fix: 修复消息去重误判并统一空字符串去重语义

electron/main/worker/import/tempDb.test.ts

@@ -0,0 +1,25 @@
+import assert from 'node:assert/strict'
+import test from 'node:test'
+
+import { generateMessageKey } from './tempDb'
+
+test('空字符串内容在写库归一化前后应生成同一个去重 key', () => {
+  const timestamp = 1710000000
+  const senderPlatformId = 'user-1'
+  const parsedContent = ''
+
+  const keyBeforePersist = generateMessageKey(timestamp, senderPlatformId, parsedContent)
+  const keyAfterPersist = generateMessageKey(timestamp, senderPlatformId, parsedContent || null)
+
+  assert.equal(keyAfterPersist, keyBeforePersist)
+})
+
+test('同秒同发送者但等长不同内容应生成不同去重 key', () => {
+  const timestamp = 1710000000
+  const senderPlatformId = 'user-1'
+
+  const firstKey = generateMessageKey(timestamp, senderPlatformId, '你好啊')
+  const secondKey = generateMessageKey(timestamp, senderPlatformId, '再见呀')
+
+  assert.notEqual(firstKey, secondKey)
+})

electron/main/worker/import/tempDb.ts

@@ -4,6 +4,7 @@
  */
 
 import Database from 'better-sqlite3'
+import { createHash } from 'crypto'
 import * as fs from 'fs'
 import * as path from 'path'
 
@@ -84,9 +85,21 @@ export function cleanupTempDatabase(dbPath: string): void {
 
 /**
  * 生成消息去重键
- * 使用 timestamp + senderPlatformId + contentLength 作为去重标识
+ * 使用固定长度哈希作为去重标识。
+ * 直接用 content.length 会误判等长不同内容，直接用原文又会让 Set 长期持有大文本。
  */
 export function generateMessageKey(timestamp: number, senderPlatformId: string, content: string | null): string {
-  const contentLength = content ? content.length : 0
-  return `${timestamp}_${senderPlatformId}_${contentLength}`
+  // 去重 key 需要和当前写库语义一致，空字符串在存储层会被折叠成 null。
+  const normalizedContent = content || null
+  const hash = createHash('sha256')
+  hash.update(String(timestamp))
+  hash.update('\0')
+  hash.update(senderPlatformId)
+  hash.update('\0')
+  hash.update(normalizedContent === null ? 'null' : 'text')
+  hash.update('\0')
+  if (normalizedContent !== null) {
+    hash.update(normalizedContent)
+  }
+  return hash.digest('base64url')
 }