@oh-my-pi/pi-utils 17.2.9 → 17.2.10

package/src/marked.ts

@@ -0,0 +1,2 @@
+/** Behavior-compatible reimplementation of marked's used surface. */
+export * from "./marked/core";

package/src/postmortem.ts

@@ -67,6 +67,17 @@ function exitProcess(code: number): never {
 let cleanupPromise: Promise<void> | undefined;
 let stdioDisconnectRegistrations = 0;
 
+/** User-facing command printed before fatal cleanup so interrupted work can be resumed. */
+export interface FatalRecoveryHint {
+	/** Stable label identifying the recoverable session or process. */
+	label: string;
+	/** Complete shell command the user can execute to resume the interrupted work. */
+	command: string;
+}
+
+type FatalRecoveryHintProvider = () => FatalRecoveryHint | undefined;
+const fatalRecoveryHintProviders = new Set<FatalRecoveryHintProvider>();
+
 /**
  * Internal: runs all registered cleanup callbacks for the given reason.
  * Ensures each callback is invoked at most once. Handles errors and prevents reentrancy.
@@ -196,6 +207,38 @@ export function interceptUnhandledRejections(interceptor: (reason: unknown) => b
 	return () => rejectionInterceptors.delete(interceptor);
 }
 
+/**
+ * Register a synchronous recovery command to print when the process exits
+ * through an uncaught exception or unhandled rejection.
+ */
+export function registerFatalRecoveryHint(provider: FatalRecoveryHintProvider): () => void {
+	fatalRecoveryHintProviders.add(provider);
+	return () => fatalRecoveryHintProviders.delete(provider);
+}
+
+function escapeFatalHintText(value: string): string {
+	return value.replace(/[\u0000-\u001f\u007f-\u009f]/gu, char => {
+		const code = char.codePointAt(0) ?? 0;
+		return `\\u${code.toString(16).padStart(4, "0")}`;
+	});
+}
+
+function formatFatalRecoveryHints(): string {
+	const lines: string[] = [];
+	const seenCommands = new Set<string>();
+	for (const provider of fatalRecoveryHintProviders) {
+		try {
+			const hint = provider();
+			if (!hint?.command || seenCommands.has(hint.command)) continue;
+			seenCommands.add(hint.command);
+			lines.push(`  ${escapeFatalHintText(hint.label)}: ${escapeFatalHintText(hint.command)}`);
+		} catch (err) {
+			logger.warn("Fatal recovery hint provider failed", { err });
+		}
+	}
+	return lines.length > 0 ? `\n[Recovery]\n${lines.join("\n")}\n` : "";
+}
+
 function formatFatalError(label: string, err: Error): string {
 	const name = err.name || "Error";
 	const message = err.message || "(no message)";
@@ -212,7 +255,7 @@ async function exitAfterFatal(label: string, logMessage: string, err: Error, rea
 		// A revoked terminal can make stream writes raise another fatal error. Use
 		// the descriptor directly so failure stays synchronous and contained.
 		try {
-			fs.writeSync(2, formatFatalError(label, err));
+			fs.writeSync(2, `${formatFatalError(label, err)}${formatFatalRecoveryHints()}`);
 		} catch {}
 		logger.error(logMessage, { err });
 		await runCleanup(reason);

package/src/procmgr.ts

@@ -50,12 +50,19 @@ function buildSpawnEnv(shell: string): Record<string, string> {
 
 /**
  * Get shell args for the resolved shell.
- * cmd.exe takes `/c`; POSIX shells take `-c`, with `-l` unless
- * PI_BASH_NO_LOGIN / CLAUDE_BASH_NO_LOGIN is set.
+ * cmd.exe takes `/c`; PowerShell (powershell.exe / pwsh) takes
+ * `-NoLogo -Command`, with `-NoProfile` when PI_BASH_NO_LOGIN /
+ * CLAUDE_BASH_NO_LOGIN is set (profile scripts are PowerShell's login-shell
+ * analog); POSIX shells take `-c`, with `-l` unless the same env is set.
+ *
+ * Exported for tests; `env` overrides the process env gate.
  */
-function getShellArgs(shell: string): string[] {
+export function getShellArgs(shell: string, env: Record<string, string | undefined> = $env): string[] {
 	if (isCmdShell(shell)) return ["/c"];
-	const noLogin = $env.PI_BASH_NO_LOGIN || $env.CLAUDE_BASH_NO_LOGIN;
+	const noLogin = env.PI_BASH_NO_LOGIN || env.CLAUDE_BASH_NO_LOGIN;
+	if (isPowerShell(shell)) {
+		return noLogin ? ["-NoLogo", "-NoProfile", "-Command"] : ["-NoLogo", "-Command"];
+	}
 	return noLogin ? ["-c"] : ["-l", "-c"];
 }
 
@@ -65,6 +72,16 @@ export function isCmdShell(shell: string): boolean {
 	return basename === "cmd.exe" || basename === "cmd";
 }
 
+/**
+ * Whether the shell is Windows PowerShell or PowerShell Core (pwsh). Spawn
+ * paths must use `-Command`: passing the POSIX `-l -c` pair makes PowerShell
+ * parse `-l` as the command and fail with `The term '-l' is not recognized`.
+ */
+export function isPowerShell(shell: string): boolean {
+	const basename = shell.replace(/\\/g, "/").split("/").pop()?.toLowerCase();
+	return basename === "powershell.exe" || basename === "powershell" || basename === "pwsh.exe" || basename === "pwsh";
+}
+
 /**
  * Get shell prefix for wrapping commands (profilers, strace, etc.).
  */

package/src/prompt.ts

@@ -1,7 +1,7 @@
-import type { HelperDelegate, HelperOptions, Template, TemplateDelegate } from "handlebars";
-import Handlebars from "handlebars";
+import type { HelperDelegate, Template } from "./template";
+import * as Handlebars from "./template";
 
-export type { HelperDelegate, HelperOptions, Template, TemplateDelegate };
+export type { HelperDelegate, HelperOptions, Template, TemplateDelegate } from "./template";
 
 export type PromptRenderPhase = "pre-render" | "post-render";
 
@@ -525,30 +525,13 @@ export function registerPartial(name: string, fn: Template): void {
 	handlebars.registerPartial(name, fn);
 }
 
-/**
- * Handlebars' lexer greedily matches `}}}` as `CLOSE_UNESCAPED` (the close of a
- * triple-stash `{{{ ... }}}`). When a regular helper close `}}` is immediately
- * followed by a literal `}` (common in compact JSON examples like
- * `{del:{{href ...}}}`), the lexer mistakes the trailing `}}}` for a triple-close
- * and rejects the input.
- *
- * We never use triple-stash (it's redundant under `noEscape: true`), so any run
- * of 3+ closing braces is unambiguously "helper close `}}`" + "literal `}`s".
- * Inject a no-op comment between them so the lexer tokenizes the helper close
- * cleanly and treats the rest as content.
- */
-function disambiguateClosingBraces(template: string): string {
-	return template.replace(/\}\}(\}+)/g, "}}{{!---}}$1");
-}
-
 const compiledTemplateCache = new Map<string, (context: TemplateContext) => string>();
 
 export function compile(template: string): (context: TemplateContext) => string {
-	// Keyed on the raw template so repeat renders skip disambiguateClosingBraces
-	// (a full-template regex pass) as well as the Handlebars compile.
+	// Keyed on the raw template so repeat renders skip parsing.
 	const cached = compiledTemplateCache.get(template);
 	if (cached) return cached;
-	const compiled = handlebars.compile(disambiguateClosingBraces(template), { noEscape: true, strict: false }) as (
+	const compiled = handlebars.compile(template, { noEscape: true, strict: false }) as (
 		context: TemplateContext,
 	) => string;
 	compiledTemplateCache.set(template, compiled);

package/src/readability/readability.ts

@@ -0,0 +1,533 @@
+/** Behavior-compatible reimplementation of @mozilla/readability's used surface. */
+
+import type {
+	ReadabilityArticle,
+	ReadabilityDocument,
+	ReadabilityElement,
+	ReadabilityNode,
+	ReadabilityOptions,
+} from "./types";
+
+const UNLIKELY =
+	/-ad-|ai2html|banner|breadcrumbs|comment|community|combx|disqus|extra|footer|gdpr|header|legends|menu|related|remark|replies|rss|shoutbox|sidebar|skyscraper|social|sponsor|supplemental|ad-break|agegate|pagination|pager|popup/i;
+const POSSIBLE = /and|article|body|column|content|main|shadow/i;
+const POSITIVE = /article|body|content|entry|hentry|h-entry|main|page|pagination|post|text|blog|story/i;
+const NEGATIVE =
+	/-ad-|hidden|^hid$| hid$| hid |^hid |banner|comment|com-|contact|footer|gdpr|masthead|media|meta|outbrain|promo|related|scroll|share|shoutbox|sidebar|skyscraper|sponsor|shopping|tags|widget/i;
+const BYLINE = /byline|author|dateline|writtenby|p-author/i;
+const SCORE_TAGS = new Set(["SECTION", "H2", "H3", "H4", "H5", "H6", "P", "TD", "PRE"]);
+const DROP_TAGS = [
+	"form",
+	"fieldset",
+	"object",
+	"embed",
+	"footer",
+	"link",
+	"aside",
+	"iframe",
+	"input",
+	"textarea",
+	"select",
+	"button",
+];
+const UNLIKELY_ROLES = new Set(["menu", "menubar", "complementary", "navigation", "alert", "alertdialog", "dialog"]);
+const ARTICLE_TYPES =
+	/^(?:Article|AdvertiserContentArticle|NewsArticle|AnalysisNewsArticle|OpinionNewsArticle|ReportageNewsArticle|ReviewNewsArticle|Report|ScholarlyArticle|MedicalScholarlyArticle|SocialMediaPosting|BlogPosting|LiveBlogPosting|DiscussionForumPosting|TechArticle|APIReference)$/;
+const NORMALIZE = /\s{2,}/g;
+
+type Metadata = {
+	title?: string;
+	byline?: string;
+	excerpt?: string;
+	siteName?: string;
+	publishedTime?: string | null;
+};
+
+type Attempt = { element: ReadabilityElement; length: number; dir?: string | null };
+
+function elements(collection: ArrayLike<ReadabilityElement>): ReadabilityElement[] {
+	return Array.from(collection);
+}
+
+function descendants(root: ReadabilityElement): ReadabilityElement[] {
+	const result: ReadabilityElement[] = [];
+	const pending = elements(root.children).reverse();
+	while (pending.length) {
+		const node = pending.pop();
+		if (!node) continue;
+		result.push(node);
+		const children = elements(node.children);
+		for (let index = children.length - 1; index >= 0; index--) pending.push(children[index]);
+	}
+	return result;
+}
+
+function text(node: ReadabilityNode): string {
+	return (node.textContent ?? "").trim().replace(NORMALIZE, " ");
+}
+
+function matchLabel(node: ReadabilityElement): string {
+	return `${typeof node.className === "string" ? node.className : ""} ${node.id ?? ""}`;
+}
+
+function classWeight(node: ReadabilityElement): number {
+	const label = matchLabel(node);
+	return (POSITIVE.test(label) ? 25 : 0) - (NEGATIVE.test(label) ? 25 : 0);
+}
+
+function initialScore(node: ReadabilityElement): number {
+	let score = classWeight(node);
+	switch (node.tagName) {
+		case "DIV":
+			score += 5;
+			break;
+		case "PRE":
+		case "TD":
+		case "BLOCKQUOTE":
+			score += 3;
+			break;
+		case "ADDRESS":
+		case "OL":
+		case "UL":
+		case "DL":
+		case "DD":
+		case "DT":
+		case "LI":
+		case "FORM":
+			score -= 3;
+			break;
+		case "H1":
+		case "H2":
+		case "H3":
+		case "H4":
+		case "H5":
+		case "H6":
+		case "TH":
+			score -= 5;
+			break;
+	}
+	return score;
+}
+
+function linkDensity(node: ReadabilityElement): number {
+	const total = text(node).length;
+	if (!total) return 0;
+	let linked = 0;
+	for (const link of elements(node.getElementsByTagName("a"))) {
+		linked += text(link).length * ((link.getAttribute("href") ?? "").startsWith("#") ? 0.3 : 1);
+	}
+	return linked / total;
+}
+
+function visible(node: ReadabilityElement): boolean {
+	const style = node.getAttribute("style")?.toLowerCase() ?? "";
+	return (
+		!node.hasAttribute("hidden") &&
+		node.getAttribute("aria-hidden") !== "true" &&
+		!/display\s*:\s*none|visibility\s*:\s*hidden/.test(style)
+	);
+}
+
+function removeAll(root: ReadabilityNode, tags: readonly string[]): void {
+	const container = root as ReadabilityElement;
+	for (const tag of tags) {
+		for (const node of elements(container.getElementsByTagName(tag))) node.remove();
+	}
+}
+
+function entityDecode(value: string | undefined | null): string | undefined | null {
+	if (!value) return value;
+	const named: Record<string, string> = { quot: '"', amp: "&", apos: "'", lt: "<", gt: ">" };
+	return value
+		.replace(/&(quot|amp|apos|lt|gt);/g, (_, name: string) => named[name] ?? "")
+		.replace(/&#(?:x([0-9a-f]+)|(\d+));/gi, (_, hex: string | undefined, decimal: string | undefined) => {
+			const value = Number.parseInt(hex ?? decimal ?? "0", hex ? 16 : 10);
+			return String.fromCodePoint(
+				value === 0 || value > 0x10ffff || (value >= 0xd800 && value <= 0xdfff) ? 0xfffd : value,
+			);
+		});
+}
+
+function titleFromDocument(document: ReadabilityDocument): string {
+	const titleElement = elements(document.getElementsByTagName("title"))[0];
+	const original = typeof document.title === "string" ? document.title.trim() : titleElement ? text(titleElement) : "";
+	let title = original;
+	const separators = [...original.matchAll(/ [|\\/>»-] /g)];
+	if (separators.length) {
+		title = original.slice(0, separators.at(-1)?.index);
+		if (title.trim().split(/\s+/).length < 3) title = original.replace(/^[^|\\/>»-]*[|\\/>»-]/, "");
+	} else if (title.includes(": ")) {
+		const matchingHeading = elements(document.querySelectorAll("h1, h2")).some(node => text(node) === title);
+		if (!matchingHeading) {
+			const suffix = original.slice(original.lastIndexOf(":") + 1);
+			title = suffix.trim().split(/\s+/).length < 3 ? original.slice(original.indexOf(":") + 1) : suffix;
+		}
+	} else if (title.length > 150 || title.length < 15) {
+		const headings = elements(document.getElementsByTagName("h1"));
+		if (headings.length === 1) title = text(headings[0]);
+	}
+	title = title.trim().replace(NORMALIZE, " ");
+	if (title.split(/\s+/).length <= 4) return original;
+	return title;
+}
+
+function jsonLdMetadata(document: ReadabilityDocument): Metadata {
+	for (const script of elements(document.getElementsByTagName("script"))) {
+		if (script.getAttribute("type") !== "application/ld+json") continue;
+		try {
+			const decoded: unknown = JSON.parse((script.textContent ?? "").replace(/^\s*<!\[CDATA\[|\]\]>\s*$/g, ""));
+			const records = Array.isArray(decoded) ? decoded : [decoded];
+			for (const candidate of records) {
+				if (!candidate || typeof candidate !== "object") continue;
+				const record = candidate as Record<string, unknown>;
+				if (typeof record["@type"] !== "string" || !ARTICLE_TYPES.test(record["@type"])) continue;
+				const author = record.author;
+				let byline: string | undefined;
+				if (author && typeof author === "object" && !Array.isArray(author)) {
+					const name = (author as Record<string, unknown>).name;
+					if (typeof name === "string") byline = name.trim();
+				} else if (Array.isArray(author)) {
+					const names = author.flatMap(item => {
+						if (!item || typeof item !== "object") return [];
+						const name = (item as Record<string, unknown>).name;
+						return typeof name === "string" ? [name.trim()] : [];
+					});
+					if (names.length) byline = names.join(", ");
+				}
+				const publisher = record.publisher;
+				const publisherName =
+					publisher && typeof publisher === "object" ? (publisher as Record<string, unknown>).name : undefined;
+				return {
+					title:
+						typeof record.name === "string"
+							? record.name.trim()
+							: typeof record.headline === "string"
+								? record.headline.trim()
+								: undefined,
+					byline,
+					excerpt: typeof record.description === "string" ? record.description.trim() : undefined,
+					siteName: typeof publisherName === "string" ? publisherName.trim() : undefined,
+					publishedTime: typeof record.datePublished === "string" ? record.datePublished.trim() : undefined,
+				};
+			}
+		} catch {
+			// Invalid publisher data is ignored just like malformed meta markup.
+		}
+	}
+	return {};
+}
+
+function metadataFromDocument(document: ReadabilityDocument, jsonLd: Metadata): Metadata {
+	const values = new Map<string, string>();
+	for (const meta of elements(document.getElementsByTagName("meta"))) {
+		const content = meta.getAttribute("content")?.trim();
+		if (!content) continue;
+		const property = meta.getAttribute("property")?.toLowerCase().replace(/\s/g, "");
+		const name = meta.getAttribute("name")?.toLowerCase().replace(/\s/g, "").replace(/\./g, ":");
+		if (
+			property &&
+			/^(?:article|dc|dcterm|og|twitter):(?:author|creator|description|published_time|title|site_name)$/.test(
+				property,
+			)
+		)
+			values.set(property, content);
+		else if (
+			name &&
+			/^(?:(?:dc|dcterm|og|twitter|parsely|weibo:(?:article|webpage))[-:]?)?(?:author|creator|pub-date|description|title|site_name)$/.test(
+				name,
+			)
+		)
+			values.set(name, content);
+	}
+	const articleAuthor = values.get("article:author");
+	const result: Metadata = {
+		title:
+			jsonLd.title ??
+			values.get("dc:title") ??
+			values.get("dcterm:title") ??
+			values.get("og:title") ??
+			values.get("title") ??
+			values.get("twitter:title") ??
+			titleFromDocument(document),
+		byline:
+			jsonLd.byline ??
+			values.get("dc:creator") ??
+			values.get("dcterm:creator") ??
+			values.get("author") ??
+			values.get("parsely-author") ??
+			(articleAuthor && !/^https?:\/\//.test(articleAuthor) ? articleAuthor : undefined),
+		excerpt:
+			jsonLd.excerpt ??
+			values.get("dc:description") ??
+			values.get("dcterm:description") ??
+			values.get("og:description") ??
+			values.get("description") ??
+			values.get("twitter:description"),
+		siteName: jsonLd.siteName ?? values.get("og:site_name"),
+		publishedTime:
+			jsonLd.publishedTime ?? values.get("article:published_time") ?? values.get("parsely-pub-date") ?? null,
+	};
+	return {
+		title: entityDecode(result.title) ?? undefined,
+		byline: entityDecode(result.byline) ?? undefined,
+		excerpt: entityDecode(result.excerpt) ?? undefined,
+		siteName: entityDecode(result.siteName) ?? undefined,
+		publishedTime: entityDecode(result.publishedTime),
+	};
+}
+
+/** Extracts the article body and metadata from a standards-shaped document. */
+export class Readability<T = string> {
+	readonly #document: ReadabilityDocument;
+	readonly #options: ReadabilityOptions<T>;
+	readonly #scores = new Map<ReadabilityElement, number>();
+	#byline: string | undefined;
+	#lang: string | null = null;
+
+	constructor(document: ReadabilityDocument, options: ReadabilityOptions<T> = {}) {
+		this.#document = document;
+		this.#options = options;
+	}
+
+	/** Runs extraction once; the supplied document is consumed and should not be reused. */
+	parse(): ReadabilityArticle<T> | null {
+		const documentElement = this.#document.documentElement;
+		if (!documentElement) return null;
+		const max = this.#options.maxElemsToParse ?? 0;
+		if (max > 0) {
+			const count = descendants(documentElement).length + 1;
+			if (count > max) throw new Error(`Aborting parsing document; ${count} elements found`);
+		}
+		const jsonLd = this.#options.disableJSONLD ? {} : jsonLdMetadata(this.#document);
+		const metadata = metadataFromDocument(this.#document, jsonLd);
+		removeAll(this.#document, ["script", "style"]);
+		const body = this.#document.body;
+		if (!body) return null;
+		const source = body.innerHTML;
+		const attempts: Attempt[] = [];
+		for (const mode of [0, 1, 2, 3]) {
+			if (mode) body.innerHTML = source;
+			this.#scores.clear();
+			this.#byline = undefined;
+			const attempt = this.#extract(body, documentElement, metadata.title ?? "", mode);
+			if (attempt) attempts.push(attempt);
+			if (attempt && attempt.length >= (this.#options.charThreshold || 500)) break;
+		}
+		attempts.sort((left, right) => right.length - left.length);
+		const best = attempts[0];
+		if (!best?.length) return null;
+		if (!metadata.excerpt) {
+			const firstParagraph = elements(best.element.getElementsByTagName("p"))[0];
+			if (firstParagraph) metadata.excerpt = (firstParagraph.textContent ?? "").trim();
+		}
+		const contentText = best.element.textContent ?? "";
+		const serializer =
+			this.#options.serializer ?? ((node: ReadabilityNode) => (node as ReadabilityElement).innerHTML as T);
+		return {
+			title: metadata.title,
+			byline: metadata.byline ?? this.#byline,
+			dir: best.dir,
+			lang: this.#lang,
+			content: serializer(best.element),
+			textContent: contentText,
+			length: contentText.length,
+			excerpt: metadata.excerpt,
+			siteName: metadata.siteName,
+			publishedTime: metadata.publishedTime,
+		};
+	}
+
+	#extract(
+		body: ReadabilityElement,
+		documentElement: ReadabilityElement,
+		articleTitle: string,
+		mode: number,
+	): Attempt | null {
+		this.#lang = documentElement.getAttribute("lang");
+		const stripUnlikely = mode === 0;
+		const weightClasses = mode < 2;
+		const all = [documentElement, ...descendants(documentElement)];
+		const scored: ReadabilityElement[] = [];
+		let titleRemoved = false;
+		for (const node of all) {
+			if (node === documentElement || node.tagName === "BODY") continue;
+			const label = matchLabel(node);
+			if (!visible(node) || (node.getAttribute("aria-modal") === "true" && node.getAttribute("role") === "dialog")) {
+				node.remove();
+				continue;
+			}
+			if (!this.#byline && this.#isByline(node, label)) {
+				this.#byline = text(node);
+				node.remove();
+				continue;
+			}
+			if (!titleRemoved && /^(?:H1|H2)$/.test(node.tagName) && this.#similar(articleTitle, text(node)) > 0.75) {
+				titleRemoved = true;
+				node.remove();
+				continue;
+			}
+			if (
+				(stripUnlikely && UNLIKELY.test(label) && !POSSIBLE.test(label)) ||
+				UNLIKELY_ROLES.has(node.getAttribute("role") ?? "")
+			) {
+				node.remove();
+				continue;
+			}
+			if (SCORE_TAGS.has(node.tagName)) scored.push(node);
+		}
+		for (const paragraph of scored) this.#scoreParagraph(paragraph, weightClasses);
+		let top: ReadabilityElement | undefined;
+		let topScore = Number.NEGATIVE_INFINITY;
+		for (const [candidate, raw] of this.#scores) {
+			if (candidate.tagName === "BODY" || candidate.tagName === "HTML") continue;
+			const score = raw * (1 - linkDensity(candidate));
+			this.#scores.set(candidate, score);
+			if (score > topScore) {
+				top = candidate;
+				topScore = score;
+			}
+		}
+		if (!top || top.tagName === "BODY") top = body;
+		while (
+			top.parentNode &&
+			(top.parentNode as ReadabilityElement).tagName !== "BODY" &&
+			(top.parentNode as ReadabilityElement).children.length === 1
+		)
+			top = top.parentNode as ReadabilityElement;
+		const parent = top.parentNode as ReadabilityElement | null;
+		const article = this.#document.createElement("DIV");
+		const siblings = parent ? elements(parent.children) : [top];
+		const threshold = Math.max(10, (this.#scores.get(top) ?? topScore) * 0.2);
+		for (const sibling of siblings) {
+			const siblingText = text(sibling);
+			const sameClassBonus =
+				sibling.className && sibling.className === top.className ? (this.#scores.get(top) ?? 0) * 0.2 : 0;
+			const include =
+				sibling === top ||
+				(this.#scores.get(sibling) ?? 0) + sameClassBonus >= threshold ||
+				(sibling.tagName === "P" &&
+					((siblingText.length > 80 && linkDensity(sibling) < 0.25) ||
+						(siblingText.length > 0 &&
+							siblingText.length < 80 &&
+							linkDensity(sibling) === 0 &&
+							/\.(?: |$)/.test(siblingText))));
+			if (!include) continue;
+			if (["DIV", "ARTICLE", "SECTION", "P", "OL", "UL"].includes(sibling.tagName)) {
+				article.appendChild(sibling);
+				continue;
+			}
+			const replacement = this.#document.createElement("DIV");
+			for (const attribute of Array.from(sibling.attributes))
+				replacement.setAttribute(attribute.name, attribute.value);
+			while (sibling.firstChild) replacement.appendChild(sibling.firstChild);
+			article.appendChild(replacement);
+		}
+		this.#clean(article, mode < 3);
+		const page = this.#document.createElement("DIV");
+		page.id = "readability-page-1";
+		page.className = "page";
+		while (article.firstChild) page.appendChild(article.firstChild);
+		article.appendChild(page);
+		const content = text(article);
+		let dir: string | null | undefined;
+		let ancestor: ReadabilityNode | null = top;
+		while (ancestor) {
+			if ((ancestor as ReadabilityElement).getAttribute) {
+				dir = (ancestor as ReadabilityElement).getAttribute("dir");
+				if (dir) break;
+			}
+			ancestor = ancestor.parentNode;
+		}
+		return { element: article, length: content.length, dir };
+	}
+
+	#scoreParagraph(node: ReadabilityElement, weightClasses: boolean): void {
+		const content = text(node);
+		if (content.length < 25) return;
+		const score =
+			1 +
+			content.split(/[\u002c\u060c\ufe50\ufe10\ufe11\u2e41\u2e34\u2e32\uff0c]/).length +
+			Math.min(Math.floor(content.length / 100), 3);
+		let ancestor = node.parentNode;
+		for (let level = 0; ancestor && level < 5; level++, ancestor = ancestor.parentNode) {
+			const element = ancestor as ReadabilityElement;
+			if (!element.tagName || !element.parentNode || !(element.parentNode as ReadabilityElement).tagName) continue;
+			const baseline =
+				this.#scores.get(element) ?? initialScore(element) - (weightClasses ? 0 : classWeight(element));
+			const divisor = level === 0 ? 1 : level === 1 ? 2 : level * 3;
+			this.#scores.set(element, baseline + score / divisor);
+		}
+	}
+
+	#clean(root: ReadabilityElement, conditional: boolean): void {
+		removeAll(root, DROP_TAGS);
+		for (const heading of elements(root.querySelectorAll("h1, h2, h3, h4, h5, h6"))) {
+			if (classWeight(heading) < 0 || linkDensity(heading) > 0.33) heading.remove();
+		}
+		if (conditional) {
+			for (const node of elements(root.querySelectorAll("table, ul, div"))) {
+				if (node === root) continue;
+				const nodeText = text(node);
+				const paragraphs = node.getElementsByTagName("p").length;
+				const images = node.getElementsByTagName("img").length;
+				const inputs = node.getElementsByTagName("input").length;
+				if (
+					classWeight(node) < 0 ||
+					(!nodeText && !images) ||
+					(nodeText.split(",").length < 10 &&
+						((images > paragraphs && paragraphs > 0) ||
+							inputs > Math.floor(paragraphs / 3) ||
+							linkDensity(node) > 0.5))
+				)
+					node.remove();
+			}
+		}
+		for (const paragraph of elements(root.getElementsByTagName("p"))) {
+			if (!text(paragraph) && !paragraph.querySelector("img, embed, object, iframe")) paragraph.remove();
+		}
+		for (const node of [root, ...descendants(root)]) {
+			if (!this.#options.keepClasses) {
+				const preserved = (this.#options.classesToPreserve ?? []).filter(name =>
+					node.className.split(/\s+/).includes(name),
+				);
+				if (node.id === "readability-page-1") preserved.unshift("page");
+				if (preserved.length) node.className = [...new Set(preserved)].join(" ");
+				else node.removeAttribute("class");
+			}
+			for (const attr of [
+				"style",
+				"align",
+				"background",
+				"bgcolor",
+				"border",
+				"cellpadding",
+				"cellspacing",
+				"frame",
+				"hspace",
+				"rules",
+				"valign",
+				"vspace",
+			])
+				node.removeAttribute(attr);
+		}
+	}
+
+	#isByline(node: ReadabilityElement, label: string): boolean {
+		const value = text(node);
+		return (
+			value.length > 0 &&
+			value.length < 100 &&
+			(node.getAttribute("rel") === "author" ||
+				(node.getAttribute("itemprop") ?? "").includes("author") ||
+				BYLINE.test(label))
+		);
+	}
+
+	#similar(left: string, right: string): number {
+		const leftTokens = left.toLowerCase().split(/\W+/).filter(Boolean);
+		const rightTokens = right.toLowerCase().split(/\W+/).filter(Boolean);
+		if (!leftTokens.length || !rightTokens.length) return 0;
+		const unmatched = rightTokens.filter(token => !leftTokens.includes(token));
+		return 1 - unmatched.join(" ").length / rightTokens.join(" ").length;
+	}
+}