@tiptap/core 3.6.6 → 3.7.0

package/src/utilities/markdown/parseIndentedBlocks.ts

@@ -0,0 +1,193 @@
+/**
+ * @fileoverview Utility for parsing indented markdown blocks with hierarchical nesting.
+ *
+ * This utility handles the complex logic of parsing markdown blocks that can contain
+ * nested content based on indentation levels, maintaining proper hierarchical structure
+ * for lists, task lists, and other indented block types.
+ */
+
+export interface ParsedBlock {
+  type: string
+  raw: string
+  mainContent: string
+  indentLevel: number
+  nestedContent?: string
+  nestedTokens?: any[]
+  [key: string]: any
+}
+
+export interface BlockParserConfig {
+  /** Regex pattern to match block items */
+  itemPattern: RegExp
+  /** Function to extract data from regex match */
+  extractItemData: (match: RegExpMatchArray) => {
+    mainContent: string
+    indentLevel: number
+    [key: string]: any
+  }
+  /** Function to create the final token */
+  createToken: (data: any, nestedTokens?: any[]) => ParsedBlock
+  /** Base indentation to remove from nested content (default: 2 spaces) */
+  baseIndentSize?: number
+  /**
+   * Custom parser for nested content. If provided, this will be called instead
+   * of the default lexer.blockTokens() for parsing nested content.
+   * This allows recursive parsing of the same block type.
+   */
+  customNestedParser?: (dedentedContent: string) => any[] | undefined
+}
+
+/**
+ * Parses markdown text into hierarchical indented blocks with proper nesting.
+ *
+ * This utility handles:
+ * - Line-by-line parsing with pattern matching
+ * - Hierarchical nesting based on indentation levels
+ * - Nested content collection and parsing
+ * - Empty line handling
+ * - Content dedenting for nested blocks
+ *
+ * The key difference from flat parsing is that this maintains the hierarchical
+ * structure where nested items become `nestedTokens` of their parent items,
+ * rather than being flattened into a single array.
+ *
+ * @param src - The markdown source text to parse
+ * @param config - Configuration object defining how to parse and create tokens
+ * @param lexer - Markdown lexer for parsing nested content
+ * @returns Parsed result with hierarchical items, or undefined if no matches
+ *
+ * @example
+ * ```ts
+ * const result = parseIndentedBlocks(src, {
+ *   itemPattern: /^(\s*)([-+*])\s+\[([ xX])\]\s+(.*)$/,
+ *   extractItemData: (match) => ({
+ *     indentLevel: match[1].length,
+ *     mainContent: match[4],
+ *     checked: match[3].toLowerCase() === 'x'
+ *   }),
+ *   createToken: (data, nestedTokens) => ({
+ *     type: 'taskItem',
+ *     checked: data.checked,
+ *     text: data.mainContent,
+ *     nestedTokens
+ *   })
+ * }, lexer)
+ * ```
+ */
+export function parseIndentedBlocks(
+  src: string,
+  config: BlockParserConfig,
+  lexer: {
+    inlineTokens: (src: string) => any[]
+    blockTokens: (src: string) => any[]
+  },
+):
+  | {
+      items: ParsedBlock[]
+      raw: string
+    }
+  | undefined {
+  const lines = src.split('\n')
+  const items: ParsedBlock[] = []
+  let totalRaw = ''
+  let i = 0
+  const baseIndentSize = config.baseIndentSize || 2
+
+  while (i < lines.length) {
+    const currentLine = lines[i]
+    const itemMatch = currentLine.match(config.itemPattern)
+
+    if (!itemMatch) {
+      // Not a matching item - stop if we have items, otherwise this isn't our block type
+      if (items.length > 0) {
+        break
+      } else if (currentLine.trim() === '') {
+        i += 1
+        continue
+      } else {
+        return undefined
+      }
+    }
+
+    const itemData = config.extractItemData(itemMatch)
+    const { indentLevel, mainContent } = itemData
+    totalRaw = `${totalRaw}${currentLine}\n`
+
+    // Collect content for this item (including nested items)
+    const itemContent = [mainContent] // Start with the main text
+    i += 1
+
+    // Look ahead for nested content (indented more than current item)
+    while (i < lines.length) {
+      const nextLine = lines[i]
+
+      if (nextLine.trim() === '') {
+        // Empty line - might be end of nested content
+        const nextNonEmptyIndex = lines.slice(i + 1).findIndex(l => l.trim() !== '')
+        if (nextNonEmptyIndex === -1) {
+          // No more content
+          break
+        }
+
+        const nextNonEmpty = lines[i + 1 + nextNonEmptyIndex]
+        const nextIndent = nextNonEmpty.match(/^(\s*)/)?.[1]?.length || 0
+
+        if (nextIndent > indentLevel) {
+          // Nested content continues after empty line
+          itemContent.push(nextLine)
+          totalRaw = `${totalRaw}${nextLine}\n`
+          i += 1
+          continue
+        } else {
+          // End of nested content
+          break
+        }
+      }
+
+      const nextIndent = nextLine.match(/^(\s*)/)?.[1]?.length || 0
+
+      if (nextIndent > indentLevel) {
+        // This is nested content for the current item
+        itemContent.push(nextLine)
+        totalRaw = `${totalRaw}${nextLine}\n`
+        i += 1
+      } else {
+        // Same or less indentation - this belongs to parent level
+        break
+      }
+    }
+
+    // Parse nested content if present
+    let nestedTokens: any[] | undefined
+    const nestedContent = itemContent.slice(1)
+
+    if (nestedContent.length > 0) {
+      // Remove the base indentation from nested content
+      const dedentedNested = nestedContent
+        .map(nestedLine => nestedLine.slice(indentLevel + baseIndentSize)) // Remove base indent + 2 spaces
+        .join('\n')
+
+      if (dedentedNested.trim()) {
+        // Use custom nested parser if provided, otherwise fall back to default
+        if (config.customNestedParser) {
+          nestedTokens = config.customNestedParser(dedentedNested)
+        } else {
+          nestedTokens = lexer.blockTokens(dedentedNested)
+        }
+      }
+    }
+
+    // Create the token using the provided factory function
+    const token = config.createToken(itemData, nestedTokens)
+    items.push(token)
+  }
+
+  if (items.length === 0) {
+    return undefined
+  }
+
+  return {
+    items,
+    raw: totalRaw.trim(),
+  }
+}

package/src/utilities/markdown/renderNestedMarkdownContent.ts

@@ -0,0 +1,94 @@
+import type { JSONContent } from '@tiptap/core'
+
+/**
+ * @fileoverview Utility functions for rendering nested content in markdown.
+ *
+ * This module provides reusable utilities for extensions that need to render
+ * content with a prefix on the main line and properly indented nested content.
+ */
+
+/**
+ * Utility function for rendering content with a main line prefix and nested indented content.
+ *
+ * This function handles the common pattern of rendering content with:
+ * 1. A main line with a prefix (like "- " for lists, "> " for blockquotes, etc.)
+ * 2. Nested content that gets indented properly
+ *
+ * @param node - The ProseMirror node representing the content
+ * @param h - The markdown renderer helper
+ * @param prefixOrGenerator - Either a string prefix or a function that generates the prefix from context
+ * @param ctx - Optional context object (used when prefixOrGenerator is a function)
+ * @returns The rendered markdown string
+ *
+ * @example
+ * ```ts
+ * // For a bullet list item with static prefix
+ * return renderNestedMarkdownContent(node, h, '- ')
+ *
+ * // For a task item with static prefix
+ * const prefix = `- [${node.attrs?.checked ? 'x' : ' '}] `
+ * return renderNestedMarkdownContent(node, h, prefix)
+ *
+ * // For a blockquote with static prefix
+ * return renderNestedMarkdownContent(node, h, '> ')
+ *
+ * // For content with dynamic prefix based on context
+ * return renderNestedMarkdownContent(node, h, ctx => {
+ *   if (ctx.parentType === 'orderedList') {
+ *     return `${ctx.index + 1}. `
+ *   }
+ *   return '- '
+ * }, ctx)
+ *
+ * // Custom extension example
+ * const CustomContainer = Node.create({
+ *   name: 'customContainer',
+ *   // ... other config
+ *   markdown: {
+ *     render: (node, h) => {
+ *       const type = node.attrs?.type || 'info'
+ *       return renderNestedMarkdownContent(node, h, `[${type}] `)
+ *     }
+ *   }
+ * })
+ * ```
+ */
+export function renderNestedMarkdownContent(
+  node: JSONContent,
+  h: {
+    renderChildren: (nodes: JSONContent[]) => string
+    indent: (text: string) => string
+  },
+  prefixOrGenerator: string | ((ctx: any) => string),
+  ctx?: any,
+): string {
+  if (!node || !Array.isArray(node.content)) {
+    return ''
+  }
+
+  // Determine the prefix based on the input
+  const prefix = typeof prefixOrGenerator === 'function' ? prefixOrGenerator(ctx) : prefixOrGenerator
+
+  const [content, ...children] = node.content
+
+  // Render the main content (typically a paragraph)
+  const mainContent = h.renderChildren([content])
+  const output = [`${prefix}${mainContent}`]
+
+  // Handle nested children with proper indentation
+  if (children && children.length > 0) {
+    children.forEach(child => {
+      const childContent = h.renderChildren([child])
+      if (childContent) {
+        // Split the child content by lines and indent each line
+        const indentedChild = childContent
+          .split('\n')
+          .map(line => (line ? h.indent(line) : ''))
+          .join('\n')
+        output.push(indentedChild)
+      }
+    })
+  }
+
+  return output.join('\n')
+}

package/dist/jsx-runtime/jsx-runtime.d.cts

@@ -1,23 +0,0 @@
-type Attributes = Record<string, any>;
-type DOMOutputSpecElement = 0 | Attributes | DOMOutputSpecArray;
-/**
- * Better describes the output of a `renderHTML` function in prosemirror
- * @see https://prosemirror.net/docs/ref/#model.DOMOutputSpec
- */
-type DOMOutputSpecArray = [string] | [string, Attributes] | [string, 0] | [string, Attributes, 0] | [string, Attributes, DOMOutputSpecArray | 0] | [string, DOMOutputSpecArray];
-declare namespace JSX {
-    type Element = DOMOutputSpecArray;
-    interface IntrinsicElements {
-        [key: string]: any;
-    }
-    interface ElementChildrenAttribute {
-        children: unknown;
-    }
-}
-type JSXRenderer = (tag: 'slot' | string | ((props?: Attributes) => DOMOutputSpecArray | DOMOutputSpecElement), props?: Attributes, ...children: JSXRenderer[]) => DOMOutputSpecArray | DOMOutputSpecElement;
-declare function Fragment(props: {
-    children: JSXRenderer[];
-}): JSXRenderer[];
-declare const h: JSXRenderer;
-
-export { type Attributes, type DOMOutputSpecArray, type DOMOutputSpecElement, Fragment, JSX, type JSXRenderer, h as createElement, h, h as jsx, h as jsxDEV, h as jsxs };

package/dist/jsx-runtime/jsx-runtime.d.ts

@@ -1,23 +0,0 @@
-type Attributes = Record<string, any>;
-type DOMOutputSpecElement = 0 | Attributes | DOMOutputSpecArray;
-/**
- * Better describes the output of a `renderHTML` function in prosemirror
- * @see https://prosemirror.net/docs/ref/#model.DOMOutputSpec
- */
-type DOMOutputSpecArray = [string] | [string, Attributes] | [string, 0] | [string, Attributes, 0] | [string, Attributes, DOMOutputSpecArray | 0] | [string, DOMOutputSpecArray];
-declare namespace JSX {
-    type Element = DOMOutputSpecArray;
-    interface IntrinsicElements {
-        [key: string]: any;
-    }
-    interface ElementChildrenAttribute {
-        children: unknown;
-    }
-}
-type JSXRenderer = (tag: 'slot' | string | ((props?: Attributes) => DOMOutputSpecArray | DOMOutputSpecElement), props?: Attributes, ...children: JSXRenderer[]) => DOMOutputSpecArray | DOMOutputSpecElement;
-declare function Fragment(props: {
-    children: JSXRenderer[];
-}): JSXRenderer[];
-declare const h: JSXRenderer;
-
-export { type Attributes, type DOMOutputSpecArray, type DOMOutputSpecElement, Fragment, JSX, type JSXRenderer, h as createElement, h, h as jsx, h as jsxDEV, h as jsxs };