feat(insights): fix DBSCAN, Persian embeddings crash, D3 physics layouts, and D3 node not found runtime error

memento-note/lib/clip/analyze-clip.ts

@@ -0,0 +1,86 @@
+import { getSystemConfig } from '@/lib/config'
+import { getChatProvider } from '@/lib/ai/factory'
+
+export interface ClipAnalysis {
+  title: string
+  summary: string
+  tags: string[]
+  readingTimeMinutes: number
+}
+
+function parseAnalysisJson(raw: string): ClipAnalysis | null {
+  const trimmed = raw.trim()
+  const jsonMatch = trimmed.match(/\{[\s\S]*\}/)
+  if (!jsonMatch) return null
+  try {
+    const parsed = JSON.parse(jsonMatch[0]) as Partial<ClipAnalysis>
+    const tags = Array.isArray(parsed.tags)
+      ? parsed.tags.filter((t): t is string => typeof t === 'string').slice(0, 5)
+      : []
+    const readingTime = typeof parsed.readingTimeMinutes === 'number'
+      ? Math.max(1, Math.min(120, Math.round(parsed.readingTimeMinutes)))
+      : 5
+    return {
+      title: typeof parsed.title === 'string' && parsed.title.trim() ? parsed.title.trim().slice(0, 200) : 'Web clip',
+      summary: typeof parsed.summary === 'string' ? parsed.summary.trim().slice(0, 800) : '',
+      tags,
+      readingTimeMinutes: readingTime,
+    }
+  } catch {
+    return null
+  }
+}
+
+function estimateReadingMinutes(text: string): number {
+  const words = text.split(/\s+/).filter(Boolean).length
+  return Math.max(1, Math.round(words / 200))
+}
+
+export async function analyzeClipContent(params: {
+  url: string
+  title: string
+  textContent: string
+}): Promise<ClipAnalysis> {
+  const excerpt = params.textContent.slice(0, 6000)
+  const fallbackReading = estimateReadingMinutes(params.textContent)
+
+  try {
+    const config = await getSystemConfig()
+    const provider = getChatProvider(config)
+    const prompt = `You analyze web articles for a personal knowledge base. URL: ${params.url}
+Page title: ${params.title}
+
+Content excerpt:
+${excerpt}
+
+Respond with ONLY valid JSON (no markdown):
+{
+  "title": "concise improved title",
+  "summary": "max 3 sentences in the same language as the content",
+  "tags": ["tag1", "tag2"],
+  "readingTimeMinutes": ${fallbackReading}
+}
+
+Rules: tags max 5, short lowercase labels, summary factual.`
+
+    const raw = await provider.generateText(prompt)
+    const parsed = parseAnalysisJson(raw)
+    if (parsed) {
+      if (!parsed.title) parsed.title = params.title || 'Web clip'
+      if (!parsed.summary && params.textContent) {
+        parsed.summary = params.textContent.slice(0, 400)
+      }
+      if (parsed.tags.length === 0) parsed.tags = []
+      return parsed
+    }
+  } catch (error) {
+    console.error('[ClipAnalyze] AI failed:', error)
+  }
+
+  return {
+    title: params.title || 'Web clip',
+    summary: params.textContent.slice(0, 400),
+    tags: [],
+    readingTimeMinutes: fallbackReading,
+  }
+}