feat: 口头禅分析

2026-04-23 18:19:01 +08:00 · 2025-11-27 22:17:06 +08:00
parent 00eea87a11
commit 8e40b73836
6 changed files with 267 additions and 17 deletions
--- a/electron/main/database/index.ts
+++ b/electron/main/database/index.ts
@@ -20,6 +20,8 @@ import type {
  RepeatRateItem,
  ChainLengthDistribution,
  HotRepeatContent,
+  CatchphraseAnalysis,
+  MemberCatchphrase,
 } from '../../../src/types/chat'

 // 数据库存储目录
@@ -774,6 +776,7 @@ export function getRepeatAnalysis(sessionId: string, filter?: TimeFilter): Repea
          msg.id,
          msg.sender_id as senderId,
          msg.content,
+          msg.ts,
          m.platform_id as platformId,
          m.name
        FROM message msg
@@ -786,6 +789,7 @@ export function getRepeatAnalysis(sessionId: string, filter?: TimeFilter): Repea
      id: number
      senderId: number
      content: string
+      ts: number
      platformId: string
      name: string
    }>
@@ -802,17 +806,23 @@ export function getRepeatAnalysis(sessionId: string, filter?: TimeFilter): Repea
    // 复读链长度统计
    const chainLengthCount = new Map<number, number>() // length -> count

-    // 热门复读内容统计（记录最长链的原创者）
-    const contentStats = new Map<string, { count: number; maxChainLength: number; originatorId: number }>()
+    // 热门复读内容统计（记录最长链的原创者和最近时间戳）
+    const contentStats = new Map<
+      string,
+      { count: number; maxChainLength: number; originatorId: number; lastTs: number }
+    >()

    // 滑动窗口算法
    let currentContent: string | null = null
-    let repeatChain: Array<{ senderId: number; content: string }> = []
+    let repeatChain: Array<{ senderId: number; content: string; ts: number }> = []
    let totalRepeatChains = 0
    let totalChainLength = 0 // 用于计算平均长度

    // 处理复读链的辅助函数（至少3人参与才算复读）
-    const processRepeatChain = (chain: Array<{ senderId: number; content: string }>, breakerId?: number) => {
+    const processRepeatChain = (
+      chain: Array<{ senderId: number; content: string; ts: number }>,
+      breakerId?: number
+    ) => {
      if (chain.length < 3) return

      totalRepeatChains++
@@ -835,18 +845,21 @@ export function getRepeatAnalysis(sessionId: string, filter?: TimeFilter): Repea
      // 复读链长度统计
      chainLengthCount.set(chainLength, (chainLengthCount.get(chainLength) || 0) + 1)

-      // 热门复读内容统计（记录最长链的原创者）
+      // 热门复读内容统计（记录最长链的原创者和时间戳）
      const content = chain[0].content
+      const chainTs = chain[0].ts // 复读链的时间戳（原创者发消息的时间）
      const existing = contentStats.get(content)
      if (existing) {
        existing.count++
+        // 更新最近时间戳
+        existing.lastTs = Math.max(existing.lastTs, chainTs)
        // 如果当前链更长，更新最长链信息和原创者
        if (chainLength > existing.maxChainLength) {
          existing.maxChainLength = chainLength
          existing.originatorId = originatorId
        }
      } else {
-        contentStats.set(content, { count: 1, maxChainLength: chainLength, originatorId })
+        contentStats.set(content, { count: 1, maxChainLength: chainLength, originatorId, lastTs: chainTs })
      }
    }

@@ -866,7 +879,7 @@ export function getRepeatAnalysis(sessionId: string, filter?: TimeFilter): Repea
        const lastSender = repeatChain[repeatChain.length - 1]?.senderId
        if (lastSender !== msg.senderId) {
          // 不同人发的相同内容，延续复读链
-          repeatChain.push({ senderId: msg.senderId, content })
+          repeatChain.push({ senderId: msg.senderId, content, ts: msg.ts })
        }
        // 同一人连续发相同内容，忽略（不算复读）
      } else {
@@ -875,7 +888,7 @@ export function getRepeatAnalysis(sessionId: string, filter?: TimeFilter): Repea

        // 开始新链
        currentContent = content
-        repeatChain = [{ senderId: msg.senderId, content }]
+        repeatChain = [{ senderId: msg.senderId, content, ts: msg.ts }]
      }
    }

@@ -937,6 +950,7 @@ export function getRepeatAnalysis(sessionId: string, filter?: TimeFilter): Repea
        count: stats.count,
        maxChainLength: stats.maxChainLength,
        originatorName: originatorInfo?.name || '未知',
+        lastTs: stats.lastTs,
      })
    }
    // 按最长复读链长度降序排序
@@ -959,3 +973,98 @@ export function getRepeatAnalysis(sessionId: string, filter?: TimeFilter): Repea
    db.close()
  }
 }
+
+/**
+ * 获取口头禅分析数据
+ * 统计每个成员最常说的内容（前5个）
+ * - 排除：系统消息、空消息、图片消息
+ * - 排除：过短的内容（少于2个字符）
+ */
+export function getCatchphraseAnalysis(sessionId: string, filter?: TimeFilter): CatchphraseAnalysis {
+  const db = openDatabase(sessionId)
+  if (!db) {
+    return { members: [] }
+  }
+
+  try {
+    const { clause, params } = buildTimeFilter(filter)
+
+    // 构建查询条件：排除系统消息、空消息、图片，且内容长度 >= 2
+    let whereClause = clause
+    if (whereClause.includes('WHERE')) {
+      whereClause +=
+        " AND m.name != '系统消息' AND msg.type = 0 AND msg.content IS NOT NULL AND LENGTH(TRIM(msg.content)) >= 2"
+    } else {
+      whereClause =
+        " WHERE m.name != '系统消息' AND msg.type = 0 AND msg.content IS NOT NULL AND LENGTH(TRIM(msg.content)) >= 2"
+    }
+
+    // 获取每个成员的发言内容及出现次数
+    const rows = db
+      .prepare(
+        `
+        SELECT
+          m.id as memberId,
+          m.platform_id as platformId,
+          m.name,
+          TRIM(msg.content) as content,
+          COUNT(*) as count
+        FROM message msg
+        JOIN member m ON msg.sender_id = m.id
+        ${whereClause}
+        GROUP BY m.id, TRIM(msg.content)
+        ORDER BY m.id, count DESC
+      `
+      )
+      .all(...params) as Array<{
+      memberId: number
+      platformId: string
+      name: string
+      content: string
+      count: number
+    }>
+
+    // 按成员分组，取每个成员的前3个口头禅
+    const memberMap = new Map<
+      number,
+      {
+        memberId: number
+        platformId: string
+        name: string
+        catchphrases: Array<{ content: string; count: number }>
+      }
+    >()
+
+    for (const row of rows) {
+      if (!memberMap.has(row.memberId)) {
+        memberMap.set(row.memberId, {
+          memberId: row.memberId,
+          platformId: row.platformId,
+          name: row.name,
+          catchphrases: [],
+        })
+      }
+
+      const member = memberMap.get(row.memberId)!
+      // 只保留前5个
+      if (member.catchphrases.length < 5) {
+        member.catchphrases.push({
+          content: row.content,
+          count: row.count,
+        })
+      }
+    }
+
+    // 按发言总次数排序（口头禅出现次数总和）
+    const members = Array.from(memberMap.values())
+    members.sort((a, b) => {
+      const aTotal = a.catchphrases.reduce((sum, c) => sum + c.count, 0)
+      const bTotal = b.catchphrases.reduce((sum, c) => sum + c.count, 0)
+      return bTotal - aTotal
+    })
+
+    return { members }
+  } finally {
+    db.close()
+  }
+}
--- a/electron/main/ipcMain.ts
+++ b/electron/main/ipcMain.ts
@@ -375,6 +375,21 @@ const mainIpcMain = (win: BrowserWindow) => {
      }
    }
  )
+
+  /**
+   * 获取口头禅分析数据
+   */
+  ipcMain.handle(
+    'chat:getCatchphraseAnalysis',
+    async (_, sessionId: string, filter?: { startTs?: number; endTs?: number }) => {
+      try {
+        return database.getCatchphraseAnalysis(sessionId, filter)
+      } catch (error) {
+        console.error('获取口头禅分析失败：', error)
+        return { members: [] }
+      }
+    }
+  )
 }

 export default mainIpcMain
--- a/electron/preload/index.d.ts
+++ b/electron/preload/index.d.ts
@@ -8,6 +8,7 @@ import type {
  MessageType,
  ImportProgress,
  RepeatAnalysis,
+  CatchphraseAnalysis,
 } from '../../src/types/chat'

 interface TimeFilter {
@@ -35,6 +36,7 @@ interface ChatApi {
  getSupportedFormats: () => Promise<Array<{ name: string; platform: string }>>
  onImportProgress: (callback: (progress: ImportProgress) => void) => () => void
  getRepeatAnalysis: (sessionId: string, filter?: TimeFilter) => Promise<RepeatAnalysis>
+  getCatchphraseAnalysis: (sessionId: string, filter?: TimeFilter) => Promise<CatchphraseAnalysis>
 }

 interface Api {
--- a/electron/preload/index.ts
+++ b/electron/preload/index.ts
@@ -9,6 +9,7 @@ import type {
  MessageType,
  ImportProgress,
  RepeatAnalysis,
+  CatchphraseAnalysis,
 } from '../../src/types/chat'

 // Custom APIs for renderer
@@ -160,6 +161,16 @@ const chatApi = {
  getRepeatAnalysis: (sessionId: string, filter?: { startTs?: number; endTs?: number }): Promise<RepeatAnalysis> => {
    return ipcRenderer.invoke('chat:getRepeatAnalysis', sessionId, filter)
  },
+
+  /**
+   * 获取口头禅分析数据
+   */
+  getCatchphraseAnalysis: (
+    sessionId: string,
+    filter?: { startTs?: number; endTs?: number }
+  ): Promise<CatchphraseAnalysis> => {
+    return ipcRenderer.invoke('chat:getCatchphraseAnalysis', sessionId, filter)
+  },
 }

 // Use `contextBridge` APIs to expose Electron APIs to
--- a/src/components/analysis/MembersTab.vue
+++ b/src/components/analysis/MembersTab.vue
@@ -1,6 +1,6 @@
 <script setup lang="ts">
 import { computed, ref, watch } from 'vue'
-import type { MemberActivity, MemberNameHistory, RepeatAnalysis } from '@/types/chat'
+import type { MemberActivity, MemberNameHistory, RepeatAnalysis, CatchphraseAnalysis } from '@/types/chat'
 import { MemberRankList, BarChart } from '@/components/charts'
 import type { MemberRankItem, BarChartData } from '@/components/charts'

@@ -25,7 +25,8 @@ const repeatRankMode = ref<'count' | 'rate'>('rate')
 // 转换复读数据为排行榜格式（绝对次数）
 const originatorRankData = computed<MemberRankItem[]>(() => {
  if (!repeatAnalysis.value) return []
-  const data = repeatRankMode.value === 'count' ? repeatAnalysis.value.originators : repeatAnalysis.value.originatorRates
+  const data =
+    repeatRankMode.value === 'count' ? repeatAnalysis.value.originators : repeatAnalysis.value.originatorRates
  return data.slice(0, 10).map((m) => ({
    id: m.memberId.toString(),
    name: m.name,
@@ -86,6 +87,33 @@ function truncateContent(content: string, maxLength = 30): string {
  return content.slice(0, maxLength) + '...'
 }

+// 格式化日期
+function formatDate(ts: number): string {
+  const date = new Date(ts * 1000)
+  const year = date.getFullYear()
+  const month = String(date.getMonth() + 1).padStart(2, '0')
+  const day = String(date.getDate()).padStart(2, '0')
+  return `${year}-${month}-${day}`
+}
+
+// ==================== 口头禅分析 ====================
+const catchphraseAnalysis = ref<CatchphraseAnalysis | null>(null)
+const isLoadingCatchphrase = ref(false)
+
+// 加载口头禅分析数据
+async function loadCatchphraseAnalysis() {
+  if (!props.sessionId) return
+
+  isLoadingCatchphrase.value = true
+  try {
+    catchphraseAnalysis.value = await window.chatApi.getCatchphraseAnalysis(props.sessionId, props.timeFilter)
+  } catch (error) {
+    console.error('加载口头禅分析失败:', error)
+  } finally {
+    isLoadingCatchphrase.value = false
+  }
+}
+
 // Top 10 排行榜数据
 const top10RankData = computed<MemberRankItem[]>(() => {
  return props.memberActivity.slice(0, 10).map((m) => ({
@@ -162,11 +190,12 @@ watch(
  { immediate: true }
 )

-// 监听 sessionId 和 timeFilter 变化，重新加载复读分析
+// 监听 sessionId 和 timeFilter 变化，重新加载复读分析和口头禅分析
 watch(
  () => [props.sessionId, props.timeFilter],
  () => {
    loadRepeatAnalysis()
+    loadCatchphraseAnalysis()
  },
  { immediate: true, deep: true }
 )
@@ -327,11 +356,7 @@ function formatPeriod(startTs: number, endTs: number | null): string {
              <p class="mt-0.5 text-xs text-gray-500 dark:text-gray-400">每次复读有多少人参与</p>
            </div>
            <div class="p-4">
-              <BarChart
-                v-if="chainLengthChartData.labels.length > 0"
-                :data="chainLengthChartData"
-                :height="200"
-              />
+              <BarChart v-if="chainLengthChartData.labels.length > 0" :data="chainLengthChartData" :height="200" />
              <div v-else class="py-6 text-center text-sm text-gray-400">暂无数据</div>
            </div>
          </div>
@@ -369,7 +394,11 @@ function formatPeriod(startTs: number, endTs: number | null): string {
                    {{ truncateContent(item.content) }}
                  </span>
                </div>
-                <span class="shrink-0 text-xs text-gray-500">{{ item.count }} 次</span>
+                <div class="flex shrink-0 items-center gap-2 text-xs text-gray-500">
+                  <span>{{ item.count }} 次</span>
+                  <span class="text-gray-300 dark:text-gray-600">|</span>
+                  <span>{{ formatDate(item.lastTs) }}</span>
+                </div>
              </div>
            </div>
            <div v-else class="px-4 py-6 text-center text-sm text-gray-400">暂无数据</div>
@@ -430,5 +459,68 @@ function formatPeriod(startTs: number, endTs: number | null): string {

      <div v-else class="px-5 py-8 text-center text-sm text-gray-400">该群组暂无复读记录</div>
    </div>
+
+    <!-- 口头禅分析模块 -->
+    <div class="rounded-xl border border-gray-200 bg-white shadow-sm dark:border-gray-800 dark:bg-gray-900">
+      <div class="border-b border-gray-200 px-5 py-4 dark:border-gray-800">
+        <h3 class="font-semibold text-gray-900 dark:text-white">💬 口头禅分析</h3>
+        <p class="mt-1 text-sm text-gray-500 dark:text-gray-400">
+          {{
+            isLoadingCatchphrase
+              ? '加载中...'
+              : catchphraseAnalysis
+                ? `分析了 ${catchphraseAnalysis.members.length} 位成员的高频发言`
+                : '暂无数据'
+          }}
+        </p>
+      </div>
+
+      <div v-if="isLoadingCatchphrase" class="px-5 py-8 text-center text-sm text-gray-400">正在分析口头禅数据...</div>
+
+      <div
+        v-else-if="catchphraseAnalysis && catchphraseAnalysis.members.length > 0"
+        class="divide-y divide-gray-100 dark:divide-gray-800"
+      >
+        <div
+          v-for="member in catchphraseAnalysis.members.slice(0, 20)"
+          :key="member.memberId"
+          class="flex items-start gap-4 px-5 py-4"
+        >
+          <!-- 成员名称 -->
+          <div class="w-28 shrink-0 pt-1 font-medium text-gray-900 dark:text-white">
+            {{ member.name }}
+          </div>
+
+          <!-- 口头禅列表 -->
+          <div class="flex flex-1 flex-wrap items-center gap-2">
+            <div
+              v-for="(phrase, index) in member.catchphrases"
+              :key="index"
+              class="flex items-center gap-1.5 rounded-lg px-3 py-1.5"
+              :class="
+                index === 0
+                  ? 'bg-amber-50 dark:bg-amber-900/20'
+                  : index === 1
+                    ? 'bg-gray-100 dark:bg-gray-800'
+                    : 'bg-gray-50 dark:bg-gray-800/50'
+              "
+            >
+              <span
+                class="text-sm"
+                :class="
+                  index === 0 ? 'font-medium text-amber-700 dark:text-amber-400' : 'text-gray-700 dark:text-gray-300'
+                "
+                :title="phrase.content"
+              >
+                {{ truncateContent(phrase.content, 20) }}
+              </span>
+              <span class="text-xs text-gray-400">{{ phrase.count }}次</span>
+            </div>
+          </div>
+        </div>
+      </div>
+
+      <div v-else class="px-5 py-8 text-center text-sm text-gray-400">暂无口头禅数据</div>
+    </div>
  </div>
 </template>
--- a/src/types/chat.ts
+++ b/src/types/chat.ts
@@ -217,6 +217,27 @@ export interface HotRepeatContent {
  count: number // 被复读次数
  maxChainLength: number // 最长复读链长度
  originatorName: string // 最长链的原创者名称
+  lastTs: number // 最近一次发生的时间戳（秒）
+}
+
+/**
+ * 成员口头禅项
+ */
+export interface MemberCatchphrase {
+  memberId: number
+  platformId: string
+  name: string
+  catchphrases: Array<{
+    content: string
+    count: number
+  }>
+}
+
+/**
+ * 口头禅分析结果
+ */
+export interface CatchphraseAnalysis {
+  members: MemberCatchphrase[]
 }

 /**