@sushichan044/kg-core 0.6.0 → 0.6.1

package/dist/index.mjs

@@ -1544,7 +1544,7 @@ function defineMatchRule(options) {
 }
 //#endregion
 //#region src/proofreading/rules/consistent-kanji-opening.ts
-const RULE_ID$5 = "kg/consistent-kanji-opening";
+const RULE_ID$6 = "kg/consistent-kanji-opening";
 const MESSAGE = "「{{ closed }}」と「{{ opened }}」が混在しています。文脈上の使い分けか表記ゆれかを確認してください";
 const KanjiOpeningPairSchema = v.pipe(readonlyObject({
 	closed: v.pipe(v.string(), v.nonEmpty()),
@@ -1598,7 +1598,7 @@ const DEFAULT_PAIRS = [
 * two spellings are a deliberate distinction is the author's call, so the finding is a warning.
 */
 function build$3(pairs) {
-	return defineParsedRule(RULE_ID$5, MESSAGE, (manuscript, report) => {
+	return defineParsedRule(RULE_ID$6, MESSAGE, (manuscript, report) => {
 		for (const { closed, opened } of pairs) {
 			if (!manuscript.displayText.includes(opened)) continue;
 			const index = manuscript.displayText.indexOf(closed);
@@ -1619,7 +1619,7 @@ function createConsistentKanjiOpeningRule(options = {}) {
 	const pairs = v.safeParse(PairsSchema, options.pairs);
 	if (!pairs.success) return ManuscriptResult.fail({
 		kind: "InvalidRuleOptions",
-		ruleId: RULE_ID$5,
+		ruleId: RULE_ID$6,
 		option: "pairs",
 		issues: ValidationIssue.from(pairs.issues)
 	});
@@ -1664,7 +1664,7 @@ const consistentNumeralWidthRule = () => defineMixedWidthRule({
 });
 //#endregion
 //#region src/proofreading/rules/dash.ts
-const RULE_ID$4 = "kg/dash";
+const RULE_ID$5 = "kg/dash";
 const DEFAULT_PREFERRED = "―";
 const DashCharacterSchema = v.picklist([
 	"―",
@@ -1686,7 +1686,7 @@ function build$2(preferred) {
 	return {
 		kind: "parsed",
 		meta: {
-			id: RULE_ID$4,
+			id: RULE_ID$5,
 			messages
 		},
 		check: (manuscript, context) => {
@@ -1715,7 +1715,7 @@ function createDashRule(options = {}) {
 	const preferred = v.safeParse(DashCharacterSchema, options.preferred);
 	if (!preferred.success) return ManuscriptResult.fail({
 		kind: "InvalidRuleOptions",
-		ruleId: RULE_ID$4,
+		ruleId: RULE_ID$5,
 		option: "preferred",
 		issues: ValidationIssue.from(preferred.issues)
 	});
@@ -1742,8 +1742,8 @@ const evenEllipsisRule = () => defineMatchRule({
 });
 //#endregion
 //#region src/proofreading/rules/fullwidth-japanese-punctuation.ts
-const RULE_ID$3 = "kg/fullwidth-japanese-punctuation";
-const MESSAGES$2 = {
+const RULE_ID$4 = "kg/fullwidth-japanese-punctuation";
+const MESSAGES$3 = {
 	"halfwidth-japanese-punctuation": "半角の「{{ character }}」があります。日本語の句読点・括弧・感嘆符・疑問符は全角にしてください",
 	"halfwidth-near-japanese": "日本語の前後に半角の「{{ character }}」があります。全角にすべきか確認してください"
 };
@@ -1788,8 +1788,8 @@ function isEllipsisSubstitute(text, index) {
 const fullwidthJapanesePunctuationRule = () => ({
 	kind: "parsed",
 	meta: {
-		id: RULE_ID$3,
-		messages: MESSAGES$2
+		id: RULE_ID$4,
+		messages: MESSAGES$3
 	},
 	check: (manuscript, context) => {
 		for (const line of splitDisplayLines(manuscript.displayText)) for (const { pattern, messageId, severity, requiresJapanese } of CHECKS) for (const match of line.text.matchAll(pattern)) {
@@ -1807,12 +1807,12 @@ const fullwidthJapanesePunctuationRule = () => ({
 });
 //#endregion
 //#region src/proofreading/rules/max-arabic-numeral-digits.ts
-const RULE_ID$2 = "kg/max-arabic-numeral-digits";
+const RULE_ID$3 = "kg/max-arabic-numeral-digits";
 const DEFAULT_MAX_DIGITS = 2;
 const MaxDigitsSchema = v.pipe(v.number(), v.finite(), v.integer(), v.minValue(1));
 function build$1(maxDigits) {
 	return defineMatchRule({
-		id: RULE_ID$2,
+		id: RULE_ID$3,
 		pattern: /([0-9０-９]+)(?:[.．]([0-9０-９]+))?/gu,
 		accept: (match) => (match[1]?.length ?? 0) > maxDigits || (match[2]?.length ?? 0) > maxDigits,
 		message: `${maxDigits}桁を超えるアラビア数字が使われています`
@@ -1824,7 +1824,7 @@ function createMaxArabicNumeralDigitsRule(options = {}) {
 	const maxDigits = v.safeParse(MaxDigitsSchema, options.maxDigits);
 	if (!maxDigits.success) return ManuscriptResult.fail({
 		kind: "InvalidRuleOptions",
-		ruleId: RULE_ID$2,
+		ruleId: RULE_ID$3,
 		option: "maxDigits",
 		issues: ValidationIssue.from(maxDigits.issues)
 	});
@@ -1891,8 +1891,8 @@ function leadingSpaces(text) {
 }
 //#endregion
 //#region src/proofreading/rules/paragraph-opening.ts
-const RULE_ID$1 = "kg/paragraph-opening";
-const MESSAGES$1 = {
+const RULE_ID$2 = "kg/paragraph-opening";
+const MESSAGES$2 = {
 	"leading-character": "段落の先頭には全角スペースまたは開き括弧が必要です",
 	"indent-width": "段落冒頭の字下げは全角スペース1字にしてください",
 	"unexpected-indent": "括弧から始まる段落を字下げすることはできません"
@@ -1907,8 +1907,8 @@ function build(openingBrackets) {
 	return {
 		kind: "parsed",
 		meta: {
-			id: RULE_ID$1,
-			messages: MESSAGES$1
+			id: RULE_ID$2,
+			messages: MESSAGES$2
 		},
 		check: (manuscript, context) => {
 			const report = (start, end, messageId) => {
@@ -1938,7 +1938,7 @@ function createParagraphOpeningRule(options = {}) {
 	const brackets = v.safeParse(OpeningBracketsSchema, options.openingBrackets);
 	if (!brackets.success) return ManuscriptResult.fail({
 		kind: "InvalidRuleOptions",
-		ruleId: RULE_ID$1,
+		ruleId: RULE_ID$2,
 		option: "openingBrackets",
 		issues: ValidationIssue.from(brackets.issues)
 	});
@@ -1954,8 +1954,8 @@ const punctuationBeforeClosingQuoteRule = () => defineMatchRule({
 });
 //#endregion
 //#region src/proofreading/rules/space-after-question-or-exclamation.ts
-const RULE_ID = "kg/space-after-question-or-exclamation";
-const MESSAGES = {
+const RULE_ID$1 = "kg/space-after-question-or-exclamation";
+const MESSAGES$1 = {
 	default: "感嘆符または疑問符の直後には全角スペースか閉じ括弧が必要です",
 	"space-width": "感嘆符または疑問符の直後の空白は全角スペース1字にしてください",
 	"before-closing-bracket": "閉じ括弧の直前に空白を置くことはできません"
@@ -1971,8 +1971,8 @@ const MARKS = /[？！]+/gu;
 const spaceAfterQuestionOrExclamationRule = () => ({
 	kind: "parsed",
 	meta: {
-		id: RULE_ID,
-		messages: MESSAGES
+		id: RULE_ID$1,
+		messages: MESSAGES$1
 	},
 	check: (manuscript, context) => {
 		const report = (start, end, messageId) => {
@@ -2000,22 +2000,74 @@ const spaceAfterQuestionOrExclamationRule = () => ({
 });
 //#endregion
 //#region src/proofreading/rules/variant-character.ts
+const RULE_ID = "kg/variant-character";
+const MESSAGES = {
+	default: "異体字または字形選択子が使われています",
+	"with-suggestion": "異体字または字形選択子が使われています。「{{ suggestion }}」ではありませんか？"
+};
+/**
+* Mongolian free variation selectors, variation selectors (incl. emoji presentation), IVSes: all
+* combine with a preceding base character rather than standing on their own.
+*/
+function isSelector(codePoint) {
+	return codePoint >= 6155 && codePoint <= 6157 || codePoint === 6159 || codePoint >= 65024 && codePoint <= 65039 || codePoint >= 917760 && codePoint <= 917999;
+}
 /**
-* Mongolian free variation selectors, variation selectors, CJK compatibility ideographs, IVSes.
+* CJK compatibility ideographs stand alone; each has a canonical unified form reachable via NFKC.
 */
+function isCompatibilityIdeograph(codePoint) {
+	return codePoint >= 63744 && codePoint <= 64255 || codePoint >= 194560 && codePoint <= 195103;
+}
 function isVariantCharacter(character) {
 	const codePoint = character.codePointAt(0);
 	if (codePoint === void 0) return false;
-	return codePoint >= 6155 && codePoint <= 6157 || codePoint === 6159 || codePoint >= 65024 && codePoint <= 65039 || codePoint >= 63744 && codePoint <= 64255 || codePoint >= 194560 && codePoint <= 195103 || codePoint >= 917760 && codePoint <= 917999;
+	return isSelector(codePoint) || isCompatibilityIdeograph(codePoint);
+}
+/**
+* The plain form a reader would expect: the selector stripped away, or the compatibility
+* ideograph's canonical unified form. Undefined when that plain form is identical to the original,
+* since NFKC does not touch selectors and a suggestion equal to the input helps no one.
+*/
+function suggestPlainForm(segment) {
+	const withoutSelectors = Array.from(segment).filter((character) => {
+		const codePoint = character.codePointAt(0);
+		return codePoint !== void 0 && !isSelector(codePoint);
+	}).join("");
+	if (withoutSelectors === "") return void 0;
+	const normalized = Array.from(withoutSelectors).map((character) => {
+		const codePoint = character.codePointAt(0);
+		return codePoint !== void 0 && isCompatibilityIdeograph(codePoint) ? character.normalize("NFKC") : character;
+	}).join("");
+	return normalized === segment ? void 0 : normalized;
 }
 /**
 * Scans the source rather than the display text: notation removal must not hide a variant.
 */
-const variantCharacterRule = () => defineParsedRule("kg/variant-character", "異体字または字形選択子が使われています", (manuscript, report) => {
-	for (const { index, segment } of graphemeSegmenter.segment(manuscript.source)) if (Array.from(segment).some(isVariantCharacter)) report(sourceRange(manuscript, {
-		start: index,
-		end: index + segment.length
-	}));
+const variantCharacterRule = () => ({
+	kind: "parsed",
+	meta: {
+		id: RULE_ID,
+		messages: MESSAGES
+	},
+	check: (manuscript, context) => {
+		for (const { index, segment } of graphemeSegmenter.segment(manuscript.source)) {
+			if (!Array.from(segment).some(isVariantCharacter)) continue;
+			const range = sourceRange(manuscript, {
+				start: index,
+				end: index + segment.length
+			});
+			const suggestion = suggestPlainForm(segment);
+			if (suggestion === void 0) context.report({
+				range,
+				messageId: "default"
+			});
+			else context.report({
+				range,
+				messageId: "with-suggestion",
+				data: { suggestion }
+			});
+		}
+	}
 });
 //#endregion
 //#region src/proofreading/rules/default-rules.ts

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@sushichan044/kg-core",
-  "version": "0.6.0",
+  "version": "0.6.1",
   "description": "Framework-independent parsing, composition, and proofreading for Japanese manuscripts",
   "keywords": [
     "japanese",