Uint8Contents]: <2a 00 00 00 00 00 00 00>, byteLength: 8 } }.

Classes and functions:

ES6 Class: If a callable (i.e., typeof value === "function") object is an ES6 class, it is displayed as [class ${className}] or [class (anonymous)] if the class name is not available, or [class ${className} extends ${superClassName}] if the class has a superclass and the superclass name is available.
Function: Other callables are identified as functions. A function is displayed as [${Type}: ${functionName}] if functionName is available, or [${Type} (anonymous)] otherwise. ${Type} is Function for regular functions, AsyncFunction for async functions, GeneratorFunction for generator functions, and AsyncGeneratorFunction for async generator functions. For functions that ${className} is not equal to its ${Type}, they are displayed with a ${className} suffix, e.g., [Function: anonymous] MyFn for new class MyFn extends Function {}(), or [Function (anonymous)] Object for Function.prototype, or [GeneratorFunction: anonymous] MyGenFn [GeneratorFunction] for new class MyGenFn extends function* () {}.constructor {}().

Collections:

Array: Array objects are displayed as [e1, e2, ...]. Empty slots are displayed as <${n} empty item${n === 1 ? "" : "s"}>. For subclasses of Array or typed arrays, the ${className} is displayed with its size, e.g., MyArray(4) [1, <2 empty items>, 2, { foo: "bar" }] or Uint8Array(3) [1, 2, 3].
Long Arrays: If maxArrayLength is set and the array has more elements than maxArrayLength, it is displayed as [${e1}, ${e2}, ... ${n} more item${n === 1 ? "" : "s"}], e.g., [1, 2, 3, 4, ... 2 more items].
Map: Map objects are displayed as Map(${size}) { key1 => value1, key2 => value2, ... }.
Set: Set objects are displayed as Set(${size}) { value1, value2, ... }.
WeakMap/WeakSet: WeakMap and WeakSet objects are displayed as ${className} { <items unknown> }.

Regular objects:

Module: Module objects are displayed as [Module] if its [[Prototype]] is not null, or [Module: null prototype] otherwise.
Object: Other objects are displayed as { key: value1, "non identifier key": value2, [Symbol(id)]: value3, ... }. ${className} is displayed if it’s not "Object" and no prefix is already defined, e.g., MyClass { key1: value1, key2: value2, ... }. Objects with a null [[Prototype]] are displayed as [Object: null prototype].
Object keys: String keys are displayed as key if they are valid identifiers, or "key" otherwise. Symbol keys are displayed as [Symbol(key)]. If a key is non-enumerable, it is displayed as [key] (NOTE: non-enumerable keys are only displayed if showHidden is not "none").
Getter/Setter: Getters and setters are displayed as [Getter/Setter], [Getter], or [Setter]. For example, { foo: [Getter/Setter], bar: [Getter] }. If getters is not "none", getters are further inspected, e.g., { foo: [Getter] { bar: "baz" } }. Note that getters might throw an error, and in such cases, the error message is displayed if the thrown value is an object with a message property, e.g., { foo: [Getter: <Inspection threw (error message)>] }.

Available Options

Option	Type	Default	Description
`depth`	`number`	`Infinity`	Maximum recursion depth of the object to be inspected, similar to `util.inspect`.
`indent`	`number`	`0`	Number of spaces to indent the output. If `indent` is `0`, the output is not indented.
`breakLength`	`number`	`80`	Maximum line length before breaking. This option is ignored if `indent` is `0`.
`showHidden`	`"none" \| "always" \| "exclude-meta" \| boolean`	`"none"`	Whether to show hidden (non-enumerable) properties, should be `"none"`, `"always"`, or `"exclude-meta"`. If set to `"exclude-meta"`, it behaves like `"always"` but removes common meta properties like the `length` property of arrays. For compatibility with Node.js’s `util.inspect`, `true` is also accepted as `"always"`, and `false` is also accepted as `"none"`, but it is recommended to use the string values for clarity.
`getters`	`"none" \| "get" \| "set" \| "all"`	`"none"`	Whether to inspect getters. If set to `"get"`, only getters without a corresponding setter are inspected. If set to `"set"`, only getters with a corresponding setter are inspected. If set to `"all"`, all getters are inspected.
`sorted`	`boolean`	`false`	Whether to sort the keys of objects (including `Map`s and `Set`s) in the resulting string.
`omittedKeys`	`Set<string \| symbol>`	`new Set()`	A set of keys to omit from the output. Note that this option is not recursive and only omits the top-level keys.
`quoteStyle`	`"single" \| "double" \| "backtick"`	`["double", "single", "backtick"]`	Preferred quote style for strings. If an array is provided, showify tries each quote style in order.
`quoteKeys`	`"auto" \| "always"`	`"auto"`	Whether to quote object keys. If set to `"auto"`, object keys are quoted only when necessary.
`numericSeparator`	`"none" \| string`	`"none"`	The thousands separator for numbers (including BigInts), e.g., `","` or `"_"`. If set to `"none"`, no separator is used.
`trailingComma`	`"none" \| "always" \| "auto"`	`"none"`	Whether to add a trailing comma to the last element of an array or object. If set to `"auto"`, a trailing comma is added only when the last item is on a separate line.
`arrayBracketSpacing`	`boolean`	`false`	Whether to add spaces inside the brackets of arrays.
`objectCurlySpacing`	`boolean`	`true`	Whether to add spaces inside the curly braces of objects.
`referencePointer`	`boolean`	`true`	Whether to display circular references with a reference pointer. If set to `false`, circular references are displayed as `[Circular]`.
`maxArrayLength`	`number`	`Infinity`	Maximum number of array elements to display. If an array has more elements than `maxArrayLength`, it is displayed as `[${e1}, ${e2}, ... ${n} more item${n === 1 ? "" : "s"}]`.
`maxStringLength`	`number`	`Infinity`	Maximum length of a string to display. If a string is longer than `maxStringLength`, it is displayed as `"${s}"... ${n} more character${n === 1 ? "" : "s}"`.
`colors`	`boolean`	`false`	Whether to enable ANSI colors.
`styles`	`Object`	`{}`	Custom styles for different types of values. Valid keys are `string`, `symbol`, `number`, `bigint`, `boolean`, `null`, `undefined`, `date`, `regexp`, and `special`. Valid colors are `bold`, `dim`, `reset`, `black`, `blue`, `cyan`, `gray`, `green`, `magenta`, `red`, `white`, and `yellow`.
`serializers`	`Serializer[]`	`[]`	Custom serializers for your own types. See the custom serializers section for more details.

Custom Serializers

showify supports custom serializers through the serializers option in the show() function. A serializer must implement the following interface:

interface Serializer {
  if: (value: object, options: SerializerOptions) => boolean;
  then: (
    value: object,
    options: SerializerOptions,
    expand: (value: unknown, options?: Partial<SerializerOptions>) => Node,
  ) => Node;
}

You may wonder what Node is. A Node is a tree-like structure that represents the value to be stringified:

type Node =
  | { type: "circular"; ref: object }
  | { type: "text"; value: string; ref?: object }
  | { type: "variant"; inline: Node; wrap: Node; ref?: object }
  | { type: "sequence"; values: Node[]; ref?: object }
  | { type: "between"; values: Node[]; open?: Node; close?: Node; ref?: object };

Internally, showify interacts with Nodes rather than directly with strings:

╔═══════╗   ╔═════════════════════╗   ╔════════════════════════╗
║ value ║ → ║ Build tree of nodes ║ → ║ Render nodes to string ║
╚═══════╝   ╚═════════════════════╝   ╚════════════════════════╝

serializers come into play during the second step, where the tree of nodes is built. The if function of a serializer is called with the value to be serialized. If it returns true, the then function is invoked to build the tree of nodes. The expand function is a helper that recursively serializes child values.

You can use the exported serializer helper function and the Node object to create your own serializers:

import { Node as SerializerNode, serializer, show } from "showify";

const { circular, text, variant, sequence, pair, between } = SerializerNode;

class Point {
  constructor(x, y) {
    this.x = x;
    this.y = y;
  }
}

const pointSer = serializer({
  if: (value) => value instanceof Point,
  then: (value) => text(`Point(${value.x}, ${value.y})`),
});

class Box {
  constructor(data) {
    this.data = data;
  }
}

const boxSer = serializer({
  if: (value) => value instanceof Box,
  then: (value, options, expand) =>
    sequence([text("Box("), expand(value.data), text(")")]),
});

const obj = {
  point: new Point(1, 2),
  box1: new Box({ foo: "bar" }),
  box2: new Box(new Point(3, 4)),
};

console.log(show(obj, { indent: 2, serializers: [pointSer, boxSer] }));
// {
//   point: Point(1, 2),
//   box1: Box({ foo: "bar" }),
//   box2: Box(Point(3, 4))
// }

This example shows how to use custom serializers to display Point and Box objects in a custom format. The pointSer serializer formats Point objects as Point(x, y), and the boxSer serializer formats Box objects as Box(...).

We’ve seen how to use text and sequence to build simple nodes. There are other nodes we haven’t used yet. Here’s a brief introduction to all these nodes:

circular: Represents a circular reference. Normally, you don’t need to create this node manually in a serializer. All nodes returned by a custom serializer automatically get a ref property, which helps detect circular references.
text: Represents a text node. The value property holds the text to be displayed.
variant: Represents a “variant” node. Some objects format differently depending on whether they are inline or multiline. For example, when trailingComma is "auto", a trailing comma is added only when the last item is on a separate line. This can be represented with variant(inlineNode, multilineNode).
sequence: Represents a sequence of nodes. The values property is an array of nodes to be displayed in order. Unlike between, sequence does not break lines between nodes. pair(left, right) is an alias of sequence([left, right]).
between: Represents a sequence of nodes. The values property is an array of nodes to be displayed in order. The open and close properties are optional nodes to be displayed before and after the values nodes. Don’t be confused by the name, between can be used like sequence without open and close. The key difference is that between breaks lines between nodes according to the indent and breakLength options.

So far, you might wonder how sequence and between differ. Let’s use a Pair class as an example to demonstrate the difference:

class Pair {
  constructor(left, right) {
    this.left = left;
    this.right = right;
  }
}

const badPairSer = serializer({
  if: (value) => value instanceof Pair,
  then: (value, options, expand) =>
    sequence([
      text("("),
      expand(value.left),
      text(", "),
      expand(value.right),
      text(")"),
    ]),
});

const goodPairSer = serializer({
  if: (value) => value instanceof Pair,
  then: (value, options, expand) =>
    between([
      pair(expand(value.left), text(", ")),
      expand(value.right)
    ], text("("), text(")")),
});

At first glance, badPairSer and goodPairSer seem identical, and they produce the same output for short values:

const p = new Pair("left", "right");

console.log(show(p, { indent: 2, serializers: [badPairSer] }));
// ("left", "right")
console.log(show(p, { indent: 2, serializers: [goodPairSer] }));
// ("left", "right")

However, they behave differently when the Pair object is displayed in multiline format:

const s1 = "This is a very long string that will break the line";
const s2 = "This is another very long string that will break the line";

const p = new Pair(s1, s2);

console.log(show(p, { indent: 2, serializers: [badPairSer] }));
// ("This is a very long string that will break the line", "This is another very long string that will break the line")
console.log(show(p, { indent: 2, serializers: [goodPairSer] }));
// (
//   "This is a very long string that will break the line",
//   "This is another very long string that will break the line"
// )

Although the inline format exceeds the breakLength limit, badPairSer does not break lines between its children, so the output stays on a single line. On the other hand, goodPairSer uses between, which breaks lines between its children, resulting in a multiline output.

Here’s a more interesting example where child nodes themselves can be multiline:

const obj1 = { foo: "bar", bar: "baz", baz: "quxx" };
const obj2 = { qux: "baz", baz: "bar", bar: "foo" };

const p = new Pair(obj1, obj2);

console.log(show(p, { indent: 2, serializers: [badPairSer] }));
// ({
//   foo: "bar",
//   bar: "baz",
//   baz: "quux"
// }, {
//   qux: "baz",
//   baz: "bar",
//   bar: "foo"
// })
console.log(show(p, { indent: 2, serializers: [goodPairSer] }));
// (
//   { foo: "bar", bar: "baz", baz: "quux" },
//   { qux: "baz", baz: "bar", bar: "foo" }
// )

While the outputs might be different from what you expect, the underlying principle remains the same. The badPairSer serializer uses sequence, which does not break lines between direct children. Therefore, it tries to break lines within the children themselves when possible. In contrast, goodPairSer uses between, which handles line breaks between children, so it doesn’t require further instruction for child elements to break lines.

The goodPairSer is still not “good” enough — the separator is still ", " (with a space after) when multiline, which is not what we want. We can use variant to solve this problem:

const goodPairSer = serializer({
  if: (value) => value instanceof Pair,
  then: (value, options, expand) =>
    variant(
      // The inline format
      sequence([
        text("("),
        expand(value.left),
        text(", "), // A comma with a space after
        expand(value.right),
        text(")"),
      ]),
      // The multiline format
      between([
        pair(expand(value.left), text(",")), // No space after the comma
        expand(value.right)
      ], text("("), text(")")),
    ),
});

The variant node enforces inline formatting for the first argument and multiline formatting for the second argument. Using sequence for the inline format is just a personal preference; you can use between instead, as they behave similarly in inline format.

The options argument in then function and the second optional argument in expand are almost identical to the options passed to show, which we'll refer to as SerializerOptions below. The SerializerOptions is a subset of the options passed to show, with the following differences:

indent, breakLength, and referencePointer are not available in SerializerOptions, as they only affect rendering the nodes to strings and do not interact with the process of building the node tree.
level is available in SerializerOptions, indicating the current depth of the object being inspected.
ancestors is available in SerializerOptions, which is an array of the ancestors of the current value. This is useful for detecting circular references.
An object named c is available in SerializerOptions, containing functions to colorize strings using ANSI colors. These functions are similar to chalk, such as c.bold, c.cyan, and so on, as well as c.number, c.special, etc., which are aliases for corresponding styles in the styles option. Note that when colors is false, all color functions are no-op functions that return the input string unchanged.

showify automatically handles level increment in expand, and internally triggers a signal to stop recursion when level exceeds depth, so you don't need to worry about it. Circular references are also automatically detected — the current value is added to the ancestors array, and the ref property of each node returned is set to its value to detect circular references. As shown earlier, since the tree of nodes is built in two stages, indentation and line breaking are handled by between, so you don't need to worry about those either.

Here’s an example showing how to use the options argument in then and expand:

const ser = serializer({
  if: (value, { omittedKeys }) =>
    "_tag" in value &&
    typeof value._tag === "string" &&
    // Detect if `_tag` is already omitted to avoid infinite recursion
    !omittedKeys.has("_tag"),
  then: (val, { ancestors, c, level }, expand) => {
    const tag = val._tag;
    return sequence([
      text(c.blue(tag)),
      text("("),
      expand(val, {
        // Reset `level` and force expansion the object with a depth of 1
        level: 0,
        depth: 1,
        // Omit the `_tag` key when expanding the object
        omittedKeys: new Set(["_tag"]),
        // Avoid auto-adding current value to `ancestors` to avoid circulars
        ancestors: [...ancestors],
      }),
      text(")"),
    ]);
  },
});

const obj = { _tag: "Some", value: 42, nested: { foo: { bar: "baz" } } };

console.log(show(obj, { indent: 2, serializers: [ser], colors: true }));
// Some({ value: 42, nested: { foo: [Object] } })

FAQ

Why not `util.inspect`/pretty-format/object-inspect/...?

Among the libraries that stringify JavaScript values, showify is the most complete and universal, working in any JavaScript runtime environment and producing output similar to util.inspect in Node.js.

object-inspect and pretty-format are popular libraries for stringifying JavaScript values, especially the latter, which is used by Jest’s .matchSnapshot() method. However, try printing several random objects with showify (e.g., console.log(show(await import("node:util"), { indent: 2, colors: true }))), and you’ll quickly see that showify produces much cleaner and more readable output compared to these libraries.

As of now, showify is the only library that handles auto line breaking, circular references with reference pointers, hidden properties, proper getters/setters display, and much more. showify also offers a lightweight implementation under 3kB minified+gzipped, perfect for embedding in other libraries or frameworks.

Compared to util.inspect in Node.js, showify supports most of the features util.inspect offers, along with additional options not available in util.inspect. However, there are some limitations. Since showify is designed to work universally across browsers and other JavaScript runtimes, it cannot support some Node.js-specific features like inspecting a Proxy or synchronously inspecting a Promise.

For a complete list of features, see the comparison table.

How can I achieve the exactly default output as `util.inspect` in Node.js?

While showify behaves slightly differently from util.inspect in rare cases, you can achieve similar default behavior by setting the following options:

show(value, {
  depth: 2,
  indent: 2,
  quoteStyle: ["single", "double", "backtick"],
  arrayBracketSpacing: true,
  maxArrayLength: 100,
  maxStringLength: 10000,
});

License

This project is licensed under the Mozilla Public License Version 2.0 (MPL 2.0). For details, please refer to the LICENSE file.

In addition to the open-source license, a commercial license is available for proprietary use. If you modify this library and do not wish to open-source your modifications, or if you wish to use the modified library as part of a closed-source or proprietary project, you must obtain a commercial license.

For details, see COMMERCIAL_LICENSE.md.

Name		Name	Last commit message	Last commit date
Latest commit History 10 Commits
.github/workflows		.github/workflows
packages/lite		packages/lite
src		src
test		test
.editorconfig		.editorconfig
.gitignore		.gitignore
COMMERCIAL_LICENSE.md		COMMERCIAL_LICENSE.md
LICENSE		LICENSE
README.md		README.md
commitlint.config.js		commitlint.config.js
eslint.config.js		eslint.config.js
package-lock.json		package-lock.json
package.json		package.json
prettier.config.cjs		prettier.config.cjs
screenshot.png		screenshot.png
tsconfig.build.json		tsconfig.build.json
tsconfig.eslint.json		tsconfig.eslint.json
tsconfig.json		tsconfig.json
tsconfig.test.json		tsconfig.test.json

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Repository files navigation

showify

Features

Quickstart

Installation

Comparison

Documentation

Limitations

Known differences from `util.inspect` in Node.js

A complete list of rules for stringifying JavaScript values

Available Options

Custom Serializers

FAQ

Why not `util.inspect`/pretty-format/object-inspect/...?

How can I achieve the exactly default output as `util.inspect` in Node.js?

License

About

Releases 1

Packages

Languages

License

Snowflyt/showify

Folders and files

Latest commit

History

Repository files navigation

showify

Features

Quickstart

Installation

Comparison

Documentation

Limitations

Known differences from util.inspect in Node.js

A complete list of rules for stringifying JavaScript values

Available Options

Custom Serializers

FAQ

Why not util.inspect/pretty-format/object-inspect/...?

How can I achieve the exactly default output as util.inspect in Node.js?

License

About

Topics

Resources

License

Stars

Watchers

Forks

Releases 1

Packages 0

Languages

Known differences from `util.inspect` in Node.js

Why not `util.inspect`/pretty-format/object-inspect/...?

How can I achieve the exactly default output as `util.inspect` in Node.js?

Packages