Web & HTML

ワードジョイナー

U+2060。改行を防ぐゼロ幅文字。ゼロ幅ノーブレークスペースとしてのU+FEFF(BOM)の現代的な代替です。

· Updated

What Is the Word Joiner?

The Word Joiner (WJ) is Unicode character U+2060. It is an invisible, zero-width character whose sole purpose is to prevent line breaks at the position where it is inserted. Unlike a non-breaking space (U+00A0), the word joiner has no visible width — it takes up no space in the rendered text.

The word joiner was introduced in Unicode 3.2 as a replacement for the deprecated Zero Width No-Break Space (ZWNBSP, U+FEFF) when used in a non-BOM context.

How It Works

Line-breaking algorithms in text renderers follow Unicode line-break properties. Normally, line breaks are permitted between most characters. Inserting U+2060 between two characters signals to the renderer: "do not break here." The algorithm treats the adjacent characters as though they are joined by a non-breakable bond.

"some text" + WJ + "more text"
→ "some text⁠more text" (no line break permitted between the two strings)

The WJ is analogous to CSS white-space: nowrap applied at the character level, but it operates in plain text contexts where CSS is unavailable.

When to Use the Word Joiner

Inline code in flowing text: Prevent a hyphenated or compound identifier from breaking across a line.

Abbreviations and numbers: Keep 3⁠km, fig.⁠1, or no.⁠42 together.

Non-breaking compound words: Technical terms, product names, or phrases that lose meaning if split.

After soft hyphens: Occasionally used to counteract an otherwise-permitted break point.

WJ vs. Non-Breaking Space vs. ZWNBSP

Character Code Point Width Prevents Break Common Use
Non-Breaking Space U+00A0 One space width Yes Keep words together with a visible space
Word Joiner U+2060 Zero Yes Prevent break without adding space
ZWNBSP (deprecated WJ use) U+FEFF Zero Yes (legacy) Replaced by WJ; now used only as BOM
Zero-Width Space U+200B Zero No (break opportunity) Allow breaks inside long words

Code Examples

# Inserting WJ in Python
WJ = "\u2060"

product_name = f"Django{WJ}REST{WJ}Framework"
# Prevents line breaks between the components

# Removing WJ (e.g., for search normalization)
text = "some\u2060text"
cleaned = text.replace("\u2060", "")  # "sometext"

# Or with unicodedata
import unicodedata
unicodedata.category("\u2060")  # "Cf" (Format character)
const WJ = "\u2060";
const label = `3${WJ}km`;  // "3⁠km" — no break between 3 and km

// Check for WJ
label.includes("\u2060");  // true

// Strip format characters (including WJ)
label.replace(/\p{Cf}/gu, "");  // "3km"
<!-- In HTML -->
<p>See figure&#x2060;3 for details.</p>
<!-- Or use the named entity (if defined): &#8288; -->
<p>3&#8288;km from downtown.</p>

Browser and Rendering Support

The word joiner is well-supported in modern browsers and text rendering engines. In HTML, it can also be written as &#x2060; or &#8288;. Some content management systems and WYSIWYG editors strip invisible characters — in those cases, use CSS white-space: nowrap on a <span> instead.

  • U+200D Zero Width Joiner (ZWJ): Joins emoji into sequences (e.g., family emoji); also prevents breaks.
  • U+2011 Non-Breaking Hyphen: A hyphen that prevents line breaks.
  • U+00AD Soft Hyphen: The opposite — a suggested break point, displayed as a hyphen only if the line actually breaks there.

Quick Facts

Property Value
Code point U+2060
Name WORD JOINER
Category Cf (Format character)
Width Zero (invisible, no advance width)
Function Prevents line break at insertion point
HTML entity &#x2060; or &#8288;
Unicode version introduced 3.2
Replaces U+FEFF used as ZWNBSP (now deprecated for this purpose)

関連用語

Web & HTML のその他の用語

Content-Type 文字セット

レスポンスの文字エンコーディングを宣言するHTTPヘッダーパラメータ(Content-Type: text/html; charset=utf-8)。ドキュメント内のエンコーディング宣言より優先されます。

CSS content プロパティ

::beforeおよび::after疑似要素でUnicodeエスケープを使って生成コンテンツを挿入するCSSプロパティ:content: '\2713'は✓を挿入します。

CSS Text Direction

CSS properties (direction, writing-mode, unicode-bidi) controlling text layout direction. Works with Unicode …

HTML エンティティ

HTMLで文字をテキスト表現する方式。3つの形式:名前(&amp;)・十進数(&#38;)・16進数(&#x26;)。HTMLの構文と衝突する文字に必須です。

JavaScript Intl API

ECMAScript Internationalization API providing locale-aware string comparison (Collator), number formatting (NumberFormat), date …

Punycode

Unicode ドメイン名をxn--プレフィックス付きのASCII文字列に変換するASCII互換エンコーディング。münchen.de → xn--mnchen-3ya.de。

Unicode in CSS

CSS supports Unicode via escape sequences (\2713 for ✓), the content property …

XML 文字参照

XMLバージョンの数値文字参照:&#x2713;または&#10003;。XMLには名前付きエンティティが5個(&amp; &lt; &gt; &quot; &apos;)しかありませんが、HTML5は2,231個あります。

テキスト表示

デフォルトの絵文字表示の代わりに、通常は異体字セレクター15(U+FE0E)を使って文字をモノクロのテキストグリフでレンダリングすること。

パーセントエンコーディング (URL エンコーディング)

URLの非ASCII文字と予約文字を各バイトを%XXで置き換えてエンコードします。まずUTF-8に変換し、各バイトをパーセントエンコードします:é → %C3%A9。