@goodandready/dsh-tts 0.4.7 → 0.4.9

package/lib/cache.js

@@ -19,6 +19,28 @@ export function createSpeechCache({ root, maxBytes }) {
     return ready
   }
 
+  // L1 In-Memory RAM Cache (max 50 entries or 10MB total buffer size)
+  const l1Map = new Map()
+  const L1_MAX_ITEMS = 50
+  const L1_MAX_BYTES = 10 * 1024 * 1024
+
+  function putL1(key, mime, audio) {
+    const buf = Buffer.isBuffer(audio) ? audio : Buffer.from(audio)
+    l1Map.delete(key) // reset insertion order
+    l1Map.set(key, { mime, audio: buf, at: Date.now(), size: buf.length })
+
+    let totalSize = 0
+    for (const item of l1Map.values()) totalSize += item.size
+
+    while (l1Map.size > L1_MAX_ITEMS || totalSize > L1_MAX_BYTES) {
+      const oldestKey = l1Map.keys().next().value
+      if (!oldestKey) break
+      const oldest = l1Map.get(oldestKey)
+      if (oldest) totalSize -= oldest.size
+      l1Map.delete(oldestKey)
+    }
+  }
+
   async function listEntries() {
     await ensure()
     const out = []
@@ -44,12 +66,19 @@ export function createSpeechCache({ root, maxBytes }) {
   async function evict() {
     const limit = typeof maxBytes === 'function' ? maxBytes() : Number(maxBytes)
     if (!(limit > 0)) return
-    const entries = (await listEntries()).sort((a, b) => a.at - b.at)
+    const rawEntries = await listEntries()
+    const entries = rawEntries.map((e) => {
+      const k = path.basename(e.full, AUDIO_EXT)
+      const l1 = l1Map.get(k)
+      return { ...e, key: k, at: (l1 && typeof l1.at === 'number') ? l1.at : e.at }
+    }).sort((a, b) => a.at - b.at)
+
     let total = entries.reduce((acc, e) => acc + e.size, 0)
     for (const e of entries) {
       if (total <= limit) break
       await fs.promises.rm(e.full, { force: true })
       await fs.promises.rm(path.join(dir, e.meta), { force: true }).catch(() => {})
+      l1Map.delete(e.key)
       total -= e.size
     }
   }
@@ -58,15 +87,28 @@ export function createSpeechCache({ root, maxBytes }) {
     async put(key, mime, audio) {
       await ensure()
       const buf = Buffer.isBuffer(audio) ? audio : Buffer.from(audio)
+      putL1(key, mime, buf)
       await fs.promises.writeFile(path.join(dir, key + AUDIO_EXT), buf)
       await fs.promises.writeFile(path.join(dir, key + '.json'), JSON.stringify({ mime, at: Date.now() }))
       await evict()
     },
 
     async get(key) {
+      // 1. L1 RAM Hit
+      if (l1Map.has(key)) {
+        const hit = l1Map.get(key)
+        const now = Date.now()
+        l1Map.delete(key) // refresh LRU order
+        l1Map.set(key, { ...hit, at: now })
+        fs.promises.writeFile(path.join(dir, key + '.json'), JSON.stringify({ mime: hit.mime, at: now })).catch(() => {})
+        return { mime: hit.mime || 'audio/mpeg', audio: hit.audio }
+      }
+
+      // 2. L2 Disk Hit
       try {
         const meta = JSON.parse(await fs.promises.readFile(path.join(dir, key + '.json'), 'utf8'))
         const audio = await fs.promises.readFile(path.join(dir, key + AUDIO_EXT))
+        putL1(key, meta.mime, audio)
         // Touch LRU asynchronously so reads are never blocked.
         fs.promises.writeFile(path.join(dir, key + '.json'), JSON.stringify({ mime: meta.mime, at: Date.now() })).catch(() => {})
         return { mime: meta.mime || 'audio/mpeg', audio }
@@ -76,6 +118,7 @@ export function createSpeechCache({ root, maxBytes }) {
     },
 
     async clear() {
+      l1Map.clear()
       const entries = await listEntries()
       for (const e of entries) {
         await fs.promises.rm(e.full, { force: true })

package/lib/client-src/20-player.js

@@ -248,7 +248,70 @@
       return () => window.removeEventListener('dsh-voice:speaking', onSpeaking)
     }
 
-    async function pollPending() {
+        // SSE Realtime Stream Connection with Exponential Backoff & Jitter
+    let sseAttempt = 0
+    let sseTimer = null
+    let sseSource = null
+
+    function connectSseStream() {
+      if (typeof window === 'undefined' || typeof window.EventSource === 'undefined') return
+      if (sseSource) {
+        try { sseSource.close() } catch {}
+        sseSource = null
+      }
+      try {
+        sseSource = new EventSource('/dsh-tts/stream')
+        sseSource.onopen = () => {
+          sseAttempt = 0
+        }
+        sseSource.addEventListener('utterance', (e) => {
+          sseAttempt = 0
+          try {
+            const item = JSON.parse(e.data)
+            if (item && item.id) {
+              player.after = item.id
+              if (!seenIds.has(item.id)) {
+                seenIds.add(item.id)
+                if (seenIds.size > 500) {
+                  const first = seenIds.values().next().value
+                  seenIds.delete(first)
+                }
+                if (item.text) recent.add(item.text)
+                player.queue.push(item)
+                drainQueue()
+              }
+            }
+          } catch {}
+        })
+        sseSource.addEventListener('chime', (e) => {
+          sseAttempt = 0
+          try {
+            const item = JSON.parse(e.data)
+            if (item) {
+              player.queue.push(item)
+              drainQueue()
+            }
+          } catch {}
+        })
+        sseSource.onerror = () => {
+          if (sseSource) {
+            try { sseSource.close() } catch {}
+            sseSource = null
+          }
+          sseAttempt++
+          const baseDelay = Math.min(10000, 500 * Math.pow(2, sseAttempt - 1))
+          const jitter = baseDelay * (Math.random() * 0.6 - 0.3)
+          const delay = Math.max(200, Math.floor(baseDelay + jitter))
+          if (sseTimer) clearTimeout(sseTimer)
+          sseTimer = setTimeout(() => {
+            connectSseStream()
+          }, delay)
+        }
+      } catch {}
+    }
+
+async function pollPending() {
+        connectSseStream()
       try {
         const st = await fetch('/dsh-tts/status', { cache: 'no-store' })
         if (!st.ok) return

package/lib/client-src/30-locales.js

@@ -361,6 +361,9 @@
       'enableItDictionary': '内置 IT 术语发音校正',
       'enableItDictionaryHint': '自动纠正 SQL、Nginx、K8s、Docker、API、JSON、YAML 等专有名词发音。',
       'loadItDictionary': '填入常用 IT 术语',
+      'exportJson': '📥 导出 JSON',
+      'importJson': '📤 导入 JSON',
+      'invalidJsonFormat': '无效的字典 JSON 格式',
       'previewRule': '试听发音',
       'messengerIntegrationTitle': '即时通讯联动 (Telegram, Discord)',
       'messengerIntegrationHint': '通过 @goodandready/dsh-messenger-gateway 向聊天群组发送语音回复。',

package/lib/client-src/40-editors.js

@@ -164,6 +164,8 @@
     function PronEditor(props) {
       const t = props.t || ((key) => key)
       const rows = Array.isArray(props.value) ? props.value : []
+      const fileInputRef = React.useRef(null)
+
       const change = (i, patch) => {
         props.onChange(rows.map((r, k) => (k === i ? Object.assign({}, r, patch) : r)))
       }
@@ -175,7 +177,60 @@
         props.onChange(rows.concat(toAdd))
       }
 
+      const exportJson = () => {
+        const data = JSON.stringify(rows, null, 2)
+        const blob = new Blob([data], { type: 'application/json' })
+        const url = URL.createObjectURL(blob)
+        const a = document.createElement('a')
+        a.href = url
+        a.download = 'pronunciation-dictionary.json'
+        document.body.appendChild(a)
+        a.click()
+        document.body.removeChild(a)
+        setTimeout(() => { try { URL.revokeObjectURL(url) } catch {} }, 1000)
+      }
+
+      const triggerImport = () => {
+        if (fileInputRef.current) fileInputRef.current.click()
+      }
+
+      const handleImportFile = (e) => {
+        const file = e.target.files && e.target.files[0]
+        if (!file) return
+        const reader = new FileReader()
+        reader.onload = (event) => {
+          try {
+            const parsed = JSON.parse(event.target.result)
+            if (!Array.isArray(parsed)) {
+              if (typeof window !== 'undefined' && window.alert) window.alert(t('invalidJsonFormat'))
+              return
+            }
+            const importedRules = parsed.map((item) => ({
+              from: String(item.from || ''),
+              to: String(item.to || ''),
+              whole: !!item.whole,
+              lang: String(item.lang || ''),
+            })).filter((r) => r.from || r.to)
+
+            const existingKeys = new Set(rows.map((r) => String(r.from).toLowerCase()))
+            const newItems = importedRules.filter((r) => !existingKeys.has(String(r.from).toLowerCase()))
+            props.onChange(rows.concat(newItems))
+          } catch (err) {
+            if (typeof window !== 'undefined' && window.alert) window.alert(t('invalidJsonFormat'))
+          }
+          e.target.value = ''
+        }
+        reader.readAsText(file)
+      }
+
       return React.createElement('div', { className: 'dts-wrap-in' },
+        React.createElement('input', {
+          type: 'file',
+          ref: fileInputRef,
+          accept: '.json',
+          style: { display: 'none' },
+          onChange: handleImportFile,
+        }),
         rows.map((rule, i) => React.createElement('div', { className: 'dts-row', key: i },
           React.createElement('input', {
             className: 'dts-input dts-grow', placeholder: t('pronFrom'),
@@ -209,6 +264,8 @@
         React.createElement('div', { className: 'dts-row' },
           React.createElement('button', { type: 'button', className: 'dts-mini', disabled: !props.writable, onClick: add }, '+'),
           React.createElement('button', { type: 'button', className: 'dts-link', disabled: !props.writable, onClick: loadIt }, t('loadItDictionary')),
+          React.createElement('button', { type: 'button', className: 'dts-link', disabled: !props.writable, onClick: exportJson }, t('exportJson')),
+          React.createElement('button', { type: 'button', className: 'dts-link', disabled: !props.writable, onClick: triggerImport }, t('importJson')),
           React.createElement('span', { className: 'dts-sub' }, t('pronHint')),
         ),
       )

package/lib/client.js

@@ -387,7 +387,70 @@ window.__ModuleLoader__.load({
       return () => window.removeEventListener('dsh-voice:speaking', onSpeaking)
     }
 
-    async function pollPending() {
+        // SSE Realtime Stream Connection with Exponential Backoff & Jitter
+    let sseAttempt = 0
+    let sseTimer = null
+    let sseSource = null
+
+    function connectSseStream() {
+      if (typeof window === 'undefined' || typeof window.EventSource === 'undefined') return
+      if (sseSource) {
+        try { sseSource.close() } catch {}
+        sseSource = null
+      }
+      try {
+        sseSource = new EventSource('/dsh-tts/stream')
+        sseSource.onopen = () => {
+          sseAttempt = 0
+        }
+        sseSource.addEventListener('utterance', (e) => {
+          sseAttempt = 0
+          try {
+            const item = JSON.parse(e.data)
+            if (item && item.id) {
+              player.after = item.id
+              if (!seenIds.has(item.id)) {
+                seenIds.add(item.id)
+                if (seenIds.size > 500) {
+                  const first = seenIds.values().next().value
+                  seenIds.delete(first)
+                }
+                if (item.text) recent.add(item.text)
+                player.queue.push(item)
+                drainQueue()
+              }
+            }
+          } catch {}
+        })
+        sseSource.addEventListener('chime', (e) => {
+          sseAttempt = 0
+          try {
+            const item = JSON.parse(e.data)
+            if (item) {
+              player.queue.push(item)
+              drainQueue()
+            }
+          } catch {}
+        })
+        sseSource.onerror = () => {
+          if (sseSource) {
+            try { sseSource.close() } catch {}
+            sseSource = null
+          }
+          sseAttempt++
+          const baseDelay = Math.min(10000, 500 * Math.pow(2, sseAttempt - 1))
+          const jitter = baseDelay * (Math.random() * 0.6 - 0.3)
+          const delay = Math.max(200, Math.floor(baseDelay + jitter))
+          if (sseTimer) clearTimeout(sseTimer)
+          sseTimer = setTimeout(() => {
+            connectSseStream()
+          }, delay)
+        }
+      } catch {}
+    }
+
+async function pollPending() {
+        connectSseStream()
       try {
         const st = await fetch('/dsh-tts/status', { cache: 'no-store' })
         if (!st.ok) return
@@ -861,6 +924,9 @@ window.__ModuleLoader__.load({
       'enableItDictionary': '内置 IT 术语发音校正',
       'enableItDictionaryHint': '自动纠正 SQL、Nginx、K8s、Docker、API、JSON、YAML 等专有名词发音。',
       'loadItDictionary': '填入常用 IT 术语',
+      'exportJson': '📥 导出 JSON',
+      'importJson': '📤 导入 JSON',
+      'invalidJsonFormat': '无效的字典 JSON 格式',
       'previewRule': '试听发音',
       'messengerIntegrationTitle': '即时通讯联动 (Telegram, Discord)',
       'messengerIntegrationHint': '通过 @goodandready/dsh-messenger-gateway 向聊天群组发送语音回复。',
@@ -1035,6 +1101,8 @@ window.__ModuleLoader__.load({
     function PronEditor(props) {
       const t = props.t || ((key) => key)
       const rows = Array.isArray(props.value) ? props.value : []
+      const fileInputRef = React.useRef(null)
+
       const change = (i, patch) => {
         props.onChange(rows.map((r, k) => (k === i ? Object.assign({}, r, patch) : r)))
       }
@@ -1046,7 +1114,60 @@ window.__ModuleLoader__.load({
         props.onChange(rows.concat(toAdd))
       }
 
+      const exportJson = () => {
+        const data = JSON.stringify(rows, null, 2)
+        const blob = new Blob([data], { type: 'application/json' })
+        const url = URL.createObjectURL(blob)
+        const a = document.createElement('a')
+        a.href = url
+        a.download = 'pronunciation-dictionary.json'
+        document.body.appendChild(a)
+        a.click()
+        document.body.removeChild(a)
+        setTimeout(() => { try { URL.revokeObjectURL(url) } catch {} }, 1000)
+      }
+
+      const triggerImport = () => {
+        if (fileInputRef.current) fileInputRef.current.click()
+      }
+
+      const handleImportFile = (e) => {
+        const file = e.target.files && e.target.files[0]
+        if (!file) return
+        const reader = new FileReader()
+        reader.onload = (event) => {
+          try {
+            const parsed = JSON.parse(event.target.result)
+            if (!Array.isArray(parsed)) {
+              if (typeof window !== 'undefined' && window.alert) window.alert(t('invalidJsonFormat'))
+              return
+            }
+            const importedRules = parsed.map((item) => ({
+              from: String(item.from || ''),
+              to: String(item.to || ''),
+              whole: !!item.whole,
+              lang: String(item.lang || ''),
+            })).filter((r) => r.from || r.to)
+
+            const existingKeys = new Set(rows.map((r) => String(r.from).toLowerCase()))
+            const newItems = importedRules.filter((r) => !existingKeys.has(String(r.from).toLowerCase()))
+            props.onChange(rows.concat(newItems))
+          } catch (err) {
+            if (typeof window !== 'undefined' && window.alert) window.alert(t('invalidJsonFormat'))
+          }
+          e.target.value = ''
+        }
+        reader.readAsText(file)
+      }
+
       return React.createElement('div', { className: 'dts-wrap-in' },
+        React.createElement('input', {
+          type: 'file',
+          ref: fileInputRef,
+          accept: '.json',
+          style: { display: 'none' },
+          onChange: handleImportFile,
+        }),
         rows.map((rule, i) => React.createElement('div', { className: 'dts-row', key: i },
           React.createElement('input', {
             className: 'dts-input dts-grow', placeholder: t('pronFrom'),
@@ -1080,6 +1201,8 @@ window.__ModuleLoader__.load({
         React.createElement('div', { className: 'dts-row' },
           React.createElement('button', { type: 'button', className: 'dts-mini', disabled: !props.writable, onClick: add }, '+'),
           React.createElement('button', { type: 'button', className: 'dts-link', disabled: !props.writable, onClick: loadIt }, t('loadItDictionary')),
+          React.createElement('button', { type: 'button', className: 'dts-link', disabled: !props.writable, onClick: exportJson }, t('exportJson')),
+          React.createElement('button', { type: 'button', className: 'dts-link', disabled: !props.writable, onClick: triggerImport }, t('importJson')),
           React.createElement('span', { className: 'dts-sub' }, t('pronHint')),
         ),
       )

package/lib/index.js

@@ -550,28 +550,11 @@ export function apply(ctx, config) {
     const raw = cur && Array.isArray(cur.parts) ? cur.parts.join('\n') : ''
     const text = cleanText(raw, cfg)
     if (!text) return
-    const id = `u${++utteranceSeq}`
-    const controller = new AbortController()
-    const timer = setTimeout(() => controller.abort(), cfg.timeoutMs)
-    synthesize(text, cfg, controller.signal, (cur && cur.role) || 'reply')
-      .then((out) => {
-        enqueue({
-          id,
-          sessionId: sid,
-          provider: out.provider,
-          mime: out.mime,
-          audioBase64: Buffer.from(out.audio).toString('base64'),
-          tookMs: out.tookMs,
-        })
-      })
-      .catch((err) => {
-        enqueue({
-          id,
-          sessionId: sid,
-          error: String(err && err.message || err),
-        })
-      })
-      .finally(() => clearTimeout(timer))
+    const pieces = splitSentences(text, cfg.sentenceChars)
+    const activeRole = (cur && cur.role) || 'reply'
+    for (const piece of pieces) {
+      speakPiece(sid, piece, cfg, activeRole)
+    }
   }), 'dsh-tts: speak finished replies')
 
 

package/lib/routes.js

@@ -226,7 +226,10 @@ export function registerHttpRoutes(ctx, api) {
         return
       }
       const cfg = live()
-      const phrase = String(cfg.language || '').toLowerCase().startsWith('ru') ? 'Проверка голоса.' : 'Voice check.'
+      const lang = String(cfg.language || '').toLowerCase()
+      let phrase = 'Voice check.'
+      if (lang.startsWith('zh')) phrase = '语音合成测试。'
+      else if (lang.startsWith('ru')) phrase = 'Проверка голоса.'
       try {
         const probe = Object.assign({}, cfg, {
           chain: [{ provider: body.provider || '', model: body.model || '', voice: body.voice || '' }],

package/lib/text.js

@@ -170,7 +170,7 @@ export function protectAbbreviations(text) {
     const rx = new RegExp('(?<=^|[\\s"«(])' + escaped + '(?=[\\s"»)\\]]|$)', 'gi')
     res = res.replace(rx, (m) => m.replace(/\./g, '\u2024'))
   }
-  // Decimal numbers and versions: e.g. 3.14, v0.4.6, 192.168.1.111
+  // Decimal numbers and versions: e.g. 3.14, v0.4.6, 192.168.0.1
   res = res.replace(/(?<=[vV]?\d+)\.(?=\d+)/g, '\u2024')
   // List numbering at line/word starts: 1. item, 2. item
   res = res.replace(/(?<=^|[\s\n])(\d+)\.(?=\s+[A-Za-zА-Яа-я\u4E00-\u9FFF])/g, '$1\u2024')

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@goodandready/dsh-tts",
-  "version": "0.4.7",
+  "version": "0.4.9",
   "description": "Text-to-speech for DeepSeek Harness: speak agent replies in the Web UI with a provider fallback chain (OpenAI, ElevenLabs, Google, Azure, Groq, Deepgram, OpenRouter, Edge, Piper, eSpeak).",
   "license": "MIT",
   "type": "module",