@flamingo-stack/openframe-frontend-core 0.0.508 → 0.0.509

package/src/components/chat/utils/chunk-parser.ts

@@ -1,262 +0,0 @@
-/**
- * Utilities for parsing NATS chunks into structured actions
- */
-
-import { MESSAGE_TYPE, type ChunkData, type ParsedChunkAction, type ToolExecutionSegment, type PendingToolCallData } from '../types'
-
-function normalizeToolCalls(raw: unknown): PendingToolCallData[] {
-  if (!Array.isArray(raw)) return []
-  return raw
-    .filter((item): item is Record<string, any> => !!item && typeof item === 'object')
-    .map((item) => ({
-      toolExecutionRequestId: String(item.toolExecutionRequestId ?? ''),
-      toolName: String(item.toolName ?? ''),
-      toolTitle: typeof item.toolTitle === 'string' ? item.toolTitle : undefined,
-      toolExplanation: typeof item.toolExplanation === 'string' ? item.toolExplanation : undefined,
-      toolType: typeof item.toolType === 'string' ? item.toolType : undefined,
-      requiresApproval: item.requiresApproval === true,
-      approvalType: typeof item.approvalType === 'string' ? item.approvalType : null,
-      toolCallArguments:
-        item.toolCallArguments && typeof item.toolCallArguments === 'object'
-          ? (item.toolCallArguments as Record<string, any>)
-          : null,
-    }))
-}
-
-/**
- * Parse a raw NATS chunk into a structured action
- */
-export function parseChunkToAction(chunk: unknown): ParsedChunkAction | null {
-  if (!chunk || typeof chunk !== 'object') return null
-  
-  const data = chunk as ChunkData
-  const type = String(data.type || '')
-
-  switch (type) {
-    case MESSAGE_TYPE.MESSAGE_START:
-      return { action: 'message_start' }
-      
-    case MESSAGE_TYPE.MESSAGE_END:
-      return { action: 'message_end' }
-      
-    case MESSAGE_TYPE.AI_METADATA: {
-      const providerName = data.providerName || data.provider
-      if (typeof data.modelName === 'string' && typeof providerName === 'string') {
-        return {
-          action: 'metadata',
-          modelDisplayName: data.modelDisplayName,
-          modelName: data.modelName,
-          providerName,
-          contextWindow: typeof data.contextWindow === 'number' ? data.contextWindow : 0,
-        }
-      }
-      return null
-    }
-    
-    case MESSAGE_TYPE.TEXT:
-      if (typeof data.text === 'string') {
-        return { action: 'text', text: data.text }
-      }
-      return null
-
-    case MESSAGE_TYPE.THINKING:
-      if (typeof data.text === 'string') {
-        return { action: 'thinking', text: data.text }
-      }
-      return null
-
-    case MESSAGE_TYPE.GUIDE:
-      if (typeof data.text === 'string') {
-        return { action: 'guide', text: data.text }
-      }
-      return null
-
-    case MESSAGE_TYPE.EXECUTING_TOOL:
-      return {
-        action: 'tool_execution',
-        segment: {
-          type: 'tool_execution',
-          data: {
-            type: 'EXECUTING_TOOL',
-            integratedToolType: data.integratedToolType || '',
-            toolFunction: data.toolFunction || '',
-            toolTitle: typeof data.title === 'string' ? data.title : undefined,
-            toolExplanation: typeof data.toolExplanation === 'string' ? data.toolExplanation : undefined,
-            parameters: data.parameters,
-            toolExecutionRequestId: typeof data.toolExecutionRequestId === 'string' ? data.toolExecutionRequestId : undefined,
-          }
-        }
-      }
-
-    case MESSAGE_TYPE.EXECUTED_TOOL:
-      return {
-        action: 'tool_execution',
-        segment: {
-          type: 'tool_execution',
-          data: {
-            type: 'EXECUTED_TOOL',
-            integratedToolType: data.integratedToolType || '',
-            toolFunction: data.toolFunction || '',
-            toolTitle: typeof data.title === 'string' ? data.title : undefined,
-            parameters: data.parameters,
-            result: data.result,
-            success: data.success,
-            toolExecutionRequestId: typeof data.toolExecutionRequestId === 'string' ? data.toolExecutionRequestId : undefined,
-          }
-        }
-      }
-      
-    case MESSAGE_TYPE.APPROVAL_REQUEST: {
-      const requestId = data.approvalRequestId || data.approval_request_id || ''
-      const approvalType = data.approvalType || 'USER'
-      const toolCalls = normalizeToolCalls(data.toolCalls)
-
-      if (toolCalls.length > 0) {
-        return {
-          action: 'approval_batch',
-          requestId,
-          approvalType,
-          toolCalls,
-        }
-      }
-
-      return {
-        action: 'approval_request',
-        requestId,
-        command: data.command || '',
-        explanation: data.explanation,
-        approvalType,
-      }
-    }
-      
-    case MESSAGE_TYPE.APPROVAL_RESULT: {
-      // The realtime NATS chunk carries the resolver's name as `displayName`; the
-      // persisted GraphQL message exposes the same value as `resolvedByName`. Accept
-      // either so realtime and history-replay render "by {name}" identically.
-      const resolvedByName =
-        typeof data.resolvedByName === 'string'
-          ? data.resolvedByName
-          : typeof data.displayName === 'string'
-            ? data.displayName
-            : undefined
-      return {
-        action: 'approval_result',
-        requestId: data.approvalRequestId || data.approval_request_id || '',
-        approved: data.approved === true,
-        approvalType: data.approvalType || 'CLIENT',
-        resolvedByName,
-      }
-    }
-      
-    case MESSAGE_TYPE.ERROR:
-      return {
-        action: 'error',
-        error: data.error || 'An error occurred',
-        details: data.details,
-      }
-      
-    case MESSAGE_TYPE.MESSAGE_REQUEST:
-      return {
-        action: 'message_request',
-        text: String(data.text || ''),
-        ownerType: typeof data.ownerType === 'string' ? data.ownerType : undefined,
-        displayName: typeof data.displayName === 'string' ? data.displayName : undefined,
-        userId: typeof data.userId === 'string' ? data.userId : undefined,
-        // Entity-context refs the user attached to this message (backend
-        // `MESSAGE_REQUEST` chunk → `contextItems: [{ type, id }]`). The wire
-        // shape carries no label; the host resolves display text + icon. `label`
-        // is REQUIRED on `ChatContextItem`, so fall back to the id (same as the
-        // historical path in `process-historical-messages.ts`) — keeps both
-        // message sources producing an identical shape.
-        contextItems: Array.isArray(data.contextItems)
-          ? (data.contextItems as Array<{ type?: unknown; id?: unknown }>)
-              .filter((it) => typeof it?.type === 'string' && typeof it?.id === 'string')
-              .map((it) => ({ type: it.type as string, id: it.id as string, label: it.id as string }))
-          : undefined,
-      }
-
-    case MESSAGE_TYPE.TOKEN_USAGE:
-      return {
-        action: 'token_usage',
-        data: {
-          inputTokensSize: data.inputTokensSize ?? 0,
-          outputTokensSize: data.outputTokensSize ?? 0,
-          totalTokensSize: data.totalTokensSize ?? 0,
-          contextSize: data.contextSize ?? 0,
-        },
-      }
-
-    case MESSAGE_TYPE.CONTEXT_COMPACTION_START:
-      return { action: 'context_compaction_start' }
-
-    case MESSAGE_TYPE.CONTEXT_COMPACTION_END:
-      return {
-        action: 'context_compaction_end',
-        summary: typeof data.text === 'string' ? data.text : undefined,
-      }
-
-    case MESSAGE_TYPE.SYSTEM:
-      if (typeof data.text === 'string') {
-        return { action: 'system', text: data.text }
-      }
-      return null
-
-    case MESSAGE_TYPE.DIRECT_MESSAGE:
-      if (typeof data.text === 'string') {
-        return {
-          action: 'direct_message',
-          text: data.text,
-          ownerType: typeof data.ownerType === 'string' ? data.ownerType : undefined,
-          displayName: typeof data.displayName === 'string' ? data.displayName : undefined,
-          userId: typeof data.userId === 'string' ? data.userId : undefined,
-        }
-      }
-      return null
-
-    case MESSAGE_TYPE.DIALOG_CLOSED:
-      return { action: 'dialog_closed' }
-
-    default:
-      return null
-  }
-}
-
-/**
- * Check if a chunk represents a control message (start/end)
- */
-export function isControlChunk(chunk: unknown): boolean {
-  if (!chunk || typeof chunk !== 'object') return false
-  const data = chunk as ChunkData
-  const type = String(data.type || '')
-  return type === MESSAGE_TYPE.MESSAGE_START || type === MESSAGE_TYPE.MESSAGE_END
-}
-
-/**
- * Check if a chunk represents an error
- */
-export function isErrorChunk(chunk: unknown): boolean {
-  if (!chunk || typeof chunk !== 'object') return false
-  const data = chunk as ChunkData
-  return data.type === MESSAGE_TYPE.ERROR
-}
-
-/**
- * Check if a chunk represents metadata
- */
-export function isMetadataChunk(chunk: unknown): boolean {
-  if (!chunk || typeof chunk !== 'object') return false
-  const data = chunk as ChunkData
-  return data.type === MESSAGE_TYPE.AI_METADATA
-}
-
-/**
- * Extract text content from a chunk if it's a text chunk
- */
-export function extractTextFromChunk(chunk: unknown): string | null {
-  if (!chunk || typeof chunk !== 'object') return null
-  const data = chunk as ChunkData
-  if (data.type === MESSAGE_TYPE.TEXT && typeof data.text === 'string') {
-    return data.text
-  }
-  return null
-}

package/src/components/ui/.simple-markdown-renderer.md

@@ -1,52 +0,0 @@
-<!-- source-hash: 9a47236bd9fd4f90eece958d3708b946 -->
-A client-side React component that renders Markdown content with GFM support, syntax highlighting, raw HTML passthrough, and a layered XSS sanitization pipeline designed for safe rendering of LLM-generated output.
-
-## Key Components
-
-### Rehype Plugins
-
-- **`rehypeStripUnsafe()`** — Custom HAST sanitizer that runs after `rehype-raw`. Strips dangerous elements (`<script>`, `<style>`, `<object>`, etc.), removes `on*` event handler attributes, and blocks `javascript:` / `data:` URI schemes on all URL-bearing attributes including multi-candidate `srcset`.
-
-- **`srcsetHasUnsafeCandidate(srcset)`** — Helper that splits `srcset` attribute values by comma and validates each URL candidate individually to prevent injection via secondary candidates.
-
-- **`escapeUnknownHtmlT...` (truncated)** — Pre-pass that escapes unrecognized HTML-like tags in raw markdown text before `rehype-raw` processes them, preventing React warnings/errors from LLM-emitted pseudo-XML tags like `<ticket>` or `<their>`.
-
-### URL Transform
-
-- **`cardAwareUrlTransform(url, key)`** — Extends `react-markdown`'s default URL sanitizer to allow `card://` and `mention://` internal schemes on `href` attributes only (used by `remarkCardLinks` and `remarkMentionChips` plugins), while keeping all other dangerous schemes blocked.
-
-### Constants
-
-- **`SAFE_HTML_TAGS`** — Allowlist of permitted HTML tag names forwarded to React. Explicitly excludes `<video>` (handled by a separate pipeline component).
-- **`TAG_LIKE_REGEX`** — ReDoS-safe regex (CodeQL-hardened, bounded quantifiers) for matching HTML-like tags in markdown source.
-
-## Usage Example
-
-```typescript
-import { SimpleMarkdownRenderer } from './simple-markdown-renderer';
-
-// Basic rendering with safe LLM output
-<SimpleMarkdownRenderer content="## Hello\n\nThis is **markdown** with `code`." />
-
-// With authenticated image sources and link resolution
-<SimpleMarkdownRenderer
-  content={llmMarkdownOutput}
-  resolveLink={(href) => ({ url: href, label: href })}
-/>
-```
-
-## Security Model
-
-The sanitization runs in two layers:
-
-```mermaid
-graph TD
-    A[Raw Markdown] --> B[escapeUnknownHtmlTags pre-pass]
-    B --> C[remark-gfm / remark-breaks]
-    C --> D[rehype-raw]
-    D --> E[rehypeStripUnsafe HAST sanitizer]
-    E --> F[rehype-highlight]
-    F --> G[Safe React Output]
-```
-
-> **Source:** [`simple-markdown-renderer.tsx`](https://github.com/flamingo-stack/openframe-oss-lib/blob/main/simple-markdown-renderer.tsx)