hanlint 0.0.7 → 0.0.9

package/data/translationese.json

@@ -60,6 +60,16 @@
       "why": "번역투다. 목적어나 `~은` 으로 되돌린다",
       "source": "urimal.org 김민지 2020"
     },
+    {
+      "pattern": "에 대한",
+      "why": "번역투다. `문제점에 대한 이해` 는 `문제점을 아는 것` 이다. 목적어로 되돌린다",
+      "source": "urimal.org 김민지 2020"
+    },
+    {
+      "pattern": "에 관한",
+      "why": "번역투다. `설정에 관한 설명` 은 `설정 설명` 이다. 목적어나 `~은` 으로 되돌린다",
+      "source": "urimal.org 김민지 2020"
+    },
     {
       "pattern": "(?:을|를) 위(?:해|하여|해서)",
       "why": "번역투다. `여행을 가기 위해` 는 `여행 가려고` 다",

package/data/verbTails.txt

@@ -54,3 +54,20 @@
 된
 될
 돼
+# 실측 2026-08-29: 말뭉치에서 `있던`, `힘껏`, `유난히`, `되었는지`, `올바른` 이 명사로 세어졌다. 과거 관형형, 부사 파생,
+# 의문형 연결, 관형형 꼬리다. `교수이자` 의 이자와 `알고자` 의 고자와 `은지` 는 대출이자, 신고자, 김은지 같은 명사와
+# 표층으로 안 갈려 넣지 않는다 (검증에서 걸렸다).
+던
+껏
+히
+는지
+른
+# 실측 2026-08-29: `새로운 CPU 매니저 스태틱 정책 옵션` 의 새로운 이 명사로 세어졌다. ㅂ 불규칙 형용사의 관형형 꼬리다.
+# 다운 은 카운트다운 과 갈리지 않아 넣지 않는다.
+로운
+라운
+려운
+거운
+벼운
+러운
+쉬운

package/data/version.json

@@ -1 +1 @@
-{"version": "0.0.7"}
+{"version": "0.0.9"}

package/data/writingBrief.schema.json

@@ -0,0 +1,64 @@
+{
+  "$schema": "https://json-schema.org/draft/2020-12/schema",
+  "$id": "https://hanlint.dev/schemas/writingBrief.schema.json",
+  "title": "hanlint writing brief",
+  "type": "object",
+  "additionalProperties": false,
+  "required": [
+    "version",
+    "preset",
+    "reader",
+    "task",
+    "facts",
+    "mustInclude",
+    "allowedNumbers",
+    "forbidden",
+    "length"
+  ],
+  "properties": {
+    "version": { "const": 1 },
+    "preset": {
+      "enum": ["blog", "report", "docs", "guide", "essay", "fiction", "encyclopedia"]
+    },
+    "reader": { "type": "string", "minLength": 1 },
+    "task": { "type": "string", "minLength": 1 },
+    "facts": {
+      "type": "array",
+      "minItems": 1,
+      "items": {
+        "type": "object",
+        "additionalProperties": false,
+        "required": ["id", "statement"],
+        "properties": {
+          "id": { "type": "string", "pattern": "^F[1-9][0-9]*$" },
+          "statement": { "type": "string", "minLength": 1 }
+        }
+      }
+    },
+    "mustInclude": {
+      "type": "array",
+      "minItems": 1,
+      "uniqueItems": true,
+      "items": { "type": "string", "minLength": 1 }
+    },
+    "allowedNumbers": {
+      "type": "array",
+      "uniqueItems": true,
+      "items": { "type": "string", "pattern": "^[0-9]+(?:[.][0-9]+)*$" }
+    },
+    "forbidden": {
+      "type": "array",
+      "uniqueItems": true,
+      "items": { "type": "string", "minLength": 1 }
+    },
+    "length": {
+      "type": "object",
+      "additionalProperties": false,
+      "required": ["min", "max"],
+      "properties": {
+        "min": { "type": "integer", "minimum": 1 },
+        "max": { "type": "integer", "minimum": 1 }
+      }
+    }
+  }
+}

package/data/writingBriefV2.schema.json

@@ -0,0 +1,116 @@
+{
+  "$schema": "https://json-schema.org/draft/2020-12/schema",
+  "$id": "https://hanlint.dev/schemas/writingBriefV2.schema.json",
+  "title": "hanlint evidence writing brief",
+  "type": "object",
+  "additionalProperties": false,
+  "required": [
+    "version",
+    "preset",
+    "reader",
+    "task",
+    "facts",
+    "mustInclude",
+    "allowedNumbers",
+    "forbidden",
+    "length",
+    "evidence"
+  ],
+  "properties": {
+    "version": { "const": 2 },
+    "preset": {
+      "enum": ["blog", "report", "docs", "guide", "essay", "fiction", "encyclopedia"]
+    },
+    "reader": { "type": "string", "minLength": 1 },
+    "task": { "type": "string", "minLength": 1 },
+    "facts": {
+      "type": "array",
+      "minItems": 1,
+      "items": {
+        "type": "object",
+        "additionalProperties": false,
+        "required": ["id", "statement"],
+        "properties": {
+          "id": { "type": "string", "pattern": "^F[1-9][0-9]*$" },
+          "statement": { "type": "string", "minLength": 1 }
+        }
+      }
+    },
+    "mustInclude": {
+      "type": "array",
+      "minItems": 1,
+      "uniqueItems": true,
+      "items": { "type": "string", "minLength": 1 }
+    },
+    "allowedNumbers": {
+      "type": "array",
+      "uniqueItems": true,
+      "items": { "type": "string", "pattern": "^[0-9]+(?:[.][0-9]+)*$" }
+    },
+    "forbidden": {
+      "type": "array",
+      "uniqueItems": true,
+      "items": { "type": "string", "minLength": 1 }
+    },
+    "length": {
+      "type": "object",
+      "additionalProperties": false,
+      "required": ["min", "max"],
+      "properties": {
+        "min": { "type": "integer", "minimum": 1 },
+        "max": { "type": "integer", "minimum": 1 }
+      }
+    },
+    "evidence": {
+      "type": "array",
+      "minItems": 1,
+      "items": {
+        "type": "object",
+        "additionalProperties": false,
+        "required": [
+          "id",
+          "factIds",
+          "sourceUrl",
+          "revision",
+          "checkedAt",
+          "locator",
+          "excerpt",
+          "excerptSha256",
+          "license",
+          "reviewStatus"
+        ],
+        "properties": {
+          "id": { "type": "string", "pattern": "^E[1-9][0-9]*$" },
+          "factIds": {
+            "type": "array",
+            "minItems": 1,
+            "uniqueItems": true,
+            "items": { "type": "string", "pattern": "^F[1-9][0-9]*$" }
+          },
+          "sourceUrl": { "type": "string", "format": "uri", "pattern": "^https?://" },
+          "revision": {
+            "type": ["string", "null"],
+            "minLength": 1,
+            "not": {
+              "pattern": "^(?:[Hh][Ee][Aa][Dd]|[Ll][Aa][Tt][Ee][Ss][Tt]|[Mm][Aa][Ii][Nn]|[Mm][Aa][Ss][Tt][Ee][Rr])$"
+            }
+          },
+          "checkedAt": {
+            "type": ["string", "null"],
+            "format": "date-time",
+            "pattern": "^[0-9]{4}-[0-9]{2}-[0-9]{2}T[0-9]{2}:[0-9]{2}:[0-9]{2}Z$"
+          },
+          "locator": { "type": "string", "minLength": 1 },
+          "excerpt": { "type": "string", "minLength": 1, "maxLength": 1000 },
+          "excerptSha256": { "type": "string", "pattern": "^[0-9a-f]{64}$" },
+          "license": { "type": "string", "minLength": 1 },
+          "reviewStatus": { "enum": ["unreviewed", "humanVerified"] }
+        },
+        "anyOf": [
+          { "properties": { "revision": { "type": "string", "minLength": 1 } } },
+          { "properties": { "checkedAt": { "type": "string" } } }
+        ]
+      }
+    }
+  }
+}

package/data/writingTrial.schema.json

@@ -0,0 +1,36 @@
+{
+  "$schema": "https://json-schema.org/draft/2020-12/schema",
+  "$id": "https://hanlint.dev/schemas/writingTrial.schema.json",
+  "title": "hanlint writing trial",
+  "type": "object",
+  "additionalProperties": false,
+  "required": ["version", "id", "brief", "baseline", "candidate"],
+  "properties": {
+    "version": { "const": 1 },
+    "id": { "type": "string", "minLength": 1 },
+    "brief": {
+      "oneOf": [
+        { "$ref": "writingBrief.schema.json" },
+        { "$ref": "writingBriefV2.schema.json" }
+      ]
+    },
+    "baseline": { "$ref": "#/$defs/generation" },
+    "candidate": { "$ref": "#/$defs/generation" }
+  },
+  "$defs": {
+    "sha256": { "type": "string", "pattern": "^[0-9a-f]{64}$" },
+    "generation": {
+      "type": "object",
+      "additionalProperties": false,
+      "required": ["strategyId", "modelId", "modelSha256", "promptSha256", "outputSha256", "text"],
+      "properties": {
+        "strategyId": { "type": "string", "minLength": 1 },
+        "modelId": { "type": "string", "minLength": 1 },
+        "modelSha256": { "$ref": "#/$defs/sha256" },
+        "promptSha256": { "$ref": "#/$defs/sha256" },
+        "outputSha256": { "$ref": "#/$defs/sha256" },
+        "text": { "type": "string", "minLength": 1 }
+      }
+    }
+  }
+}

package/koglType1.LICENSE.md

@@ -0,0 +1,25 @@
+# 국립국어원 파생 데이터 라이선스 (공공누리 제1유형)
+
+아래 두 파일은 국립국어원이 공공누리 제1유형 (출처 표시) 으로 공개한 자료에서 왔다. 나머지 hanlint
+코드와 데이터의 MIT 와는 별개다.
+
+## `learningVocabulary.tsv`
+
+- 원저작자: 국립국어원
+- 자료: 한국어 학습용 어휘 목록 (5,965개)
+- 원본: <https://www.korean.go.kr/front/etcData/etcDataView.do?etc_seq=70&mn_id=46&pageIndex=9>
+- 받은 날: 2026-08-30
+- 라이선스: [공공누리 제1유형 출처 표시](https://www.kogl.or.kr/info/licenseType1.do)
+- 변경: cp949 를 UTF-8 로 옮기고 열 이름을 바꿨다. 낱말, 품사, 풀이, 등급의 내용은 바꾸지 않았다.
+  만드는 절차와 원본 필드는 `learningVocabularySource.toml` 이 소유한다.
+
+## `easyWords.toml`
+
+- 원저작자: 국립국어원
+- 자료: 다듬은 말 (<https://www.korean.go.kr/front/imprv/refineList.do>) 과 표준국어대사전
+  (<https://stdict.korean.go.kr>) 의 동의어 정보
+- 라이선스: [공공누리 제1유형 출처 표시](https://www.kogl.or.kr/info/licenseType1.do)
+- 변경: 어려운 낱말과 쉬운 말의 짝만 뽑아 TOML 로 옮겼다.
+
+공공누리 제1유형은 출처를 밝히면 상업적 이용과 변형을 포함해 자유롭게 쓸 수 있다. 이 파일이 그
+출처 표시다.

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "hanlint",
-  "version": "0.0.7",
-  "description": "한국어 글에서 AI 와 사람이 반복해서 어기는 결함을 결정적으로 잡는 린터. 의존성 0",
+  "version": "0.0.9",
+  "description": "한국어 글 검사기. Korean prose linter for Markdown: translationese, noun piles, double passives, weak structure. Zero dependencies",
   "type": "module",
   "bin": {
     "hanlint": "bin/hanlint.js"
@@ -14,22 +14,34 @@
     "bin",
     "src",
     "data",
+    "LICENSE",
+    "koglType1.LICENSE.md",
     "README.md"
   ],
   "engines": {
     "node": ">=18"
   },
   "scripts": {
-    "test": "node --test test/"
+    "test": "node --test \"test/*.test.js\""
+  },
+  "devDependencies": {
+    "pyproc": "0.0.23"
   },
   "keywords": [
     "korean",
-    "lint",
+    "korean-language",
+    "hangul",
+    "linter",
+    "prose",
+    "prose-linter",
     "writing",
     "readability",
-    "markdown"
+    "markdown",
+    "style-checker",
+    "proofreading",
+    "text-linter"
   ],
-  "license": "MIT",
+  "license": "SEE LICENSE IN koglType1.LICENSE.md",
   "repository": {
     "type": "git",
     "url": "git+https://github.com/eddmpython/hanlint.git",

package/src/analysis/grammar/ending.js

@@ -0,0 +1,336 @@
+// @ts-check
+/** 파이썬 `analysis/grammar/ending.py`와 `predicate.py`의 투영. 종결 서술어를 풀고 세 문체로 짠다. */
+import { loadLines } from "../../data/load.js";
+import * as hangul from "./hangul.js";
+
+export const VERB = "동사";
+export const ADJECTIVE = "형용사";
+export const COPULA = "계사";
+export const PRESENT = "현재";
+export const PAST = "과거";
+export const FUTURE = "미래";
+export const DECLARATIVE = "평서";
+export const QUESTION = "의문";
+export const KKA_QUESTION = "까의문";
+export const IMPERATIVE = "명령";
+export const PROPOSITIVE = "청유";
+export const HAPNIDA = "합니다";
+export const HANDA = "한다";
+export const HAEYO = "해요";
+export const REGISTERS = [HAPNIDA, HANDA, HAEYO];
+
+const HAE_IRREGULAR = new Map([
+  ["그렇", "그래"],
+  ["이렇", "이래"],
+  ["저렇", "저래"],
+  ["어떻", "어때"],
+]);
+const AUX_LINKS = ["지는", "지도", "지만", "지"];
+
+let adjectiveCache = null;
+let rieulCache = null;
+let iVerbCache = null;
+let irregularCache = null;
+
+function adjectiveStems() {
+  if (!adjectiveCache) adjectiveCache = new Set(loadLines("adjectiveStems.txt"));
+  return adjectiveCache;
+}
+
+function rieulStems() {
+  if (!rieulCache) rieulCache = new Set(loadLines("rieulStems.txt"));
+  return rieulCache;
+}
+
+function iVerbStems() {
+  if (!iVerbCache) iVerbCache = new Set(loadLines("iVerbStems.txt"));
+  return iVerbCache;
+}
+
+function irregularStems() {
+  if (!irregularCache) {
+    irregularCache = loadLines("irregularStems.txt")
+      .map((line) => line.split("\t"))
+      .sort((a, b) => b[0].length - a[0].length);
+  }
+  return irregularCache;
+}
+
+/** @param {string} stem */
+export function isAdjective(stem) {
+  return adjectiveStems().has(stem);
+}
+
+/** @param {string} stem */
+function irregularClass(stem) {
+  for (const [known, kind] of irregularStems()) {
+    if (stem === known || stem.endsWith(known)) return kind;
+  }
+  return null;
+}
+
+/** @param {string} base */
+function restoreRieul(base) {
+  const last = hangul.lastSyllable(base);
+  if (last === null || hangul.finalOf(last) !== hangul.NONE) return base;
+  const candidate = base.slice(0, -1) + hangul.withFinal(last, hangul.RIEUL);
+  return rieulStems().has(candidate) ? candidate : base;
+}
+
+/** @param {string} base */
+function stripEu(base) {
+  if (base.length >= 2 && base.endsWith("으") && hangul.isSyllable(base.at(-2)) && hangul.finalOf(base.at(-2)) !== hangul.NONE) {
+    return base.slice(0, -1);
+  }
+  return base;
+}
+
+/** @param {string} base */
+function tenseOf(base) {
+  const last = hangul.lastSyllable(base);
+  if (base.endsWith("겠")) return FUTURE;
+  if (last !== null && last !== "있" && hangul.finalOf(last) === hangul.SSANGSIOT) return PAST;
+  return PRESENT;
+}
+
+/** @param {string} base @param {string | null} previous */
+function kindOf(base, previous) {
+  if (base === "않" && previous) {
+    for (const link of AUX_LINKS) {
+      if (previous.endsWith(link) && previous.length > link.length) {
+        return isAdjective(previous.slice(0, -link.length)) ? ADJECTIVE : VERB;
+      }
+    }
+  }
+  return isAdjective(base) ? ADJECTIVE : VERB;
+}
+
+/**
+ * @typedef {object} Predicate
+ * @property {string} base
+ * @property {string} kind
+ * @property {string} tense
+ * @property {string} mood
+ * @property {boolean} explicitCopula
+ */
+
+/** @param {string} base @param {string} kind @param {string} [tense] @param {string} [mood] @param {boolean} [explicitCopula] @returns {Predicate} */
+function predicate(base, kind, tense = PRESENT, mood = DECLARATIVE, explicitCopula = false) {
+  return { base, kind, tense, mood, explicitCopula };
+}
+
+/** @param {string} base @param {string | null} previous @param {string} mood */
+function classifyPredicate(base, previous, mood) {
+  if (!base) return null;
+  const tense = tenseOf(base);
+  const kind = tense !== PRESENT ? VERB : kindOf(base, previous);
+  return predicate(base, kind, tense, mood);
+}
+
+/** @param {string} word @param {number} offset @param {number} final */
+function endsWithFinal(word, offset, final) {
+  if (word.length < offset) return false;
+  const ch = word.at(-offset);
+  return hangul.isSyllable(ch) && hangul.finalOf(ch) === final;
+}
+
+/** @param {string} word @param {number} offset */
+function stemBefore(word, offset) {
+  const ch = word.at(-offset);
+  return word.slice(0, -offset) + hangul.withFinal(ch, hangul.NONE);
+}
+
+/** @param {string} noun @param {string} mood */
+function copulaOrIVerb(noun, mood) {
+  const stem = noun + "이";
+  return iVerbStems().has(stem) ? predicate(stem, VERB, PRESENT, mood) : predicate(noun, COPULA, PRESENT, mood);
+}
+
+/** 문장 끝 한글 어절 하나를 푼다. @param {string} word @param {string | null} [previous] @returns {Predicate | null} */
+export function parsePredicate(word, previous = null) {
+  if (!word || !hangul.isSyllable(word.at(-1))) return null;
+  if ((word === "다" || word === "입니다") && previous) return predicate("", COPULA);
+  if (word === "아닙니다" || word === "아니다") return predicate("아니", ADJECTIVE);
+  if (word.endsWith("입니다") && word.length > 3) return copulaOrIVerb(word.slice(0, -3), DECLARATIVE);
+  if (word.endsWith("입니까") && word.length > 3) return copulaOrIVerb(word.slice(0, -3), QUESTION);
+  if (word.endsWith("습니다")) return classifyPredicate(word.slice(0, -3), previous, DECLARATIVE);
+  if (word.endsWith("습니까")) return classifyPredicate(word.slice(0, -3), previous, QUESTION);
+  if (word.endsWith("니다") && endsWithFinal(word, 3, hangul.BIEUP)) {
+    return classifyPredicate(restoreRieul(stemBefore(word, 3)), previous, DECLARATIVE);
+  }
+  if (word.endsWith("니까") && endsWithFinal(word, 3, hangul.BIEUP)) {
+    return classifyPredicate(restoreRieul(stemBefore(word, 3)), previous, QUESTION);
+  }
+  if (word.endsWith("까요") && word.length > 2) return predicate(word.slice(0, -1), VERB, PRESENT, KKA_QUESTION);
+  if (word.endsWith("십시오") && word.length > 3) {
+    return predicate(restoreRieul(stripEu(word.slice(0, -3))), VERB, PRESENT, IMPERATIVE);
+  }
+  if (word.endsWith("세요") && word.length > 2) {
+    return predicate(restoreRieul(stripEu(word.slice(0, -2))), VERB, PRESENT, IMPERATIVE);
+  }
+  if (word.endsWith("읍시다")) return predicate(word.slice(0, -3), VERB, PRESENT, PROPOSITIVE);
+  if (word.endsWith("시다") && endsWithFinal(word, 3, hangul.BIEUP)) {
+    return predicate(restoreRieul(stemBefore(word, 3)), VERB, PRESENT, PROPOSITIVE);
+  }
+  if (word.endsWith("는다") && word.length > 2) return predicate(word.slice(0, -2), VERB);
+  if (word.endsWith("는가") && word.length > 2) return classifyPredicate(word.slice(0, -2), previous, QUESTION);
+  if (word.endsWith("자") && word.length > 1) {
+    const stem = restoreRieul(word.slice(0, -1));
+    if (stem.endsWith("하") || stem.endsWith("보") || rieulStems().has(stem)) {
+      return predicate(stem, VERB, PRESENT, PROPOSITIVE);
+    }
+  }
+  if (word.endsWith("다") && word.length > 1) {
+    const stem = word.slice(0, -1);
+    const last = stem.at(-1);
+    if (hangul.isSyllable(last) && hangul.finalOf(last) === hangul.NIEUN && !word.endsWith("이다")) {
+      return predicate(restoreRieul(stemBefore(stem, 1)), VERB);
+    }
+    const tense = tenseOf(stem);
+    if (tense !== PRESENT) return predicate(stem, VERB, tense);
+    if (word.endsWith("이다") && word.length > 2) return predicate(word.slice(0, -2), COPULA, PRESENT, DECLARATIVE, true);
+    if (stem === "않") return predicate(stem, kindOf(stem, previous));
+    if (isAdjective(stem)) return predicate(stem, ADJECTIVE);
+    return predicate(stem, COPULA);
+  }
+  return null;
+}
+
+/** 어간에 아/어를 붙인 꼴. @param {string} stem */
+export function conjugate(stem) {
+  if (stem === "아니") return "아니에";
+  if (stem.endsWith("하")) return stem.slice(0, -1) + "해";
+  const last = hangul.lastSyllable(stem);
+  if (last === null) return stem + "어";
+  const [, vowel, final] = hangul.split(last);
+  const kind = irregularClass(stem);
+  if (kind === "르" && stem.length >= 2 && hangul.isSyllable(stem.at(-2))) {
+    const before = stem.at(-2);
+    const bright = hangul.BRIGHT.has(hangul.vowelOf(before));
+    return stem.slice(0, -2) + hangul.withFinal(before, hangul.RIEUL) + (bright ? "라" : "러");
+  }
+  if (kind === "ㅂ") return stem.slice(0, -1) + hangul.withFinal(last, hangul.NONE) + "워";
+  if (kind === "ㅂ와") return stem.slice(0, -1) + hangul.withFinal(last, hangul.NONE) + "와";
+  if (kind === "ㄷ") return stem.slice(0, -1) + hangul.withFinal(last, hangul.RIEUL) + (hangul.BRIGHT.has(vowel) ? "아" : "어");
+  if (kind === "ㅅ") return stem.slice(0, -1) + hangul.withFinal(last, hangul.NONE) + (hangul.BRIGHT.has(vowel) ? "아" : "어");
+  if (kind === "ㅎ") {
+    for (const [known, changed] of HAE_IRREGULAR) {
+      if (stem.endsWith(known)) return stem.slice(0, -known.length) + changed;
+    }
+    const opened = hangul.withFinal(last, hangul.NONE);
+    const target = new Map([
+      [hangul.A, hangul.AE],
+      [hangul.YA, hangul.YAE],
+      [hangul.EO, hangul.E],
+      [hangul.YEO, hangul.YE],
+    ]).get(vowel) ?? hangul.AE;
+    return stem.slice(0, -1) + hangul.withVowel(opened, target);
+  }
+  if (final !== hangul.NONE) return stem + (hangul.BRIGHT.has(vowel) ? "아" : "어");
+  if ([hangul.A, hangul.YA, hangul.EO, hangul.YEO, hangul.AE, hangul.E, hangul.YE].includes(vowel)) return stem;
+  if (vowel === hangul.OH) return stem.slice(0, -1) + hangul.withVowel(last, hangul.WA);
+  if (vowel === hangul.U) return stem.slice(0, -1) + hangul.withVowel(last, hangul.WO);
+  if (vowel === hangul.OE) return stem.slice(0, -1) + hangul.withVowel(last, hangul.WAE);
+  if (vowel === hangul.IH) return stem.slice(0, -1) + hangul.withVowel(last, hangul.YEO);
+  if (vowel === hangul.EU) {
+    const before = stem.length >= 2 && hangul.isSyllable(stem.at(-2)) ? stem.at(-2) : null;
+    const bright = before !== null && hangul.BRIGHT.has(hangul.vowelOf(before));
+    return stem.slice(0, -1) + hangul.withVowel(last, bright ? hangul.A : hangul.EO);
+  }
+  return stem + "어";
+}
+
+/** @param {string} stem */
+function hasFinal(stem) {
+  const last = hangul.lastSyllable(stem);
+  return last !== null ? hangul.finalOf(last) : hangul.MIEUM;
+}
+
+/** @param {string} stem */
+function polite(stem) {
+  const final = hasFinal(stem);
+  if (final === hangul.NONE || final === hangul.RIEUL) {
+    return stem.slice(0, -1) + hangul.withFinal(stem.at(-1), hangul.BIEUP) + "니다";
+  }
+  return stem + "습니다";
+}
+
+/** @param {string} stem */
+function plainVerb(stem) {
+  const final = hasFinal(stem);
+  if (final === hangul.NONE || final === hangul.RIEUL) {
+    return stem.slice(0, -1) + hangul.withFinal(stem.at(-1), hangul.NIEUN) + "다";
+  }
+  return stem + "는다";
+}
+
+/** @param {string} stem @param {string} ending */
+function withEu(stem, ending) {
+  const final = hasFinal(stem);
+  if (final === hangul.NONE) return stem + ending;
+  if (final === hangul.RIEUL) {
+    const base = ending === "세요" ? stem.slice(0, -1) + hangul.withFinal(stem.at(-1), hangul.NONE) : stem;
+    return base + ending;
+  }
+  const kind = irregularClass(stem);
+  const last = stem.at(-1);
+  if (kind === "ㅂ" || kind === "ㅂ와") return stem.slice(0, -1) + hangul.withFinal(last, hangul.NONE) + "우" + ending;
+  if (kind === "ㄷ") return stem.slice(0, -1) + hangul.withFinal(last, hangul.RIEUL) + "으" + ending;
+  if (kind === "ㅅ") return stem.slice(0, -1) + hangul.withFinal(last, hangul.NONE) + "으" + ending;
+  return stem + "으" + ending;
+}
+
+/** 서술어를 target 문체로 짠다. @param {Predicate} value @param {string} target */
+export function render(value, target) {
+  const { base, kind, tense, mood } = value;
+  if (mood === KKA_QUESTION) return target === HANDA ? base : base + "요";
+  if (tense !== PRESENT) {
+    if (target === HAPNIDA) return base + (mood === QUESTION ? "습니까" : "습니다");
+    if (target === HANDA) return base + (mood === QUESTION ? "는가" : "다");
+    return base + "어요";
+  }
+  if (kind === COPULA) {
+    if (!base) return { [HAPNIDA]: "입니다", [HANDA]: "다", [HAEYO]: "예요" }[target];
+    const final = hasFinal(base);
+    if (target === HAPNIDA) return base + (mood === QUESTION ? "입니까" : "입니다");
+    if (target === HANDA) {
+      if (mood === QUESTION) return base + "인가";
+      return base + (value.explicitCopula || final !== hangul.NONE ? "이다" : "다");
+    }
+    return base + (final !== hangul.NONE ? "이에요" : "예요");
+  }
+  if (mood === IMPERATIVE) return target === HANDA ? withEu(base, "라") : withEu(base, "세요");
+  if (mood === PROPOSITIVE) {
+    if (target === HAPNIDA) {
+      const final = hasFinal(base);
+      if (final === hangul.NONE || final === hangul.RIEUL) {
+        return base.slice(0, -1) + hangul.withFinal(base.at(-1), hangul.BIEUP) + "시다";
+      }
+      return base + "읍시다";
+    }
+    if (target === HANDA) return base + "자";
+    return conjugate(base) + "요";
+  }
+  if (target === HAPNIDA) {
+    const form = polite(base);
+    return mood === QUESTION ? form.slice(0, -2) + "니까" : form;
+  }
+  if (target === HANDA) {
+    if (kind === ADJECTIVE) {
+      if (mood === QUESTION) {
+        const final = hasFinal(base);
+        if (final === hangul.NONE || final === hangul.RIEUL) {
+          return base.slice(0, -1) + hangul.withFinal(base.at(-1), hangul.NIEUN) + "가";
+        }
+        return base + "은가";
+      }
+      return base + "다";
+    }
+    if (mood === QUESTION) {
+      const stem = hasFinal(base) === hangul.RIEUL ? base.slice(0, -1) + hangul.withFinal(base.at(-1), hangul.NONE) : base;
+      return stem + "는가";
+    }
+    return plainVerb(base);
+  }
+  return conjugate(base) + "요";
+}