@intlayer/engine 9.5.10 → 9.5.12

package/dist/cjs/scan/scanWebsite.cjs

@@ -1,7 +1,9 @@
 Object.defineProperty(exports, Symbol.toStringTag, { value: 'Module' });
+const require_scan_fetchText = require('./fetchText.cjs');
 const require_scan_calculateScore = require('./calculateScore.cjs');
+const require_scan_detection_detectTechnologies = require('./detection/detectTechnologies.cjs');
 const require_scan_parseHtml = require('./parseHtml.cjs');
-const require_scan_checks = require('./checks.cjs');
+const require_scan_runScanChecks = require('./runScanChecks.cjs');
 let _intlayer_config_colors = require("@intlayer/config/colors");
 let _intlayer_config_logger = require("@intlayer/config/logger");
 
@@ -21,6 +23,22 @@ const logDeepScanRecommendation = () => {
 	]);
 };
 /**
+* Read, in the page context, which technology globals exist and the version
+* paths they expose. Serialized by puppeteer: must stay self-contained.
+*/
+const readRuntimeGlobals = (globalNames, versionPaths) => {
+	const pageWindow = window;
+	const globalVersions = {};
+	for (const versionPath of versionPaths) {
+		const value = versionPath.split(".").reduce((current, key) => current && typeof current === "object" ? current[key] : void 0, pageWindow);
+		if (typeof value === "string") globalVersions[versionPath] = value;
+	}
+	return {
+		globals: globalNames.filter((name) => pageWindow[name] !== void 0),
+		globalVersions
+	};
+};
+/**
 * Render the page with a locally installed `puppeteer` to capture
 * client-rendered content, the accurate transfer size, and lazy-loaded chunks.
 *
@@ -56,11 +74,13 @@ const runDeepScan = async (targetUrl, userAgent, timeoutMs) => {
 		await page.setExtraHTTPHeaders({ "Accept-Language": "en-US,en;q=0.9" });
 		const origin = new URL(targetUrl).origin;
 		const jsResponseMap = /* @__PURE__ */ new Map();
+		const requestUrls = [];
 		let totalPageSize = 0;
 		const pendingResponses = [];
 		page.on("response", (response) => {
 			pendingResponses.push((async () => {
 				try {
+					requestUrls.push(response.url());
 					if (response.status() !== 200) return;
 					const buffer = await response.buffer();
 					totalPageSize += buffer.length;
@@ -85,35 +105,27 @@ const runDeepScan = async (targetUrl, userAgent, timeoutMs) => {
 			isMainBundle: mainBundleUrls.has(url),
 			content
 		}));
+		const { globals: globalNames, versionGlobals } = require_scan_detection_detectTechnologies.getTechnologyGlobalNames();
+		const { globals, globalVersions } = await page.evaluate(readRuntimeGlobals, globalNames, versionGlobals).catch(() => ({
+			globals: [],
+			globalVersions: {}
+		}));
+		const cookies = await page.cookies().catch(() => []);
 		return {
 			html,
 			totalPageSize,
-			chunks
+			chunks,
+			requestUrls,
+			runtimeSignals: {
+				globals,
+				globalVersions,
+				storageKeys: cookies.map(({ name }) => name)
+			}
 		};
 	} finally {
 		if (browser) await browser.close();
 	}
 };
-/** Fetch the raw HTML document, measuring its byte size. */
-const fetchHtml = async (url, userAgent, timeoutMs) => {
-	const controller = new AbortController();
-	const timer = setTimeout(() => controller.abort(), timeoutMs);
-	try {
-		const response = await fetch(url, {
-			headers: {
-				"User-Agent": userAgent,
-				"Accept-Language": "en-US,en;q=0.9"
-			},
-			signal: controller.signal
-		});
-		return {
-			html: await response.text(),
-			finalUrl: response.url || url
-		};
-	} finally {
-		clearTimeout(timer);
-	}
-};
 /**
 * Fetch every eagerly-loaded script. Same-origin scripts keep their content so
 * their locale weight can be analyzed; third-party scripts only contribute to
@@ -154,49 +166,43 @@ const fetchScripts = async (scriptUrls, origin, userAgent) => {
 const scanWebsite = async (targetUrl, options = {}) => {
 	const { deep = true, timeoutMs = DEFAULT_TIMEOUT_MS, userAgent = DEFAULT_USER_AGENT } = options;
 	const origin = new URL(targetUrl).origin;
-	let mode = "basic";
-	let html;
-	let totalPageSize;
-	let chunks;
+	const fetchOptions = {
+		userAgent,
+		timeoutMs
+	};
 	const deepResult = deep ? await runDeepScan(targetUrl, userAgent, timeoutMs) : null;
-	if (deepResult) {
-		mode = "deep";
-		html = deepResult.html;
-		chunks = deepResult.chunks;
-		totalPageSize = deepResult.totalPageSize;
-	} else {
+	let pageInput;
+	if (deepResult) pageInput = deepResult;
+	else {
 		if (deep) logDeepScanRecommendation();
-		const { html: fetchedHtml, finalUrl } = await fetchHtml(targetUrl, userAgent, timeoutMs);
-		html = fetchedHtml;
-		const scriptUrls = require_scan_parseHtml.extractScriptUrls(fetchedHtml, finalUrl);
-		const { chunks: fetchedChunks, scriptBytes } = await fetchScripts(scriptUrls, origin, userAgent);
-		chunks = fetchedChunks;
-		totalPageSize = require_scan_parseHtml.byteLength(fetchedHtml) + scriptBytes;
+		const response = await require_scan_fetchText.fetchText(targetUrl, fetchOptions);
+		if (!response?.ok) throw new Error(`Failed to fetch ${targetUrl}${response ? ` (HTTP ${response.status})` : ""}`);
+		const scriptUrls = require_scan_parseHtml.extractScriptUrls(response.text, response.finalUrl);
+		const { chunks, scriptBytes } = await fetchScripts(scriptUrls, origin, userAgent);
+		pageInput = {
+			html: response.text,
+			chunks,
+			totalPageSize: require_scan_parseHtml.byteLength(response.text) + scriptBytes
+		};
 	}
-	const htmlSize = require_scan_parseHtml.byteLength(html);
 	const events = [];
-	const localesSet = /* @__PURE__ */ new Set();
-	const { langTag } = require_scan_checks.checkHtmlAttributes(html, targetUrl, events);
-	require_scan_checks.checkCanonical(html, targetUrl, events);
-	require_scan_checks.checkLinguisticStructure(html, targetUrl, localesSet, events);
-	require_scan_checks.checkUrlStructure(require_scan_parseHtml.extractAnchors(html), origin, targetUrl, events);
-	const bundle = require_scan_checks.checkBundleContent(chunks, html, langTag, targetUrl, totalPageSize, events);
-	await require_scan_checks.checkRobots(origin, localesSet, userAgent, events);
-	await require_scan_checks.checkSitemap(origin, localesSet, userAgent, events);
+	const checksResult = await require_scan_runScanChecks.runScanChecks({
+		targetUrl,
+		...pageInput
+	}, (event) => events.push(event), fetchOptions);
 	const rawScore = events.reduce((score, event) => require_scan_calculateScore.mutateScore(score, event), {
 		score: 0,
 		totalScore: 0
 	});
 	return {
 		url: targetUrl,
-		mode,
-		totalPageSize,
-		htmlSize,
+		mode: deepResult ? "deep" : "basic",
+		totalPageSize: pageInput.totalPageSize,
+		htmlSize: require_scan_parseHtml.byteLength(pageInput.html),
 		score: require_scan_calculateScore.toScorePercent(rawScore),
 		rawScore,
 		events,
-		locales: Array.from(localesSet),
-		bundle
+		...checksResult
 	};
 };
 

package/dist/cjs/scan/scanWebsite.cjs.map

@@ -1 +1 @@
-{"version":3,"file":"scanWebsite.cjs","names":["colorize","GREY","GREY_LIGHT","extractScriptUrls","byteLength","checkHtmlAttributes","extractAnchors","checkBundleContent","checkRobots","checkSitemap","mutateScore","toScorePercent"],"sources":["../../../src/scan/scanWebsite.ts"],"sourcesContent":["import { GREY, GREY_LIGHT } from '@intlayer/config/colors';\nimport { colorize, logger } from '@intlayer/config/logger';\nimport { mutateScore, type Score, toScorePercent } from './calculateScore';\nimport {\n  checkBundleContent,\n  checkCanonical,\n  checkHtmlAttributes,\n  checkLinguisticStructure,\n  checkRobots,\n  checkSitemap,\n  checkUrlStructure,\n} from './checks';\nimport { byteLength, extractAnchors, extractScriptUrls } from './parseHtml';\nimport type {\n  BundleChunkInput,\n  ScanEvent,\n  ScanOptions,\n  ScanResult,\n} from './types';\n\nconst DEFAULT_USER_AGENT =\n  'Mozilla/5.0 (compatible; IntlayerScanBot/1.0; +https://intlayer.org)';\nconst DEFAULT_TIMEOUT_MS = 30_000;\n\n/**\n * Log a recommendation to install `puppeteer` for a deeper scan. Mirrors the\n * style used by other optional-dependency hints across the CLI.\n */\nconst logDeepScanRecommendation = (): void => {\n  logger([\n    colorize('Recommended: Install', GREY),\n    colorize('puppeteer', GREY_LIGHT),\n    colorize(\n      'package to enable a deeper scan (renders client-side content & lazy-loaded chunks). See documentation:',\n      GREY\n    ),\n    colorize('https://intlayer.org/doc/concept/cli#scan', GREY_LIGHT),\n  ]);\n};\n\n/** Outcome of a render-based deep scan. */\ntype DeepScanResult = {\n  html: string;\n  totalPageSize: number;\n  chunks: BundleChunkInput[];\n};\n\n/**\n * Render the page with a locally installed `puppeteer` to capture\n * client-rendered content, the accurate transfer size, and lazy-loaded chunks.\n *\n * `puppeteer` is imported dynamically through a non-literal specifier so it is\n * never bundled and stays an optional dependency: when it is absent the import\n * rejects and the caller falls back to the basic scan.\n *\n * @returns The deep-scan result, or `null` when `puppeteer` is unavailable.\n */\nconst runDeepScan = async (\n  targetUrl: string,\n  userAgent: string,\n  timeoutMs: number\n): Promise<DeepScanResult | null> => {\n  const moduleName = 'puppeteer';\n  // Optional dependency resolved at runtime — typed as `any` because it is not\n  // a declared dependency of this package.\n  let puppeteer: any;\n  try {\n    const mod: any = await import(moduleName);\n    puppeteer = mod.default ?? mod;\n  } catch {\n    return null;\n  }\n\n  let browser: any;\n  try {\n    browser = await puppeteer.launch({\n      headless: true,\n      executablePath: process.env.PUPPETEER_EXECUTABLE_PATH,\n      args: [\n        '--no-sandbox',\n        '--disable-setuid-sandbox',\n        '--disable-dev-shm-usage',\n        '--disable-gpu',\n      ],\n    });\n\n    const page = await browser.newPage();\n    await page.setUserAgent(userAgent);\n    await page.setExtraHTTPHeaders({ 'Accept-Language': 'en-US,en;q=0.9' });\n\n    const origin = new URL(targetUrl).origin;\n    const jsResponseMap = new Map<string, string>();\n    let totalPageSize = 0;\n    const pendingResponses: Promise<void>[] = [];\n\n    page.on('response', (response: any) => {\n      pendingResponses.push(\n        (async () => {\n          try {\n            if (response.status() !== 200) return;\n            const buffer = await response.buffer();\n            totalPageSize += buffer.length;\n\n            const responseUrl: string = response.url();\n            const contentType: string =\n              response.headers()['content-type'] ?? '';\n            const isJavaScript =\n              contentType.includes('javascript') ||\n              /\\.(js|mjs|cjs)(\\?|$)/.test(responseUrl);\n            if (isJavaScript && responseUrl.startsWith(origin)) {\n              jsResponseMap.set(responseUrl, buffer.toString('utf-8'));\n            }\n          } catch {\n            /* response already consumed or aborted */\n          }\n        })()\n      );\n    });\n\n    await page.goto(targetUrl, {\n      waitUntil: 'domcontentloaded',\n      timeout: timeoutMs,\n    });\n    await page\n      .waitForNetworkIdle({ idleTime: 1000, timeout: 10_000 })\n      .catch(() => {\n        /* ok if it never fully idles */\n      });\n    await Promise.allSettled(pendingResponses);\n\n    const html: string = await page.content();\n    const mainBundleUrls = new Set(extractScriptUrls(html, targetUrl));\n    const chunks: BundleChunkInput[] = Array.from(jsResponseMap.entries()).map(\n      ([url, content]) => ({\n        url,\n        isMainBundle: mainBundleUrls.has(url),\n        content,\n      })\n    );\n\n    return { html, totalPageSize, chunks };\n  } finally {\n    if (browser) await browser.close();\n  }\n};\n\n/** Fetch the raw HTML document, measuring its byte size. */\nconst fetchHtml = async (\n  url: string,\n  userAgent: string,\n  timeoutMs: number\n): Promise<{ html: string; finalUrl: string }> => {\n  const controller = new AbortController();\n  const timer = setTimeout(() => controller.abort(), timeoutMs);\n  try {\n    const response = await fetch(url, {\n      headers: { 'User-Agent': userAgent, 'Accept-Language': 'en-US,en;q=0.9' },\n      signal: controller.signal,\n    });\n    const html = await response.text();\n    return { html, finalUrl: response.url || url };\n  } finally {\n    clearTimeout(timer);\n  }\n};\n\n/**\n * Fetch every eagerly-loaded script. Same-origin scripts keep their content so\n * their locale weight can be analyzed; third-party scripts only contribute to\n * the measured page size (their locale-like keys cause false positives).\n */\nconst fetchScripts = async (\n  scriptUrls: string[],\n  origin: string,\n  userAgent: string\n): Promise<{ chunks: BundleChunkInput[]; scriptBytes: number }> => {\n  const chunks: BundleChunkInput[] = [];\n  let scriptBytes = 0;\n\n  await Promise.all(\n    scriptUrls.map(async (scriptUrl) => {\n      try {\n        const response = await fetch(scriptUrl, {\n          headers: { 'User-Agent': userAgent },\n        });\n        if (!response.ok) return;\n        const content = await response.text();\n        scriptBytes += byteLength(content);\n        if (scriptUrl.startsWith(origin)) {\n          chunks.push({ url: scriptUrl, isMainBundle: true, content });\n        }\n      } catch {\n        /* ignore unreachable scripts */\n      }\n    })\n  );\n\n  return { chunks, scriptBytes };\n};\n\n/**\n * Scan a single web page for i18n/SEO health and bundle weight.\n *\n * In `deep` mode (default) the page is rendered with a locally installed\n * `puppeteer`; when `puppeteer` is missing the scan transparently falls back to\n * a `basic` fetch-based pass and logs a recommendation to install it.\n *\n * @param targetUrl - The absolute URL to scan.\n * @param options - {@link ScanOptions} controlling depth, timeout and UA.\n * @returns The {@link ScanResult} including score, page size and per-check events.\n */\nexport const scanWebsite = async (\n  targetUrl: string,\n  options: ScanOptions = {}\n): Promise<ScanResult> => {\n  const {\n    deep = true,\n    timeoutMs = DEFAULT_TIMEOUT_MS,\n    userAgent = DEFAULT_USER_AGENT,\n  } = options;\n\n  const origin = new URL(targetUrl).origin;\n\n  let mode: ScanResult['mode'] = 'basic';\n  let html: string;\n  let totalPageSize: number;\n  let chunks: BundleChunkInput[];\n\n  const deepResult = deep\n    ? await runDeepScan(targetUrl, userAgent, timeoutMs)\n    : null;\n\n  if (deepResult) {\n    mode = 'deep';\n    html = deepResult.html;\n    chunks = deepResult.chunks;\n    totalPageSize = deepResult.totalPageSize;\n  } else {\n    if (deep) logDeepScanRecommendation();\n    const { html: fetchedHtml, finalUrl } = await fetchHtml(\n      targetUrl,\n      userAgent,\n      timeoutMs\n    );\n    html = fetchedHtml;\n    const scriptUrls = extractScriptUrls(fetchedHtml, finalUrl);\n    const { chunks: fetchedChunks, scriptBytes } = await fetchScripts(\n      scriptUrls,\n      origin,\n      userAgent\n    );\n    chunks = fetchedChunks;\n    totalPageSize = byteLength(fetchedHtml) + scriptBytes;\n  }\n\n  const htmlSize = byteLength(html);\n\n  const events: ScanEvent[] = [];\n  const localesSet = new Set<string>();\n\n  const { langTag } = checkHtmlAttributes(html, targetUrl, events);\n  checkCanonical(html, targetUrl, events);\n  checkLinguisticStructure(html, targetUrl, localesSet, events);\n  checkUrlStructure(extractAnchors(html), origin, targetUrl, events);\n  const bundle = checkBundleContent(\n    chunks,\n    html,\n    langTag,\n    targetUrl,\n    totalPageSize,\n    events\n  );\n  await checkRobots(origin, localesSet, userAgent, events);\n  await checkSitemap(origin, localesSet, userAgent, events);\n\n  const rawScore = events.reduce<Score>(\n    (score, event) => mutateScore(score, event),\n    { score: 0, totalScore: 0 }\n  );\n\n  return {\n    url: targetUrl,\n    mode,\n    totalPageSize,\n    htmlSize,\n    score: toScorePercent(rawScore),\n    rawScore,\n    events,\n    locales: Array.from(localesSet),\n    bundle,\n  };\n};\n"],"mappings":";;;;;;;;AAoBA,MAAM,qBACJ;AACF,MAAM,qBAAqB;;;;;AAM3B,MAAM,kCAAwC;CAC5C,oCAAO;MACLA,kCAAS,wBAAwBC,4BAAI;MACrCD,kCAAS,aAAaE,kCAAU;MAChCF,kCACE,0GACAC,4BACF;MACAD,kCAAS,6CAA6CE,kCAAU;CAClE,CAAC;AACH;;;;;;;;;;;AAmBA,MAAM,cAAc,OAClB,WACA,WACA,cACmC;CACnC,MAAM,aAAa;CAGnB,IAAI;CACJ,IAAI;EACF,MAAM,MAAW,MAAM,OAAO;EAC9B,YAAY,IAAI,WAAW;CAC7B,QAAQ;EACN,OAAO;CACT;CAEA,IAAI;CACJ,IAAI;EACF,UAAU,MAAM,UAAU,OAAO;GAC/B,UAAU;GACV,gBAAgB,QAAQ,IAAI;GAC5B,MAAM;IACJ;IACA;IACA;IACA;GACF;EACF,CAAC;EAED,MAAM,OAAO,MAAM,QAAQ,QAAQ;EACnC,MAAM,KAAK,aAAa,SAAS;EACjC,MAAM,KAAK,oBAAoB,EAAE,mBAAmB,iBAAiB,CAAC;EAEtE,MAAM,SAAS,IAAI,IAAI,SAAS,CAAC,CAAC;EAClC,MAAM,gCAAgB,IAAI,IAAoB;EAC9C,IAAI,gBAAgB;EACpB,MAAM,mBAAoC,CAAC;EAE3C,KAAK,GAAG,aAAa,aAAkB;GACrC,iBAAiB,MACd,YAAY;IACX,IAAI;KACF,IAAI,SAAS,OAAO,MAAM,KAAK;KAC/B,MAAM,SAAS,MAAM,SAAS,OAAO;KACrC,iBAAiB,OAAO;KAExB,MAAM,cAAsB,SAAS,IAAI;KAMzC,MAJE,SAAS,QAAQ,CAAC,CAAC,mBAAmB,GAE3B,CAAC,SAAS,YAAY,KACjC,uBAAuB,KAAK,WAAW,MACrB,YAAY,WAAW,MAAM,GAC/C,cAAc,IAAI,aAAa,OAAO,SAAS,OAAO,CAAC;IAE3D,QAAQ,CAER;GACF,EAAC,CAAE,CACL;EACF,CAAC;EAED,MAAM,KAAK,KAAK,WAAW;GACzB,WAAW;GACX,SAAS;EACX,CAAC;EACD,MAAM,KACH,mBAAmB;GAAE,UAAU;GAAM,SAAS;EAAO,CAAC,CAAC,CACvD,YAAY,CAEb,CAAC;EACH,MAAM,QAAQ,WAAW,gBAAgB;EAEzC,MAAM,OAAe,MAAM,KAAK,QAAQ;EACxC,MAAM,iBAAiB,IAAI,IAAIC,yCAAkB,MAAM,SAAS,CAAC;EACjE,MAAM,SAA6B,MAAM,KAAK,cAAc,QAAQ,CAAC,CAAC,CAAC,KACpE,CAAC,KAAK,cAAc;GACnB;GACA,cAAc,eAAe,IAAI,GAAG;GACpC;EACF,EACF;EAEA,OAAO;GAAE;GAAM;GAAe;EAAO;CACvC,UAAU;EACR,IAAI,SAAS,MAAM,QAAQ,MAAM;CACnC;AACF;;AAGA,MAAM,YAAY,OAChB,KACA,WACA,cACgD;CAChD,MAAM,aAAa,IAAI,gBAAgB;CACvC,MAAM,QAAQ,iBAAiB,WAAW,MAAM,GAAG,SAAS;CAC5D,IAAI;EACF,MAAM,WAAW,MAAM,MAAM,KAAK;GAChC,SAAS;IAAE,cAAc;IAAW,mBAAmB;GAAiB;GACxE,QAAQ,WAAW;EACrB,CAAC;EAED,OAAO;GAAE,YADU,SAAS,KAAK;GAClB,UAAU,SAAS,OAAO;EAAI;CAC/C,UAAU;EACR,aAAa,KAAK;CACpB;AACF;;;;;;AAOA,MAAM,eAAe,OACnB,YACA,QACA,cACiE;CACjE,MAAM,SAA6B,CAAC;CACpC,IAAI,cAAc;CAElB,MAAM,QAAQ,IACZ,WAAW,IAAI,OAAO,cAAc;EAClC,IAAI;GACF,MAAM,WAAW,MAAM,MAAM,WAAW,EACtC,SAAS,EAAE,cAAc,UAAU,EACrC,CAAC;GACD,IAAI,CAAC,SAAS,IAAI;GAClB,MAAM,UAAU,MAAM,SAAS,KAAK;GACpC,eAAeC,kCAAW,OAAO;GACjC,IAAI,UAAU,WAAW,MAAM,GAC7B,OAAO,KAAK;IAAE,KAAK;IAAW,cAAc;IAAM;GAAQ,CAAC;EAE/D,QAAQ,CAER;CACF,CAAC,CACH;CAEA,OAAO;EAAE;EAAQ;CAAY;AAC/B;;;;;;;;;;;;AAaA,MAAa,cAAc,OACzB,WACA,UAAuB,CAAC,MACA;CACxB,MAAM,EACJ,OAAO,MACP,YAAY,oBACZ,YAAY,uBACV;CAEJ,MAAM,SAAS,IAAI,IAAI,SAAS,CAAC,CAAC;CAElC,IAAI,OAA2B;CAC/B,IAAI;CACJ,IAAI;CACJ,IAAI;CAEJ,MAAM,aAAa,OACf,MAAM,YAAY,WAAW,WAAW,SAAS,IACjD;CAEJ,IAAI,YAAY;EACd,OAAO;EACP,OAAO,WAAW;EAClB,SAAS,WAAW;EACpB,gBAAgB,WAAW;CAC7B,OAAO;EACL,IAAI,MAAM,0BAA0B;EACpC,MAAM,EAAE,MAAM,aAAa,aAAa,MAAM,UAC5C,WACA,WACA,SACF;EACA,OAAO;EACP,MAAM,aAAaD,yCAAkB,aAAa,QAAQ;EAC1D,MAAM,EAAE,QAAQ,eAAe,gBAAgB,MAAM,aACnD,YACA,QACA,SACF;EACA,SAAS;EACT,gBAAgBC,kCAAW,WAAW,IAAI;CAC5C;CAEA,MAAM,WAAWA,kCAAW,IAAI;CAEhC,MAAM,SAAsB,CAAC;CAC7B,MAAM,6BAAa,IAAI,IAAY;CAEnC,MAAM,EAAE,YAAYC,wCAAoB,MAAM,WAAW,MAAM;CAC/D,mCAAe,MAAM,WAAW,MAAM;CACtC,6CAAyB,MAAM,WAAW,YAAY,MAAM;CAC5D,sCAAkBC,sCAAe,IAAI,GAAG,QAAQ,WAAW,MAAM;CACjE,MAAM,SAASC,uCACb,QACA,MACA,SACA,WACA,eACA,MACF;CACA,MAAMC,gCAAY,QAAQ,YAAY,WAAW,MAAM;CACvD,MAAMC,iCAAa,QAAQ,YAAY,WAAW,MAAM;CAExD,MAAM,WAAW,OAAO,QACrB,OAAO,UAAUC,wCAAY,OAAO,KAAK,GAC1C;EAAE,OAAO;EAAG,YAAY;CAAE,CAC5B;CAEA,OAAO;EACL,KAAK;EACL;EACA;EACA;EACA,OAAOC,2CAAe,QAAQ;EAC9B;EACA;EACA,SAAS,MAAM,KAAK,UAAU;EAC9B;CACF;AACF"}
+{"version":3,"file":"scanWebsite.cjs","names":["colorize","GREY","GREY_LIGHT","extractScriptUrls","getTechnologyGlobalNames","byteLength","fetchText","runScanChecks","mutateScore","toScorePercent"],"sources":["../../../src/scan/scanWebsite.ts"],"sourcesContent":["import { GREY, GREY_LIGHT } from '@intlayer/config/colors';\nimport { colorize, logger } from '@intlayer/config/logger';\nimport { mutateScore, type Score, toScorePercent } from './calculateScore';\nimport { getTechnologyGlobalNames } from './detection/detectTechnologies';\nimport { fetchText } from './fetchText';\nimport { byteLength, extractScriptUrls } from './parseHtml';\nimport { runScanChecks, type ScanPageInput } from './runScanChecks';\nimport type {\n  BundleChunkInput,\n  ScanEvent,\n  ScanOptions,\n  ScanResult,\n} from './types';\n\nconst DEFAULT_USER_AGENT =\n  'Mozilla/5.0 (compatible; IntlayerScanBot/1.0; +https://intlayer.org)';\nconst DEFAULT_TIMEOUT_MS = 30_000;\n\n/**\n * Log a recommendation to install `puppeteer` for a deeper scan. Mirrors the\n * style used by other optional-dependency hints across the CLI.\n */\nconst logDeepScanRecommendation = (): void => {\n  logger([\n    colorize('Recommended: Install', GREY),\n    colorize('puppeteer', GREY_LIGHT),\n    colorize(\n      'package to enable a deeper scan (renders client-side content & lazy-loaded chunks). See documentation:',\n      GREY\n    ),\n    colorize('https://intlayer.org/doc/concept/cli#scan', GREY_LIGHT),\n  ]);\n};\n\n/** Outcome of a render-based deep scan. */\ntype DeepScanResult = Required<Omit<ScanPageInput, 'targetUrl'>>;\n\n/**\n * Read, in the page context, which technology globals exist and the version\n * paths they expose. Serialized by puppeteer: must stay self-contained.\n */\nconst readRuntimeGlobals = (\n  globalNames: string[],\n  versionPaths: string[]\n): { globals: string[]; globalVersions: Record<string, string> } => {\n  const pageWindow = window as unknown as Record<string, unknown>;\n  const globalVersions: Record<string, string> = {};\n  for (const versionPath of versionPaths) {\n    const value = versionPath\n      .split('.')\n      .reduce<unknown>(\n        (current, key) =>\n          current && typeof current === 'object'\n            ? (current as Record<string, unknown>)[key]\n            : undefined,\n        pageWindow\n      );\n    if (typeof value === 'string') globalVersions[versionPath] = value;\n  }\n  return {\n    globals: globalNames.filter((name) => pageWindow[name] !== undefined),\n    globalVersions,\n  };\n};\n\n/**\n * Render the page with a locally installed `puppeteer` to capture\n * client-rendered content, the accurate transfer size, and lazy-loaded chunks.\n *\n * `puppeteer` is imported dynamically through a non-literal specifier so it is\n * never bundled and stays an optional dependency: when it is absent the import\n * rejects and the caller falls back to the basic scan.\n *\n * @returns The deep-scan result, or `null` when `puppeteer` is unavailable.\n */\nconst runDeepScan = async (\n  targetUrl: string,\n  userAgent: string,\n  timeoutMs: number\n): Promise<DeepScanResult | null> => {\n  const moduleName = 'puppeteer';\n  // Optional dependency resolved at runtime — typed as `any` because it is not\n  // a declared dependency of this package.\n  let puppeteer: any;\n  try {\n    const mod: any = await import(moduleName);\n    puppeteer = mod.default ?? mod;\n  } catch {\n    return null;\n  }\n\n  let browser: any;\n  try {\n    browser = await puppeteer.launch({\n      headless: true,\n      executablePath: process.env.PUPPETEER_EXECUTABLE_PATH,\n      args: [\n        '--no-sandbox',\n        '--disable-setuid-sandbox',\n        '--disable-dev-shm-usage',\n        '--disable-gpu',\n      ],\n    });\n\n    const page = await browser.newPage();\n    await page.setUserAgent(userAgent);\n    await page.setExtraHTTPHeaders({ 'Accept-Language': 'en-US,en;q=0.9' });\n\n    const origin = new URL(targetUrl).origin;\n    const jsResponseMap = new Map<string, string>();\n    const requestUrls: string[] = [];\n    let totalPageSize = 0;\n    const pendingResponses: Promise<void>[] = [];\n\n    page.on('response', (response: any) => {\n      pendingResponses.push(\n        (async () => {\n          try {\n            requestUrls.push(response.url());\n            if (response.status() !== 200) return;\n            const buffer = await response.buffer();\n            totalPageSize += buffer.length;\n\n            const responseUrl: string = response.url();\n            const contentType: string =\n              response.headers()['content-type'] ?? '';\n            const isJavaScript =\n              contentType.includes('javascript') ||\n              /\\.(js|mjs|cjs)(\\?|$)/.test(responseUrl);\n            if (isJavaScript && responseUrl.startsWith(origin)) {\n              jsResponseMap.set(responseUrl, buffer.toString('utf-8'));\n            }\n          } catch {\n            /* response already consumed or aborted */\n          }\n        })()\n      );\n    });\n\n    await page.goto(targetUrl, {\n      waitUntil: 'domcontentloaded',\n      timeout: timeoutMs,\n    });\n    await page\n      .waitForNetworkIdle({ idleTime: 1000, timeout: 10_000 })\n      .catch(() => {\n        /* ok if it never fully idles */\n      });\n    await Promise.allSettled(pendingResponses);\n\n    const html: string = await page.content();\n    const mainBundleUrls = new Set(extractScriptUrls(html, targetUrl));\n    const chunks: BundleChunkInput[] = Array.from(jsResponseMap.entries()).map(\n      ([url, content]) => ({\n        url,\n        isMainBundle: mainBundleUrls.has(url),\n        content,\n      })\n    );\n\n    const { globals: globalNames, versionGlobals } = getTechnologyGlobalNames();\n    const { globals, globalVersions } = await page\n      .evaluate(readRuntimeGlobals, globalNames, versionGlobals)\n      .catch(() => ({ globals: [], globalVersions: {} }));\n    const cookies: { name: string }[] = await page.cookies().catch(() => []);\n\n    return {\n      html,\n      totalPageSize,\n      chunks,\n      requestUrls,\n      runtimeSignals: {\n        globals,\n        globalVersions,\n        storageKeys: cookies.map(({ name }) => name),\n      },\n    };\n  } finally {\n    if (browser) await browser.close();\n  }\n};\n\n/**\n * Fetch every eagerly-loaded script. Same-origin scripts keep their content so\n * their locale weight can be analyzed; third-party scripts only contribute to\n * the measured page size (their locale-like keys cause false positives).\n */\nconst fetchScripts = async (\n  scriptUrls: string[],\n  origin: string,\n  userAgent: string\n): Promise<{ chunks: BundleChunkInput[]; scriptBytes: number }> => {\n  const chunks: BundleChunkInput[] = [];\n  let scriptBytes = 0;\n\n  await Promise.all(\n    scriptUrls.map(async (scriptUrl) => {\n      try {\n        const response = await fetch(scriptUrl, {\n          headers: { 'User-Agent': userAgent },\n        });\n        if (!response.ok) return;\n        const content = await response.text();\n        scriptBytes += byteLength(content);\n        if (scriptUrl.startsWith(origin)) {\n          chunks.push({ url: scriptUrl, isMainBundle: true, content });\n        }\n      } catch {\n        /* ignore unreachable scripts */\n      }\n    })\n  );\n\n  return { chunks, scriptBytes };\n};\n\n/**\n * Scan a single web page for i18n/SEO health and bundle weight.\n *\n * In `deep` mode (default) the page is rendered with a locally installed\n * `puppeteer`; when `puppeteer` is missing the scan transparently falls back to\n * a `basic` fetch-based pass and logs a recommendation to install it.\n *\n * @param targetUrl - The absolute URL to scan.\n * @param options - {@link ScanOptions} controlling depth, timeout and UA.\n * @returns The {@link ScanResult} including score, page size and per-check events.\n */\nexport const scanWebsite = async (\n  targetUrl: string,\n  options: ScanOptions = {}\n): Promise<ScanResult> => {\n  const {\n    deep = true,\n    timeoutMs = DEFAULT_TIMEOUT_MS,\n    userAgent = DEFAULT_USER_AGENT,\n  } = options;\n\n  const origin = new URL(targetUrl).origin;\n  const fetchOptions = { userAgent, timeoutMs };\n\n  const deepResult = deep\n    ? await runDeepScan(targetUrl, userAgent, timeoutMs)\n    : null;\n\n  let pageInput: Omit<ScanPageInput, 'targetUrl'>;\n\n  if (deepResult) {\n    pageInput = deepResult;\n  } else {\n    if (deep) logDeepScanRecommendation();\n    const response = await fetchText(targetUrl, fetchOptions);\n    if (!response?.ok) {\n      throw new Error(\n        `Failed to fetch ${targetUrl}${response ? ` (HTTP ${response.status})` : ''}`\n      );\n    }\n    const scriptUrls = extractScriptUrls(response.text, response.finalUrl);\n    const { chunks, scriptBytes } = await fetchScripts(\n      scriptUrls,\n      origin,\n      userAgent\n    );\n    pageInput = {\n      html: response.text,\n      chunks,\n      totalPageSize: byteLength(response.text) + scriptBytes,\n    };\n  }\n\n  const events: ScanEvent[] = [];\n  const checksResult = await runScanChecks(\n    { targetUrl, ...pageInput },\n    (event) => events.push(event),\n    fetchOptions\n  );\n\n  const rawScore = events.reduce<Score>(\n    (score, event) => mutateScore(score, event),\n    { score: 0, totalScore: 0 }\n  );\n\n  return {\n    url: targetUrl,\n    mode: deepResult ? 'deep' : 'basic',\n    totalPageSize: pageInput.totalPageSize,\n    htmlSize: byteLength(pageInput.html),\n    score: toScorePercent(rawScore),\n    rawScore,\n    events,\n    ...checksResult,\n  };\n};\n"],"mappings":";;;;;;;;;;AAcA,MAAM,qBACJ;AACF,MAAM,qBAAqB;;;;;AAM3B,MAAM,kCAAwC;CAC5C,oCAAO;MACLA,kCAAS,wBAAwBC,4BAAI;MACrCD,kCAAS,aAAaE,kCAAU;MAChCF,kCACE,0GACAC,4BACF;MACAD,kCAAS,6CAA6CE,kCAAU;CAClE,CAAC;AACH;;;;;AASA,MAAM,sBACJ,aACA,iBACkE;CAClE,MAAM,aAAa;CACnB,MAAM,iBAAyC,CAAC;CAChD,KAAK,MAAM,eAAe,cAAc;EACtC,MAAM,QAAQ,YACX,MAAM,GAAG,CAAC,CACV,QACE,SAAS,QACR,WAAW,OAAO,YAAY,WACzB,QAAoC,OACrC,QACN,UACF;EACF,IAAI,OAAO,UAAU,UAAU,eAAe,eAAe;CAC/D;CACA,OAAO;EACL,SAAS,YAAY,QAAQ,SAAS,WAAW,UAAU,MAAS;EACpE;CACF;AACF;;;;;;;;;;;AAYA,MAAM,cAAc,OAClB,WACA,WACA,cACmC;CACnC,MAAM,aAAa;CAGnB,IAAI;CACJ,IAAI;EACF,MAAM,MAAW,MAAM,OAAO;EAC9B,YAAY,IAAI,WAAW;CAC7B,QAAQ;EACN,OAAO;CACT;CAEA,IAAI;CACJ,IAAI;EACF,UAAU,MAAM,UAAU,OAAO;GAC/B,UAAU;GACV,gBAAgB,QAAQ,IAAI;GAC5B,MAAM;IACJ;IACA;IACA;IACA;GACF;EACF,CAAC;EAED,MAAM,OAAO,MAAM,QAAQ,QAAQ;EACnC,MAAM,KAAK,aAAa,SAAS;EACjC,MAAM,KAAK,oBAAoB,EAAE,mBAAmB,iBAAiB,CAAC;EAEtE,MAAM,SAAS,IAAI,IAAI,SAAS,CAAC,CAAC;EAClC,MAAM,gCAAgB,IAAI,IAAoB;EAC9C,MAAM,cAAwB,CAAC;EAC/B,IAAI,gBAAgB;EACpB,MAAM,mBAAoC,CAAC;EAE3C,KAAK,GAAG,aAAa,aAAkB;GACrC,iBAAiB,MACd,YAAY;IACX,IAAI;KACF,YAAY,KAAK,SAAS,IAAI,CAAC;KAC/B,IAAI,SAAS,OAAO,MAAM,KAAK;KAC/B,MAAM,SAAS,MAAM,SAAS,OAAO;KACrC,iBAAiB,OAAO;KAExB,MAAM,cAAsB,SAAS,IAAI;KAMzC,MAJE,SAAS,QAAQ,CAAC,CAAC,mBAAmB,GAE3B,CAAC,SAAS,YAAY,KACjC,uBAAuB,KAAK,WAAW,MACrB,YAAY,WAAW,MAAM,GAC/C,cAAc,IAAI,aAAa,OAAO,SAAS,OAAO,CAAC;IAE3D,QAAQ,CAER;GACF,EAAC,CAAE,CACL;EACF,CAAC;EAED,MAAM,KAAK,KAAK,WAAW;GACzB,WAAW;GACX,SAAS;EACX,CAAC;EACD,MAAM,KACH,mBAAmB;GAAE,UAAU;GAAM,SAAS;EAAO,CAAC,CAAC,CACvD,YAAY,CAEb,CAAC;EACH,MAAM,QAAQ,WAAW,gBAAgB;EAEzC,MAAM,OAAe,MAAM,KAAK,QAAQ;EACxC,MAAM,iBAAiB,IAAI,IAAIC,yCAAkB,MAAM,SAAS,CAAC;EACjE,MAAM,SAA6B,MAAM,KAAK,cAAc,QAAQ,CAAC,CAAC,CAAC,KACpE,CAAC,KAAK,cAAc;GACnB;GACA,cAAc,eAAe,IAAI,GAAG;GACpC;EACF,EACF;EAEA,MAAM,EAAE,SAAS,aAAa,mBAAmBC,mEAAyB;EAC1E,MAAM,EAAE,SAAS,mBAAmB,MAAM,KACvC,SAAS,oBAAoB,aAAa,cAAc,CAAC,CACzD,aAAa;GAAE,SAAS,CAAC;GAAG,gBAAgB,CAAC;EAAE,EAAE;EACpD,MAAM,UAA8B,MAAM,KAAK,QAAQ,CAAC,CAAC,YAAY,CAAC,CAAC;EAEvE,OAAO;GACL;GACA;GACA;GACA;GACA,gBAAgB;IACd;IACA;IACA,aAAa,QAAQ,KAAK,EAAE,WAAW,IAAI;GAC7C;EACF;CACF,UAAU;EACR,IAAI,SAAS,MAAM,QAAQ,MAAM;CACnC;AACF;;;;;;AAOA,MAAM,eAAe,OACnB,YACA,QACA,cACiE;CACjE,MAAM,SAA6B,CAAC;CACpC,IAAI,cAAc;CAElB,MAAM,QAAQ,IACZ,WAAW,IAAI,OAAO,cAAc;EAClC,IAAI;GACF,MAAM,WAAW,MAAM,MAAM,WAAW,EACtC,SAAS,EAAE,cAAc,UAAU,EACrC,CAAC;GACD,IAAI,CAAC,SAAS,IAAI;GAClB,MAAM,UAAU,MAAM,SAAS,KAAK;GACpC,eAAeC,kCAAW,OAAO;GACjC,IAAI,UAAU,WAAW,MAAM,GAC7B,OAAO,KAAK;IAAE,KAAK;IAAW,cAAc;IAAM;GAAQ,CAAC;EAE/D,QAAQ,CAER;CACF,CAAC,CACH;CAEA,OAAO;EAAE;EAAQ;CAAY;AAC/B;;;;;;;;;;;;AAaA,MAAa,cAAc,OACzB,WACA,UAAuB,CAAC,MACA;CACxB,MAAM,EACJ,OAAO,MACP,YAAY,oBACZ,YAAY,uBACV;CAEJ,MAAM,SAAS,IAAI,IAAI,SAAS,CAAC,CAAC;CAClC,MAAM,eAAe;EAAE;EAAW;CAAU;CAE5C,MAAM,aAAa,OACf,MAAM,YAAY,WAAW,WAAW,SAAS,IACjD;CAEJ,IAAI;CAEJ,IAAI,YACF,YAAY;MACP;EACL,IAAI,MAAM,0BAA0B;EACpC,MAAM,WAAW,MAAMC,iCAAU,WAAW,YAAY;EACxD,IAAI,CAAC,UAAU,IACb,MAAM,IAAI,MACR,mBAAmB,YAAY,WAAW,UAAU,SAAS,OAAO,KAAK,IAC3E;EAEF,MAAM,aAAaH,yCAAkB,SAAS,MAAM,SAAS,QAAQ;EACrE,MAAM,EAAE,QAAQ,gBAAgB,MAAM,aACpC,YACA,QACA,SACF;EACA,YAAY;GACV,MAAM,SAAS;GACf;GACA,eAAeE,kCAAW,SAAS,IAAI,IAAI;EAC7C;CACF;CAEA,MAAM,SAAsB,CAAC;CAC7B,MAAM,eAAe,MAAME,yCACzB;EAAE;EAAW,GAAG;CAAU,IACzB,UAAU,OAAO,KAAK,KAAK,GAC5B,YACF;CAEA,MAAM,WAAW,OAAO,QACrB,OAAO,UAAUC,wCAAY,OAAO,KAAK,GAC1C;EAAE,OAAO;EAAG,YAAY;CAAE,CAC5B;CAEA,OAAO;EACL,KAAK;EACL,MAAM,aAAa,SAAS;EAC5B,eAAe,UAAU;EACzB,UAAUH,kCAAW,UAAU,IAAI;EACnC,OAAOI,2CAAe,QAAQ;EAC9B;EACA;EACA,GAAG;CACL;AACF"}

package/dist/cjs/scan/sitemap.cjs

@@ -0,0 +1,109 @@
+Object.defineProperty(exports, Symbol.toStringTag, { value: 'Module' });
+const require_scan_fetchText = require('./fetchText.cjs');
+const require_scan_robots = require('./robots.cjs');
+
+//#region src/scan/sitemap.ts
+const MAX_CHILD_SITEMAPS = 5;
+const MAX_ENTRIES = 5e3;
+/** Decode the XML entities that can appear in sitemap URLs. */
+const decodeXmlEntities = (text) => text.replace(/&amp;/g, "&").replace(/&lt;/g, "<").replace(/&gt;/g, ">").replace(/&quot;/g, "\"").replace(/&apos;/g, "'");
+/** Read an attribute of an XML tag, accepting both quote styles. */
+const readXmlAttribute = (tag, name) => {
+	const match = tag.match(new RegExp(`\\b${name}\\s*=\\s*("([^"]*)"|'([^']*)')`, "i"));
+	return match ? decodeXmlEntities(match[2] ?? match[3] ?? "") : void 0;
+};
+/** Extract every `<loc>` value of an XML fragment. */
+const extractLocs = (xml) => Array.from(xml.matchAll(/<loc>\s*([\s\S]*?)\s*<\/loc>/gi), (match) => decodeXmlEntities((match[1] ?? "").replace(/^<!\[CDATA\[|\]\]>$/g, "")));
+/**
+* Parse a sitemap or sitemap index (regex based, no XML dependency). Handles
+* `<xhtml:link rel="alternate" hreflang>` alternates in any namespace prefix.
+*/
+const parseSitemap = (xml) => {
+	if (/<sitemapindex\b/i.test(xml)) return {
+		entries: [],
+		childSitemapUrls: extractLocs(xml)
+	};
+	const entries = [];
+	for (const [urlBlock] of xml.matchAll(/<url\b[\s\S]*?<\/url>/gi)) {
+		const loc = extractLocs(urlBlock)[0];
+		if (!loc) continue;
+		const alternates = [];
+		for (const [linkTag] of urlBlock.matchAll(/<(?:\w+:)?link\b[^>]*>/gi)) {
+			const hreflang = readXmlAttribute(linkTag, "hreflang");
+			const href = readXmlAttribute(linkTag, "href");
+			if (hreflang && href) alternates.push({
+				hreflang,
+				href
+			});
+		}
+		entries.push({
+			loc,
+			alternates
+		});
+	}
+	return {
+		entries,
+		childSitemapUrls: []
+	};
+};
+/**
+* Fetch the given sitemaps, following sitemap indexes one level deep, and
+* gather their `<url>` entries.
+*
+* @param sitemapUrls - Sitemaps to read (from `robots.txt`, else defaults).
+* @param fetchOptions - Network options, including the SSRF guard.
+*/
+const collectSitemapEntries = async (sitemapUrls, fetchOptions) => {
+	const fetchedSitemapUrls = [];
+	const entries = [];
+	const visitedUrls = /* @__PURE__ */ new Set();
+	const queue = [...sitemapUrls];
+	let childSitemapCount = 0;
+	while (queue.length > 0 && entries.length < MAX_ENTRIES) {
+		const sitemapUrl = queue.shift();
+		if (visitedUrls.has(sitemapUrl)) continue;
+		visitedUrls.add(sitemapUrl);
+		const response = await require_scan_fetchText.fetchText(sitemapUrl, fetchOptions);
+		if (!response?.ok || !/<(urlset|sitemapindex)\b/i.test(response.text)) continue;
+		fetchedSitemapUrls.push(sitemapUrl);
+		const parsedSitemap = parseSitemap(response.text);
+		entries.push(...parsedSitemap.entries);
+		for (const childUrl of parsedSitemap.childSitemapUrls) {
+			if (childSitemapCount >= MAX_CHILD_SITEMAPS) break;
+			childSitemapCount++;
+			queue.push(childUrl);
+		}
+	}
+	return {
+		fetchedSitemapUrls,
+		entries: entries.slice(0, MAX_ENTRIES),
+		isTruncated: entries.length >= MAX_ENTRIES || queue.length > 0
+	};
+};
+/** Default sitemap locations tried when `robots.txt` declares none. */
+const getDefaultSitemapUrls = (origin) => [`${origin}/sitemap.xml`, `${origin}/sitemap_index.xml`];
+/**
+* List the page URLs of a site from its sitemaps (primary `<loc>` and every
+* hreflang alternate), used to seed a multi-page audit.
+*
+* @returns De-duplicated URLs, or `[targetUrl]` when no sitemap is found.
+*/
+const discoverSitemapUrls = async (targetUrl, fetchOptions) => {
+	const { origin } = new URL(targetUrl);
+	const robots = await require_scan_fetchText.fetchText(`${origin}/robots.txt`, fetchOptions);
+	const declaredSitemaps = robots?.ok ? require_scan_robots.parseRobots(robots.text).sitemapUrls : [];
+	const { entries } = await collectSitemapEntries(declaredSitemaps.length > 0 ? declaredSitemaps : getDefaultSitemapUrls(origin), fetchOptions);
+	const urls = /* @__PURE__ */ new Set();
+	for (const { loc, alternates } of entries) {
+		urls.add(loc);
+		for (const { href } of alternates) urls.add(href);
+	}
+	return urls.size > 0 ? Array.from(urls) : [targetUrl];
+};
+
+//#endregion
+exports.collectSitemapEntries = collectSitemapEntries;
+exports.discoverSitemapUrls = discoverSitemapUrls;
+exports.getDefaultSitemapUrls = getDefaultSitemapUrls;
+exports.parseSitemap = parseSitemap;
+//# sourceMappingURL=sitemap.cjs.map

package/dist/cjs/scan/sitemap.cjs.map

@@ -0,0 +1 @@
+{"version":3,"file":"sitemap.cjs","names":["fetchText","parseRobots"],"sources":["../../../src/scan/sitemap.ts"],"sourcesContent":["import { fetchText, type ScanFetchOptions } from './fetchText';\nimport { parseRobots } from './robots';\n\n/** A `<url>` entry of a sitemap with its hreflang alternates. */\nexport type SitemapEntry = {\n  loc: string;\n  alternates: { hreflang: string; href: string }[];\n};\n\n/** Result of {@link parseSitemap}. */\nexport type ParsedSitemap = {\n  /** `<url>` entries (empty for a sitemap index). */\n  entries: SitemapEntry[];\n  /** Child sitemap URLs, when the document is a `<sitemapindex>`. */\n  childSitemapUrls: string[];\n};\n\n/** Result of {@link collectSitemapEntries}. */\nexport type CollectedSitemaps = {\n  /** Sitemap URLs that answered with a valid document. */\n  fetchedSitemapUrls: string[];\n  entries: SitemapEntry[];\n  /** Whether the entry limit was reached before reading every sitemap. */\n  isTruncated: boolean;\n};\n\nconst MAX_CHILD_SITEMAPS = 5;\nconst MAX_ENTRIES = 5_000;\n\n/** Decode the XML entities that can appear in sitemap URLs. */\nconst decodeXmlEntities = (text: string): string =>\n  text\n    .replace(/&amp;/g, '&')\n    .replace(/&lt;/g, '<')\n    .replace(/&gt;/g, '>')\n    .replace(/&quot;/g, '\"')\n    .replace(/&apos;/g, \"'\");\n\n/** Read an attribute of an XML tag, accepting both quote styles. */\nconst readXmlAttribute = (tag: string, name: string): string | undefined => {\n  const match = tag.match(\n    new RegExp(`\\\\b${name}\\\\s*=\\\\s*(\"([^\"]*)\"|'([^']*)')`, 'i')\n  );\n  return match ? decodeXmlEntities(match[2] ?? match[3] ?? '') : undefined;\n};\n\n/** Extract every `<loc>` value of an XML fragment. */\nconst extractLocs = (xml: string): string[] =>\n  Array.from(xml.matchAll(/<loc>\\s*([\\s\\S]*?)\\s*<\\/loc>/gi), (match) =>\n    decodeXmlEntities((match[1] ?? '').replace(/^<!\\[CDATA\\[|\\]\\]>$/g, ''))\n  );\n\n/**\n * Parse a sitemap or sitemap index (regex based, no XML dependency). Handles\n * `<xhtml:link rel=\"alternate\" hreflang>` alternates in any namespace prefix.\n */\nexport const parseSitemap = (xml: string): ParsedSitemap => {\n  if (/<sitemapindex\\b/i.test(xml)) {\n    return { entries: [], childSitemapUrls: extractLocs(xml) };\n  }\n\n  const entries: SitemapEntry[] = [];\n  for (const [urlBlock] of xml.matchAll(/<url\\b[\\s\\S]*?<\\/url>/gi)) {\n    const loc = extractLocs(urlBlock)[0];\n    if (!loc) continue;\n\n    const alternates: SitemapEntry['alternates'] = [];\n    for (const [linkTag] of urlBlock.matchAll(/<(?:\\w+:)?link\\b[^>]*>/gi)) {\n      const hreflang = readXmlAttribute(linkTag, 'hreflang');\n      const href = readXmlAttribute(linkTag, 'href');\n      if (hreflang && href) alternates.push({ hreflang, href });\n    }\n    entries.push({ loc, alternates });\n  }\n\n  return { entries, childSitemapUrls: [] };\n};\n\n/**\n * Fetch the given sitemaps, following sitemap indexes one level deep, and\n * gather their `<url>` entries.\n *\n * @param sitemapUrls - Sitemaps to read (from `robots.txt`, else defaults).\n * @param fetchOptions - Network options, including the SSRF guard.\n */\nexport const collectSitemapEntries = async (\n  sitemapUrls: string[],\n  fetchOptions: ScanFetchOptions\n): Promise<CollectedSitemaps> => {\n  const fetchedSitemapUrls: string[] = [];\n  const entries: SitemapEntry[] = [];\n  const visitedUrls = new Set<string>();\n  const queue = [...sitemapUrls];\n  let childSitemapCount = 0;\n\n  while (queue.length > 0 && entries.length < MAX_ENTRIES) {\n    const sitemapUrl = queue.shift() as string;\n    if (visitedUrls.has(sitemapUrl)) continue;\n    visitedUrls.add(sitemapUrl);\n\n    const response = await fetchText(sitemapUrl, fetchOptions);\n    if (!response?.ok || !/<(urlset|sitemapindex)\\b/i.test(response.text)) {\n      continue;\n    }\n\n    fetchedSitemapUrls.push(sitemapUrl);\n    const parsedSitemap = parseSitemap(response.text);\n    entries.push(...parsedSitemap.entries);\n\n    for (const childUrl of parsedSitemap.childSitemapUrls) {\n      if (childSitemapCount >= MAX_CHILD_SITEMAPS) break;\n      childSitemapCount++;\n      queue.push(childUrl);\n    }\n  }\n\n  return {\n    fetchedSitemapUrls,\n    entries: entries.slice(0, MAX_ENTRIES),\n    isTruncated: entries.length >= MAX_ENTRIES || queue.length > 0,\n  };\n};\n\n/** Default sitemap locations tried when `robots.txt` declares none. */\nexport const getDefaultSitemapUrls = (origin: string): string[] => [\n  `${origin}/sitemap.xml`,\n  `${origin}/sitemap_index.xml`,\n];\n\n/**\n * List the page URLs of a site from its sitemaps (primary `<loc>` and every\n * hreflang alternate), used to seed a multi-page audit.\n *\n * @returns De-duplicated URLs, or `[targetUrl]` when no sitemap is found.\n */\nexport const discoverSitemapUrls = async (\n  targetUrl: string,\n  fetchOptions: ScanFetchOptions\n): Promise<string[]> => {\n  const { origin } = new URL(targetUrl);\n  const robots = await fetchText(`${origin}/robots.txt`, fetchOptions);\n  const declaredSitemaps = robots?.ok\n    ? parseRobots(robots.text).sitemapUrls\n    : [];\n\n  const { entries } = await collectSitemapEntries(\n    declaredSitemaps.length > 0\n      ? declaredSitemaps\n      : getDefaultSitemapUrls(origin),\n    fetchOptions\n  );\n\n  const urls = new Set<string>();\n  for (const { loc, alternates } of entries) {\n    urls.add(loc);\n    for (const { href } of alternates) urls.add(href);\n  }\n\n  return urls.size > 0 ? Array.from(urls) : [targetUrl];\n};\n"],"mappings":";;;;;AA0BA,MAAM,qBAAqB;AAC3B,MAAM,cAAc;;AAGpB,MAAM,qBAAqB,SACzB,KACG,QAAQ,UAAU,GAAG,CAAC,CACtB,QAAQ,SAAS,GAAG,CAAC,CACrB,QAAQ,SAAS,GAAG,CAAC,CACrB,QAAQ,WAAW,IAAG,CAAC,CACvB,QAAQ,WAAW,GAAG;;AAG3B,MAAM,oBAAoB,KAAa,SAAqC;CAC1E,MAAM,QAAQ,IAAI,MAChB,IAAI,OAAO,MAAM,KAAK,iCAAiC,GAAG,CAC5D;CACA,OAAO,QAAQ,kBAAkB,MAAM,MAAM,MAAM,MAAM,EAAE,IAAI;AACjE;;AAGA,MAAM,eAAe,QACnB,MAAM,KAAK,IAAI,SAAS,gCAAgC,IAAI,UAC1D,mBAAmB,MAAM,MAAM,GAAE,CAAE,QAAQ,wBAAwB,EAAE,CAAC,CACxE;;;;;AAMF,MAAa,gBAAgB,QAA+B;CAC1D,IAAI,mBAAmB,KAAK,GAAG,GAC7B,OAAO;EAAE,SAAS,CAAC;EAAG,kBAAkB,YAAY,GAAG;CAAE;CAG3D,MAAM,UAA0B,CAAC;CACjC,KAAK,MAAM,CAAC,aAAa,IAAI,SAAS,yBAAyB,GAAG;EAChE,MAAM,MAAM,YAAY,QAAQ,CAAC,CAAC;EAClC,IAAI,CAAC,KAAK;EAEV,MAAM,aAAyC,CAAC;EAChD,KAAK,MAAM,CAAC,YAAY,SAAS,SAAS,0BAA0B,GAAG;GACrE,MAAM,WAAW,iBAAiB,SAAS,UAAU;GACrD,MAAM,OAAO,iBAAiB,SAAS,MAAM;GAC7C,IAAI,YAAY,MAAM,WAAW,KAAK;IAAE;IAAU;GAAK,CAAC;EAC1D;EACA,QAAQ,KAAK;GAAE;GAAK;EAAW,CAAC;CAClC;CAEA,OAAO;EAAE;EAAS,kBAAkB,CAAC;CAAE;AACzC;;;;;;;;AASA,MAAa,wBAAwB,OACnC,aACA,iBAC+B;CAC/B,MAAM,qBAA+B,CAAC;CACtC,MAAM,UAA0B,CAAC;CACjC,MAAM,8BAAc,IAAI,IAAY;CACpC,MAAM,QAAQ,CAAC,GAAG,WAAW;CAC7B,IAAI,oBAAoB;CAExB,OAAO,MAAM,SAAS,KAAK,QAAQ,SAAS,aAAa;EACvD,MAAM,aAAa,MAAM,MAAM;EAC/B,IAAI,YAAY,IAAI,UAAU,GAAG;EACjC,YAAY,IAAI,UAAU;EAE1B,MAAM,WAAW,MAAMA,iCAAU,YAAY,YAAY;EACzD,IAAI,CAAC,UAAU,MAAM,CAAC,4BAA4B,KAAK,SAAS,IAAI,GAClE;EAGF,mBAAmB,KAAK,UAAU;EAClC,MAAM,gBAAgB,aAAa,SAAS,IAAI;EAChD,QAAQ,KAAK,GAAG,cAAc,OAAO;EAErC,KAAK,MAAM,YAAY,cAAc,kBAAkB;GACrD,IAAI,qBAAqB,oBAAoB;GAC7C;GACA,MAAM,KAAK,QAAQ;EACrB;CACF;CAEA,OAAO;EACL;EACA,SAAS,QAAQ,MAAM,GAAG,WAAW;EACrC,aAAa,QAAQ,UAAU,eAAe,MAAM,SAAS;CAC/D;AACF;;AAGA,MAAa,yBAAyB,WAA6B,CACjE,GAAG,OAAO,eACV,GAAG,OAAO,mBACZ;;;;;;;AAQA,MAAa,sBAAsB,OACjC,WACA,iBACsB;CACtB,MAAM,EAAE,WAAW,IAAI,IAAI,SAAS;CACpC,MAAM,SAAS,MAAMA,iCAAU,GAAG,OAAO,cAAc,YAAY;CACnE,MAAM,mBAAmB,QAAQ,KAC7BC,gCAAY,OAAO,IAAI,CAAC,CAAC,cACzB,CAAC;CAEL,MAAM,EAAE,YAAY,MAAM,sBACxB,iBAAiB,SAAS,IACtB,mBACA,sBAAsB,MAAM,GAChC,YACF;CAEA,MAAM,uBAAO,IAAI,IAAY;CAC7B,KAAK,MAAM,EAAE,KAAK,gBAAgB,SAAS;EACzC,KAAK,IAAI,GAAG;EACZ,KAAK,MAAM,EAAE,UAAU,YAAY,KAAK,IAAI,IAAI;CAClD;CAEA,OAAO,KAAK,OAAO,IAAI,MAAM,KAAK,IAAI,IAAI,CAAC,SAAS;AACtD"}

package/dist/cjs/utils/chunkJSON.cjs

@@ -21,7 +21,8 @@ const setAtPath = (root, path, value) => {
 	let current = root;
 	for (let i = 0; i < path.length - 1; i++) {
 		const key = path[i];
-		const isNextIndex = typeof path[i + 1] === "number";
+		if (key === void 0) continue;
+		const isNextIndex = typeof nextKey === "number";
 		if (typeof key === "number") {
 			if (!Array.isArray(current)) throw new Error(`Expected array at path segment ${i}`);
 			if (current[key] === void 0) current[key] = isNextIndex ? [] : {};
@@ -33,6 +34,7 @@ const setAtPath = (root, path, value) => {
 		}
 	}
 	const last = path[path.length - 1];
+	if (last === void 0) return;
 	if (typeof last === "number") {
 		if (!Array.isArray(current)) throw new Error(`Expected array at final segment`);
 		current[last] = value;
@@ -91,11 +93,14 @@ const flattenToPatches = (value, maxCharsPerChunk, path = [], seen = /* @__PURE_
 				path: [...currentPath, "__total"],
 				value: parts.length
 			});
-			for (let i = 0; i < parts.length; i++) patches.push({
-				op: "set",
-				path: [...currentPath, String(i + 1)],
-				value: parts[i]
-			});
+			for (let i = 0; i < parts.length; i++) {
+				const part = parts[i];
+				if (part !== void 0) patches.push({
+					op: "set",
+					path: [...currentPath, String(i + 1)],
+					value: part
+				});
+			}
 			return;
 		}
 		if (currentValue === null || typeof currentValue !== "object") {
@@ -108,8 +113,14 @@ const flattenToPatches = (value, maxCharsPerChunk, path = [], seen = /* @__PURE_
 		}
 		if (seen.has(currentValue)) throw new Error("Cannot serialize circular structures to JSON.");
 		seen.add(currentValue);
-		if (Array.isArray(currentValue)) for (let i = 0; i < currentValue.length; i++) walk(currentValue[i], [...currentPath, i]);
-		else for (const key of Object.keys(currentValue)) walk(currentValue[key], [...currentPath, key]);
+		if (Array.isArray(currentValue)) for (let i = 0; i < currentValue.length; i++) {
+			const item = currentValue[i];
+			if (item !== void 0) walk(item, [...currentPath, i]);
+		}
+		else for (const key of Object.keys(currentValue)) {
+			const childVal = currentValue[key];
+			if (childVal !== void 0) walk(childVal, [...currentPath, key]);
+		}
 		seen.delete(currentValue);
 	};
 	walk(value, path);
@@ -237,10 +248,13 @@ const reconstructFromSingleChunk = (chunk) => {
 const assembleJSON = (chunks) => {
 	if (!chunks || chunks.length === 0) throw new Error("No chunks provided.");
 	const sorted = [...chunks].sort((a, b) => a.index - b.index);
-	const { checksum, rootType } = sorted[0];
+	const firstChunk = sorted[0];
+	if (!firstChunk) throw new Error("No chunks provided.");
+	const { checksum, rootType } = firstChunk;
 	const schemaVersion = 1;
 	for (let i = 0; i < sorted.length; i++) {
 		const chunk = sorted[i];
+		if (!chunk) continue;
 		if (chunk.schemaVersion !== schemaVersion) {
 			console.error("Unsupported schemaVersion.", {
 				cause: chunk,
@@ -326,6 +340,7 @@ const assembleJSON = (chunks) => {
 	const reconciled = reconcileSplitNodes(root);
 	for (let i = 0; i < sorted.length; i++) {
 		const chunk = sorted[i];
+		if (!chunk) continue;
 		if (chunk.total !== sorted.length) throw new Error(`Chunk total does not match provided count. Expected ${sorted.length}, but chunk ${i} has total=${chunk.total}`);
 	}
 	return reconciled;

package/dist/cjs/utils/chunkJSON.cjs.map

@@ -1 +1 @@
-{"version":3,"file":"chunkJSON.cjs","names":["getChunk"],"sources":["../../../src/utils/chunkJSON.ts"],"sourcesContent":["/**\n * Split & reassemble JSON by character budget.\n * - Measures serialized size using JSON.stringify(..).length (characters).\n * - Ensures each chunk is itself valid JSON.\n * - Very large strings are split into safe pieces using getChunk and re-concatenated on assemble.\n * - Protects against circular structures (JSON can't serialize those anyway).\n */\n\nimport { getChunk } from './getChunk';\n\ntype JSONPrimitive = string | number | boolean | null;\ntype JSONValue = JSONPrimitive | JSONObject | JSONArray;\n\nexport type JSONObject = {\n  [k: string]: JSONValue;\n};\n\ntype JSONArray = JSONValue[];\n\ntype Path = Array<string | number>;\n\ntype SetPatch = { op: 'set'; path: Path; value: JSONValue };\ntype StrAppendPatch = {\n  op: 'str-append';\n  path: Path;\n  value: string; // part of a longer string\n  index: number;\n  total: number;\n};\ntype Patch = SetPatch | StrAppendPatch;\n\ntype RootType = 'object' | 'array';\n\nexport type JsonChunk = {\n  schemaVersion: 1;\n  index: number;\n  total: number;\n  rootType: RootType;\n  checksum: string; // hash of the full original JSON string\n  entries: Patch[];\n};\n\nconst isObject = (val: unknown): val is Record<string, unknown> => {\n  return typeof val === 'object' && val !== null && !Array.isArray(val);\n};\n\nconst computeDjb2 = (str: string): string => {\n  // Simple 32-bit hash; deterministic & fast\n  let hash = 5381;\n\n  for (let i = 0; i < str.length; i++) {\n    hash = ((hash << 5) + hash) ^ str.charCodeAt(i);\n  }\n\n  // convert to unsigned hex\n  return (hash >>> 0).toString(16).padStart(8, '0');\n};\n\nconst setAtPath = (root: any, path: Path, value: JSONValue) => {\n  let current = root;\n\n  for (let i = 0; i < path.length - 1; i++) {\n    const key = path[i];\n    const nextKey = path[i + 1];\n    const isNextIndex = typeof nextKey === 'number';\n\n    if (typeof key === 'number') {\n      if (!Array.isArray(current)) {\n        throw new Error(`Expected array at path segment ${i}`);\n      }\n\n      if (current[key] === undefined) {\n        current[key] = isNextIndex ? [] : {};\n      }\n\n      current = current[key];\n    } else {\n      if (!isObject(current)) {\n        throw new Error(`Expected object at path segment ${i}`);\n      }\n\n      if (!(key in current)) {\n        (current as any)[key] = isNextIndex ? [] : {};\n      }\n\n      current = (current as any)[key];\n    }\n  }\n\n  const last = path[path.length - 1];\n\n  if (typeof last === 'number') {\n    if (!Array.isArray(current)) {\n      throw new Error(`Expected array at final segment`);\n    }\n\n    current[last] = value as any;\n  } else {\n    if (!isObject(current)) {\n      throw new Error(`Expected object at final segment`);\n    }\n\n    (current as any)[last] = value as any;\n  }\n};\n\nconst pathKey = (path: Path): string => {\n  // stable key for grouping string parts\n  return JSON.stringify(path);\n};\n\n/**\n * Split a string into parts using getChunk with a charLength budget per part.\n */\nconst splitStringByBudget = (\n  str: string,\n  maxCharsPerPart: number\n): string[] => {\n  if (maxCharsPerPart <= 0) {\n    throw new Error('maxChars must be > 0');\n  }\n\n  const output: string[] = [];\n  let offset = 0;\n\n  while (offset < str.length) {\n    const part = getChunk(str, {\n      charStart: offset,\n      charLength: maxCharsPerPart,\n    });\n\n    if (!part) break;\n\n    output.push(part);\n    offset += part.length;\n  }\n\n  return output;\n};\n\n/**\n * Flatten JSON into patches (leaf writes). Strings too large to fit in a single\n * chunk are yielded as multiple str-append patches.\n */\nconst flattenToPatches = (\n  value: JSONValue,\n  maxCharsPerChunk: number,\n  path: Path = [],\n  seen = new WeakSet<object>()\n): Patch[] => {\n  // Conservative per-entry cap so a single entry fits into an empty chunk with envelope overhead.\n  // (Envelope ~ a few hundred chars; we keep a comfortable margin.)\n  const maxStringPiece = Math.max(\n    1,\n    Math.floor((maxCharsPerChunk - 400) * 0.8)\n  );\n\n  const patches: Patch[] = [];\n\n  const walk = (currentValue: JSONValue, currentPath: Path) => {\n    if (typeof currentValue === 'string') {\n      // If the serialized patch wouldn't fit, split the string into multiple parts\n      // and encode as a split-node sentinel with numbered keys.\n      const testPatch: SetPatch = {\n        op: 'set',\n        path: currentPath,\n        value: currentValue,\n      };\n      const testLen = JSON.stringify(testPatch).length + 150; // margin\n\n      if (testLen <= maxCharsPerChunk) {\n        patches.push(testPatch);\n        return;\n      }\n\n      // Use getChunk-based splitting to produce stable parts\n      const parts = splitStringByBudget(currentValue, maxStringPiece);\n\n      // Emit split-node metadata and parts as individual leaf writes\n      patches.push({\n        op: 'set',\n        path: [...currentPath, '__splittedType'],\n        value: 'string',\n      });\n      patches.push({\n        op: 'set',\n        path: [...currentPath, '__total'],\n        value: parts.length,\n      });\n\n      for (let i = 0; i < parts.length; i++) {\n        patches.push({\n          op: 'set',\n          path: [...currentPath, String(i + 1)],\n          value: parts[i],\n        });\n      }\n\n      return;\n    }\n\n    if (currentValue === null || typeof currentValue !== 'object') {\n      patches.push({ op: 'set', path: currentPath, value: currentValue });\n      return;\n    }\n\n    if (seen.has(currentValue as object)) {\n      throw new Error('Cannot serialize circular structures to JSON.');\n    }\n\n    seen.add(currentValue as object);\n\n    if (Array.isArray(currentValue)) {\n      for (let i = 0; i < currentValue.length; i++) {\n        walk(currentValue[i] as JSONValue, [...currentPath, i]);\n      }\n    } else {\n      for (const key of Object.keys(currentValue)) {\n        walk((currentValue as JSONObject)[key], [...currentPath, key]);\n      }\n    }\n\n    seen.delete(currentValue as object);\n  };\n\n  walk(value, path);\n  return patches;\n};\n\n/**\n * Split JSON into chunks constrained by character count of serialized chunk.\n */\nexport const chunkJSON = (\n  value: JSONObject | JSONArray,\n  maxChars: number\n): JsonChunk[] => {\n  if (!isObject(value) && !Array.isArray(value)) {\n    throw new Error('Root must be an object or array.');\n  }\n\n  if (maxChars < 500) {\n    // You can lower this if you truly need; recommended to keep some envelope headroom.\n    throw new Error('maxChars is too small. Use at least 500 characters.');\n  }\n\n  const rootType: RootType = Array.isArray(value) ? 'array' : 'object';\n  let sourceString: string;\n\n  try {\n    sourceString = JSON.stringify(value);\n  } catch {\n    // Provide a deterministic error message for circular refs\n    throw new Error('Cannot serialize circular structures to JSON.');\n  }\n\n  const checksum = computeDjb2(sourceString);\n  const allPatches = flattenToPatches(value as JSONValue, maxChars);\n\n  const chunks: JsonChunk[] = [];\n  let currentChunk: JsonChunk = {\n    schemaVersion: 1,\n    index: 0, // provisional\n    total: 0, // provisional\n    rootType,\n    checksum,\n    entries: [],\n  };\n\n  const emptyEnvelopeSize = JSON.stringify({\n    ...currentChunk,\n    entries: [],\n  }).length;\n\n  const tryFlush = () => {\n    if (currentChunk.entries.length > 0) {\n      chunks.push(currentChunk);\n      currentChunk = {\n        schemaVersion: 1,\n        index: 0,\n        total: 0,\n        rootType,\n        checksum,\n        entries: [],\n      };\n    }\n  };\n\n  for (const patch of allPatches) {\n    // Would adding this patch exceed maxChars?\n    const withPatchSize =\n      emptyEnvelopeSize +\n      JSON.stringify(currentChunk.entries).length + // current entries array\n      (currentChunk.entries.length ? 1 : 0) + // possible comma\n      JSON.stringify(patch).length;\n\n    if (withPatchSize <= maxChars) {\n      currentChunk.entries.push(patch);\n    } else {\n      // Start a new chunk if current has items\n      if (currentChunk.entries.length > 0) {\n        tryFlush();\n      }\n\n      // Ensure single patch fits into an empty chunk\n      const singleSize = emptyEnvelopeSize + JSON.stringify([patch]).length;\n\n      if (singleSize > maxChars) {\n        // This should only happen for massive strings, which we pre-split;\n        // if it happens for a non-string, we cannot split further.\n        throw new Error(\n          'A single entry exceeds maxChars and cannot be split. Reduce entry size or increase maxChars.'\n        );\n      }\n\n      currentChunk.entries.push(patch);\n    }\n  }\n\n  tryFlush();\n\n  // Ensure at least one chunk exists (even for empty root)\n  if (chunks.length === 0) {\n    chunks.push({\n      schemaVersion: 1,\n      index: 0,\n      total: 0, // provisional\n      rootType,\n      checksum,\n      entries: [],\n    });\n  }\n\n  // Finalize indices & totals\n  const totalChunks = chunks.length;\n\n  chunks.forEach((chunk, index) => {\n    chunk.index = index;\n    chunk.total = totalChunks;\n  });\n\n  return chunks;\n};\n\n/**\n * Reassemble JSON from chunks.\n * - Validates checksums and indices.\n * - Applies 'set' patches and merges string pieces from 'str-append'.\n */\n/**\n * Reconstruct content from a single chunk without validation.\n * Useful for processing individual chunks in a pipeline where you don't have all chunks yet.\n * Note: This will only reconstruct the partial content contained in this chunk.\n */\nexport const reconstructFromSingleChunk = (\n  chunk: JsonChunk\n): JSONObject | JSONArray => {\n  const root: any = chunk.rootType === 'array' ? [] : {};\n\n  // Apply all 'set' patches from this chunk\n  for (const entry of chunk.entries) {\n    if (entry.op === 'set') {\n      setAtPath(root, entry.path, entry.value);\n    }\n  }\n\n  // Reconcile split-node sentinels for strings/arrays\n  // When reconstructing from a single chunk, we may have incomplete split nodes\n  const reconcileSplitNodes = (node: any): any => {\n    if (node === null || typeof node !== 'object') return node;\n\n    if (Array.isArray(node)) {\n      for (let i = 0; i < node.length; i++) {\n        node[i] = reconcileSplitNodes(node[i]);\n      }\n      return node;\n    }\n\n    // string split-node\n    if ((node as any)['__splittedType'] === 'string') {\n      const total = (node as any)['__total'];\n\n      if (typeof total !== 'number' || total <= 0) {\n        // Invalid split node, return as-is\n        return node;\n      }\n\n      const parts: string[] = [];\n      let hasAllParts = true;\n\n      for (let i = 1; i <= total; i++) {\n        const piece = (node as any)[String(i)];\n\n        if (typeof piece !== 'string') {\n          hasAllParts = false;\n          break;\n        }\n\n        parts.push(piece);\n      }\n\n      // Only reconstruct if we have all parts, otherwise return the node as-is\n      if (hasAllParts) {\n        return parts.join('');\n      }\n\n      return node;\n    }\n\n    // array split-node (optional support)\n    if ((node as any)['__splittedType'] === 'array') {\n      const total = (node as any)['__total'];\n\n      if (typeof total !== 'number' || total < 0) {\n        // Invalid split node, return as-is\n        return node;\n      }\n\n      const output: any[] = [];\n      let hasAllParts = true;\n\n      for (let i = 1; i <= total; i++) {\n        const slice = (node as any)[String(i)];\n\n        if (!Array.isArray(slice)) {\n          hasAllParts = false;\n          break;\n        }\n\n        for (let j = 0; j < slice.length; j++) {\n          output.push(reconcileSplitNodes(slice[j]));\n        }\n      }\n\n      // Only reconstruct if we have all parts\n      if (hasAllParts) {\n        return output;\n      }\n\n      return node;\n    }\n\n    // walk normal object\n    for (const key of Object.keys(node)) {\n      node[key] = reconcileSplitNodes(node[key]);\n    }\n\n    return node;\n  };\n\n  return reconcileSplitNodes(root);\n};\n\nexport const assembleJSON = (chunks: JsonChunk[]): JSONObject | JSONArray => {\n  if (!chunks || chunks.length === 0) {\n    throw new Error('No chunks provided.');\n  }\n\n  // Basic validation & sort\n  const sorted = [...chunks].sort((a, b) => a.index - b.index);\n  const { checksum, rootType } = sorted[0];\n  const schemaVersion = 1;\n\n  for (let i = 0; i < sorted.length; i++) {\n    const chunk = sorted[i];\n\n    if (chunk.schemaVersion !== schemaVersion) {\n      console.error('Unsupported schemaVersion.', {\n        cause: chunk,\n        schemaVersion,\n      });\n      throw new Error('Unsupported schemaVersion.');\n    }\n\n    if (chunk.rootType !== rootType) {\n      console.error('Chunks rootType mismatch.', {\n        cause: chunk,\n        rootType,\n      });\n      throw new Error('Chunks rootType mismatch.');\n    }\n\n    if (chunk.checksum !== checksum) {\n      console.error('Chunks checksum mismatch (different source objects?).', {\n        cause: chunk,\n        checksum,\n      });\n      throw new Error('Chunks checksum mismatch (different source objects?).');\n    }\n\n    if (chunk.index !== i) {\n      console.error('Chunk indices are not contiguous or sorted.', {\n        cause: chunk,\n        index: chunk.index,\n        i,\n      });\n      throw new Error('Chunk indices are not contiguous or sorted.');\n    }\n\n    // Defer total check until after reconstruction to prefer more specific errors\n  }\n\n  const root: any = rootType === 'array' ? [] : {};\n\n  // Collect string parts by path\n  const stringParts = new Map<\n    string,\n    { path: Path; total: number; received: StrAppendPatch[] }\n  >();\n\n  const applySet = (patch: SetPatch) =>\n    setAtPath(root, patch.path, patch.value);\n\n  for (const chunk of sorted) {\n    for (const entry of chunk.entries) {\n      if (entry.op === 'set') {\n        applySet(entry);\n      } else {\n        const key = pathKey(entry.path);\n        const record = stringParts.get(key) ?? {\n          path: entry.path,\n          total: entry.total,\n          received: [],\n        };\n\n        if (record.total !== entry.total) {\n          throw new Error('Inconsistent string part totals for a path.');\n        }\n\n        record.received.push(entry);\n        stringParts.set(key, record);\n      }\n    }\n  }\n\n  // Stitch strings\n  for (const { path, total, received } of stringParts.values()) {\n    if (received.length !== total) {\n      throw new Error('Missing string parts for a path; incomplete chunk set.');\n    }\n\n    received.sort((a, b) => a.index - b.index);\n    const fullString = received.map((part) => part.value).join('');\n    setAtPath(root, path, fullString);\n  }\n\n  // Reconcile split-node sentinels for strings/arrays after all patches applied\n  const reconcileSplitNodes = (node: any): any => {\n    if (node === null || typeof node !== 'object') return node;\n\n    if (Array.isArray(node)) {\n      for (let i = 0; i < node.length; i++) {\n        node[i] = reconcileSplitNodes(node[i]);\n      }\n      return node;\n    }\n\n    // string split-node\n    if ((node as any)['__splittedType'] === 'string') {\n      const total = (node as any)['__total'];\n\n      if (typeof total !== 'number' || total <= 0) {\n        throw new Error('Invalid split-node total for a path.');\n      }\n\n      const parts: string[] = [];\n\n      for (let i = 1; i <= total; i++) {\n        const piece = (node as any)[String(i)];\n\n        if (typeof piece !== 'string') {\n          throw new Error(\n            'Missing string parts for a path; incomplete chunk set.'\n          );\n        }\n\n        parts.push(piece);\n      }\n\n      return parts.join('');\n    }\n\n    // array split-node (optional support)\n    if ((node as any)['__splittedType'] === 'array') {\n      const total = (node as any)['__total'];\n\n      if (typeof total !== 'number' || total < 0) {\n        throw new Error('Invalid split-node total for a path.');\n      }\n\n      const output: any[] = [];\n\n      for (let i = 1; i <= total; i++) {\n        const slice = (node as any)[String(i)];\n\n        if (!Array.isArray(slice)) {\n          throw new Error(\n            'Missing string parts for a path; incomplete chunk set.'\n          );\n        }\n\n        for (let j = 0; j < slice.length; j++) {\n          output.push(reconcileSplitNodes(slice[j]));\n        }\n      }\n\n      return output;\n    }\n\n    // walk normal object\n    for (const key of Object.keys(node)) {\n      node[key] = reconcileSplitNodes(node[key]);\n    }\n\n    return node;\n  };\n\n  const reconciled = reconcileSplitNodes(root);\n\n  // Now validate totals match provided count\n  for (let i = 0; i < sorted.length; i++) {\n    const chunk = sorted[i];\n\n    if (chunk.total !== sorted.length) {\n      throw new Error(\n        `Chunk total does not match provided count. Expected ${sorted.length}, but chunk ${i} has total=${chunk.total}`\n      );\n    }\n  }\n\n  return reconciled;\n};\n\n/* -------------------------------------------\n * Example usage\n * -------------------------------------------\nconst big: JSONObject = {\n  title: \"Document\",\n  content: \"…a very very long text…\",\n  items: Array.from({ length: 2000 }, (_, i) => ({ id: i, label: `Item ${i}` }))\n};\n\n// Split to ~16k-char chunks\nconst chunks = chunkJSON(big, 16_000);\n\n// Send each `chunks[i]` as JSON to your backend.\n// Later, reassemble:\nconst restored = assembleJSON(chunks);\nconsole.log(JSON.stringify(restored) === JSON.stringify(big)); // true\n*/\n"],"mappings":";;;;;;;;;;;AA0CA,MAAM,YAAY,QAAiD;CACjE,OAAO,OAAO,QAAQ,YAAY,QAAQ,QAAQ,CAAC,MAAM,QAAQ,GAAG;AACtE;AAEA,MAAM,eAAe,QAAwB;CAE3C,IAAI,OAAO;CAEX,KAAK,IAAI,IAAI,GAAG,IAAI,IAAI,QAAQ,KAC9B,QAAS,QAAQ,KAAK,OAAQ,IAAI,WAAW,CAAC;CAIhD,QAAQ,SAAS,EAAC,CAAE,SAAS,EAAE,CAAC,CAAC,SAAS,GAAG,GAAG;AAClD;AAEA,MAAM,aAAa,MAAW,MAAY,UAAqB;CAC7D,IAAI,UAAU;CAEd,KAAK,IAAI,IAAI,GAAG,IAAI,KAAK,SAAS,GAAG,KAAK;EACxC,MAAM,MAAM,KAAK;EAEjB,MAAM,cAAc,OADJ,KAAK,IAAI,OACc;EAEvC,IAAI,OAAO,QAAQ,UAAU;GAC3B,IAAI,CAAC,MAAM,QAAQ,OAAO,GACxB,MAAM,IAAI,MAAM,kCAAkC,GAAG;GAGvD,IAAI,QAAQ,SAAS,QACnB,QAAQ,OAAO,cAAc,CAAC,IAAI,CAAC;GAGrC,UAAU,QAAQ;EACpB,OAAO;GACL,IAAI,CAAC,SAAS,OAAO,GACnB,MAAM,IAAI,MAAM,mCAAmC,GAAG;GAGxD,IAAI,EAAE,OAAO,UACX,AAAC,QAAgB,OAAO,cAAc,CAAC,IAAI,CAAC;GAG9C,UAAW,QAAgB;EAC7B;CACF;CAEA,MAAM,OAAO,KAAK,KAAK,SAAS;CAEhC,IAAI,OAAO,SAAS,UAAU;EAC5B,IAAI,CAAC,MAAM,QAAQ,OAAO,GACxB,MAAM,IAAI,MAAM,iCAAiC;EAGnD,QAAQ,QAAQ;CAClB,OAAO;EACL,IAAI,CAAC,SAAS,OAAO,GACnB,MAAM,IAAI,MAAM,kCAAkC;EAGpD,AAAC,QAAgB,QAAQ;CAC3B;AACF;AAEA,MAAM,WAAW,SAAuB;CAEtC,OAAO,KAAK,UAAU,IAAI;AAC5B;;;;AAKA,MAAM,uBACJ,KACA,oBACa;CACb,IAAI,mBAAmB,GACrB,MAAM,IAAI,MAAM,sBAAsB;CAGxC,MAAM,SAAmB,CAAC;CAC1B,IAAI,SAAS;CAEb,OAAO,SAAS,IAAI,QAAQ;EAC1B,MAAM,OAAOA,gCAAS,KAAK;GACzB,WAAW;GACX,YAAY;EACd,CAAC;EAED,IAAI,CAAC,MAAM;EAEX,OAAO,KAAK,IAAI;EAChB,UAAU,KAAK;CACjB;CAEA,OAAO;AACT;;;;;AAMA,MAAM,oBACJ,OACA,kBACA,OAAa,CAAC,GACd,uBAAO,IAAI,QAAgB,MACf;CAGZ,MAAM,iBAAiB,KAAK,IAC1B,GACA,KAAK,OAAO,mBAAmB,OAAO,EAAG,CAC3C;CAEA,MAAM,UAAmB,CAAC;CAE1B,MAAM,QAAQ,cAAyB,gBAAsB;EAC3D,IAAI,OAAO,iBAAiB,UAAU;GAGpC,MAAM,YAAsB;IAC1B,IAAI;IACJ,MAAM;IACN,OAAO;GACT;GAGA,IAFgB,KAAK,UAAU,SAAS,CAAC,CAAC,SAAS,OAEpC,kBAAkB;IAC/B,QAAQ,KAAK,SAAS;IACtB;GACF;GAGA,MAAM,QAAQ,oBAAoB,cAAc,cAAc;GAG9D,QAAQ,KAAK;IACX,IAAI;IACJ,MAAM,CAAC,GAAG,aAAa,gBAAgB;IACvC,OAAO;GACT,CAAC;GACD,QAAQ,KAAK;IACX,IAAI;IACJ,MAAM,CAAC,GAAG,aAAa,SAAS;IAChC,OAAO,MAAM;GACf,CAAC;GAED,KAAK,IAAI,IAAI,GAAG,IAAI,MAAM,QAAQ,KAChC,QAAQ,KAAK;IACX,IAAI;IACJ,MAAM,CAAC,GAAG,aAAa,OAAO,IAAI,CAAC,CAAC;IACpC,OAAO,MAAM;GACf,CAAC;GAGH;EACF;EAEA,IAAI,iBAAiB,QAAQ,OAAO,iBAAiB,UAAU;GAC7D,QAAQ,KAAK;IAAE,IAAI;IAAO,MAAM;IAAa,OAAO;GAAa,CAAC;GAClE;EACF;EAEA,IAAI,KAAK,IAAI,YAAsB,GACjC,MAAM,IAAI,MAAM,+CAA+C;EAGjE,KAAK,IAAI,YAAsB;EAE/B,IAAI,MAAM,QAAQ,YAAY,GAC5B,KAAK,IAAI,IAAI,GAAG,IAAI,aAAa,QAAQ,KACvC,KAAK,aAAa,IAAiB,CAAC,GAAG,aAAa,CAAC,CAAC;OAGxD,KAAK,MAAM,OAAO,OAAO,KAAK,YAAY,GACxC,KAAM,aAA4B,MAAM,CAAC,GAAG,aAAa,GAAG,CAAC;EAIjE,KAAK,OAAO,YAAsB;CACpC;CAEA,KAAK,OAAO,IAAI;CAChB,OAAO;AACT;;;;AAKA,MAAa,aACX,OACA,aACgB;CAChB,IAAI,CAAC,SAAS,KAAK,KAAK,CAAC,MAAM,QAAQ,KAAK,GAC1C,MAAM,IAAI,MAAM,kCAAkC;CAGpD,IAAI,WAAW,KAEb,MAAM,IAAI,MAAM,qDAAqD;CAGvE,MAAM,WAAqB,MAAM,QAAQ,KAAK,IAAI,UAAU;CAC5D,IAAI;CAEJ,IAAI;EACF,eAAe,KAAK,UAAU,KAAK;CACrC,QAAQ;EAEN,MAAM,IAAI,MAAM,+CAA+C;CACjE;CAEA,MAAM,WAAW,YAAY,YAAY;CACzC,MAAM,aAAa,iBAAiB,OAAoB,QAAQ;CAEhE,MAAM,SAAsB,CAAC;CAC7B,IAAI,eAA0B;EAC5B,eAAe;EACf,OAAO;EACP,OAAO;EACP;EACA;EACA,SAAS,CAAC;CACZ;CAEA,MAAM,oBAAoB,KAAK,UAAU;EACvC,GAAG;EACH,SAAS,CAAC;CACZ,CAAC,CAAC,CAAC;CAEH,MAAM,iBAAiB;EACrB,IAAI,aAAa,QAAQ,SAAS,GAAG;GACnC,OAAO,KAAK,YAAY;GACxB,eAAe;IACb,eAAe;IACf,OAAO;IACP,OAAO;IACP;IACA;IACA,SAAS,CAAC;GACZ;EACF;CACF;CAEA,KAAK,MAAM,SAAS,YAQlB,IALE,oBACA,KAAK,UAAU,aAAa,OAAO,CAAC,CAAC,UACpC,aAAa,QAAQ,SAAS,IAAI,KACnC,KAAK,UAAU,KAAK,CAAC,CAAC,UAEH,UACnB,aAAa,QAAQ,KAAK,KAAK;MAC1B;EAEL,IAAI,aAAa,QAAQ,SAAS,GAChC,SAAS;EAMX,IAFmB,oBAAoB,KAAK,UAAU,CAAC,KAAK,CAAC,CAAC,CAAC,SAE9C,UAGf,MAAM,IAAI,MACR,8FACF;EAGF,aAAa,QAAQ,KAAK,KAAK;CACjC;CAGF,SAAS;CAGT,IAAI,OAAO,WAAW,GACpB,OAAO,KAAK;EACV,eAAe;EACf,OAAO;EACP,OAAO;EACP;EACA;EACA,SAAS,CAAC;CACZ,CAAC;CAIH,MAAM,cAAc,OAAO;CAE3B,OAAO,SAAS,OAAO,UAAU;EAC/B,MAAM,QAAQ;EACd,MAAM,QAAQ;CAChB,CAAC;CAED,OAAO;AACT;;;;;;;;;;;AAYA,MAAa,8BACX,UAC2B;CAC3B,MAAM,OAAY,MAAM,aAAa,UAAU,CAAC,IAAI,CAAC;CAGrD,KAAK,MAAM,SAAS,MAAM,SACxB,IAAI,MAAM,OAAO,OACf,UAAU,MAAM,MAAM,MAAM,MAAM,KAAK;CAM3C,MAAM,uBAAuB,SAAmB;EAC9C,IAAI,SAAS,QAAQ,OAAO,SAAS,UAAU,OAAO;EAEtD,IAAI,MAAM,QAAQ,IAAI,GAAG;GACvB,KAAK,IAAI,IAAI,GAAG,IAAI,KAAK,QAAQ,KAC/B,KAAK,KAAK,oBAAoB,KAAK,EAAE;GAEvC,OAAO;EACT;EAGA,IAAK,KAAa,sBAAsB,UAAU;GAChD,MAAM,QAAS,KAAa;GAE5B,IAAI,OAAO,UAAU,YAAY,SAAS,GAExC,OAAO;GAGT,MAAM,QAAkB,CAAC;GACzB,IAAI,cAAc;GAElB,KAAK,IAAI,IAAI,GAAG,KAAK,OAAO,KAAK;IAC/B,MAAM,QAAS,KAAa,OAAO,CAAC;IAEpC,IAAI,OAAO,UAAU,UAAU;KAC7B,cAAc;KACd;IACF;IAEA,MAAM,KAAK,KAAK;GAClB;GAGA,IAAI,aACF,OAAO,MAAM,KAAK,EAAE;GAGtB,OAAO;EACT;EAGA,IAAK,KAAa,sBAAsB,SAAS;GAC/C,MAAM,QAAS,KAAa;GAE5B,IAAI,OAAO,UAAU,YAAY,QAAQ,GAEvC,OAAO;GAGT,MAAM,SAAgB,CAAC;GACvB,IAAI,cAAc;GAElB,KAAK,IAAI,IAAI,GAAG,KAAK,OAAO,KAAK;IAC/B,MAAM,QAAS,KAAa,OAAO,CAAC;IAEpC,IAAI,CAAC,MAAM,QAAQ,KAAK,GAAG;KACzB,cAAc;KACd;IACF;IAEA,KAAK,IAAI,IAAI,GAAG,IAAI,MAAM,QAAQ,KAChC,OAAO,KAAK,oBAAoB,MAAM,EAAE,CAAC;GAE7C;GAGA,IAAI,aACF,OAAO;GAGT,OAAO;EACT;EAGA,KAAK,MAAM,OAAO,OAAO,KAAK,IAAI,GAChC,KAAK,OAAO,oBAAoB,KAAK,IAAI;EAG3C,OAAO;CACT;CAEA,OAAO,oBAAoB,IAAI;AACjC;AAEA,MAAa,gBAAgB,WAAgD;CAC3E,IAAI,CAAC,UAAU,OAAO,WAAW,GAC/B,MAAM,IAAI,MAAM,qBAAqB;CAIvC,MAAM,SAAS,CAAC,GAAG,MAAM,CAAC,CAAC,MAAM,GAAG,MAAM,EAAE,QAAQ,EAAE,KAAK;CAC3D,MAAM,EAAE,UAAU,aAAa,OAAO;CACtC,MAAM,gBAAgB;CAEtB,KAAK,IAAI,IAAI,GAAG,IAAI,OAAO,QAAQ,KAAK;EACtC,MAAM,QAAQ,OAAO;EAErB,IAAI,MAAM,kBAAkB,eAAe;GACzC,QAAQ,MAAM,8BAA8B;IAC1C,OAAO;IACP;GACF,CAAC;GACD,MAAM,IAAI,MAAM,4BAA4B;EAC9C;EAEA,IAAI,MAAM,aAAa,UAAU;GAC/B,QAAQ,MAAM,6BAA6B;IACzC,OAAO;IACP;GACF,CAAC;GACD,MAAM,IAAI,MAAM,2BAA2B;EAC7C;EAEA,IAAI,MAAM,aAAa,UAAU;GAC/B,QAAQ,MAAM,yDAAyD;IACrE,OAAO;IACP;GACF,CAAC;GACD,MAAM,IAAI,MAAM,uDAAuD;EACzE;EAEA,IAAI,MAAM,UAAU,GAAG;GACrB,QAAQ,MAAM,+CAA+C;IAC3D,OAAO;IACP,OAAO,MAAM;IACb;GACF,CAAC;GACD,MAAM,IAAI,MAAM,6CAA6C;EAC/D;CAGF;CAEA,MAAM,OAAY,aAAa,UAAU,CAAC,IAAI,CAAC;CAG/C,MAAM,8BAAc,IAAI,IAGtB;CAEF,MAAM,YAAY,UAChB,UAAU,MAAM,MAAM,MAAM,MAAM,KAAK;CAEzC,KAAK,MAAM,SAAS,QAClB,KAAK,MAAM,SAAS,MAAM,SACxB,IAAI,MAAM,OAAO,OACf,SAAS,KAAK;MACT;EACL,MAAM,MAAM,QAAQ,MAAM,IAAI;EAC9B,MAAM,SAAS,YAAY,IAAI,GAAG,KAAK;GACrC,MAAM,MAAM;GACZ,OAAO,MAAM;GACb,UAAU,CAAC;EACb;EAEA,IAAI,OAAO,UAAU,MAAM,OACzB,MAAM,IAAI,MAAM,6CAA6C;EAG/D,OAAO,SAAS,KAAK,KAAK;EAC1B,YAAY,IAAI,KAAK,MAAM;CAC7B;CAKJ,KAAK,MAAM,EAAE,MAAM,OAAO,cAAc,YAAY,OAAO,GAAG;EAC5D,IAAI,SAAS,WAAW,OACtB,MAAM,IAAI,MAAM,wDAAwD;EAG1E,SAAS,MAAM,GAAG,MAAM,EAAE,QAAQ,EAAE,KAAK;EACzC,MAAM,aAAa,SAAS,KAAK,SAAS,KAAK,KAAK,CAAC,CAAC,KAAK,EAAE;EAC7D,UAAU,MAAM,MAAM,UAAU;CAClC;CAGA,MAAM,uBAAuB,SAAmB;EAC9C,IAAI,SAAS,QAAQ,OAAO,SAAS,UAAU,OAAO;EAEtD,IAAI,MAAM,QAAQ,IAAI,GAAG;GACvB,KAAK,IAAI,IAAI,GAAG,IAAI,KAAK,QAAQ,KAC/B,KAAK,KAAK,oBAAoB,KAAK,EAAE;GAEvC,OAAO;EACT;EAGA,IAAK,KAAa,sBAAsB,UAAU;GAChD,MAAM,QAAS,KAAa;GAE5B,IAAI,OAAO,UAAU,YAAY,SAAS,GACxC,MAAM,IAAI,MAAM,sCAAsC;GAGxD,MAAM,QAAkB,CAAC;GAEzB,KAAK,IAAI,IAAI,GAAG,KAAK,OAAO,KAAK;IAC/B,MAAM,QAAS,KAAa,OAAO,CAAC;IAEpC,IAAI,OAAO,UAAU,UACnB,MAAM,IAAI,MACR,wDACF;IAGF,MAAM,KAAK,KAAK;GAClB;GAEA,OAAO,MAAM,KAAK,EAAE;EACtB;EAGA,IAAK,KAAa,sBAAsB,SAAS;GAC/C,MAAM,QAAS,KAAa;GAE5B,IAAI,OAAO,UAAU,YAAY,QAAQ,GACvC,MAAM,IAAI,MAAM,sCAAsC;GAGxD,MAAM,SAAgB,CAAC;GAEvB,KAAK,IAAI,IAAI,GAAG,KAAK,OAAO,KAAK;IAC/B,MAAM,QAAS,KAAa,OAAO,CAAC;IAEpC,IAAI,CAAC,MAAM,QAAQ,KAAK,GACtB,MAAM,IAAI,MACR,wDACF;IAGF,KAAK,IAAI,IAAI,GAAG,IAAI,MAAM,QAAQ,KAChC,OAAO,KAAK,oBAAoB,MAAM,EAAE,CAAC;GAE7C;GAEA,OAAO;EACT;EAGA,KAAK,MAAM,OAAO,OAAO,KAAK,IAAI,GAChC,KAAK,OAAO,oBAAoB,KAAK,IAAI;EAG3C,OAAO;CACT;CAEA,MAAM,aAAa,oBAAoB,IAAI;CAG3C,KAAK,IAAI,IAAI,GAAG,IAAI,OAAO,QAAQ,KAAK;EACtC,MAAM,QAAQ,OAAO;EAErB,IAAI,MAAM,UAAU,OAAO,QACzB,MAAM,IAAI,MACR,uDAAuD,OAAO,OAAO,cAAc,EAAE,aAAa,MAAM,OAC1G;CAEJ;CAEA,OAAO;AACT"}
+{"version":3,"file":"chunkJSON.cjs","names":["getChunk"],"sources":["../../../src/utils/chunkJSON.ts"],"sourcesContent":["/**\n * Split & reassemble JSON by character budget.\n * - Measures serialized size using JSON.stringify(..).length (characters).\n * - Ensures each chunk is itself valid JSON.\n * - Very large strings are split into safe pieces using getChunk and re-concatenated on assemble.\n * - Protects against circular structures (JSON can't serialize those anyway).\n */\n\nimport { getChunk } from './getChunk';\n\ntype JSONPrimitive = string | number | boolean | null;\ntype JSONValue = JSONPrimitive | JSONObject | JSONArray;\n\nexport type JSONObject = {\n  [k: string]: JSONValue;\n};\n\ntype JSONArray = JSONValue[];\n\ntype Path = Array<string | number>;\n\ntype SetPatch = { op: 'set'; path: Path; value: JSONValue };\ntype StrAppendPatch = {\n  op: 'str-append';\n  path: Path;\n  value: string; // part of a longer string\n  index: number;\n  total: number;\n};\ntype Patch = SetPatch | StrAppendPatch;\n\ntype RootType = 'object' | 'array';\n\nexport type JsonChunk = {\n  schemaVersion: 1;\n  index: number;\n  total: number;\n  rootType: RootType;\n  checksum: string; // hash of the full original JSON string\n  entries: Patch[];\n};\n\nconst isObject = (val: unknown): val is Record<string, unknown> => {\n  return typeof val === 'object' && val !== null && !Array.isArray(val);\n};\n\nconst computeDjb2 = (str: string): string => {\n  // Simple 32-bit hash; deterministic & fast\n  let hash = 5381;\n\n  for (let i = 0; i < str.length; i++) {\n    hash = ((hash << 5) + hash) ^ str.charCodeAt(i);\n  }\n\n  // convert to unsigned hex\n  return (hash >>> 0).toString(16).padStart(8, '0');\n};\n\nconst setAtPath = (root: any, path: Path, value: JSONValue) => {\n  let current = root;\n\n  for (let i = 0; i < path.length - 1; i++) {\n    const key = path[i];\n    if (key === undefined) continue;\n\n    const isNextIndex = typeof nextKey === 'number';\n\n    if (typeof key === 'number') {\n      if (!Array.isArray(current)) {\n        throw new Error(`Expected array at path segment ${i}`);\n      }\n\n      if (current[key] === undefined) {\n        current[key] = isNextIndex ? [] : {};\n      }\n\n      current = current[key];\n    } else {\n      if (!isObject(current)) {\n        throw new Error(`Expected object at path segment ${i}`);\n      }\n\n      if (!(key in current)) {\n        (current as Record<string, any>)[key] = isNextIndex ? [] : {};\n      }\n\n      current = (current as Record<string, any>)[key];\n    }\n  }\n\n  const last = path[path.length - 1];\n\n  if (last === undefined) return;\n\n  if (typeof last === 'number') {\n    if (!Array.isArray(current)) {\n      throw new Error(`Expected array at final segment`);\n    }\n\n    current[last] = value as any;\n  } else {\n    if (!isObject(current)) {\n      throw new Error(`Expected object at final segment`);\n    }\n\n    (current as Record<string, any>)[last] = value as any;\n  }\n};\n\nconst pathKey = (path: Path): string => {\n  // stable key for grouping string parts\n  return JSON.stringify(path);\n};\n\n/**\n * Split a string into parts using getChunk with a charLength budget per part.\n */\nconst splitStringByBudget = (\n  str: string,\n  maxCharsPerPart: number\n): string[] => {\n  if (maxCharsPerPart <= 0) {\n    throw new Error('maxChars must be > 0');\n  }\n\n  const output: string[] = [];\n  let offset = 0;\n\n  while (offset < str.length) {\n    const part = getChunk(str, {\n      charStart: offset,\n      charLength: maxCharsPerPart,\n    });\n\n    if (!part) break;\n\n    output.push(part);\n    offset += part.length;\n  }\n\n  return output;\n};\n\n/**\n * Flatten JSON into patches (leaf writes). Strings too large to fit in a single\n * chunk are yielded as multiple str-append patches.\n */\nconst flattenToPatches = (\n  value: JSONValue,\n  maxCharsPerChunk: number,\n  path: Path = [],\n  seen = new WeakSet<object>()\n): Patch[] => {\n  // Conservative per-entry cap so a single entry fits into an empty chunk with envelope overhead.\n  // (Envelope ~ a few hundred chars; we keep a comfortable margin.)\n  const maxStringPiece = Math.max(\n    1,\n    Math.floor((maxCharsPerChunk - 400) * 0.8)\n  );\n\n  const patches: Patch[] = [];\n\n  const walk = (currentValue: JSONValue, currentPath: Path) => {\n    if (typeof currentValue === 'string') {\n      // If the serialized patch wouldn't fit, split the string into multiple parts\n      // and encode as a split-node sentinel with numbered keys.\n      const testPatch: SetPatch = {\n        op: 'set',\n        path: currentPath,\n        value: currentValue,\n      };\n      const testLen = JSON.stringify(testPatch).length + 150; // margin\n\n      if (testLen <= maxCharsPerChunk) {\n        patches.push(testPatch);\n        return;\n      }\n\n      // Use getChunk-based splitting to produce stable parts\n      const parts = splitStringByBudget(currentValue, maxStringPiece);\n\n      // Emit split-node metadata and parts as individual leaf writes\n      patches.push({\n        op: 'set',\n        path: [...currentPath, '__splittedType'],\n        value: 'string',\n      });\n      patches.push({\n        op: 'set',\n        path: [...currentPath, '__total'],\n        value: parts.length,\n      });\n\n      for (let i = 0; i < parts.length; i++) {\n        const part = parts[i];\n\n        if (part !== undefined) {\n          patches.push({\n            op: 'set',\n            path: [...currentPath, String(i + 1)],\n            value: part,\n          });\n        }\n      }\n\n      return;\n    }\n\n    if (currentValue === null || typeof currentValue !== 'object') {\n      patches.push({ op: 'set', path: currentPath, value: currentValue });\n      return;\n    }\n\n    if (seen.has(currentValue as object)) {\n      throw new Error('Cannot serialize circular structures to JSON.');\n    }\n\n    seen.add(currentValue as object);\n\n    if (Array.isArray(currentValue)) {\n      for (let i = 0; i < currentValue.length; i++) {\n        const item = currentValue[i];\n        if (item !== undefined) {\n          walk(item as JSONValue, [...currentPath, i]);\n        }\n      }\n    } else {\n      for (const key of Object.keys(currentValue)) {\n        const childVal = (currentValue as JSONObject)[key];\n\n        if (childVal !== undefined) {\n          walk(childVal, [...currentPath, key]);\n        }\n      }\n    }\n\n    seen.delete(currentValue as object);\n  };\n\n  walk(value, path);\n  return patches;\n};\n\n/**\n * Split JSON into chunks constrained by character count of serialized chunk.\n */\nexport const chunkJSON = (\n  value: JSONObject | JSONArray,\n  maxChars: number\n): JsonChunk[] => {\n  if (!isObject(value) && !Array.isArray(value)) {\n    throw new Error('Root must be an object or array.');\n  }\n\n  if (maxChars < 500) {\n    // You can lower this if you truly need; recommended to keep some envelope headroom.\n    throw new Error('maxChars is too small. Use at least 500 characters.');\n  }\n\n  const rootType: RootType = Array.isArray(value) ? 'array' : 'object';\n  let sourceString: string;\n\n  try {\n    sourceString = JSON.stringify(value);\n  } catch {\n    // Provide a deterministic error message for circular refs\n    throw new Error('Cannot serialize circular structures to JSON.');\n  }\n\n  const checksum = computeDjb2(sourceString);\n  const allPatches = flattenToPatches(value as JSONValue, maxChars);\n\n  const chunks: JsonChunk[] = [];\n  let currentChunk: JsonChunk = {\n    schemaVersion: 1,\n    index: 0, // provisional\n    total: 0, // provisional\n    rootType,\n    checksum,\n    entries: [],\n  };\n\n  const emptyEnvelopeSize = JSON.stringify({\n    ...currentChunk,\n    entries: [],\n  }).length;\n\n  const tryFlush = () => {\n    if (currentChunk.entries.length > 0) {\n      chunks.push(currentChunk);\n      currentChunk = {\n        schemaVersion: 1,\n        index: 0,\n        total: 0,\n        rootType,\n        checksum,\n        entries: [],\n      };\n    }\n  };\n\n  for (const patch of allPatches) {\n    // Would adding this patch exceed maxChars?\n    const withPatchSize =\n      emptyEnvelopeSize +\n      JSON.stringify(currentChunk.entries).length + // current entries array\n      (currentChunk.entries.length ? 1 : 0) + // possible comma\n      JSON.stringify(patch).length;\n\n    if (withPatchSize <= maxChars) {\n      currentChunk.entries.push(patch);\n    } else {\n      // Start a new chunk if current has items\n      if (currentChunk.entries.length > 0) {\n        tryFlush();\n      }\n\n      // Ensure single patch fits into an empty chunk\n      const singleSize = emptyEnvelopeSize + JSON.stringify([patch]).length;\n\n      if (singleSize > maxChars) {\n        // This should only happen for massive strings, which we pre-split;\n        // if it happens for a non-string, we cannot split further.\n        throw new Error(\n          'A single entry exceeds maxChars and cannot be split. Reduce entry size or increase maxChars.'\n        );\n      }\n\n      currentChunk.entries.push(patch);\n    }\n  }\n\n  tryFlush();\n\n  // Ensure at least one chunk exists (even for empty root)\n  if (chunks.length === 0) {\n    chunks.push({\n      schemaVersion: 1,\n      index: 0,\n      total: 0, // provisional\n      rootType,\n      checksum,\n      entries: [],\n    });\n  }\n\n  // Finalize indices & totals\n  const totalChunks = chunks.length;\n\n  chunks.forEach((chunk, index) => {\n    chunk.index = index;\n    chunk.total = totalChunks;\n  });\n\n  return chunks;\n};\n\n/**\n * Reassemble JSON from chunks.\n * - Validates checksums and indices.\n * - Applies 'set' patches and merges string pieces from 'str-append'.\n */\n/**\n * Reconstruct content from a single chunk without validation.\n * Useful for processing individual chunks in a pipeline where you don't have all chunks yet.\n * Note: This will only reconstruct the partial content contained in this chunk.\n */\nexport const reconstructFromSingleChunk = (\n  chunk: JsonChunk\n): JSONObject | JSONArray => {\n  const root: any = chunk.rootType === 'array' ? [] : {};\n\n  // Apply all 'set' patches from this chunk\n  for (const entry of chunk.entries) {\n    if (entry.op === 'set') {\n      setAtPath(root, entry.path, entry.value);\n    }\n  }\n\n  // Reconcile split-node sentinels for strings/arrays\n  // When reconstructing from a single chunk, we may have incomplete split nodes\n  const reconcileSplitNodes = (node: any): any => {\n    if (node === null || typeof node !== 'object') return node;\n\n    if (Array.isArray(node)) {\n      for (let i = 0; i < node.length; i++) {\n        node[i] = reconcileSplitNodes(node[i]);\n      }\n      return node;\n    }\n\n    // string split-node\n    if ((node as any)['__splittedType'] === 'string') {\n      const total = (node as any)['__total'];\n\n      if (typeof total !== 'number' || total <= 0) {\n        // Invalid split node, return as-is\n        return node;\n      }\n\n      const parts: string[] = [];\n      let hasAllParts = true;\n\n      for (let i = 1; i <= total; i++) {\n        const piece = (node as any)[String(i)];\n\n        if (typeof piece !== 'string') {\n          hasAllParts = false;\n          break;\n        }\n\n        parts.push(piece);\n      }\n\n      // Only reconstruct if we have all parts, otherwise return the node as-is\n      if (hasAllParts) {\n        return parts.join('');\n      }\n\n      return node;\n    }\n\n    // array split-node (optional support)\n    if ((node as any)['__splittedType'] === 'array') {\n      const total = (node as any)['__total'];\n\n      if (typeof total !== 'number' || total < 0) {\n        // Invalid split node, return as-is\n        return node;\n      }\n\n      const output: any[] = [];\n      let hasAllParts = true;\n\n      for (let i = 1; i <= total; i++) {\n        const slice = (node as any)[String(i)];\n\n        if (!Array.isArray(slice)) {\n          hasAllParts = false;\n          break;\n        }\n\n        for (let j = 0; j < slice.length; j++) {\n          output.push(reconcileSplitNodes(slice[j]));\n        }\n      }\n\n      // Only reconstruct if we have all parts\n      if (hasAllParts) {\n        return output;\n      }\n\n      return node;\n    }\n\n    // walk normal object\n    for (const key of Object.keys(node)) {\n      node[key] = reconcileSplitNodes(node[key]);\n    }\n\n    return node;\n  };\n\n  return reconcileSplitNodes(root);\n};\n\nexport const assembleJSON = (chunks: JsonChunk[]): JSONObject | JSONArray => {\n  if (!chunks || chunks.length === 0) {\n    throw new Error('No chunks provided.');\n  }\n\n  // Basic validation & sort\n  const sorted = [...chunks].sort((a, b) => a.index - b.index);\n  const firstChunk = sorted[0];\n\n  if (!firstChunk) {\n    throw new Error('No chunks provided.');\n  }\n\n  const { checksum, rootType } = firstChunk;\n  const schemaVersion = 1;\n\n  for (let i = 0; i < sorted.length; i++) {\n    const chunk = sorted[i];\n\n    if (!chunk) continue;\n\n    if (chunk.schemaVersion !== schemaVersion) {\n      console.error('Unsupported schemaVersion.', {\n        cause: chunk,\n        schemaVersion,\n      });\n      throw new Error('Unsupported schemaVersion.');\n    }\n\n    if (chunk.rootType !== rootType) {\n      console.error('Chunks rootType mismatch.', {\n        cause: chunk,\n        rootType,\n      });\n      throw new Error('Chunks rootType mismatch.');\n    }\n\n    if (chunk.checksum !== checksum) {\n      console.error('Chunks checksum mismatch (different source objects?).', {\n        cause: chunk,\n        checksum,\n      });\n      throw new Error('Chunks checksum mismatch (different source objects?).');\n    }\n\n    if (chunk.index !== i) {\n      console.error('Chunk indices are not contiguous or sorted.', {\n        cause: chunk,\n        index: chunk.index,\n        i,\n      });\n      throw new Error('Chunk indices are not contiguous or sorted.');\n    }\n\n    // Defer total check until after reconstruction to prefer more specific errors\n  }\n\n  const root: any = rootType === 'array' ? [] : {};\n\n  // Collect string parts by path\n  const stringParts = new Map<\n    string,\n    { path: Path; total: number; received: StrAppendPatch[] }\n  >();\n\n  const applySet = (patch: SetPatch) =>\n    setAtPath(root, patch.path, patch.value);\n\n  for (const chunk of sorted) {\n    for (const entry of chunk.entries) {\n      if (entry.op === 'set') {\n        applySet(entry);\n      } else {\n        const key = pathKey(entry.path);\n        const record = stringParts.get(key) ?? {\n          path: entry.path,\n          total: entry.total,\n          received: [],\n        };\n\n        if (record.total !== entry.total) {\n          throw new Error('Inconsistent string part totals for a path.');\n        }\n\n        record.received.push(entry);\n        stringParts.set(key, record);\n      }\n    }\n  }\n\n  // Stitch strings\n  for (const { path, total, received } of stringParts.values()) {\n    if (received.length !== total) {\n      throw new Error('Missing string parts for a path; incomplete chunk set.');\n    }\n\n    received.sort((a, b) => a.index - b.index);\n    const fullString = received.map((part) => part.value).join('');\n    setAtPath(root, path, fullString);\n  }\n\n  // Reconcile split-node sentinels for strings/arrays after all patches applied\n  const reconcileSplitNodes = (node: any): any => {\n    if (node === null || typeof node !== 'object') return node;\n\n    if (Array.isArray(node)) {\n      for (let i = 0; i < node.length; i++) {\n        node[i] = reconcileSplitNodes(node[i]);\n      }\n      return node;\n    }\n\n    // string split-node\n    if ((node as any)['__splittedType'] === 'string') {\n      const total = (node as any)['__total'];\n\n      if (typeof total !== 'number' || total <= 0) {\n        throw new Error('Invalid split-node total for a path.');\n      }\n\n      const parts: string[] = [];\n\n      for (let i = 1; i <= total; i++) {\n        const piece = (node as any)[String(i)];\n\n        if (typeof piece !== 'string') {\n          throw new Error(\n            'Missing string parts for a path; incomplete chunk set.'\n          );\n        }\n\n        parts.push(piece);\n      }\n\n      return parts.join('');\n    }\n\n    // array split-node (optional support)\n    if ((node as any)['__splittedType'] === 'array') {\n      const total = (node as any)['__total'];\n\n      if (typeof total !== 'number' || total < 0) {\n        throw new Error('Invalid split-node total for a path.');\n      }\n\n      const output: any[] = [];\n\n      for (let i = 1; i <= total; i++) {\n        const slice = (node as any)[String(i)];\n\n        if (!Array.isArray(slice)) {\n          throw new Error(\n            'Missing string parts for a path; incomplete chunk set.'\n          );\n        }\n\n        for (let j = 0; j < slice.length; j++) {\n          output.push(reconcileSplitNodes(slice[j]));\n        }\n      }\n\n      return output;\n    }\n\n    // walk normal object\n    for (const key of Object.keys(node)) {\n      node[key] = reconcileSplitNodes(node[key]);\n    }\n\n    return node;\n  };\n\n  const reconciled = reconcileSplitNodes(root);\n\n  // Now validate totals match provided count\n  for (let i = 0; i < sorted.length; i++) {\n    const chunk = sorted[i];\n\n    if (!chunk) continue;\n\n    if (chunk.total !== sorted.length) {\n      throw new Error(\n        `Chunk total does not match provided count. Expected ${sorted.length}, but chunk ${i} has total=${chunk.total}`\n      );\n    }\n  }\n\n  return reconciled;\n};\n\n/* -------------------------------------------\n * Example usage\n * -------------------------------------------\nconst big: JSONObject = {\n  title: \"Document\",\n  content: \"…a very very long text…\",\n  items: Array.from({ length: 2000 }, (_, i) => ({ id: i, label: `Item ${i}` }))\n};\n\n// Split to ~16k-char chunks\nconst chunks = chunkJSON(big, 16_000);\n\n// Send each `chunks[i]` as JSON to your backend.\n// Later, reassemble:\nconst restored = assembleJSON(chunks);\nconsole.log(JSON.stringify(restored) === JSON.stringify(big)); // true\n*/\n"],"mappings":";;;;;;;;;;;AA0CA,MAAM,YAAY,QAAiD;CACjE,OAAO,OAAO,QAAQ,YAAY,QAAQ,QAAQ,CAAC,MAAM,QAAQ,GAAG;AACtE;AAEA,MAAM,eAAe,QAAwB;CAE3C,IAAI,OAAO;CAEX,KAAK,IAAI,IAAI,GAAG,IAAI,IAAI,QAAQ,KAC9B,QAAS,QAAQ,KAAK,OAAQ,IAAI,WAAW,CAAC;CAIhD,QAAQ,SAAS,EAAC,CAAE,SAAS,EAAE,CAAC,CAAC,SAAS,GAAG,GAAG;AAClD;AAEA,MAAM,aAAa,MAAW,MAAY,UAAqB;CAC7D,IAAI,UAAU;CAEd,KAAK,IAAI,IAAI,GAAG,IAAI,KAAK,SAAS,GAAG,KAAK;EACxC,MAAM,MAAM,KAAK;EACjB,IAAI,QAAQ,QAAW;EAEvB,MAAM,cAAc,OAAO,YAAY;EAEvC,IAAI,OAAO,QAAQ,UAAU;GAC3B,IAAI,CAAC,MAAM,QAAQ,OAAO,GACxB,MAAM,IAAI,MAAM,kCAAkC,GAAG;GAGvD,IAAI,QAAQ,SAAS,QACnB,QAAQ,OAAO,cAAc,CAAC,IAAI,CAAC;GAGrC,UAAU,QAAQ;EACpB,OAAO;GACL,IAAI,CAAC,SAAS,OAAO,GACnB,MAAM,IAAI,MAAM,mCAAmC,GAAG;GAGxD,IAAI,EAAE,OAAO,UACX,AAAC,QAAgC,OAAO,cAAc,CAAC,IAAI,CAAC;GAG9D,UAAW,QAAgC;EAC7C;CACF;CAEA,MAAM,OAAO,KAAK,KAAK,SAAS;CAEhC,IAAI,SAAS,QAAW;CAExB,IAAI,OAAO,SAAS,UAAU;EAC5B,IAAI,CAAC,MAAM,QAAQ,OAAO,GACxB,MAAM,IAAI,MAAM,iCAAiC;EAGnD,QAAQ,QAAQ;CAClB,OAAO;EACL,IAAI,CAAC,SAAS,OAAO,GACnB,MAAM,IAAI,MAAM,kCAAkC;EAGpD,AAAC,QAAgC,QAAQ;CAC3C;AACF;AAEA,MAAM,WAAW,SAAuB;CAEtC,OAAO,KAAK,UAAU,IAAI;AAC5B;;;;AAKA,MAAM,uBACJ,KACA,oBACa;CACb,IAAI,mBAAmB,GACrB,MAAM,IAAI,MAAM,sBAAsB;CAGxC,MAAM,SAAmB,CAAC;CAC1B,IAAI,SAAS;CAEb,OAAO,SAAS,IAAI,QAAQ;EAC1B,MAAM,OAAOA,gCAAS,KAAK;GACzB,WAAW;GACX,YAAY;EACd,CAAC;EAED,IAAI,CAAC,MAAM;EAEX,OAAO,KAAK,IAAI;EAChB,UAAU,KAAK;CACjB;CAEA,OAAO;AACT;;;;;AAMA,MAAM,oBACJ,OACA,kBACA,OAAa,CAAC,GACd,uBAAO,IAAI,QAAgB,MACf;CAGZ,MAAM,iBAAiB,KAAK,IAC1B,GACA,KAAK,OAAO,mBAAmB,OAAO,EAAG,CAC3C;CAEA,MAAM,UAAmB,CAAC;CAE1B,MAAM,QAAQ,cAAyB,gBAAsB;EAC3D,IAAI,OAAO,iBAAiB,UAAU;GAGpC,MAAM,YAAsB;IAC1B,IAAI;IACJ,MAAM;IACN,OAAO;GACT;GAGA,IAFgB,KAAK,UAAU,SAAS,CAAC,CAAC,SAAS,OAEpC,kBAAkB;IAC/B,QAAQ,KAAK,SAAS;IACtB;GACF;GAGA,MAAM,QAAQ,oBAAoB,cAAc,cAAc;GAG9D,QAAQ,KAAK;IACX,IAAI;IACJ,MAAM,CAAC,GAAG,aAAa,gBAAgB;IACvC,OAAO;GACT,CAAC;GACD,QAAQ,KAAK;IACX,IAAI;IACJ,MAAM,CAAC,GAAG,aAAa,SAAS;IAChC,OAAO,MAAM;GACf,CAAC;GAED,KAAK,IAAI,IAAI,GAAG,IAAI,MAAM,QAAQ,KAAK;IACrC,MAAM,OAAO,MAAM;IAEnB,IAAI,SAAS,QACX,QAAQ,KAAK;KACX,IAAI;KACJ,MAAM,CAAC,GAAG,aAAa,OAAO,IAAI,CAAC,CAAC;KACpC,OAAO;IACT,CAAC;GAEL;GAEA;EACF;EAEA,IAAI,iBAAiB,QAAQ,OAAO,iBAAiB,UAAU;GAC7D,QAAQ,KAAK;IAAE,IAAI;IAAO,MAAM;IAAa,OAAO;GAAa,CAAC;GAClE;EACF;EAEA,IAAI,KAAK,IAAI,YAAsB,GACjC,MAAM,IAAI,MAAM,+CAA+C;EAGjE,KAAK,IAAI,YAAsB;EAE/B,IAAI,MAAM,QAAQ,YAAY,GAC5B,KAAK,IAAI,IAAI,GAAG,IAAI,aAAa,QAAQ,KAAK;GAC5C,MAAM,OAAO,aAAa;GAC1B,IAAI,SAAS,QACX,KAAK,MAAmB,CAAC,GAAG,aAAa,CAAC,CAAC;EAE/C;OAEA,KAAK,MAAM,OAAO,OAAO,KAAK,YAAY,GAAG;GAC3C,MAAM,WAAY,aAA4B;GAE9C,IAAI,aAAa,QACf,KAAK,UAAU,CAAC,GAAG,aAAa,GAAG,CAAC;EAExC;EAGF,KAAK,OAAO,YAAsB;CACpC;CAEA,KAAK,OAAO,IAAI;CAChB,OAAO;AACT;;;;AAKA,MAAa,aACX,OACA,aACgB;CAChB,IAAI,CAAC,SAAS,KAAK,KAAK,CAAC,MAAM,QAAQ,KAAK,GAC1C,MAAM,IAAI,MAAM,kCAAkC;CAGpD,IAAI,WAAW,KAEb,MAAM,IAAI,MAAM,qDAAqD;CAGvE,MAAM,WAAqB,MAAM,QAAQ,KAAK,IAAI,UAAU;CAC5D,IAAI;CAEJ,IAAI;EACF,eAAe,KAAK,UAAU,KAAK;CACrC,QAAQ;EAEN,MAAM,IAAI,MAAM,+CAA+C;CACjE;CAEA,MAAM,WAAW,YAAY,YAAY;CACzC,MAAM,aAAa,iBAAiB,OAAoB,QAAQ;CAEhE,MAAM,SAAsB,CAAC;CAC7B,IAAI,eAA0B;EAC5B,eAAe;EACf,OAAO;EACP,OAAO;EACP;EACA;EACA,SAAS,CAAC;CACZ;CAEA,MAAM,oBAAoB,KAAK,UAAU;EACvC,GAAG;EACH,SAAS,CAAC;CACZ,CAAC,CAAC,CAAC;CAEH,MAAM,iBAAiB;EACrB,IAAI,aAAa,QAAQ,SAAS,GAAG;GACnC,OAAO,KAAK,YAAY;GACxB,eAAe;IACb,eAAe;IACf,OAAO;IACP,OAAO;IACP;IACA;IACA,SAAS,CAAC;GACZ;EACF;CACF;CAEA,KAAK,MAAM,SAAS,YAQlB,IALE,oBACA,KAAK,UAAU,aAAa,OAAO,CAAC,CAAC,UACpC,aAAa,QAAQ,SAAS,IAAI,KACnC,KAAK,UAAU,KAAK,CAAC,CAAC,UAEH,UACnB,aAAa,QAAQ,KAAK,KAAK;MAC1B;EAEL,IAAI,aAAa,QAAQ,SAAS,GAChC,SAAS;EAMX,IAFmB,oBAAoB,KAAK,UAAU,CAAC,KAAK,CAAC,CAAC,CAAC,SAE9C,UAGf,MAAM,IAAI,MACR,8FACF;EAGF,aAAa,QAAQ,KAAK,KAAK;CACjC;CAGF,SAAS;CAGT,IAAI,OAAO,WAAW,GACpB,OAAO,KAAK;EACV,eAAe;EACf,OAAO;EACP,OAAO;EACP;EACA;EACA,SAAS,CAAC;CACZ,CAAC;CAIH,MAAM,cAAc,OAAO;CAE3B,OAAO,SAAS,OAAO,UAAU;EAC/B,MAAM,QAAQ;EACd,MAAM,QAAQ;CAChB,CAAC;CAED,OAAO;AACT;;;;;;;;;;;AAYA,MAAa,8BACX,UAC2B;CAC3B,MAAM,OAAY,MAAM,aAAa,UAAU,CAAC,IAAI,CAAC;CAGrD,KAAK,MAAM,SAAS,MAAM,SACxB,IAAI,MAAM,OAAO,OACf,UAAU,MAAM,MAAM,MAAM,MAAM,KAAK;CAM3C,MAAM,uBAAuB,SAAmB;EAC9C,IAAI,SAAS,QAAQ,OAAO,SAAS,UAAU,OAAO;EAEtD,IAAI,MAAM,QAAQ,IAAI,GAAG;GACvB,KAAK,IAAI,IAAI,GAAG,IAAI,KAAK,QAAQ,KAC/B,KAAK,KAAK,oBAAoB,KAAK,EAAE;GAEvC,OAAO;EACT;EAGA,IAAK,KAAa,sBAAsB,UAAU;GAChD,MAAM,QAAS,KAAa;GAE5B,IAAI,OAAO,UAAU,YAAY,SAAS,GAExC,OAAO;GAGT,MAAM,QAAkB,CAAC;GACzB,IAAI,cAAc;GAElB,KAAK,IAAI,IAAI,GAAG,KAAK,OAAO,KAAK;IAC/B,MAAM,QAAS,KAAa,OAAO,CAAC;IAEpC,IAAI,OAAO,UAAU,UAAU;KAC7B,cAAc;KACd;IACF;IAEA,MAAM,KAAK,KAAK;GAClB;GAGA,IAAI,aACF,OAAO,MAAM,KAAK,EAAE;GAGtB,OAAO;EACT;EAGA,IAAK,KAAa,sBAAsB,SAAS;GAC/C,MAAM,QAAS,KAAa;GAE5B,IAAI,OAAO,UAAU,YAAY,QAAQ,GAEvC,OAAO;GAGT,MAAM,SAAgB,CAAC;GACvB,IAAI,cAAc;GAElB,KAAK,IAAI,IAAI,GAAG,KAAK,OAAO,KAAK;IAC/B,MAAM,QAAS,KAAa,OAAO,CAAC;IAEpC,IAAI,CAAC,MAAM,QAAQ,KAAK,GAAG;KACzB,cAAc;KACd;IACF;IAEA,KAAK,IAAI,IAAI,GAAG,IAAI,MAAM,QAAQ,KAChC,OAAO,KAAK,oBAAoB,MAAM,EAAE,CAAC;GAE7C;GAGA,IAAI,aACF,OAAO;GAGT,OAAO;EACT;EAGA,KAAK,MAAM,OAAO,OAAO,KAAK,IAAI,GAChC,KAAK,OAAO,oBAAoB,KAAK,IAAI;EAG3C,OAAO;CACT;CAEA,OAAO,oBAAoB,IAAI;AACjC;AAEA,MAAa,gBAAgB,WAAgD;CAC3E,IAAI,CAAC,UAAU,OAAO,WAAW,GAC/B,MAAM,IAAI,MAAM,qBAAqB;CAIvC,MAAM,SAAS,CAAC,GAAG,MAAM,CAAC,CAAC,MAAM,GAAG,MAAM,EAAE,QAAQ,EAAE,KAAK;CAC3D,MAAM,aAAa,OAAO;CAE1B,IAAI,CAAC,YACH,MAAM,IAAI,MAAM,qBAAqB;CAGvC,MAAM,EAAE,UAAU,aAAa;CAC/B,MAAM,gBAAgB;CAEtB,KAAK,IAAI,IAAI,GAAG,IAAI,OAAO,QAAQ,KAAK;EACtC,MAAM,QAAQ,OAAO;EAErB,IAAI,CAAC,OAAO;EAEZ,IAAI,MAAM,kBAAkB,eAAe;GACzC,QAAQ,MAAM,8BAA8B;IAC1C,OAAO;IACP;GACF,CAAC;GACD,MAAM,IAAI,MAAM,4BAA4B;EAC9C;EAEA,IAAI,MAAM,aAAa,UAAU;GAC/B,QAAQ,MAAM,6BAA6B;IACzC,OAAO;IACP;GACF,CAAC;GACD,MAAM,IAAI,MAAM,2BAA2B;EAC7C;EAEA,IAAI,MAAM,aAAa,UAAU;GAC/B,QAAQ,MAAM,yDAAyD;IACrE,OAAO;IACP;GACF,CAAC;GACD,MAAM,IAAI,MAAM,uDAAuD;EACzE;EAEA,IAAI,MAAM,UAAU,GAAG;GACrB,QAAQ,MAAM,+CAA+C;IAC3D,OAAO;IACP,OAAO,MAAM;IACb;GACF,CAAC;GACD,MAAM,IAAI,MAAM,6CAA6C;EAC/D;CAGF;CAEA,MAAM,OAAY,aAAa,UAAU,CAAC,IAAI,CAAC;CAG/C,MAAM,8BAAc,IAAI,IAGtB;CAEF,MAAM,YAAY,UAChB,UAAU,MAAM,MAAM,MAAM,MAAM,KAAK;CAEzC,KAAK,MAAM,SAAS,QAClB,KAAK,MAAM,SAAS,MAAM,SACxB,IAAI,MAAM,OAAO,OACf,SAAS,KAAK;MACT;EACL,MAAM,MAAM,QAAQ,MAAM,IAAI;EAC9B,MAAM,SAAS,YAAY,IAAI,GAAG,KAAK;GACrC,MAAM,MAAM;GACZ,OAAO,MAAM;GACb,UAAU,CAAC;EACb;EAEA,IAAI,OAAO,UAAU,MAAM,OACzB,MAAM,IAAI,MAAM,6CAA6C;EAG/D,OAAO,SAAS,KAAK,KAAK;EAC1B,YAAY,IAAI,KAAK,MAAM;CAC7B;CAKJ,KAAK,MAAM,EAAE,MAAM,OAAO,cAAc,YAAY,OAAO,GAAG;EAC5D,IAAI,SAAS,WAAW,OACtB,MAAM,IAAI,MAAM,wDAAwD;EAG1E,SAAS,MAAM,GAAG,MAAM,EAAE,QAAQ,EAAE,KAAK;EACzC,MAAM,aAAa,SAAS,KAAK,SAAS,KAAK,KAAK,CAAC,CAAC,KAAK,EAAE;EAC7D,UAAU,MAAM,MAAM,UAAU;CAClC;CAGA,MAAM,uBAAuB,SAAmB;EAC9C,IAAI,SAAS,QAAQ,OAAO,SAAS,UAAU,OAAO;EAEtD,IAAI,MAAM,QAAQ,IAAI,GAAG;GACvB,KAAK,IAAI,IAAI,GAAG,IAAI,KAAK,QAAQ,KAC/B,KAAK,KAAK,oBAAoB,KAAK,EAAE;GAEvC,OAAO;EACT;EAGA,IAAK,KAAa,sBAAsB,UAAU;GAChD,MAAM,QAAS,KAAa;GAE5B,IAAI,OAAO,UAAU,YAAY,SAAS,GACxC,MAAM,IAAI,MAAM,sCAAsC;GAGxD,MAAM,QAAkB,CAAC;GAEzB,KAAK,IAAI,IAAI,GAAG,KAAK,OAAO,KAAK;IAC/B,MAAM,QAAS,KAAa,OAAO,CAAC;IAEpC,IAAI,OAAO,UAAU,UACnB,MAAM,IAAI,MACR,wDACF;IAGF,MAAM,KAAK,KAAK;GAClB;GAEA,OAAO,MAAM,KAAK,EAAE;EACtB;EAGA,IAAK,KAAa,sBAAsB,SAAS;GAC/C,MAAM,QAAS,KAAa;GAE5B,IAAI,OAAO,UAAU,YAAY,QAAQ,GACvC,MAAM,IAAI,MAAM,sCAAsC;GAGxD,MAAM,SAAgB,CAAC;GAEvB,KAAK,IAAI,IAAI,GAAG,KAAK,OAAO,KAAK;IAC/B,MAAM,QAAS,KAAa,OAAO,CAAC;IAEpC,IAAI,CAAC,MAAM,QAAQ,KAAK,GACtB,MAAM,IAAI,MACR,wDACF;IAGF,KAAK,IAAI,IAAI,GAAG,IAAI,MAAM,QAAQ,KAChC,OAAO,KAAK,oBAAoB,MAAM,EAAE,CAAC;GAE7C;GAEA,OAAO;EACT;EAGA,KAAK,MAAM,OAAO,OAAO,KAAK,IAAI,GAChC,KAAK,OAAO,oBAAoB,KAAK,IAAI;EAG3C,OAAO;CACT;CAEA,MAAM,aAAa,oBAAoB,IAAI;CAG3C,KAAK,IAAI,IAAI,GAAG,IAAI,OAAO,QAAQ,KAAK;EACtC,MAAM,QAAQ,OAAO;EAErB,IAAI,CAAC,OAAO;EAEZ,IAAI,MAAM,UAAU,OAAO,QACzB,MAAM,IAAI,MACR,uDAAuD,OAAO,OAAO,cAAc,EAAE,aAAa,MAAM,OAC1G;CAEJ;CAEA,OAAO;AACT"}

package/dist/cjs/utils/getChunk.cjs

@@ -9,13 +9,13 @@ const getChunk = (text, options = {}) => {
 	const getCharIndexOfLineStart = (lineNumber) => {
 		if (lineNumber <= 0) return 0;
 		let idx = 0;
-		for (let i = 0; i < Math.min(lineNumber, lines.length); i++) idx += lines[i].length;
+		for (let i = 0; i < Math.min(lineNumber, lines.length); i++) idx += lines[i]?.length ?? 0;
 		return idx;
 	};
 	const getCharIndexOfLineEnd = (lineNumber) => {
 		if (lineNumber >= lines.length) return text.length;
 		const line = lines[lineNumber];
-		return getCharIndexOfLineStart(lineNumber) + line.length;
+		return getCharIndexOfLineStart(lineNumber) + (line?.length ?? 0);
 	};
 	let effectiveStart = 0;
 	let effectiveEnd = text.length;

package/dist/cjs/utils/getChunk.cjs.map

@@ -1 +1 @@
-{"version":3,"file":"getChunk.cjs","names":["splitTextByLines"],"sources":["../../../src/utils/getChunk.ts"],"sourcesContent":["import { splitTextByLines } from './splitTextByLine';\n\ntype TrunkOptions = {\n  lineStart?: number;\n  lineLength?: number;\n  charStart?: number;\n  charLength?: number;\n};\n\nexport const getChunk = (text: string, options: TrunkOptions = {}): string => {\n  const { lineStart, lineLength, charStart, charLength } = options;\n\n  // Fast-path: if no filters were provided, return the whole text\n  if (\n    lineStart === undefined &&\n    lineLength === undefined &&\n    charStart === undefined &&\n    charLength === undefined\n  ) {\n    return text;\n  }\n\n  // ---------------------------------------------------------------------------\n  // Utility helpers to convert between line numbers and absolute char indices\n  // ---------------------------------------------------------------------------\n  const lines = splitTextByLines(text);\n\n  const getCharIndexOfLineStart = (lineNumber: number): number => {\n    if (lineNumber <= 0) return 0;\n    // Sum the length of every previous line\n    let idx = 0;\n    for (let i = 0; i < Math.min(lineNumber, lines.length); i++) {\n      idx += lines[i].length;\n    }\n    return idx;\n  };\n\n  const getCharIndexOfLineEnd = (lineNumber: number): number => {\n    // If the requested line number exceeds the number of lines, clamp to the last character\n    if (lineNumber >= lines.length) {\n      return text.length;\n    }\n    const line = lines[lineNumber];\n    const lineEnd = getCharIndexOfLineStart(lineNumber) + line.length;\n\n    return lineEnd;\n  };\n\n  // ---------------------------------------------------------------------------\n  // Compute the effective (inclusive) charStart/charEnd for the requested slice\n  // ---------------------------------------------------------------------------\n  let effectiveStart = 0; // inclusive\n  let effectiveEnd = text.length; // exclusive\n\n  // Apply line boundaries if provided\n  if (lineStart !== undefined) {\n    effectiveStart = Math.max(\n      effectiveStart,\n      getCharIndexOfLineStart(lineStart)\n    );\n  }\n\n  // Apply character boundaries if provided\n  if (charStart !== undefined) {\n    effectiveStart = Math.max(effectiveStart, charStart);\n  }\n\n  // Apply line length boundary from lineStart (or 0)\n  if (lineLength !== undefined) {\n    const endLine = (lineStart ?? 0) + lineLength - 1;\n    effectiveEnd = Math.min(effectiveEnd, getCharIndexOfLineEnd(endLine));\n  }\n\n  // Apply character length boundary from effectiveStart\n  if (charLength !== undefined) {\n    effectiveEnd = Math.min(\n      effectiveEnd,\n      (charStart ?? effectiveStart) + charLength\n    );\n  }\n\n  // If bounds do not overlap, return empty string\n  if (effectiveStart >= effectiveEnd) {\n    return '';\n  }\n\n  return text.slice(effectiveStart, effectiveEnd);\n};\n"],"mappings":";;;;AASA,MAAa,YAAY,MAAc,UAAwB,CAAC,MAAc;CAC5E,MAAM,EAAE,WAAW,YAAY,WAAW,eAAe;CAGzD,IACE,cAAc,UACd,eAAe,UACf,cAAc,UACd,eAAe,QAEf,OAAO;CAMT,MAAM,QAAQA,+CAAiB,IAAI;CAEnC,MAAM,2BAA2B,eAA+B;EAC9D,IAAI,cAAc,GAAG,OAAO;EAE5B,IAAI,MAAM;EACV,KAAK,IAAI,IAAI,GAAG,IAAI,KAAK,IAAI,YAAY,MAAM,MAAM,GAAG,KACtD,OAAO,MAAM,EAAE,CAAC;EAElB,OAAO;CACT;CAEA,MAAM,yBAAyB,eAA+B;EAE5D,IAAI,cAAc,MAAM,QACtB,OAAO,KAAK;EAEd,MAAM,OAAO,MAAM;EAGnB,OAFgB,wBAAwB,UAAU,IAAI,KAAK;CAG7D;CAKA,IAAI,iBAAiB;CACrB,IAAI,eAAe,KAAK;CAGxB,IAAI,cAAc,QAChB,iBAAiB,KAAK,IACpB,gBACA,wBAAwB,SAAS,CACnC;CAIF,IAAI,cAAc,QAChB,iBAAiB,KAAK,IAAI,gBAAgB,SAAS;CAIrD,IAAI,eAAe,QAAW;EAC5B,MAAM,WAAW,aAAa,KAAK,aAAa;EAChD,eAAe,KAAK,IAAI,cAAc,sBAAsB,OAAO,CAAC;CACtE;CAGA,IAAI,eAAe,QACjB,eAAe,KAAK,IAClB,eACC,aAAa,kBAAkB,UAClC;CAIF,IAAI,kBAAkB,cACpB,OAAO;CAGT,OAAO,KAAK,MAAM,gBAAgB,YAAY;AAChD"}
+{"version":3,"file":"getChunk.cjs","names":["splitTextByLines"],"sources":["../../../src/utils/getChunk.ts"],"sourcesContent":["import { splitTextByLines } from './splitTextByLine';\n\ntype TrunkOptions = {\n  lineStart?: number;\n  lineLength?: number;\n  charStart?: number;\n  charLength?: number;\n};\n\nexport const getChunk = (text: string, options: TrunkOptions = {}): string => {\n  const { lineStart, lineLength, charStart, charLength } = options;\n\n  // Fast-path: if no filters were provided, return the whole text\n  if (\n    lineStart === undefined &&\n    lineLength === undefined &&\n    charStart === undefined &&\n    charLength === undefined\n  ) {\n    return text;\n  }\n\n  // ---------------------------------------------------------------------------\n  // Utility helpers to convert between line numbers and absolute char indices\n  // ---------------------------------------------------------------------------\n  const lines = splitTextByLines(text);\n\n  const getCharIndexOfLineStart = (lineNumber: number): number => {\n    if (lineNumber <= 0) return 0;\n    // Sum the length of every previous line\n    let idx = 0;\n    for (let i = 0; i < Math.min(lineNumber, lines.length); i++) {\n      idx += lines[i]?.length ?? 0;\n    }\n    return idx;\n  };\n\n  const getCharIndexOfLineEnd = (lineNumber: number): number => {\n    // If the requested line number exceeds the number of lines, clamp to the last character\n    if (lineNumber >= lines.length) {\n      return text.length;\n    }\n    const line = lines[lineNumber];\n    const lineEnd = getCharIndexOfLineStart(lineNumber) + (line?.length ?? 0);\n\n    return lineEnd;\n  };\n\n  // ---------------------------------------------------------------------------\n  // Compute the effective (inclusive) charStart/charEnd for the requested slice\n  // ---------------------------------------------------------------------------\n  let effectiveStart = 0; // inclusive\n  let effectiveEnd = text.length; // exclusive\n\n  // Apply line boundaries if provided\n  if (lineStart !== undefined) {\n    effectiveStart = Math.max(\n      effectiveStart,\n      getCharIndexOfLineStart(lineStart)\n    );\n  }\n\n  // Apply character boundaries if provided\n  if (charStart !== undefined) {\n    effectiveStart = Math.max(effectiveStart, charStart);\n  }\n\n  // Apply line length boundary from lineStart (or 0)\n  if (lineLength !== undefined) {\n    const endLine = (lineStart ?? 0) + lineLength - 1;\n    effectiveEnd = Math.min(effectiveEnd, getCharIndexOfLineEnd(endLine));\n  }\n\n  // Apply character length boundary from effectiveStart\n  if (charLength !== undefined) {\n    effectiveEnd = Math.min(\n      effectiveEnd,\n      (charStart ?? effectiveStart) + charLength\n    );\n  }\n\n  // If bounds do not overlap, return empty string\n  if (effectiveStart >= effectiveEnd) {\n    return '';\n  }\n\n  return text.slice(effectiveStart, effectiveEnd);\n};\n"],"mappings":";;;;AASA,MAAa,YAAY,MAAc,UAAwB,CAAC,MAAc;CAC5E,MAAM,EAAE,WAAW,YAAY,WAAW,eAAe;CAGzD,IACE,cAAc,UACd,eAAe,UACf,cAAc,UACd,eAAe,QAEf,OAAO;CAMT,MAAM,QAAQA,+CAAiB,IAAI;CAEnC,MAAM,2BAA2B,eAA+B;EAC9D,IAAI,cAAc,GAAG,OAAO;EAE5B,IAAI,MAAM;EACV,KAAK,IAAI,IAAI,GAAG,IAAI,KAAK,IAAI,YAAY,MAAM,MAAM,GAAG,KACtD,OAAO,MAAM,EAAE,EAAE,UAAU;EAE7B,OAAO;CACT;CAEA,MAAM,yBAAyB,eAA+B;EAE5D,IAAI,cAAc,MAAM,QACtB,OAAO,KAAK;EAEd,MAAM,OAAO,MAAM;EAGnB,OAFgB,wBAAwB,UAAU,KAAK,MAAM,UAAU;CAGzE;CAKA,IAAI,iBAAiB;CACrB,IAAI,eAAe,KAAK;CAGxB,IAAI,cAAc,QAChB,iBAAiB,KAAK,IACpB,gBACA,wBAAwB,SAAS,CACnC;CAIF,IAAI,cAAc,QAChB,iBAAiB,KAAK,IAAI,gBAAgB,SAAS;CAIrD,IAAI,eAAe,QAAW;EAC5B,MAAM,WAAW,aAAa,KAAK,aAAa;EAChD,eAAe,KAAK,IAAI,cAAc,sBAAsB,OAAO,CAAC;CACtE;CAGA,IAAI,eAAe,QACjB,eAAe,KAAK,IAClB,eACC,aAAa,kBAAkB,UAClC;CAIF,IAAI,kBAAkB,cACpB,OAAO;CAGT,OAAO,KAAK,MAAM,gBAAgB,YAAY;AAChD"}