@lijian-ui/dsh-file-manager 0.2.2 → 0.2.4

package/src/client/preview/markdown.ts

@@ -1,351 +1,351 @@
-/**
- * A compact markdown renderer for the preview panel: headings, paragraphs,
- * fenced + inline code, bold/italic, links/images, lists, blockquotes, hr,
- * and tables. All HTML is escaped before transformation — the output only
- * ever contains the renderer's own tags. Pure and exported for tests.
- * @module dsh-filemgr/client/preview/markdown
- */
-
-/** Escape HTML special characters. */
-export function escapeHtml(text: string): string {
-  return text
-    .replace(/&/g, '&')
-    .replace(/</g, '&lt;')
-    .replace(/>/g, '&gt;')
-    .replace(/"/g, '&quot;')
-    .replace(/'/g, '&#39;')
-}
-
-/** How one image src resolves against the markdown file's location. */
-export type MarkdownImageResolution =
-  /** Scheme URL or fragment: the browser resolves it as-is. */
-  | { kind: 'absolute' }
-  /** Workspace-relative target: resolved path plus any ?query#fragment suffix. */
-  | { kind: 'relative'; path: string; suffix: string }
-  /** `..` escaped the project root: the image must be dropped. */
-  | { kind: 'escape' }
-
-/** Directory of a workspace-relative file path ('' when at the root). */
-function dirOf(filePath: string): string {
-  const slash = filePath.lastIndexOf('/')
-  return slash === -1 ? '' : filePath.slice(0, slash)
-}
-
-/** Collapse . and .. segments; null when .. escapes the base. */
-function normalizeRelPath(rel: string): string | null {
-  const out: string[] = []
-  for (const part of rel.split('/')) {
-    if (part === '' || part === '.') continue
-    if (part === '..') {
-      if (out.length === 0) return null
-      out.pop()
-      continue
-    }
-    out.push(part)
-  }
-  return out.join('/')
-}
-
-/** Percent-decode a path portion (best effort; never throws). */
-function decodePathPart(raw: string): string {
-  try {
-    return decodeURIComponent(raw)
-  } catch {
-    return raw
-  }
-}
-
-/**
- * Resolve one markdown image src against the markdown file's location:
- * - Absolute URLs (http/https/data:/...) and fragment-only srcs are left to
- *   the browser ('absolute').
- * - Root-relative srcs (/img.png) resolve from the project root; other
- *   relative srcs resolve against the file's directory. `..` escaping the
- *   project root is rejected ('escape').
- * - The path portion is percent-decoded (markdown authors encode spaces in
- *   filenames) and any ?query#fragment suffix is preserved verbatim, so
- *   cache-busting srcs like ./img.png?v=2 still fetch img.png.
- */
-export function resolveMarkdownImage(filePath: string, src: string): MarkdownImageResolution {
-  const trimmed = src.trim()
-  if (trimmed === '' || trimmed.startsWith('#')) return { kind: 'absolute' }
-  if (/^[a-zA-Z][a-zA-Z0-9+.-]*:/.test(trimmed)) return { kind: 'absolute' }
-  const decoded = decodePathPart(trimmed)
-  const q = decoded.indexOf('?')
-  const h = decoded.indexOf('#')
-  let cut = decoded.length
-  if (q !== -1) cut = Math.min(cut, q)
-  if (h !== -1) cut = Math.min(cut, h)
-  const pathPart = decoded.slice(0, cut)
-  const suffix = decoded.slice(cut)
-  const base = pathPart.startsWith('/') ? '' : dirOf(filePath)
-  const joined = base === '' ? pathPart : `${base}/${pathPart}`
-  const normalized = normalizeRelPath(joined)
-  if (normalized === null) return { kind: 'escape' }
-  return { kind: 'relative', path: normalized, suffix }
-}
-
-/** Options controlling markdown rendering. */
-export interface MarkdownRenderOptions {
-  /**
-   * Rewrite image srcs before they are emitted. Return the URL to use, or
-   * null to drop the image (alt text only). Relative workspace paths are
-   * typically resolved to absolute URLs here.
-   */
-  resolveImageSrc?: (src: string) => string | null
-}
-
-/**
- * Guard a raw link/image target against dangerous protocols. Returns the
- * (trimmed) raw string when safe, else null. Only these schemes are allowed:
- * http:, https:, mailto: and fragment anchors (#...). Scheme-less relative
- * paths (./ ../ / and plain filenames) pass through unchanged. Anything with
- * a scheme outside the allow-list — javascript:, data:, vbscript:, etc. —
- * is rejected so the value never reaches dangerouslySetInnerHTML.
- */
-export function safeUrl(raw: string): string | null {
-  const trimmed = raw.trim()
-  if (trimmed === '') return null
-  if (trimmed.startsWith('#')) return trimmed
-  const scheme = /^([a-zA-Z][a-zA-Z0-9+.-]*):/.exec(trimmed)
-  if (scheme === null) return trimmed
-  const name = scheme[1].toLowerCase()
-  return name === 'http' || name === 'https' || name === 'mailto' ? trimmed : null
-}
-
-/** Inline pass: code spans, bold, italic, images, links. */
-export function renderInline(text: string, options?: MarkdownRenderOptions): string {
-  let out = ''
-  let i = 0
-  const n = text.length
-  while (i < n) {
-    const char = text[i]
-    // Fenced inline code first.
-    if (char === '`') {
-      const end = text.indexOf('`', i + 1)
-      if (end !== -1) {
-        out += `<code>${escapeHtml(text.slice(i + 1, end))}</code>`
-        i = end + 1
-        continue
-      }
-    }
-    // Image ![alt](src)
-    if (char === '!' && text[i + 1] === '[') {
-      const close = text.indexOf('](', i + 2)
-      if (close !== -1) {
-        const parenEnd = text.indexOf(')', close + 2)
-        if (parenEnd !== -1) {
-          const alt = text.slice(i + 2, close)
-          const src = text.slice(close + 2, parenEnd)
-          const safe = safeUrl(src)
-          if (safe === null) {
-            // Unsafe image target: drop the img, keep the alt text.
-            out += escapeHtml(alt)
-          } else {
-            let target: string | null = safe
-            if (options?.resolveImageSrc !== undefined) {
-              target = options.resolveImageSrc(safe)
-            }
-            if (target === null) {
-              // The resolver dropped the image (unresolvable path): alt only.
-              out += escapeHtml(alt)
-            } else {
-              const srcEsc = escapeHtml(target).replace(/\s+/g, '%20')
-              out += `<img alt="${escapeHtml(alt)}" src="${srcEsc}" />`
-            }
-          }
-          i = parenEnd + 1
-          continue
-        }
-      }
-    }
-    // Link [text](href)
-    if (char === '[') {
-      const close = text.indexOf('](', i + 1)
-      if (close !== -1) {
-        const parenEnd = text.indexOf(')', close + 2)
-        if (parenEnd !== -1) {
-          const label = text.slice(i + 1, close)
-          const href = text.slice(close + 2, parenEnd)
-          const safe = safeUrl(href)
-          if (safe === null) {
-            // Unsafe link target: render the label as plain text, no <a>.
-            out += renderInline(label, options)
-          } else {
-            out += `<a href="${escapeHtml(safe)}" target="_blank" rel="noopener noreferrer">${renderInline(label, options)}</a>`
-          }
-          i = parenEnd + 1
-          continue
-        }
-      }
-    }
-    // Bold **text**
-    if (char === '*' && text[i + 1] === '*') {
-      const end = text.indexOf('**', i + 2)
-      if (end !== -1) {
-        out += `<strong>${renderInline(text.slice(i + 2, end), options)}</strong>`
-        i = end + 2
-        continue
-      }
-    }
-    // Italic *text*
-    if (char === '*' && text[i - 1] !== '*' && text[i + 1] !== '*') {
-      const end = text.indexOf('*', i + 1)
-      if (end !== -1 && text[end + 1] !== '*') {
-        out += `<em>${renderInline(text.slice(i + 1, end), options)}</em>`
-        i = end + 1
-        continue
-      }
-    }
-    // Strikethrough ~~text~~
-    if (char === '~' && text[i + 1] === '~') {
-      const end = text.indexOf('~~', i + 2)
-      if (end !== -1) {
-        out += `<del>${renderInline(text.slice(i + 2, end), options)}</del>`
-        i = end + 2
-        continue
-      }
-    }
-    out += escapeHtml(char)
-    i += 1
-  }
-  return out
-}
-
-/** Render a markdown document to HTML (block pass). */
-export function renderMarkdown(source: string, options?: MarkdownRenderOptions): string {
-  const lines = source.replace(/\r\n/g, '\n').split('\n')
-  const out: string[] = []
-  let i = 0
-  const n = lines.length
-
-  const flushParagraph = (buffer: string[]): void => {
-    if (buffer.length === 0) return
-    out.push(`<p>${renderInline(buffer.join("\n"), options)}</p>`)
-    buffer.length = 0
-  }
-
-  let paragraph: string[] = []
-  while (i < n) {
-    const line = lines[i]
-
-    // Fenced code block.
-    const fence = /^```([\w+-]*)\s*$/.exec(line)
-    if (fence !== null) {
-      flushParagraph(paragraph)
-      const lang = fence[1] ?? ''
-      i += 1
-      const code: string[] = []
-      while (i < n && !/^```\s*$/.test(lines[i])) {
-        code.push(lines[i])
-        i += 1
-      }
-      i += 1 // closing fence
-      const langAttr = lang === '' ? '' : ` class="language-${escapeHtml(lang)}"`
-      out.push(`<pre${langAttr}><code>${escapeHtml(code.join('\n'))}</code></pre>`)
-      continue
-    }
-
-    // Heading.
-    const heading = /^(#{1,6})\s+(.*)$/.exec(line)
-    if (heading !== null) {
-      flushParagraph(paragraph)
-      const level = heading[1].length
-      out.push(`<h${level}>${renderInline(heading[2] ?? '', options)}</h${level}>`)
-      i += 1
-      continue
-    }
-
-    // Horizontal rule.
-    if (/^\s*(---+|\*\*\*+|___+)\s*$/.test(line)) {
-      flushParagraph(paragraph)
-      out.push('<hr />')
-      i += 1
-      continue
-    }
-
-    // Table: header row then separator row.
-    if (line.includes('|') && i + 1 < n && /^\s*\|?[\s:|-]+\|?\s*$/.test(lines[i + 1]) && lines[i + 1].includes('-')) {
-      flushParagraph(paragraph)
-      const headerCells = splitTableRow(line)
-      i += 2
-      const rows: string[][] = []
-      while (i < n && lines[i].includes('|')) {
-        rows.push(splitTableRow(lines[i]))
-        i += 1
-      }
-      out.push('<table>')
-      out.push(`<thead><tr>${headerCells.map((cell) => `<th>${renderInline(cell, options)}</th>`).join('')}</tr></thead>`)
-      if (rows.length > 0) {
-        out.push(`<tbody>${rows.map((row) => `<tr>${row.map((cell) => `<td>${renderInline(cell, options)}</td>`).join('')}</tr>`).join('')}</tbody>`)
-      }
-      out.push('</table>')
-      continue
-    }
-
-    // Blockquote (one level).
-    const quote = /^>\s?(.*)$/.exec(line)
-    if (quote !== null) {
-      flushParagraph(paragraph)
-      const body: string[] = []
-      while (i < n) {
-        const q = /^>\s?(.*)$/.exec(lines[i])
-        if (q === null) break
-        body.push(q[1] ?? '')
-        i += 1
-      }
-      out.push(`<blockquote><p>${body.map((line) => renderInline(line, options)).join('<br />')}</p></blockquote>`)
-      continue
-    }
-
-    // Unordered list.
-    const ul = /^\s*([-*+])\s+(.*)$/.exec(line)
-    if (ul !== null) {
-      flushParagraph(paragraph)
-      const items: string[] = []
-      while (i < n) {
-        const item = /^\s*([-*+])\s+(.*)$/.exec(lines[i])
-        if (item === null) break
-        items.push(`<li>${renderInline(item[2] ?? '', options)}</li>`)
-        i += 1
-      }
-      out.push(`<ul>${items.join('')}</ul>`)
-      continue
-    }
-
-    // Ordered list.
-    const ol = /^\s*\d+[.)]\s+(.*)$/.exec(line)
-    if (ol !== null) {
-      flushParagraph(paragraph)
-      const items: string[] = []
-      while (i < n) {
-        const item = /^\s*\d+[.)]\s+(.*)$/.exec(lines[i])
-        if (item === null) break
-        items.push(`<li>${renderInline(item[1] ?? '', options)}</li>`)
-        i += 1
-      }
-      out.push(`<ol>${items.join('')}</ol>`)
-      continue
-    }
-
-    // Blank line: flush the paragraph.
-    if (line.trim() === '') {
-      flushParagraph(paragraph)
-      i += 1
-      continue
-    }
-
-    paragraph.push(line)
-    i += 1
-  }
-  flushParagraph(paragraph)
-  return out.join('\n')
-}
-
-/** Split one table row into cells (respecting the leading/trailing pipes). */
-function splitTableRow(line: string): string[] {
-  const trimmed = line.trim()
-  const inner = trimmed.startsWith('|') ? trimmed.slice(1) : trimmed
-  const withoutTrailing = inner.endsWith('|') ? inner.slice(0, -1) : inner
-  return withoutTrailing.split('|').map((cell) => cell.trim())
-}
+/**
+ * A compact markdown renderer for the preview panel: headings, paragraphs,
+ * fenced + inline code, bold/italic, links/images, lists, blockquotes, hr,
+ * and tables. All HTML is escaped before transformation — the output only
+ * ever contains the renderer's own tags. Pure and exported for tests.
+ * @module dsh-filemgr/client/preview/markdown
+ */
+
+/** Escape HTML special characters. */
+export function escapeHtml(text: string): string {
+  return text
+    .replace(/&/g, '&amp;')
+    .replace(/</g, '&lt;')
+    .replace(/>/g, '&gt;')
+    .replace(/"/g, '&quot;')
+    .replace(/'/g, '&#39;')
+}
+
+/** How one image src resolves against the markdown file's location. */
+export type MarkdownImageResolution =
+  /** Scheme URL or fragment: the browser resolves it as-is. */
+  | { kind: 'absolute' }
+  /** Workspace-relative target: resolved path plus any ?query#fragment suffix. */
+  | { kind: 'relative'; path: string; suffix: string }
+  /** `..` escaped the project root: the image must be dropped. */
+  | { kind: 'escape' }
+
+/** Directory of a workspace-relative file path ('' when at the root). */
+function dirOf(filePath: string): string {
+  const slash = filePath.lastIndexOf('/')
+  return slash === -1 ? '' : filePath.slice(0, slash)
+}
+
+/** Collapse . and .. segments; null when .. escapes the base. */
+function normalizeRelPath(rel: string): string | null {
+  const out: string[] = []
+  for (const part of rel.split('/')) {
+    if (part === '' || part === '.') continue
+    if (part === '..') {
+      if (out.length === 0) return null
+      out.pop()
+      continue
+    }
+    out.push(part)
+  }
+  return out.join('/')
+}
+
+/** Percent-decode a path portion (best effort; never throws). */
+function decodePathPart(raw: string): string {
+  try {
+    return decodeURIComponent(raw)
+  } catch {
+    return raw
+  }
+}
+
+/**
+ * Resolve one markdown image src against the markdown file's location:
+ * - Absolute URLs (http/https/data:/...) and fragment-only srcs are left to
+ *   the browser ('absolute').
+ * - Root-relative srcs (/img.png) resolve from the project root; other
+ *   relative srcs resolve against the file's directory. `..` escaping the
+ *   project root is rejected ('escape').
+ * - The path portion is percent-decoded (markdown authors encode spaces in
+ *   filenames) and any ?query#fragment suffix is preserved verbatim, so
+ *   cache-busting srcs like ./img.png?v=2 still fetch img.png.
+ */
+export function resolveMarkdownImage(filePath: string, src: string): MarkdownImageResolution {
+  const trimmed = src.trim()
+  if (trimmed === '' || trimmed.startsWith('#')) return { kind: 'absolute' }
+  if (/^[a-zA-Z][a-zA-Z0-9+.-]*:/.test(trimmed)) return { kind: 'absolute' }
+  const decoded = decodePathPart(trimmed)
+  const q = decoded.indexOf('?')
+  const h = decoded.indexOf('#')
+  let cut = decoded.length
+  if (q !== -1) cut = Math.min(cut, q)
+  if (h !== -1) cut = Math.min(cut, h)
+  const pathPart = decoded.slice(0, cut)
+  const suffix = decoded.slice(cut)
+  const base = pathPart.startsWith('/') ? '' : dirOf(filePath)
+  const joined = base === '' ? pathPart : `${base}/${pathPart}`
+  const normalized = normalizeRelPath(joined)
+  if (normalized === null) return { kind: 'escape' }
+  return { kind: 'relative', path: normalized, suffix }
+}
+
+/** Options controlling markdown rendering. */
+export interface MarkdownRenderOptions {
+  /**
+   * Rewrite image srcs before they are emitted. Return the URL to use, or
+   * null to drop the image (alt text only). Relative workspace paths are
+   * typically resolved to absolute URLs here.
+   */
+  resolveImageSrc?: (src: string) => string | null
+}
+
+/**
+ * Guard a raw link/image target against dangerous protocols. Returns the
+ * (trimmed) raw string when safe, else null. Only these schemes are allowed:
+ * http:, https:, mailto: and fragment anchors (#...). Scheme-less relative
+ * paths (./ ../ / and plain filenames) pass through unchanged. Anything with
+ * a scheme outside the allow-list — javascript:, data:, vbscript:, etc. —
+ * is rejected so the value never reaches dangerouslySetInnerHTML.
+ */
+export function safeUrl(raw: string): string | null {
+  const trimmed = raw.trim()
+  if (trimmed === '') return null
+  if (trimmed.startsWith('#')) return trimmed
+  const scheme = /^([a-zA-Z][a-zA-Z0-9+.-]*):/.exec(trimmed)
+  if (scheme === null) return trimmed
+  const name = scheme[1].toLowerCase()
+  return name === 'http' || name === 'https' || name === 'mailto' ? trimmed : null
+}
+
+/** Inline pass: code spans, bold, italic, images, links. */
+export function renderInline(text: string, options?: MarkdownRenderOptions): string {
+  let out = ''
+  let i = 0
+  const n = text.length
+  while (i < n) {
+    const char = text[i]
+    // Fenced inline code first.
+    if (char === '`') {
+      const end = text.indexOf('`', i + 1)
+      if (end !== -1) {
+        out += `<code>${escapeHtml(text.slice(i + 1, end))}</code>`
+        i = end + 1
+        continue
+      }
+    }
+    // Image ![alt](src)
+    if (char === '!' && text[i + 1] === '[') {
+      const close = text.indexOf('](', i + 2)
+      if (close !== -1) {
+        const parenEnd = text.indexOf(')', close + 2)
+        if (parenEnd !== -1) {
+          const alt = text.slice(i + 2, close)
+          const src = text.slice(close + 2, parenEnd)
+          const safe = safeUrl(src)
+          if (safe === null) {
+            // Unsafe image target: drop the img, keep the alt text.
+            out += escapeHtml(alt)
+          } else {
+            let target: string | null = safe
+            if (options?.resolveImageSrc !== undefined) {
+              target = options.resolveImageSrc(safe)
+            }
+            if (target === null) {
+              // The resolver dropped the image (unresolvable path): alt only.
+              out += escapeHtml(alt)
+            } else {
+              const srcEsc = escapeHtml(target).replace(/\s+/g, '%20')
+              out += `<img alt="${escapeHtml(alt)}" src="${srcEsc}" />`
+            }
+          }
+          i = parenEnd + 1
+          continue
+        }
+      }
+    }
+    // Link [text](href)
+    if (char === '[') {
+      const close = text.indexOf('](', i + 1)
+      if (close !== -1) {
+        const parenEnd = text.indexOf(')', close + 2)
+        if (parenEnd !== -1) {
+          const label = text.slice(i + 1, close)
+          const href = text.slice(close + 2, parenEnd)
+          const safe = safeUrl(href)
+          if (safe === null) {
+            // Unsafe link target: render the label as plain text, no <a>.
+            out += renderInline(label, options)
+          } else {
+            out += `<a href="${escapeHtml(safe)}" target="_blank" rel="noopener noreferrer">${renderInline(label, options)}</a>`
+          }
+          i = parenEnd + 1
+          continue
+        }
+      }
+    }
+    // Bold **text**
+    if (char === '*' && text[i + 1] === '*') {
+      const end = text.indexOf('**', i + 2)
+      if (end !== -1) {
+        out += `<strong>${renderInline(text.slice(i + 2, end), options)}</strong>`
+        i = end + 2
+        continue
+      }
+    }
+    // Italic *text*
+    if (char === '*' && text[i - 1] !== '*' && text[i + 1] !== '*') {
+      const end = text.indexOf('*', i + 1)
+      if (end !== -1 && text[end + 1] !== '*') {
+        out += `<em>${renderInline(text.slice(i + 1, end), options)}</em>`
+        i = end + 1
+        continue
+      }
+    }
+    // Strikethrough ~~text~~
+    if (char === '~' && text[i + 1] === '~') {
+      const end = text.indexOf('~~', i + 2)
+      if (end !== -1) {
+        out += `<del>${renderInline(text.slice(i + 2, end), options)}</del>`
+        i = end + 2
+        continue
+      }
+    }
+    out += escapeHtml(char)
+    i += 1
+  }
+  return out
+}
+
+/** Render a markdown document to HTML (block pass). */
+export function renderMarkdown(source: string, options?: MarkdownRenderOptions): string {
+  const lines = source.replace(/\r\n/g, '\n').split('\n')
+  const out: string[] = []
+  let i = 0
+  const n = lines.length
+
+  const flushParagraph = (buffer: string[]): void => {
+    if (buffer.length === 0) return
+    out.push(`<p>${renderInline(buffer.join("\n"), options)}</p>`)
+    buffer.length = 0
+  }
+
+  let paragraph: string[] = []
+  while (i < n) {
+    const line = lines[i]
+
+    // Fenced code block.
+    const fence = /^```([\w+-]*)\s*$/.exec(line)
+    if (fence !== null) {
+      flushParagraph(paragraph)
+      const lang = fence[1] ?? ''
+      i += 1
+      const code: string[] = []
+      while (i < n && !/^```\s*$/.test(lines[i])) {
+        code.push(lines[i])
+        i += 1
+      }
+      i += 1 // closing fence
+      const langAttr = lang === '' ? '' : ` class="language-${escapeHtml(lang)}"`
+      out.push(`<pre${langAttr}><code>${escapeHtml(code.join('\n'))}</code></pre>`)
+      continue
+    }
+
+    // Heading.
+    const heading = /^(#{1,6})\s+(.*)$/.exec(line)
+    if (heading !== null) {
+      flushParagraph(paragraph)
+      const level = heading[1].length
+      out.push(`<h${level}>${renderInline(heading[2] ?? '', options)}</h${level}>`)
+      i += 1
+      continue
+    }
+
+    // Horizontal rule.
+    if (/^\s*(---+|\*\*\*+|___+)\s*$/.test(line)) {
+      flushParagraph(paragraph)
+      out.push('<hr />')
+      i += 1
+      continue
+    }
+
+    // Table: header row then separator row.
+    if (line.includes('|') && i + 1 < n && /^\s*\|?[\s:|-]+\|?\s*$/.test(lines[i + 1]) && lines[i + 1].includes('-')) {
+      flushParagraph(paragraph)
+      const headerCells = splitTableRow(line)
+      i += 2
+      const rows: string[][] = []
+      while (i < n && lines[i].includes('|')) {
+        rows.push(splitTableRow(lines[i]))
+        i += 1
+      }
+      out.push('<table>')
+      out.push(`<thead><tr>${headerCells.map((cell) => `<th>${renderInline(cell, options)}</th>`).join('')}</tr></thead>`)
+      if (rows.length > 0) {
+        out.push(`<tbody>${rows.map((row) => `<tr>${row.map((cell) => `<td>${renderInline(cell, options)}</td>`).join('')}</tr>`).join('')}</tbody>`)
+      }
+      out.push('</table>')
+      continue
+    }
+
+    // Blockquote (one level).
+    const quote = /^>\s?(.*)$/.exec(line)
+    if (quote !== null) {
+      flushParagraph(paragraph)
+      const body: string[] = []
+      while (i < n) {
+        const q = /^>\s?(.*)$/.exec(lines[i])
+        if (q === null) break
+        body.push(q[1] ?? '')
+        i += 1
+      }
+      out.push(`<blockquote><p>${body.map((line) => renderInline(line, options)).join('<br />')}</p></blockquote>`)
+      continue
+    }
+
+    // Unordered list.
+    const ul = /^\s*([-*+])\s+(.*)$/.exec(line)
+    if (ul !== null) {
+      flushParagraph(paragraph)
+      const items: string[] = []
+      while (i < n) {
+        const item = /^\s*([-*+])\s+(.*)$/.exec(lines[i])
+        if (item === null) break
+        items.push(`<li>${renderInline(item[2] ?? '', options)}</li>`)
+        i += 1
+      }
+      out.push(`<ul>${items.join('')}</ul>`)
+      continue
+    }
+
+    // Ordered list.
+    const ol = /^\s*\d+[.)]\s+(.*)$/.exec(line)
+    if (ol !== null) {
+      flushParagraph(paragraph)
+      const items: string[] = []
+      while (i < n) {
+        const item = /^\s*\d+[.)]\s+(.*)$/.exec(lines[i])
+        if (item === null) break
+        items.push(`<li>${renderInline(item[1] ?? '', options)}</li>`)
+        i += 1
+      }
+      out.push(`<ol>${items.join('')}</ol>`)
+      continue
+    }
+
+    // Blank line: flush the paragraph.
+    if (line.trim() === '') {
+      flushParagraph(paragraph)
+      i += 1
+      continue
+    }
+
+    paragraph.push(line)
+    i += 1
+  }
+  flushParagraph(paragraph)
+  return out.join('\n')
+}
+
+/** Split one table row into cells (respecting the leading/trailing pipes). */
+function splitTableRow(line: string): string[] {
+  const trimmed = line.trim()
+  const inner = trimmed.startsWith('|') ? trimmed.slice(1) : trimmed
+  const withoutTrailing = inner.endsWith('|') ? inner.slice(0, -1) : inner
+  return withoutTrailing.split('|').map((cell) => cell.trim())
+}