add free edge tts engine

application/lib/calibre/ebooks/conversion/plugins/recipe_input.py

@@ -80,18 +80,19 @@ def convert(self, recipes, opts, file_ext, log, output_dir, fs):
         self.aborted_articles = []
         self.failed_downloads = []
         for recipe in recipes:
+            indexFile = None
             try:
                 ro = recipe(opts, log, output_dir, fs, feed_index_start=feed_index_start)
                 if recipeNum > 1: #只有单独推送才使用recipe的封面或报头
                     ro.cover_url = None
                     ro.masthead_url = None
-                ro.download()
+                indexFile = ro.download()
             except Exception as e:
                 msg = 'Failed to execute recipe "{}": {}'.format(recipe.title, e)
                 log.warning(msg)
                 continue
 
-            if ro.feed_objects:
+            if indexFile and ro.feed_objects:
                 feed_index_start += len(ro.feed_objects)
                 self.feeds.extend(ro.feed_objects)
                 self.aborted_articles.extend(ro.aborted_articles)

application/lib/ebook_tts/engines/__init__.py

@@ -1,10 +1,11 @@
 #!/usr/bin/env python3
 # -*- coding:utf-8 -*-
 from .tts_base import TTSBase
-from .azure import AzureTTS
+from .azure import AzureTTS, EdgeTTSFree
 from .google import GoogleWebTTSFree, GoogleTextToSpeech
 builtin_tts_engines = {
     GoogleWebTTSFree.name: GoogleWebTTSFree,
     GoogleTextToSpeech.name: GoogleTextToSpeech,
+    EdgeTTSFree.name: EdgeTTSFree,
     AzureTTS.name: AzureTTS,
 }

application/lib/ebook_tts/engines/azure.py

@@ -6,6 +6,10 @@
 from urllib.parse import urljoin
 from urlopener import UrlOpener
 from .tts_base import TTSBase
+try:
+    import edge_tts
+except:
+    edge_tts = None
 
 #键为BCP-47语种代码，值为语音名字列表
 azuretts_languages = {
@@ -203,7 +207,7 @@ class AzureTTS(TTSBase):
     default_timeout = 60
     #https://learn.microsoft.com/en-us/azure/ai-services/speech-service/speech-services-quotas-and-limits
     request_interval = 3  #20 transactions per 60 seconds
-    #每段音频不能超过10分钟，所以对于中文，大约2000字，因为大约1500 word
+    #每段音频不能超过10分钟，对于中文，大约2000字，因为大约1500 word
     max_len_per_request = 1000
     languages = azuretts_languages
     regions = azure_regions
@@ -233,7 +237,7 @@ def voice_list(self):
         else:
             return {'status': UrlOpener.CodeMap(resp.status_code)}
 
-    #文本转换为语音，
+    #文本转换为语音，返回(mime, audio)
     #支持的音频格式参考： 
     #<https://learn.microsoft.com/en-us/azure/ai-services/speech-service/rest-text-to-speech?tabs=streaming#audio-outputs>
     def tts(self, text):
@@ -251,4 +255,65 @@ def tts(self, text):
             raise Exception(self.opener.CodeMap(resp.status_code))
 
 
+class EdgeTTSFree(TTSBase):
+    name = 'EdgeTTS(Free)'
+    alias = 'Microsoft Edge TTS (Free)'
+    need_api_key = False
+    api_key_hint = ''
+    default_api_host = ''
+    default_timeout = 60
+    #https://learn.microsoft.com/en-us/azure/ai-services/speech-service/speech-services-quotas-and-limits
+    request_interval = 10  #20 transactions per 60 seconds
+    #每段音频不能超过10分钟，对于中文，大约2000字，因为大约1500 word
+    max_len_per_request = 1000
+    languages = azuretts_languages
+    regions = {}
+    engine_url = ''
+    region_url = ''
+    voice_url = ''
+    language_url = ''
+
+    def __init__(self, params):
+        import asyncio
+        super().__init__(params)
+        try:
+            self.eventLoop = asyncio.get_event_loop()
+        except Exception as e:
+            if str(e).startswith('There is no current event loop in thread'):
+                self.eventLoop = asyncio.new_event_loop()
+                asyncio.set_event_loop(self.eventLoop)
+            else:
+                default_log.warning(f'asyncio.get_event_loop failed: {e}')
+
+    #获取支持的语音列表
+    #或者可以直接到网页去查询
+    #https://learn.microsoft.com/en-us/azure/ai-services/speech-service/language-support?tabs=tts
+    def voice_list(self):
+        return edge_tts.list_voices()
+
+    #文本转换为语音，返回(mime, audio)
+    #支持的音频格式参考： 
+    #<https://learn.microsoft.com/en-us/azure/ai-services/speech-service/rest-text-to-speech?tabs=streaming#audio-outputs>
+    def tts(self, text):
+        if not self.eventLoop:
+            raise Exception('Cannot create event_loop')
+
+        #将异步函数包装为同步调用
+        return self.eventLoop.run_until_complete(self.async_get_tts_audio(text))
 
+    #调用edge_tts的异步函数
+    async def async_get_tts_audio(self, text):
+        tts = edge_tts.Communicate(text,
+            voice=self.voice,
+            pitch=self.prosody_attributes['pitch'].get(self.pitch, '+0Hz'),
+            rate=self.prosody_attributes['rate'].get(self.rate, '+0%'),
+            volume=self.prosody_attributes['volume'].get(self.volume, '+0%'))
+        lines = []
+        try:
+            async for chunk in tts.stream():
+                if chunk["type"] == "audio":
+                    lines.append(chunk["data"])
+        except Exception as e:
+            default_log.warning(f'EdgeTTSFree.async_get_tts_audio failed: {e}')
+            return '', b''
+        return 'audio/mpeg', b''.join(lines)

application/lib/ebook_tts/engines/tts_base.py

@@ -21,9 +21,9 @@ class TTSBase:
 
     #语音语调的允许常量值列表，除了使用常量值，也可以使用一个正负数值，比如 100%, +1.5, -30.00% 等
     prosody_attributes = {
-        'rate': ('x-slow', 'slow', 'medium', 'fast', 'x-fast'),
-        'pitch': ('x-low', 'low', 'medium', 'high', 'x-high'),
-        'volume': ('silent', 'x-soft', 'soft', 'medium', 'loud', 'x-loud')
+        'rate': {'x-slow': '-50%', 'slow': '-25%', 'medium': '+0%', 'fast': '+25%', 'x-fast': '+50%'},
+        'pitch': {'x-low': '-50Hz', 'low': '-25Hz', 'medium': '+0Hz', 'high': '+25Hz', 'x-high': '+50Hz'},
+        'volume': {'silent': '-100%', 'x-soft': '-50%', 'soft': '-25%', 'medium': '+0%', 'loud': '+25%', 'x-loud': '+50%'}
     }
 
     def __init__(self, params):

application/work/worker.py

@@ -177,6 +177,7 @@ def mp3cat_path():
     return execFile
 
 #合并TTS生成的音频片段
+#返回 (ext, audio)
 def MergeAudioSegment(roList):
     audioDirs = [ro.tts.get('audio_dir') for ro in roList if ro.tts.get('audio_dir')]
     ret = ('', None)
@@ -259,6 +260,6 @@ def MergeAudioSegment(roList):
         try:
             shutil.rmtree(dir_)
         except Exception as e:
-            default.log.debug(f"An error occurred while deleting '{item}': {e}")
+            default_log.debug(f"An error occurred while deleting '{dir_}': {e}")
 
     return ret

tools/update_req.py

@@ -31,7 +31,8 @@ def new_secret_key(length=12):
     ('six', '~=1.16.0'),
     ('feedparser', '~=6.0.11'),
     ('qrcode', '~=7.4.2'),
-    ('#gtts', '~=2.5.1'),
+    ('gtts', '~=2.5.1'),
+    ('edge-tts', '~=6.1.11'),
 ]
 
 REQ_DB = {