drop custom SSML support

README.md

@@ -61,35 +61,11 @@ You must first check the available voices with the `--list-voices` option:
 
 ### Custom SSML
 
-It is possible to send Microsoft's text-to-speech servers a custom SSML document which would allow greater customization of the speech. 
+Support for custom SSML has been removed since 5.0.0 because Microsoft has taken the initiative to prevent it from working. You cannot use custom SSML anymore.
-
-Information about the SSML format can be found here on Microsoft's own website: https://docs.microsoft.com/en-us/azure/cognitive-services/speech-service/speech-synthesis-markup
-
-As a short example, if you want to apply the following SSML document and play it back using `edge-tts`.
-
-```
-<speak version="1.0" xmlns="http://www.w3.org/2001/10/synthesis"
-       xmlns:mstts="https://www.w3.org/2001/mstts" xml:lang="en-US">
-    <voice name="en-US-AriaNeural">
-        <mstts:express-as style="cheerful">
-            That'd be just amazing!
-        </mstts:express-as>
-    </voice>
-</speak>
-```
-
-It would be easiest to do the following:
-
-1. Create a file called `custom_ssml.xml` with the above content.
-2. Run the following command:
-
-       $ edge-tts --custom-ssml --file custom_ssml.xml --write-media amazing.mp3
-
-3. Voila!
 
 ### Changing pitch, rate, volume, etc.
 
-It is possible to make minor changes to the generated speech without resorting to custom SSML. However, you must note that you couldn't use the `--custom-ssml` option with the `--pitch`, `--rate`, `--volume`, etc. options.
+It is possible to make minor changes to the generated speech.
 
     $ edge-tts --pitch=-10Hz --text "Hello, world!" --write-media hello_with_pitch_down.mp3
     $ edge-tts --rate=0.5 --text "Hello, world!" --write-media hello_with_rate_halved.mp3

setup.cfg

@@ -1,6 +1,6 @@
 [metadata]
 name = edge-tts
-version = 4.0.11
+version = 5.0.0
 author = rany
 author_email = ranygh@riseup.net
 description = Microsoft Edge's TTS

src/edge_tts/communicate.py

@@ -213,7 +213,6 @@ class Communicate:
         pitch="+0Hz",
         rate="+0%",
         volume="+0%",
-        customspeak=False,
         proxy=None,
     ):
         """
@@ -223,11 +222,10 @@ class Communicate:
             messages (str or list): A list of SSML strings or a single text.
             boundery_type (int): The type of boundary to use. 0 for none, 1 for word_boundary, 2 for sentence_boundary.
             codec (str): The codec to use.
-            voice (str): The voice to use (only applicable to non-customspeak).
+            voice (str): The voice to use.
-            pitch (str): The pitch to use (only applicable to non-customspeak).
+            pitch (str): The pitch to use.
-            rate (str): The rate to use (only applicable to non-customspeak).
+            rate (str): The rate to use.
-            volume (str): The volume to use (only applicable to non-customspeak).
+            volume (str): The volume to use.
-            customspeak (bool): Whether to create the SSML or treat the messages as SSML.
 
         Yields:
             tuple: The subtitle offset, subtitle, and audio data.
@@ -244,23 +242,19 @@ class Communicate:
 
         word_boundary = str(word_boundary).lower()
 
-        if not customspeak:
+        websocket_max_size = 2**16
-            websocket_max_size = 2**16
+        overhead_per_message = (
-            overhead_per_message = (
+            len(
-                len(
+                ssml_headers_plus_data(
-                    ssml_headers_plus_data(
+                    connect_id(), self.date, mkssml("", voice, pitch, rate, volume)
-                        connect_id(), self.date, mkssml("", voice, pitch, rate, volume)
-                    )
                 )
-                + 50
-            )  # margin of error
-            messages = split_text_by_byte_length(
-                escape(remove_incompatible_characters(messages)),
-                websocket_max_size - overhead_per_message,
             )
-        else:
+            + 50
-            if isinstance(messages, str):
+        )  # margin of error
-                messages = [messages]
+        messages = split_text_by_byte_length(
+            escape(remove_incompatible_characters(messages)),
+            websocket_max_size - overhead_per_message,
+        )
 
         # Variables for the loop
         download = False
@@ -307,18 +301,13 @@ class Communicate:
                     # Send the request to the service.
                     await websocket.send_str(request)
                     # Send the message itself.
-                    if not customspeak:
+                    await websocket.send_str(
-                        await websocket.send_str(
+                        ssml_headers_plus_data(
-                            ssml_headers_plus_data(
+                            connect_id(),
-                                connect_id(),
+                            self.date,
-                                self.date,
+                            mkssml(message, voice, pitch, rate, volume),
-                                mkssml(message, voice, pitch, rate, volume),
-                            )
-                        )
-                    else:
-                        await websocket.send_str(
-                            ssml_headers_plus_data(connect_id(), self.date, message)
                         )
+                    )
 
                     # Begin listening for the response.
                     async for received in websocket:

src/edge_tts/util.py

@@ -38,7 +38,6 @@ async def _tts(args):
         args.pitch,
         args.rate,
         args.volume,
-        customspeak=args.custom_ssml,
         proxy=args.proxy,
     ):
         if i[2] is not None:
@@ -62,12 +61,6 @@ async def _main():
     group = parser.add_mutually_exclusive_group(required=True)
     group.add_argument("-t", "--text", help="what TTS will say")
     group.add_argument("-f", "--file", help="same as --text but read from file")
-    parser.add_argument(
-        "-z",
-        "--custom-ssml",
-        help="treat text as ssml to send. For more info check https://bit.ly/3fIq13S",
-        action="store_true",
-    )
     parser.add_argument(
         "-v",
         "--voice",