flappa-doormal 2.5.1 → 2.5.2

package/README.md

@@ -777,6 +777,23 @@ const rule = {
 };
 ```
 
+### Expanding composite tokens (for adding named captures)
+
+Some tokens are **composites** (e.g. `{{numbered}}`), which are great for quick signatures but less convenient when you want to add named captures (e.g. capture the number).
+
+You can expand composites back into their underlying template form:
+
+```typescript
+import { expandCompositeTokensInTemplate } from 'flappa-doormal';
+
+const base = expandCompositeTokensInTemplate('{{numbered}}');
+// base === '{{raqms}} {{dash}} '
+
+// Now you can add a named capture:
+const withCapture = base.replace('{{raqms}}', '{{raqms:num}}');
+// withCapture === '{{raqms:num}} {{dash}} '
+```
+
 ## Types
 
 ### `SplitRule`

package/dist/index.d.mts

@@ -864,6 +864,21 @@ declare const segmentPages: (pages: Page[], options: SegmentationOptions) => Seg
  * // → '{{harf}}' (unchanged - no brackets outside tokens)
  */
 declare const escapeTemplateBrackets: (pattern: string) => string;
+/**
+ * Expands any *composite* tokens (like `{{numbered}}`) into their underlying template form
+ * (like `{{raqms}} {{dash}} `).
+ *
+ * This is useful when you want to take a signature produced by `analyzeCommonLineStarts()`
+ * and turn it into an editable template where you can add named captures, e.g.:
+ *
+ * - `{{numbered}}` → `{{raqms}} {{dash}} `
+ * - then: `{{raqms:num}} {{dash}} ` to capture the number
+ *
+ * Notes:
+ * - This only expands the plain `{{token}}` form (not `{{token:name}}`).
+ * - Expansion is repeated a few times to support nested composites.
+ */
+declare const expandCompositeTokensInTemplate: (template: string) => string;
 /**
  * Token definitions mapping human-readable token names to regex patterns.
  *
@@ -1198,5 +1213,5 @@ declare const analyzeTextForRule: (text: string) => {
   detected: DetectedPattern[];
 } | null;
 //#endregion
-export { type Breakpoint, type BreakpointRule, type CommonLineStartPattern, type DetectedPattern, type ExpandResult, type LineStartAnalysisOptions, type LineStartPatternExample, type Logger, type Page, type PageRange, type Segment, type SegmentationOptions, type SplitRule, TOKEN_PATTERNS, analyzeCommonLineStarts, analyzeTextForRule, containsTokens, detectTokenPatterns, escapeRegex, escapeTemplateBrackets, expandTokens, expandTokensWithCaptures, generateTemplateFromText, getAvailableTokens, getTokenPattern, makeDiacriticInsensitive, segmentPages, suggestPatternConfig, templateToRegex };
+export { type Breakpoint, type BreakpointRule, type CommonLineStartPattern, type DetectedPattern, type ExpandResult, type LineStartAnalysisOptions, type LineStartPatternExample, type Logger, type Page, type PageRange, type Segment, type SegmentationOptions, type SplitRule, TOKEN_PATTERNS, analyzeCommonLineStarts, analyzeTextForRule, containsTokens, detectTokenPatterns, escapeRegex, escapeTemplateBrackets, expandCompositeTokensInTemplate, expandTokens, expandTokensWithCaptures, generateTemplateFromText, getAvailableTokens, getTokenPattern, makeDiacriticInsensitive, segmentPages, suggestPatternConfig, templateToRegex };
 //# sourceMappingURL=index.d.mts.map

package/dist/index.d.mts.map

@@ -1 +1 @@
-{"version":3,"file":"index.d.mts","names":[],"sources":["../src/segmentation/fuzzy.ts","../src/segmentation/types.ts","../src/segmentation/segmenter.ts","../src/segmentation/tokens.ts","../src/analysis.ts","../src/detection.ts"],"sourcesContent":[],"mappings":";;AAkEA;AA+FA;;;;;ACnIiB;AA4BG;AA8BM;AAiCC;AAwBH;;;;;;;AAoBC;AA2FzB;AAAkD;AAgIlD;;;;;AAkBA;AAqCA;AA0EY,cD/bC,WC+bqB,EAAA,CAAA,CAAA,EAAA,MAAc,EAAA,GAAA,MAAA;AA8BhD;AAiDA;;;;;AA+HA;;;;ACjPA;;;;;;;;AC1ZA;AAgQA;AA2CA;AAWA;AA2DA;AAyHA;AAuBA;AAqBA;AAgBA;;;;ACtmBA;AAkEA;AAEA;AAuRA;;AAEa,cJhMA,wBIgMA,EAAA,CAAA,IAAA,EAAA,MAAA,EAAA,GAAA,MAAA;;;;AJ/Rb;AA+FA;;;;;ACnIiB;AA4BG;AA8BM;AAiCC;AAwBH;;;;;;;AAoBC;AA2FzB;AAAkD;AAgIlD;;;;;AAkBA,KApXK,YAAA,GAoXW;EAqCJ;EA0EA,KAAA,EAAA,MAAU;AA8BtB,CAAA;AAiDA;;;;;AA+HA;;;;ACjPA;;;;;;;;AC1ZA;AAgQA;AA2CA;AAWA;AA2DA;AAyHA;AAuBA,KF3gBK,eAAA,GEkhBJ;EAcY;EAgBA,QAAA,EAAA,MAAA;;;;ACtmBb;AAkEA;AAEA;AAuRA;;;;;;;;ACnVA;AA+EA;AAgEA;AAuBA;AAiCA;;;;;;;;KJ3HK,qBAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;KAiCA,sBAAA;;;;;;;;;;;;;;;;;;;;;;;KAwBA,mBAAA;;;;;;;;;;;;;;KAeA,WAAA,GACC,eACA,kBACA,wBACA,yBACA;;;;;;;KAYD,aAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;KA+EO,SAAA;;;;;;;KAYP,eAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;YAyCS;;;;;;;;;;;;SAaH;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;KA8DC,SAAA,GAAY,cAAc,gBAAgB;;;;;;;;;;;;;KAkB1C,IAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;KAqCA,cAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;YAqCE;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;KAqCF,UAAA,YAAsB;;;;;;;;;;;;;;;;;;;;;;;;;UA8BjB,MAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;KAiDL,mBAAA;;;;;;;;UAQA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;gBA8CM;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;WAwDL;;;;;;;;;;;;;;;;KAiBD,OAAA;;;;;;;;;;;;;;;;;;;;;;;;;;SA6BD;;;;;;AA1VX;AAqCA;AA0EA;AA8BA;AAiDA;;;;;AA+HA;;;;ACjPA;;;;;;;;AC1ZA;AAgQA;AA2CA;AAWA;AA2DA;AAyHA;AAuBA;AAqBA;AAgBA;;;;ACtmBA;AAkEA;AAEA;AAuRA;;;;;;;;ACnVA;AA+EA;AAgEA;AAuBA;AAiCA;;;;;;cH2Qa,sBAAuB,iBAAiB,wBAAsB;;;;AF5Z3E;AA+FA;;;;;ACnIiB;AA4BG;AA8BM;AAiCC;AAwBH;;;;;;;AAoBC;AA2FzB;AAAkD;AAgIlD;;;;;AAkBA;AAqCA;AA0EA;AA8BA;AAiDA;;;;;AA+HA;;;;ACjPA;;;;;;;;AC1ZA;AAgQA;AA2CA;AAWA;AA2DA;AAyHA;AAuBA;AAqBA;AAgBA;;;;ACtmBA;AAkEA;AAEA;AAuRA;AACW,cD5RE,sBC4RF,EAAA,CAAA,OAAA,EAAA,MAAA,EAAA,GAAA,MAAA;;;;;;;ACpVX;AA+EA;AAgEA;AAuBA;AAiCA;;;;;;;;;;;;;;;cFiHa,gBAAgB;;;;;;;;;;;;;;;;cA2ChB;;;;;;;KAWD,YAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;cA2DC,mHAIV;;;;;;;;;;;;;;;;;;;;cAqHU;;;;;;;;;;;;;;;;;;;;;;cAuBA,uCAAmC;;;;;;;;;;;;;cAqBnC;;;;;;;;;;;;;;;cAgBA;;;AHxiBA,KI9DD,wBAAA,GJ8D8E;EA+F7E;;;;ECnIR;EA4BA,aAAA,CAAA,EAAA,MAAe;EA8Bf;EAiCA,QAAA,CAAA,EAAA,MAAA;EAwBA;EAeA,WAAA,CAAA,EAAW,MAAA;EACV;;;;EAIA,wBAAA,CAAA,EAAA,OAAA;EAAmB;AAAA;AA2FzB;AAAkD;AAgIlD;;;;EAAqE,yBAAA,CAAA,EAAA,OAAA;EAkBzD;AAqCZ;AA0EA;AA8BA;AAiDA;;EAsDkB,MAAA,CAAA,EAAA,aAAA,GAAA,OAAA;EAwDL;;AAiBb;;;;ACjPA;;;;EAAkF,UAAA,CAAA,EAAA,CAAA,IAAA,EAAA,MAAA,EAAA,MAAA,EAAA,MAAA,EAAA,GAAA,OAAA;;;;AC1ZlF;AAgQA;AA2CA;AAWA;AA2DA;AAyHA;AAuBA;AAqBA;AAgBA;;mBC9iBqB;;AAxDrB;AAkEA;AAEA;AAuRA;;EAEa,UAAA,CAAA,EAAA,OAAA,GAAA,OAAA;CACV;AAAsB,KA5Rb,uBAAA,GA4Ra;;;;ACtVb,KD4DA,sBAAA,GC5De;EA+Ed,OAAA,EAAA,MAAA;EAgEA,KAAA,EAAA,MAAA;EAuBA,QAAA,EDvGC,uBCgIb,EAxBa;AAgCd,CAAA;;;;;;;cD4Ia,iCACF,kBACE,6BACV;;;;AJhSH;AA+FA;;;;;ACnIiB;AA4BG;AA+Df,KI7GO,eAAA,GJ6Ge;EAwBtB;EAeA,KAAA,EAAA,MAAA;EACC;EACA,KAAA,EAAA,MAAA;EACA;EACA,KAAA,EAAA,MAAA;EACA;EAAmB,QAAA,EAAA,MAAA;AAAA,CAAA;AA2FzB;AAAkD;AAgIlD;;;;;AAkBA;AAqCA;AA0EA;AA8BA;AAiDA;;;;AA8GmB,cInmBN,mBJmmBM,EAAA,CAAA,IAAA,EAAA,MAAA,EAAA,GInmB6B,eJmmB7B,EAAA;AAiBnB;;;;ACjPA;;;;;;;;AC1ZA;AAgQa,cEzKA,wBFyKsB,EAAA,CAAA,IAAA,EAAA,MAAA,EAAA,QAAA,EEzK8B,eFyK9B,EAAA,EAAA,GAAA,MAAA;AA2CnC;AAWA;AA2DA;AAyHA;AAuBA;AAqBA;AAgBa,cExbA,oBFwbsF,EAAA,CAAA,QAAA,EEvbrF,eFubqF,EAAA,EAAA,GAAA;;;;ACtmBnG,CAAA;AAkEA;AAEA;AAuRA;;;;AAGyB,cC/IZ,kBD+IY,EAAA,CAAA,IAAA,EAAA,MAAA,EAAA,GAAA;;;;ECtVb,QAAA,CAAA,EAAA,MAAA;EA+EC,QAAA,EA+HC,eA9Eb,EAAA;AAeD,CAAA,GAAa,IAAA"}
+{"version":3,"file":"index.d.mts","names":[],"sources":["../src/segmentation/fuzzy.ts","../src/segmentation/types.ts","../src/segmentation/segmenter.ts","../src/segmentation/tokens.ts","../src/analysis.ts","../src/detection.ts"],"sourcesContent":[],"mappings":";;AAkEA;AA+FA;;;;;ACnIiB;AA4BG;AA8BM;AAiCC;AAwBH;;;;;;;AAoBC;AA2FzB;AAAkD;AAgIlD;;;;;AAkBA;AAqCA;AA0EY,cD/bC,WC+bqB,EAAA,CAAA,CAAA,EAAA,MAAc,EAAA,GAAA,MAAA;AA8BhD;AAiDA;;;;;AA+HA;;;;ACjPA;;;;;;;;AC1ZA;AAuOA;AAsDA;AA2CA;AAWA;AAuKA;AA6CA;AAuBA;AAqBA;AAgBA;;;;ACnqBA;AAkEA;AAEA;AAuRA;AACW,cJ/LE,wBI+LF,EAAA,CAAA,IAAA,EAAA,MAAA,EAAA,GAAA,MAAA;;;;AJ9RX;AA+FA;;;;;ACnIiB;AA4BG;AA8BM;AAiCC;AAwBH;;;;;;;AAoBC;AA2FzB;AAAkD;AAgIlD;;;;;AAkBA,KApXK,YAAA,GAoXW;EAqCJ;EA0EA,KAAA,EAAA,MAAU;AA8BtB,CAAA;AAiDA;;;;;AA+HA;;;;ACjPA;;;;;;;;AC1ZA;AAuOA;AAsDA;AA2CA;AAWA;AAuKA;AA6CA,KFjjBK,eAAA,GEijBiF;EAuBzE;EAqBA,QAAA,EAAA,MAAA;AAgBb,CAAA;;;;ACnqBA;AAkEA;AAEA;AAuRA;;;;;;;;ACnVA;AA+EA;AAgEA;AAuBA;AAiCA;;;;;;;KJ3HK,qBAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;KAiCA,sBAAA;;;;;;;;;;;;;;;;;;;;;;;KAwBA,mBAAA;;;;;;;;;;;;;;KAeA,WAAA,GACC,eACA,kBACA,wBACA,yBACA;;;;;;;KAYD,aAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;KA+EO,SAAA;;;;;;;KAYP,eAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;YAyCS;;;;;;;;;;;;SAaH;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;KA8DC,SAAA,GAAY,cAAc,gBAAgB;;;;;;;;;;;;;KAkB1C,IAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;KAqCA,cAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;YAqCE;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;KAqCF,UAAA,YAAsB;;;;;;;;;;;;;;;;;;;;;;;;;UA8BjB,MAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;KAiDL,mBAAA;;;;;;;;UAQA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;gBA8CM;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;WAwDL;;;;;;;;;;;;;;;;KAiBD,OAAA;;;;;;;;;;;;;;;;;;;;;;;;;;SA6BD;;;;;;AA1VX;AAqCA;AA0EA;AA8BA;AAiDA;;;;;AA+HA;;;;ACjPA;;;;;;;;AC1ZA;AAuOA;AAsDA;AA2CA;AAWA;AAuKA;AA6CA;AAuBA;AAqBA;AAgBA;;;;ACnqBA;AAkEA;AAEA;AAuRA;;;;;;;;ACnVA;AA+EA;AAgEA;AAuBA;AAiCA;;;;;cH2Qa,sBAAuB,iBAAiB,wBAAsB;;;;AF5Z3E;AA+FA;;;;;ACnIiB;AA4BG;AA8BM;AAiCC;AAwBH;;;;;;;AAoBC;AA2FzB;AAAkD;AAgIlD;;;;;AAkBA;AAqCA;AA0EA;AA8BA;AAiDA;;;;;AA+HA;;;;ACjPA;;;;;;;;AC1ZA;AAuOA;AAsDA;AA2CA;AAWA;AAuKA;AA6CA;AAuBA;AAqBA;AAgBA;;;;ACnqBA;AAkEA;AAEA;AAuRa,cD3RA,sBC8WZ,EAAA,CAAA,OAAA,EAAA,MAAA,EAAA,GAAA,MAAA;;;;;;;;ACtaD;AA+EA;AAgEA;AAuBA;AAiCA;;;cFwFa;;;;;;;;;;;;;;;;;;;;;;;;;;cAsDA,gBAAgB;;;;;;;;;;;;;;;;cA2ChB;;;;;;;KAWD,YAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;cAuKC,mHAIV;;;;;;;;;;;;;;;;;;;;cAyCU;;;;;;;;;;;;;;;;;;;;;;cAuBA,uCAAmC;;;;;;;;;;;;;cAqBnC;;;;;;;;;;;;;;;cAgBA;;;AHrmBA,KI9DD,wBAAA,GJ8D8E;EA+F7E;;;;ECnIR;EA4BA,aAAA,CAAA,EAAA,MAAe;EA8Bf;EAiCA,QAAA,CAAA,EAAA,MAAA;EAwBA;EAeA,WAAA,CAAA,EAAW,MAAA;EACV;;;;EAIA,wBAAA,CAAA,EAAA,OAAA;EAAmB;AAAA;AA2FzB;AAAkD;AAgIlD;;;;EAAqE,yBAAA,CAAA,EAAA,OAAA;EAkBzD;AAqCZ;AA0EA;AA8BA;AAiDA;;EAsDkB,MAAA,CAAA,EAAA,aAAA,GAAA,OAAA;EAwDL;;AAiBb;;;;ACjPA;;;;EAAkF,UAAA,CAAA,EAAA,CAAA,IAAA,EAAA,MAAA,EAAA,MAAA,EAAA,MAAA,EAAA,GAAA,OAAA;;;;AC1ZlF;AAuOA;AAsDA;AA2CA;AAWA;AAuKA;AA6CA;AAuBA;AAqBA;AAgBA;mBC3mBqB;;;AAxDrB;AAkEA;AAEA;AAuRA;EACW,UAAA,CAAA,EAAA,OAAA,GAAA,OAAA;CACE;AACV,KA5RS,uBAAA,GA4RT;EAAsB,IAAA,EAAA,MAAA;;;KA1Rb,sBAAA;EC5DA,OAAA,EAAA,MAAA;EA+EC,KAAA,EAAA,MAAA;EAgEA,QAAA,EDhFC,uBC+Fb,EAAA;AAQD,CAAA;AAiCA;;;;;;cD4Ia,iCACF,kBACE,6BACV;;;;AJhSH;AA+FA;;;;;ACnIiB;AA4BG;AA+Df,KI7GO,eAAA,GJ6Ge;EAwBtB;EAeA,KAAA,EAAA,MAAA;EACC;EACA,KAAA,EAAA,MAAA;EACA;EACA,KAAA,EAAA,MAAA;EACA;EAAmB,QAAA,EAAA,MAAA;AAAA,CAAA;AA2FzB;AAAkD;AAgIlD;;;;;AAkBA;AAqCA;AA0EA;AA8BA;AAiDA;;;;AA8GmB,cInmBN,mBJmmBM,EAAA,CAAA,IAAA,EAAA,MAAA,EAAA,GInmB6B,eJmmB7B,EAAA;AAiBnB;;;;ACjPA;;;;;;;;AC1ZA;AAuOa,cEhJA,wBF6JZ,EAAA,CAAA,IAAA,EAAA,MAAA,EAAA,QAAA,EE7JgE,eF6JhE,EAAA,EAAA,GAAA,MAAA;AAyCD;AA2CA;AAWA;AAuKA;AA6CA;AAuBA;AAqBa,cEreA,oBFqesD,EAAA,CAAA,QAAA,EEperD,eFoeqD,EAAA,EAAA,GAAA;EAgBtD,WAAA,EAAA,gBAAsF,GAAA,iBAAA;;;;ACnqBnG;AAkEA;AAEA;AAuRA;;;AAGG,cC/IU,kBD+IV,EAAA,CAAA,IAAA,EAAA,MAAA,EAAA,GAAA;EAAsB,QAAA,EAAA,MAAA;;;;ECtVb,QAAA,EA8ME,eA9Ma,EAAA;AA+E3B,CAAA,GAAa,IAAA"}

package/dist/index.mjs

@@ -974,7 +974,7 @@ const escapeTemplateBrackets = (pattern) => {
 		return `\\${bracket}`;
 	});
 };
-const RUMUZ_ATOM = `(?:خت|خغ|بخ|عخ|مق|مت|عس|سي|كن|مد|قد|خد|فد|دل|كد|غد|صد|دت|تم|فق|دق|[خرزيمنصسدفلتقع]|(?<![\\u0660-\\u0669])٤(?![\\u0660-\\u0669]))`;
+const RUMUZ_ATOM = `(?:خت|خغ|بخ|عخ|مق|مت|عس|سي|سن|كن|مد|قد|خد|فد|دل|كد|غد|صد|دت|تم|فق|دق|[خرزيمنصسدفلتقع]|(?<![\\u0660-\\u0669])٤(?![\\u0660-\\u0669]))`;
 const RUMUZ_BLOCK = `${RUMUZ_ATOM}(?:\\s+${RUMUZ_ATOM})*`;
 const BASE_TOKENS = {
 	bab: "باب",
@@ -1012,6 +1012,31 @@ const BASE_TOKENS = {
 */
 const COMPOSITE_TOKENS = { numbered: "{{raqms}} {{dash}} " };
 /**
+* Expands any *composite* tokens (like `{{numbered}}`) into their underlying template form
+* (like `{{raqms}} {{dash}} `).
+*
+* This is useful when you want to take a signature produced by `analyzeCommonLineStarts()`
+* and turn it into an editable template where you can add named captures, e.g.:
+*
+* - `{{numbered}}` → `{{raqms}} {{dash}} `
+* - then: `{{raqms:num}} {{dash}} ` to capture the number
+*
+* Notes:
+* - This only expands the plain `{{token}}` form (not `{{token:name}}`).
+* - Expansion is repeated a few times to support nested composites.
+*/
+const expandCompositeTokensInTemplate = (template) => {
+	let out = template;
+	for (let i = 0; i < 10; i++) {
+		const next = out.replace(/\{\{(\w+)\}\}/g, (m, tokenName) => {
+			return COMPOSITE_TOKENS[tokenName] ?? m;
+		});
+		if (next === out) break;
+		out = next;
+	}
+	return out;
+};
+/**
 * Expands base tokens in a template string.
 * Used internally to pre-expand composite tokens.
 *
@@ -1092,6 +1117,73 @@ const containsTokens = (query) => {
 	SIMPLE_TOKEN_REGEX.lastIndex = 0;
 	return SIMPLE_TOKEN_REGEX.test(query);
 };
+const splitTemplateIntoSegments = (query) => {
+	const segments = [];
+	let lastIndex = 0;
+	TOKEN_WITH_CAPTURE_REGEX.lastIndex = 0;
+	let match;
+	while ((match = TOKEN_WITH_CAPTURE_REGEX.exec(query)) !== null) {
+		if (match.index > lastIndex) segments.push({
+			type: "text",
+			value: query.slice(lastIndex, match.index)
+		});
+		segments.push({
+			type: "token",
+			value: match[0]
+		});
+		lastIndex = match.index + match[0].length;
+	}
+	if (lastIndex < query.length) segments.push({
+		type: "text",
+		value: query.slice(lastIndex)
+	});
+	return segments;
+};
+const maybeApplyFuzzyToText = (text, fuzzyTransform) => {
+	if (fuzzyTransform && /[\u0600-\u06FF]/u.test(text)) return fuzzyTransform(text);
+	return text;
+};
+const maybeApplyFuzzyToTokenPattern = (tokenPattern, fuzzyTransform) => {
+	if (!fuzzyTransform) return tokenPattern;
+	return tokenPattern.split("|").map((part) => /[\u0600-\u06FF]/u.test(part) ? fuzzyTransform(part) : part).join("|");
+};
+const parseTokenLiteral = (literal) => {
+	TOKEN_WITH_CAPTURE_REGEX.lastIndex = 0;
+	const tokenMatch = TOKEN_WITH_CAPTURE_REGEX.exec(literal);
+	if (!tokenMatch) return null;
+	const [, tokenName, captureName] = tokenMatch;
+	return {
+		captureName,
+		tokenName
+	};
+};
+const createCaptureRegistry = (capturePrefix) => {
+	const captureNames = [];
+	const captureNameCounts = /* @__PURE__ */ new Map();
+	const register = (baseName) => {
+		const count = captureNameCounts.get(baseName) ?? 0;
+		captureNameCounts.set(baseName, count + 1);
+		const uniqueName = count === 0 ? baseName : `${baseName}_${count + 1}`;
+		const prefixedName = capturePrefix ? `${capturePrefix}${uniqueName}` : uniqueName;
+		captureNames.push(prefixedName);
+		return prefixedName;
+	};
+	return {
+		captureNames,
+		register
+	};
+};
+const expandTokenLiteral = (literal, opts) => {
+	const parsed = parseTokenLiteral(literal);
+	if (!parsed) return literal;
+	const { tokenName, captureName } = parsed;
+	if (!tokenName && captureName) return `(?<${opts.registerCapture(captureName)}>.+)`;
+	let tokenPattern = TOKEN_PATTERNS[tokenName];
+	if (!tokenPattern) return literal;
+	tokenPattern = maybeApplyFuzzyToTokenPattern(tokenPattern, opts.fuzzyTransform);
+	if (captureName) return `(?<${opts.registerCapture(captureName)}>${tokenPattern})`;
+	return tokenPattern;
+};
 /**
 * Expands template tokens with support for named captures.
 *
@@ -1129,65 +1221,19 @@ const containsTokens = (query) => {
 * // → { pattern: 'بَ?ا?بٌ?', captureNames: [], hasCaptures: false }
 */
 const expandTokensWithCaptures = (query, fuzzyTransform, capturePrefix) => {
-	const captureNames = [];
-	const captureNameCounts = /* @__PURE__ */ new Map();
-	/**
-	* Gets a unique capture name, appending _2, _3, etc. for duplicates.
-	* This prevents invalid regex with duplicate named groups.
-	*/
-	const getUniqueCaptureName = (baseName) => {
-		const count = captureNameCounts.get(baseName) ?? 0;
-		captureNameCounts.set(baseName, count + 1);
-		return count === 0 ? baseName : `${baseName}_${count + 1}`;
-	};
-	const segments = [];
-	let lastIndex = 0;
-	TOKEN_WITH_CAPTURE_REGEX.lastIndex = 0;
-	let match;
-	while ((match = TOKEN_WITH_CAPTURE_REGEX.exec(query)) !== null) {
-		if (match.index > lastIndex) segments.push({
-			type: "text",
-			value: query.slice(lastIndex, match.index)
-		});
-		segments.push({
-			type: "token",
-			value: match[0]
-		});
-		lastIndex = match.index + match[0].length;
-	}
-	if (lastIndex < query.length) segments.push({
-		type: "text",
-		value: query.slice(lastIndex)
-	});
+	const segments = splitTemplateIntoSegments(query);
+	const registry = createCaptureRegistry(capturePrefix);
 	const processedParts = segments.map((segment) => {
-		if (segment.type === "text") {
-			if (fuzzyTransform && /[\u0600-\u06FF]/.test(segment.value)) return fuzzyTransform(segment.value);
-			return segment.value;
-		}
-		TOKEN_WITH_CAPTURE_REGEX.lastIndex = 0;
-		const tokenMatch = TOKEN_WITH_CAPTURE_REGEX.exec(segment.value);
-		if (!tokenMatch) return segment.value;
-		const [, tokenName, captureName] = tokenMatch;
-		if (!tokenName && captureName) {
-			const uniqueName = getUniqueCaptureName(captureName);
-			const prefixedName = capturePrefix ? `${capturePrefix}${uniqueName}` : uniqueName;
-			captureNames.push(prefixedName);
-			return `(?<${prefixedName}>.+)`;
-		}
-		let tokenPattern = TOKEN_PATTERNS[tokenName];
-		if (!tokenPattern) return segment.value;
-		if (fuzzyTransform) tokenPattern = tokenPattern.split("|").map((part) => /[\u0600-\u06FF]/.test(part) ? fuzzyTransform(part) : part).join("|");
-		if (captureName) {
-			const uniqueName = getUniqueCaptureName(captureName);
-			const prefixedName = capturePrefix ? `${capturePrefix}${uniqueName}` : uniqueName;
-			captureNames.push(prefixedName);
-			return `(?<${prefixedName}>${tokenPattern})`;
-		}
-		return tokenPattern;
+		if (segment.type === "text") return maybeApplyFuzzyToText(segment.value, fuzzyTransform);
+		return expandTokenLiteral(segment.value, {
+			capturePrefix,
+			fuzzyTransform,
+			registerCapture: registry.register
+		});
 	});
 	return {
-		captureNames,
-		hasCaptures: captureNames.length > 0,
+		captureNames: registry.captureNames,
+		hasCaptures: registry.captureNames.length > 0,
 		pattern: processedParts.join("")
 	};
 };
@@ -2449,5 +2495,5 @@ const analyzeTextForRule = (text) => {
 };
 
 //#endregion
-export { TOKEN_PATTERNS, analyzeCommonLineStarts, analyzeTextForRule, containsTokens, detectTokenPatterns, escapeRegex, escapeTemplateBrackets, expandTokens, expandTokensWithCaptures, generateTemplateFromText, getAvailableTokens, getTokenPattern, makeDiacriticInsensitive, segmentPages, suggestPatternConfig, templateToRegex };
+export { TOKEN_PATTERNS, analyzeCommonLineStarts, analyzeTextForRule, containsTokens, detectTokenPatterns, escapeRegex, escapeTemplateBrackets, expandCompositeTokensInTemplate, expandTokens, expandTokensWithCaptures, generateTemplateFromText, getAvailableTokens, getTokenPattern, makeDiacriticInsensitive, segmentPages, suggestPatternConfig, templateToRegex };
 //# sourceMappingURL=index.mjs.map