fix: 修复嵌套 markdown fence 导致的渲染截断 (#222)
* fix: keep nested markdown fences rendered * fix: prevent thinking placeholder leaks * fix: normalize nested markdown example fences
This commit is contained in:
@@ -4,6 +4,7 @@ import { useI18n } from 'vue-i18n'
|
|||||||
import { useMessage } from 'naive-ui'
|
import { useMessage } from 'naive-ui'
|
||||||
import MarkdownIt from 'markdown-it'
|
import MarkdownIt from 'markdown-it'
|
||||||
import { handleCodeBlockCopyClick, renderHighlightedCodeBlock } from './highlight'
|
import { handleCodeBlockCopyClick, renderHighlightedCodeBlock } from './highlight'
|
||||||
|
import { repairNestedMarkdownFences } from './markdownFenceRepair'
|
||||||
import { downloadFile } from '@/api/hermes/download'
|
import { downloadFile } from '@/api/hermes/download'
|
||||||
|
|
||||||
const props = withDefaults(defineProps<{
|
const props = withDefaults(defineProps<{
|
||||||
@@ -26,7 +27,7 @@ const md: MarkdownIt = new MarkdownIt({
|
|||||||
})
|
})
|
||||||
|
|
||||||
const renderedHtml = computed(() => {
|
const renderedHtml = computed(() => {
|
||||||
let html = md.render(props.content)
|
let html = md.render(repairNestedMarkdownFences(props.content))
|
||||||
if (props.mentionNames && props.mentionNames.length > 0) {
|
if (props.mentionNames && props.mentionNames.length > 0) {
|
||||||
const escaped = props.mentionNames.map(n => n.replace(/[.*+?^${}()|[\]\\]/g, '\\$&'))
|
const escaped = props.mentionNames.map(n => n.replace(/[.*+?^${}()|[\]\\]/g, '\\$&'))
|
||||||
const re = new RegExp(`(?<=[\\s>]|^)@(${escaped.join('|')})(?=\\s|$)`, 'gi')
|
const re = new RegExp(`(?<=[\\s>]|^)@(${escaped.join('|')})(?=\\s|$)`, 'gi')
|
||||||
|
|||||||
@@ -0,0 +1,216 @@
|
|||||||
|
const MARKDOWN_FENCE_LANGUAGES = new Set(['md', 'markdown', 'mdown', 'mkd'])
|
||||||
|
|
||||||
|
type FenceInfo = {
|
||||||
|
indent: string
|
||||||
|
marker: string
|
||||||
|
fence: string
|
||||||
|
length: number
|
||||||
|
info: string
|
||||||
|
}
|
||||||
|
|
||||||
|
function parseFence(line: string): FenceInfo | null {
|
||||||
|
const match = line.match(/^( {0,3})(`{3,}|~{3,})(.*)$/)
|
||||||
|
if (!match) return null
|
||||||
|
|
||||||
|
const [, indent, fence, rawInfo = ''] = match
|
||||||
|
const marker = fence[0]
|
||||||
|
const info = rawInfo.trim()
|
||||||
|
|
||||||
|
// CommonMark permits backticks in tilde-fence info strings, but not in
|
||||||
|
// backtick-fence info strings. Keeping this distinction prevents inline-ish
|
||||||
|
// malformed backtick text from being promoted into a fence opener.
|
||||||
|
if (marker === '`' && info.includes('`')) return null
|
||||||
|
|
||||||
|
return {
|
||||||
|
indent,
|
||||||
|
marker,
|
||||||
|
fence,
|
||||||
|
length: fence.length,
|
||||||
|
info,
|
||||||
|
}
|
||||||
|
}
|
||||||
|
|
||||||
|
function serializeFence(fence: FenceInfo, length = fence.length, info = fence.info): string {
|
||||||
|
return `${fence.indent}${fence.marker.repeat(length)}${info ? ` ${info}` : ''}`
|
||||||
|
}
|
||||||
|
|
||||||
|
function isMarkdownFence(fence: FenceInfo): boolean {
|
||||||
|
const language = fence.info.split(/\s+/)[0]?.toLowerCase()
|
||||||
|
return MARKDOWN_FENCE_LANGUAGES.has(language)
|
||||||
|
}
|
||||||
|
|
||||||
|
function isClosingFence(line: string, opener: FenceInfo): boolean {
|
||||||
|
const fence = parseFence(line)
|
||||||
|
return Boolean(
|
||||||
|
fence
|
||||||
|
&& fence.marker === opener.marker
|
||||||
|
&& fence.length >= opener.length
|
||||||
|
&& fence.info === '',
|
||||||
|
)
|
||||||
|
}
|
||||||
|
|
||||||
|
function findLastNonEmptyLine(lines: string[], start = lines.length - 1): number {
|
||||||
|
let index = start
|
||||||
|
while (index >= 0 && lines[index].trim() === '') {
|
||||||
|
index -= 1
|
||||||
|
}
|
||||||
|
return index
|
||||||
|
}
|
||||||
|
|
||||||
|
function findFinalClosingFence(lines: string[], opener: FenceInfo, start: number): number {
|
||||||
|
for (let i = findLastNonEmptyLine(lines); i > start; i -= 1) {
|
||||||
|
if (isClosingFence(lines[i], opener)) {
|
||||||
|
return i
|
||||||
|
}
|
||||||
|
}
|
||||||
|
return -1
|
||||||
|
}
|
||||||
|
|
||||||
|
type OpenFence = {
|
||||||
|
marker: string
|
||||||
|
length: number
|
||||||
|
}
|
||||||
|
|
||||||
|
function canBalanceNestedFences(lines: string[], marker: string): boolean {
|
||||||
|
const stack: OpenFence[] = []
|
||||||
|
let sawFence = false
|
||||||
|
|
||||||
|
for (const line of lines) {
|
||||||
|
const fence = parseFence(line)
|
||||||
|
if (!fence || fence.marker !== marker) continue
|
||||||
|
|
||||||
|
sawFence = true
|
||||||
|
const current = stack[stack.length - 1]
|
||||||
|
if (fence.info === '' && current && fence.length >= current.length) {
|
||||||
|
stack.pop()
|
||||||
|
continue
|
||||||
|
}
|
||||||
|
|
||||||
|
// Inside a Markdown example, an unlabeled fence can be either a closing
|
||||||
|
// fence or a literal nested unlabeled example opener. If there is no nested
|
||||||
|
// opener waiting to close, treat it as the latter while evaluating a later
|
||||||
|
// candidate closing fence for the outer example.
|
||||||
|
stack.push({ marker: fence.marker, length: fence.length })
|
||||||
|
}
|
||||||
|
|
||||||
|
return sawFence && stack.length === 0
|
||||||
|
}
|
||||||
|
|
||||||
|
function findBalancedClosingFence(lines: string[], opener: FenceInfo, start: number): number {
|
||||||
|
const candidates: number[] = []
|
||||||
|
|
||||||
|
for (let i = start; i < lines.length; i += 1) {
|
||||||
|
const fence = parseFence(lines[i])
|
||||||
|
if (
|
||||||
|
fence
|
||||||
|
&& fence.marker === opener.marker
|
||||||
|
&& fence.info === ''
|
||||||
|
&& fence.length >= opener.length
|
||||||
|
) {
|
||||||
|
candidates.push(i)
|
||||||
|
}
|
||||||
|
}
|
||||||
|
|
||||||
|
for (let i = candidates.length - 1; i >= 0; i -= 1) {
|
||||||
|
const candidate = candidates[i]
|
||||||
|
if (canBalanceNestedFences(lines.slice(start, candidate), opener.marker)) {
|
||||||
|
return candidate
|
||||||
|
}
|
||||||
|
}
|
||||||
|
|
||||||
|
return candidates[0] ?? -1
|
||||||
|
}
|
||||||
|
|
||||||
|
function maxFenceLength(lines: string[], marker: string): number {
|
||||||
|
let maxLength = 0
|
||||||
|
for (const line of lines) {
|
||||||
|
const fence = parseFence(line)
|
||||||
|
if (fence?.marker === marker) {
|
||||||
|
maxLength = Math.max(maxLength, fence.length)
|
||||||
|
}
|
||||||
|
}
|
||||||
|
return maxLength
|
||||||
|
}
|
||||||
|
|
||||||
|
function promoteMarkdownExampleFences(lines: string[]): string[] {
|
||||||
|
const output: string[] = []
|
||||||
|
|
||||||
|
for (let i = 0; i < lines.length; i += 1) {
|
||||||
|
const opener = parseFence(lines[i])
|
||||||
|
if (!opener || !isMarkdownFence(opener)) {
|
||||||
|
output.push(lines[i])
|
||||||
|
continue
|
||||||
|
}
|
||||||
|
|
||||||
|
const balancedClose = findBalancedClosingFence(lines, opener, i + 1)
|
||||||
|
if (balancedClose === -1) {
|
||||||
|
output.push(lines[i])
|
||||||
|
continue
|
||||||
|
}
|
||||||
|
|
||||||
|
const body = lines.slice(i + 1, balancedClose)
|
||||||
|
const innerMaxLength = maxFenceLength(body, opener.marker)
|
||||||
|
if (innerMaxLength >= opener.length) {
|
||||||
|
const promotedLength = innerMaxLength + 1
|
||||||
|
output.push(serializeFence(opener, promotedLength))
|
||||||
|
output.push(...body)
|
||||||
|
output.push(serializeFence(opener, promotedLength, ''))
|
||||||
|
} else {
|
||||||
|
output.push(lines[i])
|
||||||
|
output.push(...body)
|
||||||
|
output.push(lines[balancedClose])
|
||||||
|
}
|
||||||
|
|
||||||
|
i = balancedClose
|
||||||
|
}
|
||||||
|
|
||||||
|
return output
|
||||||
|
}
|
||||||
|
|
||||||
|
/**
|
||||||
|
* LLMs often wrap a complete PR draft or Markdown answer in an outer
|
||||||
|
* ```md fence. Showing that outer wrapper as a code block makes the UI look
|
||||||
|
* like Markdown rendering is broken: headings, lists, and inline code remain
|
||||||
|
* literal text. Strip only that outer draft wrapper before handing content to
|
||||||
|
* markdown-it.
|
||||||
|
*
|
||||||
|
* The unwrapped draft can still contain Markdown examples that themselves
|
||||||
|
* contain fenced examples. CommonMark closes fences at the first same-marker
|
||||||
|
* line with at least the opener length, so a malformed example like
|
||||||
|
* ```md ... ```md ... ``` ... ``` must be normalized by making the example's
|
||||||
|
* outer fence longer than the literal fences inside it.
|
||||||
|
*/
|
||||||
|
export function repairNestedMarkdownFences(content: string): string {
|
||||||
|
if (!content.includes('```') && !content.includes('~~~')) return content
|
||||||
|
|
||||||
|
const lines = content.split('\n')
|
||||||
|
const output: string[] = []
|
||||||
|
let changed = false
|
||||||
|
|
||||||
|
for (let i = 0; i < lines.length; i += 1) {
|
||||||
|
const opener = parseFence(lines[i])
|
||||||
|
if (!opener || !isMarkdownFence(opener)) {
|
||||||
|
output.push(lines[i])
|
||||||
|
continue
|
||||||
|
}
|
||||||
|
|
||||||
|
const finalClose = findFinalClosingFence(lines, opener, i + 1)
|
||||||
|
if (finalClose === -1) {
|
||||||
|
output.push(lines[i])
|
||||||
|
continue
|
||||||
|
}
|
||||||
|
|
||||||
|
const lastNonEmpty = findLastNonEmptyLine(lines)
|
||||||
|
if (finalClose !== lastNonEmpty) {
|
||||||
|
output.push(lines[i])
|
||||||
|
continue
|
||||||
|
}
|
||||||
|
|
||||||
|
output.push(...promoteMarkdownExampleFences(lines.slice(i + 1, finalClose)))
|
||||||
|
output.push(...lines.slice(finalClose + 1))
|
||||||
|
changed = true
|
||||||
|
break
|
||||||
|
}
|
||||||
|
|
||||||
|
return changed ? output.join('\n') : content
|
||||||
|
}
|
||||||
@@ -14,7 +14,7 @@ const TAG_RE = /<(think|thinking|reasoning)>([\s\S]*?)<\/\1>/gi
|
|||||||
const PLACEHOLDER_PREFIX = '\u0000THKCODE'
|
const PLACEHOLDER_PREFIX = '\u0000THKCODE'
|
||||||
const PLACEHOLDER_SUFFIX = '\u0000'
|
const PLACEHOLDER_SUFFIX = '\u0000'
|
||||||
|
|
||||||
const FENCED_RE = /(```|~~~)([\s\S]*?)\1/g
|
const FENCED_RE = /(^|\n)( {0,3})(`{3,}|~{3,})[^\n]*\n[\s\S]*?\n\2\3[ \t]*(?=\n|$)/g
|
||||||
const INLINE_CODE_RE = /`[^`\n]*`/g
|
const INLINE_CODE_RE = /`[^`\n]*`/g
|
||||||
|
|
||||||
function protectCodeBlocks(input: string): { masked: string; blocks: string[] } {
|
function protectCodeBlocks(input: string): { masked: string; blocks: string[] } {
|
||||||
@@ -32,10 +32,20 @@ function protectCodeBlocks(input: string): { masked: string; blocks: string[] }
|
|||||||
|
|
||||||
function restoreCodeBlocks(text: string, blocks: string[]): string {
|
function restoreCodeBlocks(text: string, blocks: string[]): string {
|
||||||
if (blocks.length === 0) return text
|
if (blocks.length === 0) return text
|
||||||
return text.replace(
|
|
||||||
new RegExp(`${PLACEHOLDER_PREFIX}(\\d+)${PLACEHOLDER_SUFFIX}`, 'g'),
|
const placeholderRe = new RegExp(`${PLACEHOLDER_PREFIX}(\\d+)${PLACEHOLDER_SUFFIX}`, 'g')
|
||||||
(_, idx) => blocks[Number(idx)] ?? '',
|
let restored = text
|
||||||
)
|
|
||||||
|
for (let i = 0; i < blocks.length; i += 1) {
|
||||||
|
const next = restored.replace(
|
||||||
|
placeholderRe,
|
||||||
|
(_, idx) => blocks[Number(idx)] ?? '',
|
||||||
|
)
|
||||||
|
if (next === restored) break
|
||||||
|
restored = next
|
||||||
|
}
|
||||||
|
|
||||||
|
return restored
|
||||||
}
|
}
|
||||||
|
|
||||||
export function parseThinking(content: string, opts: ParseOptions): ParsedThinking {
|
export function parseThinking(content: string, opts: ParseOptions): ParsedThinking {
|
||||||
|
|||||||
@@ -86,6 +86,155 @@ describe('MarkdownRenderer', () => {
|
|||||||
expect(wrapper.find('code.hljs').text()).toContain('INFO Starting server')
|
expect(wrapper.find('code.hljs').text()).toContain('INFO Starting server')
|
||||||
})
|
})
|
||||||
|
|
||||||
|
it('renders outer markdown draft fences as markdown while preserving nested fenced examples', () => {
|
||||||
|
const wrapper = mount(MarkdownRenderer, {
|
||||||
|
props: {
|
||||||
|
content: [
|
||||||
|
'下面是可直接手动编辑的 PR draft。',
|
||||||
|
'',
|
||||||
|
'```md',
|
||||||
|
'标题: fix(chat): 保留附件在同一聊天后续轮次的上下文',
|
||||||
|
'',
|
||||||
|
'## Summary',
|
||||||
|
'',
|
||||||
|
'附件上传后,首轮 `startRun()` 的 `input` 已包含上传文件引用:',
|
||||||
|
'',
|
||||||
|
'```md',
|
||||||
|
'[File: screenshot.png](/uploaded/path)',
|
||||||
|
'```',
|
||||||
|
'',
|
||||||
|
'但本地保存的用户消息只保留 UI 可见文本。',
|
||||||
|
'',
|
||||||
|
'## Fix',
|
||||||
|
'- Preserve context.',
|
||||||
|
'```',
|
||||||
|
].join('\n'),
|
||||||
|
},
|
||||||
|
})
|
||||||
|
|
||||||
|
expect(wrapper.findAll('.hljs-code-block')).toHaveLength(1)
|
||||||
|
expect(wrapper.find('.code-lang').text()).toBe('md')
|
||||||
|
expect(wrapper.find('code.hljs').text()).toContain('[File: screenshot.png](/uploaded/path)')
|
||||||
|
expect(wrapper.find('.markdown-body').findAll('h2')).toHaveLength(2)
|
||||||
|
expect(wrapper.find('.markdown-body').find('h2').text()).toBe('Summary')
|
||||||
|
expect(wrapper.find('.markdown-body').text()).toContain('但本地保存的用户消息只保留 UI 可见文本。')
|
||||||
|
expect(wrapper.find('.markdown-body').text()).toContain('Preserve context.')
|
||||||
|
})
|
||||||
|
|
||||||
|
it('keeps markdown examples with their own nested fences intact after unwrapping a draft fence', () => {
|
||||||
|
const wrapper = mount(MarkdownRenderer, {
|
||||||
|
props: {
|
||||||
|
content: [
|
||||||
|
'```md',
|
||||||
|
'## Regression Coverage',
|
||||||
|
'',
|
||||||
|
'```md',
|
||||||
|
'下面是一个 PR draft。',
|
||||||
|
'',
|
||||||
|
'```md',
|
||||||
|
'[File: Screenshot.png](/tmp/example.png)',
|
||||||
|
'```',
|
||||||
|
'',
|
||||||
|
'## Fix',
|
||||||
|
'',
|
||||||
|
'- 后续 heading 不应被截断。',
|
||||||
|
'```',
|
||||||
|
'',
|
||||||
|
'## Local Verification',
|
||||||
|
'',
|
||||||
|
'- localhost renders after the example.',
|
||||||
|
'```',
|
||||||
|
].join('\n'),
|
||||||
|
},
|
||||||
|
})
|
||||||
|
|
||||||
|
const headings = wrapper.find('.markdown-body').findAll('h2').map(heading => heading.text())
|
||||||
|
expect(headings).toEqual(['Regression Coverage', 'Local Verification'])
|
||||||
|
expect(wrapper.findAll('.hljs-code-block')).toHaveLength(1)
|
||||||
|
|
||||||
|
const codeText = wrapper.find('code.hljs').text()
|
||||||
|
expect(codeText).toContain('下面是一个 PR draft。')
|
||||||
|
expect(codeText).toContain('```md\n[File: Screenshot.png](/tmp/example.png)\n```')
|
||||||
|
expect(codeText).toContain('## Fix')
|
||||||
|
expect(codeText).toContain('- 后续 heading 不应被截断。')
|
||||||
|
expect(wrapper.find('.markdown-body').text()).toContain('localhost renders after the example.')
|
||||||
|
})
|
||||||
|
|
||||||
|
it('keeps markdown examples with unlabeled nested fences intact', () => {
|
||||||
|
const wrapper = mount(MarkdownRenderer, {
|
||||||
|
props: {
|
||||||
|
content: [
|
||||||
|
'```md',
|
||||||
|
'## Unlabeled Fence Example',
|
||||||
|
'',
|
||||||
|
'```md',
|
||||||
|
'```',
|
||||||
|
'plain nested block',
|
||||||
|
'```',
|
||||||
|
'```',
|
||||||
|
'',
|
||||||
|
'Done outside.',
|
||||||
|
'```',
|
||||||
|
].join('\n'),
|
||||||
|
},
|
||||||
|
})
|
||||||
|
|
||||||
|
expect(wrapper.find('.markdown-body').find('h2').text()).toBe('Unlabeled Fence Example')
|
||||||
|
expect(wrapper.findAll('.hljs-code-block')).toHaveLength(1)
|
||||||
|
expect(wrapper.find('code.hljs').text()).toContain('```\nplain nested block\n```')
|
||||||
|
expect(wrapper.find('.markdown-body').text()).toContain('Done outside.')
|
||||||
|
})
|
||||||
|
|
||||||
|
it('keeps tilde-fenced markdown examples with nested tilde fences intact', () => {
|
||||||
|
const wrapper = mount(MarkdownRenderer, {
|
||||||
|
props: {
|
||||||
|
content: [
|
||||||
|
'```md',
|
||||||
|
'## Tilde Example',
|
||||||
|
'',
|
||||||
|
'~~~md',
|
||||||
|
'~~~yaml',
|
||||||
|
'ok: true',
|
||||||
|
'~~~',
|
||||||
|
'~~~',
|
||||||
|
'',
|
||||||
|
'Done outside.',
|
||||||
|
'```',
|
||||||
|
].join('\n'),
|
||||||
|
},
|
||||||
|
})
|
||||||
|
|
||||||
|
expect(wrapper.find('.markdown-body').find('h2').text()).toBe('Tilde Example')
|
||||||
|
expect(wrapper.findAll('.hljs-code-block')).toHaveLength(1)
|
||||||
|
expect(wrapper.find('code.hljs').text()).toContain('~~~yaml\nok: true\n~~~')
|
||||||
|
expect(wrapper.find('.markdown-body').text()).toContain('Done outside.')
|
||||||
|
})
|
||||||
|
|
||||||
|
it('keeps already-valid longer markdown example fences valid', () => {
|
||||||
|
const wrapper = mount(MarkdownRenderer, {
|
||||||
|
props: {
|
||||||
|
content: [
|
||||||
|
'```md',
|
||||||
|
'## Longer Fence Example',
|
||||||
|
'',
|
||||||
|
'````md',
|
||||||
|
'```ts',
|
||||||
|
'const answer = 42',
|
||||||
|
'```',
|
||||||
|
'````',
|
||||||
|
'',
|
||||||
|
'Done outside.',
|
||||||
|
'```',
|
||||||
|
].join('\n'),
|
||||||
|
},
|
||||||
|
})
|
||||||
|
|
||||||
|
expect(wrapper.find('.markdown-body').find('h2').text()).toBe('Longer Fence Example')
|
||||||
|
expect(wrapper.findAll('.hljs-code-block')).toHaveLength(1)
|
||||||
|
expect(wrapper.find('code.hljs').text()).toContain('```ts\nconst answer = 42\n```')
|
||||||
|
expect(wrapper.find('.markdown-body').text()).toContain('Done outside.')
|
||||||
|
})
|
||||||
|
|
||||||
it('copies code through the delegated click handler', async () => {
|
it('copies code through the delegated click handler', async () => {
|
||||||
const writeText = vi.mocked(navigator.clipboard.writeText)
|
const writeText = vi.mocked(navigator.clipboard.writeText)
|
||||||
const wrapper = mount(MarkdownRenderer, {
|
const wrapper = mount(MarkdownRenderer, {
|
||||||
|
|||||||
@@ -92,6 +92,24 @@ describe('parseThinking', () => {
|
|||||||
expect(r.body).toBe('text\n```\n<think>fake</think>\n```')
|
expect(r.body).toBe('text\n```\n<think>fake</think>\n```')
|
||||||
})
|
})
|
||||||
|
|
||||||
|
it('does not leak code-protection placeholders for inline mentions of markdown fences', () => {
|
||||||
|
const src = [
|
||||||
|
'Previous fix kept the outer ` ```md ` block as code.',
|
||||||
|
'',
|
||||||
|
'````md',
|
||||||
|
'下面是可直接手动编辑的 PR draft。',
|
||||||
|
'```md',
|
||||||
|
'标题',
|
||||||
|
'```',
|
||||||
|
'````',
|
||||||
|
].join('\n')
|
||||||
|
const r = parseThinking(src, { streaming: false })
|
||||||
|
expect(r.hasThinking).toBe(false)
|
||||||
|
expect(r.body).toBe(src)
|
||||||
|
expect(r.body).not.toContain('THKCODE')
|
||||||
|
expect(r.body).not.toContain('\u0000')
|
||||||
|
})
|
||||||
|
|
||||||
it('same-name nesting: inner tag absorbed into first segment (documented limitation)', () => {
|
it('same-name nesting: inner tag absorbed into first segment (documented limitation)', () => {
|
||||||
const r = parseThinking('<think>a<think>b</think>c</think>', { streaming: false })
|
const r = parseThinking('<think>a<think>b</think>c</think>', { streaming: false })
|
||||||
expect(r.segments).toEqual(['a<think>b'])
|
expect(r.segments).toEqual(['a<think>b'])
|
||||||
|
|||||||
Reference in New Issue
Block a user