@incremark/core 0.2.2 → 0.2.4

package/src/benchmark/run.ts

@@ -1,93 +0,0 @@
-import { runBenchmark } from './index'
-
-console.log('')
-console.log('🔬 Running Incremark Benchmark Suite')
-console.log('')
-
-// 不同文档长度的测试
-const documentSizes = [
-  { name: 'Short (~1KB)', length: 1000 },
-  { name: 'Medium (~5KB)', length: 5000 },
-  { name: 'Long (~10KB)', length: 10000 },
-  { name: 'Very Long (~20KB)', length: 20000 },
-]
-
-const allResults: Array<{
-  docSize: string
-  docLength: number
-  chunkSize: number
-  timeSaved: number
-  charsSaved: number
-  speedup: number
-}> = []
-
-for (const doc of documentSizes) {
-  console.log(`\n${'='.repeat(60)}`)
-  console.log(`📄 Document Size: ${doc.name} (${doc.length} chars)`)
-  console.log('='.repeat(60))
-  
-  // 测试不同 chunk 大小
-  for (const chunkSize of [10, 50]) {
-    console.log(`\n📦 Chunk size: ${chunkSize} chars\n`)
-    const result = runBenchmark({ 
-      chunkSize, 
-      iterations: 20,
-      markdownLength: doc.length
-    })
-    allResults.push({
-      docSize: doc.name,
-      docLength: doc.length,
-      chunkSize,
-      timeSaved: result.timeSaved,
-      charsSaved: result.charsSaved,
-      speedup: result.speedup
-    })
-  }
-}
-
-// 汇总报告
-console.log('\n')
-console.log('='.repeat(80))
-console.log('📈 Complete Benchmark Summary')
-console.log('='.repeat(80))
-console.log('')
-console.log('| Document Size    | Chunk | Time Saved | Chars Saved | Speedup |')
-console.log('|------------------|-------|------------|-------------|---------|')
-for (const r of allResults) {
-  console.log(`| ${r.docSize.padEnd(16)} | ${r.chunkSize.toString().padEnd(5)} | ${r.timeSaved.toFixed(1).padStart(9)}% | ${r.charsSaved.toFixed(1).padStart(10)}% | ${r.speedup.toFixed(2).padStart(6)}x |`)
-}
-console.log('')
-
-// 计算平均值
-const avgTimeSaved = allResults.reduce((sum, r) => sum + r.timeSaved, 0) / allResults.length
-const avgCharsSaved = allResults.reduce((sum, r) => sum + r.charsSaved, 0) / allResults.length
-const avgSpeedup = allResults.reduce((sum, r) => sum + r.speedup, 0) / allResults.length
-
-// 按文档大小分组计算
-const byDocSize = new Map<string, typeof allResults>()
-for (const r of allResults) {
-  if (!byDocSize.has(r.docSize)) {
-    byDocSize.set(r.docSize, [])
-  }
-  byDocSize.get(r.docSize)!.push(r)
-}
-
-console.log('-'.repeat(80))
-console.log('')
-console.log('📊 Average by Document Size:')
-console.log('')
-for (const [size, results] of byDocSize) {
-  const avgSpeed = results.reduce((sum, r) => sum + r.speedup, 0) / results.length
-  const avgTime = results.reduce((sum, r) => sum + r.timeSaved, 0) / results.length
-  console.log(`   ${size}: ${avgSpeed.toFixed(2)}x faster, ${avgTime.toFixed(1)}% time saved`)
-}
-console.log('')
-
-console.log('-'.repeat(80))
-console.log('')
-console.log(`🎯 Overall Average:`)
-console.log(`   Time Saved: ${avgTimeSaved.toFixed(1)}%`)
-console.log(`   Chars Saved: ${avgCharsSaved.toFixed(1)}%`)
-console.log(`   Speedup: ${avgSpeedup.toFixed(2)}x`)
-console.log('')
-console.log('='.repeat(80))

package/src/detector/index.test.ts

@@ -1,150 +0,0 @@
-import { describe, it, expect } from 'vitest'
-import {
-  detectFenceStart,
-  detectFenceEnd,
-  isEmptyLine,
-  isHeading,
-  isThematicBreak,
-  isListItemStart,
-  isBlockquoteStart,
-  detectContainer,
-  createInitialContext
-} from './index'
-
-describe('块检测器', () => {
-  describe('detectFenceStart', () => {
-    it('检测反引号 fence', () => {
-      const result = detectFenceStart('```js')
-      expect(result).toEqual({ char: '`', length: 3 })
-    })
-
-    it('检测波浪号 fence', () => {
-      const result = detectFenceStart('~~~python')
-      expect(result).toEqual({ char: '~', length: 3 })
-    })
-
-    it('检测更长的 fence', () => {
-      const result = detectFenceStart('`````')
-      expect(result).toEqual({ char: '`', length: 5 })
-    })
-
-    it('非 fence 返回 null', () => {
-      expect(detectFenceStart('普通文本')).toBeNull()
-      expect(detectFenceStart('``不够长')).toBeNull()
-    })
-  })
-
-  describe('detectFenceEnd', () => {
-    it('匹配相同长度', () => {
-      const context = { ...createInitialContext(), inFencedCode: true, fenceChar: '`', fenceLength: 3 }
-      expect(detectFenceEnd('```', context)).toBe(true)
-    })
-
-    it('匹配更长的结束', () => {
-      const context = { ...createInitialContext(), inFencedCode: true, fenceChar: '`', fenceLength: 3 }
-      expect(detectFenceEnd('`````', context)).toBe(true)
-    })
-
-    it('不匹配更短的', () => {
-      const context = { ...createInitialContext(), inFencedCode: true, fenceChar: '`', fenceLength: 5 }
-      expect(detectFenceEnd('```', context)).toBe(false)
-    })
-  })
-
-  describe('isEmptyLine', () => {
-    it('空行', () => {
-      expect(isEmptyLine('')).toBe(true)
-      expect(isEmptyLine('   ')).toBe(true)
-      expect(isEmptyLine('\t')).toBe(true)
-    })
-
-    it('非空行', () => {
-      expect(isEmptyLine('内容')).toBe(false)
-      expect(isEmptyLine('  a  ')).toBe(false)
-    })
-  })
-
-  describe('isHeading', () => {
-    it('有效标题', () => {
-      expect(isHeading('# H1')).toBe(true)
-      expect(isHeading('## H2')).toBe(true)
-      expect(isHeading('###### H6')).toBe(true)
-    })
-
-    it('无效标题', () => {
-      expect(isHeading('####### H7')).toBe(false) // 超过 6 级
-      expect(isHeading('#没有空格')).toBe(false)
-      expect(isHeading('普通文本')).toBe(false)
-    })
-  })
-
-  describe('isThematicBreak', () => {
-    it('有效水平线', () => {
-      expect(isThematicBreak('---')).toBe(true)
-      expect(isThematicBreak('***')).toBe(true)
-      expect(isThematicBreak('___')).toBe(true)
-      expect(isThematicBreak('-----')).toBe(true)
-    })
-
-    it('无效水平线', () => {
-      expect(isThematicBreak('--')).toBe(false)
-      expect(isThematicBreak('- - -')).toBe(false) // 有空格
-    })
-  })
-
-  describe('isListItemStart', () => {
-    it('无序列表', () => {
-      expect(isListItemStart('- item')).toEqual({ ordered: false, indent: 0 })
-      expect(isListItemStart('  * item')).toEqual({ ordered: false, indent: 2 })
-    })
-
-    it('有序列表', () => {
-      expect(isListItemStart('1. item')).toEqual({ ordered: true, indent: 0 })
-      expect(isListItemStart('  99. item')).toEqual({ ordered: true, indent: 2 })
-    })
-
-    it('非列表', () => {
-      expect(isListItemStart('普通文本')).toBeNull()
-    })
-  })
-
-  describe('isBlockquoteStart', () => {
-    it('有效引用', () => {
-      expect(isBlockquoteStart('> quote')).toBe(true)
-      expect(isBlockquoteStart('  > quote')).toBe(true)
-    })
-
-    it('无效引用', () => {
-      expect(isBlockquoteStart('    > 太多缩进')).toBe(false)
-    })
-  })
-
-  describe('detectContainer', () => {
-    it('检测容器开始', () => {
-      const result = detectContainer('::: warning')
-      expect(result).toEqual({ name: 'warning', markerLength: 3, isEnd: false })
-    })
-
-    it('检测容器结束', () => {
-      const result = detectContainer(':::')
-      expect(result).toEqual({ name: '', markerLength: 3, isEnd: true })
-    })
-
-    it('检测更长的标记', () => {
-      const result = detectContainer('::::: outer')
-      expect(result).toEqual({ name: 'outer', markerLength: 5, isEnd: false })
-    })
-
-    it('名称白名单', () => {
-      const config = { allowedNames: ['warning', 'info'] }
-      expect(detectContainer('::: warning', config)).not.toBeNull()
-      expect(detectContainer('::: danger', config)).toBeNull()
-    })
-
-    it('非容器语法', () => {
-      expect(detectContainer('普通文本')).toBeNull()
-      expect(detectContainer(':: 太短')).toBeNull()
-    })
-  })
-})
-

package/src/detector/index.ts

@@ -1,330 +0,0 @@
-/**
- * 块类型检测与边界判断
- *
- * Markdown 块级元素的识别规则
- */
-
-import type { BlockContext, ContainerConfig, ContainerMatch } from '../types'
-
-// ============ 预编译正则表达式（性能优化） ============
-
-const RE_FENCE_START = /^(\s*)((`{3,})|(~{3,}))/
-const RE_EMPTY_LINE = /^\s*$/
-const RE_HEADING = /^#{1,6}\s/
-const RE_THEMATIC_BREAK = /^(\*{3,}|-{3,}|_{3,})\s*$/
-const RE_UNORDERED_LIST = /^(\s*)([-*+])\s/
-const RE_ORDERED_LIST = /^(\s*)(\d{1,9})[.)]\s/
-const RE_BLOCKQUOTE = /^\s{0,3}>/
-const RE_HTML_BLOCK_1 = /^\s{0,3}<(script|pre|style|textarea|!--|!DOCTYPE|\?|!\[CDATA\[)/i
-const RE_HTML_BLOCK_2 = /^\s{0,3}<\/?[a-zA-Z][a-zA-Z0-9-]*(\s|>|$)/
-const RE_TABLE_DELIMITER = /^\|?\s*:?-{3,}:?\s*(\|\s*:?-{3,}:?\s*)*\|?$/
-const RE_ESCAPE_SPECIAL = /[.*+?^${}()|[\]\\]/g
-const RE_FOOTNOTE_DEFINITION = /^\[\^[^\]]+\]:\s/
-const RE_FOOTNOTE_CONTINUATION = /^(?:    |\t)/
-
-/** fence 结束模式缓存 */
-const fenceEndPatternCache = new Map<string, RegExp>()
-
-/** 容器模式缓存 */
-const containerPatternCache = new Map<string, RegExp>()
-
-// ============ 代码块检测 ============
-
-/**
- * 检测行是否是代码块 fence 开始
- */
-export function detectFenceStart(line: string): { char: string; length: number } | null {
-  const match = line.match(RE_FENCE_START)
-  if (match) {
-    const fence = match[2]
-    const char = fence[0]
-    return { char, length: fence.length }
-  }
-  return null
-}
-
-/**
- * 检测行是否是代码块 fence 结束
- */
-export function detectFenceEnd(line: string, context: BlockContext): boolean {
-  if (!context.inFencedCode || !context.fenceChar || !context.fenceLength) {
-    return false
-  }
-
-  // 使用缓存的正则表达式
-  const cacheKey = `${context.fenceChar}-${context.fenceLength}`
-  let pattern = fenceEndPatternCache.get(cacheKey)
-  if (!pattern) {
-    pattern = new RegExp(`^\\s{0,3}${context.fenceChar}{${context.fenceLength},}\\s*$`)
-    fenceEndPatternCache.set(cacheKey, pattern)
-  }
-  return pattern.test(line)
-}
-
-// ============ 行类型检测 ============
-
-/**
- * 检测是否是空行或仅包含空白字符
- */
-export function isEmptyLine(line: string): boolean {
-  return RE_EMPTY_LINE.test(line)
-}
-
-/**
- * 检测是否是标题行
- */
-export function isHeading(line: string): boolean {
-  return RE_HEADING.test(line)
-}
-
-/**
- * 检测是否是 thematic break（水平线）
- */
-export function isThematicBreak(line: string): boolean {
-  return RE_THEMATIC_BREAK.test(line.trim())
-}
-
-/**
- * 检测是否是列表项开始
- */
-export function isListItemStart(line: string): { ordered: boolean; indent: number } | null {
-  // 无序列表: - * +
-  const unordered = line.match(RE_UNORDERED_LIST)
-  if (unordered) {
-    return { ordered: false, indent: unordered[1].length }
-  }
-
-  // 有序列表: 1. 2) 等
-  const ordered = line.match(RE_ORDERED_LIST)
-  if (ordered) {
-    return { ordered: true, indent: ordered[1].length }
-  }
-
-  return null
-}
-
-/**
- * 检测是否是引用块开始
- */
-export function isBlockquoteStart(line: string): boolean {
-  return RE_BLOCKQUOTE.test(line)
-}
-
-/**
- * 检测是否是 HTML 块
- */
-export function isHtmlBlock(line: string): boolean {
-  return RE_HTML_BLOCK_1.test(line) || RE_HTML_BLOCK_2.test(line)
-}
-
-/**
- * 检测表格分隔行
- */
-export function isTableDelimiter(line: string): boolean {
-  return RE_TABLE_DELIMITER.test(line.trim())
-}
-
-// ============ 脚注检测 ============
-
-/**
- * 检测是否是脚注定义的起始行
- * 格式: [^id]: content
- * 
- * @example
- * isFootnoteDefinitionStart('[^1]: 脚注内容') // true
- * isFootnoteDefinitionStart('[^note]: 内容') // true
- * isFootnoteDefinitionStart('    缩进内容')   // false
- */
-export function isFootnoteDefinitionStart(line: string): boolean {
-  return RE_FOOTNOTE_DEFINITION.test(line)
-}
-
-/**
- * 检测是否是脚注定义的延续行（缩进行）
- * 至少4个空格或1个tab
- * 
- * @example
- * isFootnoteContinuation('    第二行')  // true
- * isFootnoteContinuation('\t第二行')    // true
- * isFootnoteContinuation('  两个空格')   // false
- */
-export function isFootnoteContinuation(line: string): boolean {
-  return RE_FOOTNOTE_CONTINUATION.test(line)
-}
-
-// ============ 容器检测 ============
-
-/**
- * 检测容器开始或结束
- *
- * 支持格式：
- * - ::: name      开始
- * - ::: name attr 开始（带属性）
- * - :::           结束
- * - :::::: name   开始（更长的标记，用于嵌套）
- */
-export function detectContainer(line: string, config?: ContainerConfig): ContainerMatch | null {
-  const marker = config?.marker || ':'
-  const minLength = config?.minMarkerLength || 3
-
-  // 使用缓存的正则表达式
-  const cacheKey = `${marker}-${minLength}`
-  let pattern = containerPatternCache.get(cacheKey)
-  if (!pattern) {
-    const escapedMarker = marker.replace(RE_ESCAPE_SPECIAL, '\\$&')
-    pattern = new RegExp(
-      `^(\\s*)(${escapedMarker}{${minLength},})(?:\\s+(\\w[\\w-]*))?(?:\\s+(.*))?\\s*$`
-    )
-    containerPatternCache.set(cacheKey, pattern)
-  }
-
-  const match = line.match(pattern)
-  if (!match) {
-    return null
-  }
-
-  const markerLength = match[2].length
-  const name = match[3] || ''
-  const isEnd = !name && !match[4]
-
-  if (!isEnd && config?.allowedNames && config.allowedNames.length > 0) {
-    if (!config.allowedNames.includes(name)) {
-      return null
-    }
-  }
-
-  return { name, markerLength, isEnd }
-}
-
-/**
- * 检测容器结束
- */
-export function detectContainerEnd(
-  line: string,
-  context: BlockContext,
-  config?: ContainerConfig
-): boolean {
-  if (!context.inContainer || !context.containerMarkerLength) {
-    return false
-  }
-
-  const result = detectContainer(line, config)
-  if (!result) {
-    return false
-  }
-
-  return result.isEnd && result.markerLength >= context.containerMarkerLength
-}
-
-// ============ 边界检测 ============
-
-/**
- * 判断两行之间是否构成块边界
- */
-export function isBlockBoundary(
-  prevLine: string,
-  currentLine: string,
-  context: BlockContext
-): boolean {
-  if (context.inFencedCode) {
-    return detectFenceEnd(currentLine, context)
-  }
-
-  if (isEmptyLine(prevLine) && !isEmptyLine(currentLine)) {
-    return true
-  }
-
-  if (isHeading(currentLine) && !isEmptyLine(prevLine)) {
-    return true
-  }
-
-  if (isThematicBreak(currentLine)) {
-    return true
-  }
-
-  if (detectFenceStart(currentLine)) {
-    return true
-  }
-
-  return false
-}
-
-// ============ 上下文管理 ============
-
-/**
- * 创建初始上下文
- */
-export function createInitialContext(): BlockContext {
-  return {
-    inFencedCode: false,
-    listDepth: 0,
-    blockquoteDepth: 0,
-    inContainer: false,
-    containerDepth: 0
-  }
-}
-
-/**
- * 更新上下文（处理一行后）
- */
-export function updateContext(
-  line: string,
-  context: BlockContext,
-  containerConfig?: ContainerConfig | boolean
-): BlockContext {
-  const newContext = { ...context }
-
-  const containerCfg =
-    containerConfig === true ? {} : containerConfig === false ? undefined : containerConfig
-
-  // 代码块优先级最高
-  if (context.inFencedCode) {
-    if (detectFenceEnd(line, context)) {
-      newContext.inFencedCode = false
-      newContext.fenceChar = undefined
-      newContext.fenceLength = undefined
-    }
-    return newContext
-  }
-
-  const fence = detectFenceStart(line)
-  if (fence) {
-    newContext.inFencedCode = true
-    newContext.fenceChar = fence.char
-    newContext.fenceLength = fence.length
-    return newContext
-  }
-
-  // 容器处理
-  if (containerCfg !== undefined) {
-    if (context.inContainer) {
-      if (detectContainerEnd(line, context, containerCfg)) {
-        newContext.containerDepth = context.containerDepth - 1
-        if (newContext.containerDepth === 0) {
-          newContext.inContainer = false
-          newContext.containerMarkerLength = undefined
-          newContext.containerName = undefined
-        }
-        return newContext
-      }
-
-      const nested = detectContainer(line, containerCfg)
-      if (nested && !nested.isEnd) {
-        newContext.containerDepth = context.containerDepth + 1
-        return newContext
-      }
-    } else {
-      const container = detectContainer(line, containerCfg)
-      if (container && !container.isEnd) {
-        newContext.inContainer = true
-        newContext.containerMarkerLength = container.markerLength
-        newContext.containerName = container.name
-        newContext.containerDepth = 1
-        return newContext
-      }
-    }
-  }
-
-  return newContext
-}
-