Kunten / kaeriten (訓点・返り点)

Kunten are the marginal annotations Japanese readers add to classical Chinese (漢文) source so that it can be read in Japanese word order. aozora recognises kaeriten (返り点) — the reading-order return marks — in their bracketed form. The recognised marks are:

• single: レ, 一, 二, 三, 四, 上, 中, 下, 甲, 乙, 丙, 丁
• Xレ compounds: 一レ, 二レ, 三レ, 上レ, 中レ, 下レ
• 送り仮名: the parenthesised （…） form

(Re-reading marks — 再読文字 like 未 / 将 / 当 — and any other kunten that do not match the above are carried as generic ［＃…］ annotations.)

A handful of late-Edo / Meiji Aozora Bunko works carry these. In real source the marks sit between characters as ［＃…］ annotations:

有［＃二］朋自遠方来［＃一］

Notation forms

Bracketed (the recognised form)

aozora recognises the bracketed form only — the mark in a ［＃…］ annotation:

有［＃二］朋自遠方来［＃一］

Renders as:

有<sup class="aozora-kaeriten">二</sup>朋自遠方来<sup class="aozora-kaeriten">一</sup>

Inline (not recognised)

A bare reading-mark glyph written directly between characters (有レ朋自遠方来) is left as plain text — the parser cannot tell a genuine 返り点 from an ordinary 一 / 上 / レ in running prose, which is exactly why the bracketed form exists. Use ［＃…］ for any mark you want recognised.

Okurigana

Kunten 送り仮名 (reading-aid kana) use the parenthesised form, also inside a ［＃…］ annotation:

有［＃（リ）］

These are classified as kaeriten nodes but are not ladder marks (they take no part in the pairing check).

AST shape

The recognised marks (single 一 二 三 四 上 中 下 甲 乙 丙 丁 レ, the Xレ compounds, and （…） okurigana) all produce one node that stores the raw mark text:

pub struct Kaeriten<'src> {
    pub mark: NonEmptyStr<'src>,   // the raw mark, e.g. "二" / "一レ" / "（リ）"
}

The renderer wraps it in <sup class="aozora-kaeriten">…</sup>. The bracketed_kaeriten_no_pair / kaeriten_outside_kanbun checks classify the mark’s family and rank from this string at diagnostic time rather than storing a typed enum.

Diagnostics

See also

• Notation overview — the orientation map for all the inline annotations.