agora-python-server-sdk 2.2.4__tar.gz → 2.3.0__tar.gz

{agora_python_server_sdk-2.2.4/agora_python_server_sdk.egg-info → agora_python_server_sdk-2.3.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: agora_python_server_sdk
-Version: 2.2.4
+Version: 2.3.0
 Summary: A Python SDK for Agora Server
 Home-page: https://github.com/AgoraIO-Extensions/Agora-Python-Server-SDK
 Classifier: Intended Audience :: Developers
@@ -19,9 +19,19 @@ Description-Content-Type: text/markdown
 - The examples are provided as very simple demonstrations and are not recommended for use in production environments.
 
 # Very Important Notice !!!
-- A process can only have one instance.
+- A process can only have one instance,and the instance created in process startup is the global instance,and released in process shutdown.
 - An instance can have multiple connections.
 - In all observers or callbacks, you must not call the SDK's own APIs, nor perform CPU-intensive tasks in the callbacks; data copying is allowed.
+- supported video codec:
+  - H264: support both encoding and decoding
+  - VP8: support both encoding and decoding
+  - VP9: support both encoding and decoding
+  - H265: only decoding is supported, encoding willl support in the future
+  - AV1: both encoding and decoding are suported but if video image's solution is less than 360p, the codec will be changed to H264
+- if you want to recive the encoded video data:
+  - set: video_subscription_options.encodedFrameOnly = 1
+  - and then register: set_encoded_video_frame_observer
+
 
 # Required Operating Systems and Python Versions
 - Supported Linux versions:
@@ -50,6 +60,157 @@ python agora_rtc/examples/example_audio_pcm_send.py --appId=xxx --channelId=xxx
 ```
 
 # Change log
+#2025.09.01  release 2.3.0
+
+Overview
+
+Agora Python SDK version 2.3.0 introduces significant enhancements for real-time audio and video communication, with particular optimizations for AI server scenarios. This release includes API improvements, automation of previously manual processes, and increased flexibility in connection configuration.
+
+New Features
+
+1. AudioScenarioAiServer Support
+
+• Added support for the AudioScenarioAiServer scenario type, now set as the default AudioScenario.
+
+• Enables configuration of different scenarios and profiles for connections within the same process.
+
+• Important: When using AI server scenario, client-side must use AIClient scenario to avoid audio abnormalities. Please consult Agora support for SDK versions supporting AIClient scenario.
+
+2. Enhanced Connection Configuration
+
+• Introduced PublishConfigure parameter in connection creation for setting:
+
+  • Scenario and profile configurations
+
+  • Audio/Video publication settings
+
+  • Other publication parameters
+
+• Added multiple observer registration methods to connections:
+
+  • RegisterLocalUserObserver
+
+  • RegisterAudioFrameObserver
+
+  • RegisterVideoFrameObserver
+
+  • RegisterVideoEncodedFrameObserver
+
+3. Stream Management Improvements
+
+• Added publication control methods:
+
+  • PublishAudio/UnpublishAudio
+
+  • PublishVideo/UnpublishVideo
+
+• Added data pushing methods:
+
+  • PushAudioPcmData/PushAudioEncodedData
+
+  • PushVideoFrame/PushVideoEncodedData
+
+• Added InterruptAudio method for supporting interruption functionality
+
+• Added IsPushToRTCCompleted method for checking push status
+
+• Added SendAudioMetaData method for audio metadata transmission
+
+4. Automation Enhancements
+
+• Eliminated manual CreateDataStream calls - now handled automatically
+
+• Automatic observer unregistration on Release() instead of manual unregistering
+
+• Internal handling of media node factory creation:
+
+  • No longer requires manual calls to newMediaNodeFactory
+
+  • Automatic handling of track creation (NewCustomAudioTrackPcm, NewCustomAudioTrackEncoded, NewCustomVideoTrack)
+
+  • Automatic management of audio data senders (NewAudioPcmDataSender)
+
+Integration Process
+
+The updated integration workflow is as follows:
+
+1. Initialize Agora Service (once per process startup):
+   config = AgoraServiceConfig()
+   agora_service = AgoraService()
+   agora_service.initialize(config)
+   
+
+2. Connection Management (can be looped for multiple connections):
+   # Create connection with configuration
+   con = agora_service.create_rtc_connection(con_config, publish_config)
+   
+   # Register observers
+   con.register_observer(conHandler)
+   con.register_audio_frame_observer(audioFrameObserver)
+   con.register_local_user_observer(localUserObserver)
+   
+   # Connect and publish
+   con.connect(token, channelName, userId)
+   con.publish_audio()  # or con.publish_video()
+   
+   # Push data
+   con.push_audio_pcm_data()  # or push_audio_encoded_data()
+   # or con.push_video_frame()/push_video_encoded_data()
+   
+   # Disconnect and release
+   con.disconnect()
+   con.release()
+   
+
+3. Release Agora Service (once during process termination):
+   agora_service.release()
+   
+
+Performance Recommendations
+
+1. For AI Scenarios:
+   • Use AudioScenarioAIServer for server-side applications
+
+   • Provides optimized performance with lower latency (20-30ms reduction on iPhone compared to chorus)
+
+   • Enhanced experience in weak network conditions
+
+   • Mandatory: Client must use AIClient scenario
+
+2. For Non-AI Scenarios:
+   • Consult Agora technical support for appropriate scenario configuration
+
+   • Ensure configuration matches specific business use cases
+
+Bug Fixes
+
+• Updated RTC SDK with 2 bug fixes
+
+Important Notes
+
+• Backward Compatibility: Review integration code for manual calls that are now automated
+
+• Scenario Matching: Ensure client-server scenario compatibility (AIServer requires AIClient)
+
+• Resource Management: Connection release now automatically handles observer unregistration
+
+• Consult Support: For specific integration guidance and upgrade assistance, contact Agora SA
+
+Summary of Core Changes
+
+Before After 2.3.0
+
+Manual CreateDataStream ✅ Automatic
+
+Manual observer unregistration ✅ Automatic on Release()
+
+Fixed per-process scenario ✅ Multi-scenario per process
+
+Client/Server scenario mismatch ❗ AIClient mandatory for AI use
+
+For detailed implementation guidance and version-specific support, please consult Agora technical support.
+
+
 ## 2025.04.28 Release 2.2.4
 -- Update: update rtc sdk from 4.4.31 to 4.4.32
 ## 2025.04.14 Release 2.2.3
@@ -69,6 +230,7 @@ python agora_rtc/examples/example_audio_pcm_send.py --appId=xxx --channelId=xxx
   - Added connection::agora_rtc_conn_enable_encryption.
 -- Additions:
   - Added connectionObserver::on_encryption_error (but not working for now, need to fix in the next monthly version 4.4.32).
+
 2025.02.26 Release 2.2.1
 --Update：
   ​- Reduced buffer size from ​180ms​ to ​100ms​ to minimize latency.
@@ -89,7 +251,6 @@ python agora_rtc/examples/example_audio_pcm_send.py --appId=xxx --channelId=xxx
   - Add the AudioMetaData interface: localuser::send_audio_meta_data. Done.
   - Add the OnAudioMetaDataReceived callback to localuserObserver::on_audio_meta_data_received. Done.
 -- Sample modifications.
-
 2024.12.17 Release 2.1.7
 --Changes:
 
@@ -117,7 +278,6 @@ python agora_rtc/examples/example_audio_pcm_send.py --appId=xxx --channelId=xxx
   -- Replaced the use of pacer with AudioConsumer for pushing PCM audio.
 - Updates:
   -- Updated the samples related to Pacer and VAD.
-
 ## 2024.12.03 release Version 2.1.5
 - Modifications:
   - LocalUser/audioTrack:
@@ -173,7 +333,6 @@ Fixed some bug.
 - Handle business operations.
 - When a user logs out, execute con.disconnect() and release the audio/video tracks and observers associated with the connection, but do not call con.release(); then put the connection back into the connection pool.
 - When the process exits, release the connection pool (release each con.release()), service/media_node_factory, and the connection pool (release each con.release()) to ensure resource release and optimal performance.
-
 ## Use of VAD
 # Source code: voice_detection.py
 # Sample code: example_audio_vad.py
@@ -204,7 +363,8 @@ Judge the value of state according to the returned state, and do corresponding p
 ### How to push the audio generated by TTS into the channel?
   # Source code: audio_consumer.py
   # Sample code: example_audio_consumer.py
-### How to release resources?
+
+### How to release resource?
 ## 如何释放资源？
     localuser.unpublish_audio(audio_track)
     localuser.unpublish_video(video_track)

{agora_python_server_sdk-2.2.4 → agora_python_server_sdk-2.3.0}/README.md

@@ -4,9 +4,19 @@
 - The examples are provided as very simple demonstrations and are not recommended for use in production environments.
 
 # Very Important Notice !!!
-- A process can only have one instance.
+- A process can only have one instance,and the instance created in process startup is the global instance,and released in process shutdown.
 - An instance can have multiple connections.
 - In all observers or callbacks, you must not call the SDK's own APIs, nor perform CPU-intensive tasks in the callbacks; data copying is allowed.
+- supported video codec:
+  - H264: support both encoding and decoding
+  - VP8: support both encoding and decoding
+  - VP9: support both encoding and decoding
+  - H265: only decoding is supported, encoding willl support in the future
+  - AV1: both encoding and decoding are suported but if video image's solution is less than 360p, the codec will be changed to H264
+- if you want to recive the encoded video data:
+  - set: video_subscription_options.encodedFrameOnly = 1
+  - and then register: set_encoded_video_frame_observer
+
 
 # Required Operating Systems and Python Versions
 - Supported Linux versions:
@@ -35,6 +45,157 @@ python agora_rtc/examples/example_audio_pcm_send.py --appId=xxx --channelId=xxx
 ```
 
 # Change log
+#2025.09.01  release 2.3.0
+
+Overview
+
+Agora Python SDK version 2.3.0 introduces significant enhancements for real-time audio and video communication, with particular optimizations for AI server scenarios. This release includes API improvements, automation of previously manual processes, and increased flexibility in connection configuration.
+
+New Features
+
+1. AudioScenarioAiServer Support
+
+• Added support for the AudioScenarioAiServer scenario type, now set as the default AudioScenario.
+
+• Enables configuration of different scenarios and profiles for connections within the same process.
+
+• Important: When using AI server scenario, client-side must use AIClient scenario to avoid audio abnormalities. Please consult Agora support for SDK versions supporting AIClient scenario.
+
+2. Enhanced Connection Configuration
+
+• Introduced PublishConfigure parameter in connection creation for setting:
+
+  • Scenario and profile configurations
+
+  • Audio/Video publication settings
+
+  • Other publication parameters
+
+• Added multiple observer registration methods to connections:
+
+  • RegisterLocalUserObserver
+
+  • RegisterAudioFrameObserver
+
+  • RegisterVideoFrameObserver
+
+  • RegisterVideoEncodedFrameObserver
+
+3. Stream Management Improvements
+
+• Added publication control methods:
+
+  • PublishAudio/UnpublishAudio
+
+  • PublishVideo/UnpublishVideo
+
+• Added data pushing methods:
+
+  • PushAudioPcmData/PushAudioEncodedData
+
+  • PushVideoFrame/PushVideoEncodedData
+
+• Added InterruptAudio method for supporting interruption functionality
+
+• Added IsPushToRTCCompleted method for checking push status
+
+• Added SendAudioMetaData method for audio metadata transmission
+
+4. Automation Enhancements
+
+• Eliminated manual CreateDataStream calls - now handled automatically
+
+• Automatic observer unregistration on Release() instead of manual unregistering
+
+• Internal handling of media node factory creation:
+
+  • No longer requires manual calls to newMediaNodeFactory
+
+  • Automatic handling of track creation (NewCustomAudioTrackPcm, NewCustomAudioTrackEncoded, NewCustomVideoTrack)
+
+  • Automatic management of audio data senders (NewAudioPcmDataSender)
+
+Integration Process
+
+The updated integration workflow is as follows:
+
+1. Initialize Agora Service (once per process startup):
+   config = AgoraServiceConfig()
+   agora_service = AgoraService()
+   agora_service.initialize(config)
+   
+
+2. Connection Management (can be looped for multiple connections):
+   # Create connection with configuration
+   con = agora_service.create_rtc_connection(con_config, publish_config)
+   
+   # Register observers
+   con.register_observer(conHandler)
+   con.register_audio_frame_observer(audioFrameObserver)
+   con.register_local_user_observer(localUserObserver)
+   
+   # Connect and publish
+   con.connect(token, channelName, userId)
+   con.publish_audio()  # or con.publish_video()
+   
+   # Push data
+   con.push_audio_pcm_data()  # or push_audio_encoded_data()
+   # or con.push_video_frame()/push_video_encoded_data()
+   
+   # Disconnect and release
+   con.disconnect()
+   con.release()
+   
+
+3. Release Agora Service (once during process termination):
+   agora_service.release()
+   
+
+Performance Recommendations
+
+1. For AI Scenarios:
+   • Use AudioScenarioAIServer for server-side applications
+
+   • Provides optimized performance with lower latency (20-30ms reduction on iPhone compared to chorus)
+
+   • Enhanced experience in weak network conditions
+
+   • Mandatory: Client must use AIClient scenario
+
+2. For Non-AI Scenarios:
+   • Consult Agora technical support for appropriate scenario configuration
+
+   • Ensure configuration matches specific business use cases
+
+Bug Fixes
+
+• Updated RTC SDK with 2 bug fixes
+
+Important Notes
+
+• Backward Compatibility: Review integration code for manual calls that are now automated
+
+• Scenario Matching: Ensure client-server scenario compatibility (AIServer requires AIClient)
+
+• Resource Management: Connection release now automatically handles observer unregistration
+
+• Consult Support: For specific integration guidance and upgrade assistance, contact Agora SA
+
+Summary of Core Changes
+
+Before After 2.3.0
+
+Manual CreateDataStream ✅ Automatic
+
+Manual observer unregistration ✅ Automatic on Release()
+
+Fixed per-process scenario ✅ Multi-scenario per process
+
+Client/Server scenario mismatch ❗ AIClient mandatory for AI use
+
+For detailed implementation guidance and version-specific support, please consult Agora technical support.
+
+
 ## 2025.04.28 Release 2.2.4
 -- Update: update rtc sdk from 4.4.31 to 4.4.32
 ## 2025.04.14 Release 2.2.3
@@ -54,6 +215,7 @@ python agora_rtc/examples/example_audio_pcm_send.py --appId=xxx --channelId=xxx
   - Added connection::agora_rtc_conn_enable_encryption.
 -- Additions:
   - Added connectionObserver::on_encryption_error (but not working for now, need to fix in the next monthly version 4.4.32).
+
 2025.02.26 Release 2.2.1
 --Update：
   ​- Reduced buffer size from ​180ms​ to ​100ms​ to minimize latency.
@@ -74,7 +236,6 @@ python agora_rtc/examples/example_audio_pcm_send.py --appId=xxx --channelId=xxx
   - Add the AudioMetaData interface: localuser::send_audio_meta_data. Done.
   - Add the OnAudioMetaDataReceived callback to localuserObserver::on_audio_meta_data_received. Done.
 -- Sample modifications.
-
 2024.12.17 Release 2.1.7
 --Changes:
 
@@ -102,7 +263,6 @@ python agora_rtc/examples/example_audio_pcm_send.py --appId=xxx --channelId=xxx
   -- Replaced the use of pacer with AudioConsumer for pushing PCM audio.
 - Updates:
   -- Updated the samples related to Pacer and VAD.
-
 ## 2024.12.03 release Version 2.1.5
 - Modifications:
   - LocalUser/audioTrack:
@@ -158,7 +318,6 @@ Fixed some bug.
 - Handle business operations.
 - When a user logs out, execute con.disconnect() and release the audio/video tracks and observers associated with the connection, but do not call con.release(); then put the connection back into the connection pool.
 - When the process exits, release the connection pool (release each con.release()), service/media_node_factory, and the connection pool (release each con.release()) to ensure resource release and optimal performance.
-
 ## Use of VAD
 # Source code: voice_detection.py
 # Sample code: example_audio_vad.py
@@ -189,7 +348,8 @@ Judge the value of state according to the returned state, and do corresponding p
 ### How to push the audio generated by TTS into the channel?
   # Source code: audio_consumer.py
   # Sample code: example_audio_consumer.py
-### How to release resources?
+
+### How to release resource?
 ## 如何释放资源？
     localuser.unpublish_audio(audio_track)
     localuser.unpublish_video(video_track)

{agora_python_server_sdk-2.2.4 → agora_python_server_sdk-2.3.0}/agora/rtc/__init__.py

@@ -36,21 +36,25 @@ def _check_download_and_extract_sdk():
     #url = "https://download.agora.io/sdk/release/agora_rtc_sdk-x86_64-linux-gnu-v4.4.30-20241024_101940-398537.zip"
     # version 2.2.0 for linux
     #url = "https://download.agora.io/sdk/release/agora_rtc_sdk-x86_64-linux-gnu-v4.4.31-20241223_111509-491956.zip"
-    url  = "https://download.agora.io/sdk/release/agora_rtc_sdk-x86_64-linux-gnu-v4.4.32-20250425_144419-675648.zip"
+    #url  = "https://download.agora.io/sdk/release/agora_rtc_sdk-x86_64-linux-gnu-v4.4.32-20250715_161625-791246.zip"
+    url = "https://download.agora.io/sdk/release/agora_rtc_sdk-x86_64-linux-gnu-v4.4.32-20250829_160340-860733.zip"
+    
         
            
     libagora_rtc_sdk_path = os.path.join(sdk_dir, "libagora_rtc_sdk.so")
     #rtc_md5 = "7031dd10d1681cd88fd89d68c5b54282"
-    rtc_md5 = "0d6c96bbebdd9b0a84b9cdf01ba6aa7f"
+    rtc_md5 = "7eb8042e43246f95f188549d8711d1bf"
     if sys.platform == 'darwin':
         #url = "https://download.agora.io/sdk/release/agora_rtc_sdk_mac_rel.v4.4.30_22472_FULL_20241024_1224_398653.zip"
         # version   2.2.0 for mac
         #url = "https://download.agora.io/sdk/release/agora_sdk_mac_v4.4.31_23136_FULL_20241223_1245_492039.zip"
-        url = "https://download.agora.io/sdk/release/agora_sdk_mac_v4.4.32_24257_FULL_20250425_1609_675722.zip"
+        #url = "https://download.agora.io/sdk/release/agora_sdk_mac_v4.4.32_24915_FULL_20250715_1710_791284.zip"
+        url = "https://download.agora.io/sdk/release/agora_sdk_mac_v4.4.32_25418_FULL_20250829_1647_860754.zip"
+        
 
         libagora_rtc_sdk_path = os.path.join(sdk_dir, "libAgoraRtcKit.dylib")
         #rtc_md5 = "ca3ca14f9e2b7d97eb2594d1f32dab9f"
-        rtc_md5 = "358be32c5dfc72192402cf49838ea942"
+        rtc_md5 = "df0ec3b5073d17dee76cc4d97c13699a"
     if arch == "aarch64" and sys.platform == 'linux':
         #url = "https://download.agora.io/sdk/release/Agora-RTC-aarch64-linux-gnu-v4.4.31-20250307_175457-603878.zip"
         url = "https://download.agora.io/sdk/release/Agora-RTC-aarch64-linux-gnu-v4.4.32-20250425_150503-675674.zip"

{agora_python_server_sdk-2.2.4 → agora_python_server_sdk-2.3.0}/agora/rtc/_ctypes_handle/_ctypes_data.py

@@ -1000,7 +1000,6 @@ class RemoteAudioTrackStatsInner(ctypes.Structure):
             stats.received_bytes
         )
 
-
 class EncodedAudioFrameInfoInner(ctypes.Structure):
     _fields_ = [
         ('speech', ctypes.c_int),
@@ -1008,7 +1007,8 @@ class EncodedAudioFrameInfoInner(ctypes.Structure):
         ('sample_rate_hz', ctypes.c_int),
         ('samples_per_channel', ctypes.c_int),
         ('send_even_if_empty', ctypes.c_int),
-        ('number_of_channels', ctypes.c_int)
+        ('number_of_channels', ctypes.c_int),
+        ('capture_time_ms', ctypes.c_int64)
     ]
 
     @staticmethod
@@ -1019,7 +1019,8 @@ class EncodedAudioFrameInfoInner(ctypes.Structure):
             info.sample_rate,
             info.samples_per_channel,
             info.send_even_if_empty,
-            info.number_of_channels
+            info.number_of_channels,
+            info.capture_time_ms
         )
 
 
@@ -1067,6 +1068,10 @@ class AudioFrameInner(ctypes.Structure):
         ("buffer", ctypes.c_void_p),
         ("render_time_ms", ctypes.c_int64),
         ("avsync_type", ctypes.c_int),
+        ("presentation_ms", ctypes.c_int64),
+        ("audio_track_number", ctypes.c_int),
+        ("rtp_timestamp", ctypes.c_uint32),
+       
         ("far_field_flag", ctypes.c_int),
         ("rms", ctypes.c_int),
         ("voice_prob", ctypes.c_int),
@@ -1084,6 +1089,9 @@ class AudioFrameInner(ctypes.Structure):
             buffer=bytearray(ctypes.string_at(self.buffer, self.samples_per_channel * self.bytes_per_sample * self.channels)),
             render_time_ms=self.render_time_ms,
             avsync_type=self.avsync_type,
+            presentation_ms=self.presentation_ms,
+            audio_track_number=self.audio_track_number,
+            rtp_timestamp=self.rtp_timestamp,
             far_field_flag=self.far_field_flag,
             rms=self.rms,
             voice_prob=self.voice_prob,
@@ -1102,6 +1110,9 @@ class AudioFrameInner(ctypes.Structure):
             ctypes.cast(frame.buffer, ctypes.c_void_p),
             frame.render_time_ms,
             frame.avsync_type,
+            frame.presentation_ms,
+            frame.audio_track_number,
+            frame.rtp_timestamp,
             frame.far_field_flag,
             frame.rms,
             frame.voice_prob,
@@ -1133,6 +1144,11 @@ class AgoraServiceConfigInner(ctypes.Structure):
 
         ('use_string_uid', ctypes.c_int),
         ('domain_limit', ctypes.c_int),
+        ('log_level', ctypes.c_int),
+        ('log_file_path', ctypes.c_char_p),
+        ('log_file_size_kb', ctypes.c_uint32),
+        ('data_dir', ctypes.c_char_p),
+        ('config_dir', ctypes.c_char_p),
     ]
 
     def get(self):
@@ -1146,7 +1162,12 @@ class AgoraServiceConfigInner(ctypes.Structure):
             channel_profile=self.channel_profile,
             audio_scenario=self.audio_scenario,
             use_string_uid=self.use_string_uid,
-            domain_limit=self.domain_limit
+            domain_limit=self.domain_limit,
+            log_level=self.log_level,
+            log_path=self.log_file_path.decode() if self.log_file_path else "",
+            log_file_size_kb=self.log_file_size_kb,
+            data_dir=self.data_dir.decode() if self.data_dir else "",
+            config_dir=self.config_dir.decode() if self.config_dir else ""
         )
 
     @staticmethod
@@ -1156,12 +1177,17 @@ class AgoraServiceConfigInner(ctypes.Structure):
             config.enable_audio_device,
             config.enable_video,
             config.context,
-            config.appid.encode(),
+            config.appid.encode() if config.appid else None,
             config.area_code,
             config.channel_profile,
             config.audio_scenario,
             config.use_string_uid,
-            config.domain_limit
+            config.domain_limit,
+            config.log_level,
+            config.log_path.encode() if config.log_path else None,
+            config.log_file_size_kb,
+            config.data_dir.encode() if config.data_dir else None,
+            config.config_dir.encode() if config.config_dir else None
         )
 
 
@@ -1177,7 +1203,8 @@ class EncodedVideoFrameInfoInner(ctypes.Structure):
         ("capture_time_ms", ctypes.c_int64),
         ("decode_time_ms", ctypes.c_int64),
         ("uid", ctypes.c_uint),
-        ("stream_type", ctypes.c_int)
+        ("stream_type", ctypes.c_int),
+        ("presentation_ms", ctypes.c_int64)
     ]
 
     def get(self):
@@ -1192,7 +1219,8 @@ class EncodedVideoFrameInfoInner(ctypes.Structure):
             capture_time_ms=self.capture_time_ms,
             decode_time_ms=self.decode_time_ms,
             uid=self.uid,
-            stream_type=self.stream_type
+            stream_type=self.stream_type,
+            presentation_ms=self.presentation_ms
         )
 
     @staticmethod
@@ -1208,7 +1236,8 @@ class EncodedVideoFrameInfoInner(ctypes.Structure):
             info.capture_time_ms,
             info.decode_time_ms,
             info.uid,
-            info.stream_type
+            info.stream_type,
+            info.presentation_ms
         )
     
 
@@ -1249,3 +1278,94 @@ class EncryptionConfigInner(ctypes.Structure):
             encryption_key,
             encryption_kdf_salt
         )
+
+
+class CapabilityItemInner(ctypes.Structure):
+    _fields_ = [
+        ("id", ctypes.c_uint8),
+        ("name", ctypes.c_char_p)
+    ]
+
+    def get(self):
+        return CapabilityItem(
+            id=self.id,
+            name=self.name.decode() if self.name else ""
+        )
+
+    @staticmethod
+    def create(item: CapabilityItem) -> 'CapabilityItemInner':
+        return CapabilityItemInner(
+            item.id,
+            item.name.encode() if item.name else None
+        )
+
+
+class CapabilityItemMapInner(ctypes.Structure):
+    _fields_ = [
+        ("item", ctypes.POINTER(CapabilityItemInner)),
+        ("size", ctypes.c_size_t)
+    ]
+
+    def get(self):
+        items = []
+        if self.item and self.size > 0:
+            for i in range(self.size):
+                items.append(self.item[i].get())
+        return CapabilityItemMap(
+            item=items,
+            size=self.size
+        )
+
+    @staticmethod
+    def create(item_map: CapabilityItemMap) -> 'CapabilityItemMapInner':
+        if item_map.item and len(item_map.item) > 0:
+            items_array = (CapabilityItemInner * len(item_map.item))()
+            for i, item in enumerate(item_map.item):
+                items_array[i] = CapabilityItemInner.create(item)
+            items_ptr = ctypes.cast(items_array, ctypes.POINTER(CapabilityItemInner))
+        else:
+            items_ptr = ctypes.POINTER(CapabilityItemInner)()
+        
+        return CapabilityItemMapInner(
+            items_ptr,
+            len(item_map.item) if item_map.item else 0
+        )
+
+
+class CapabilitiesInner(ctypes.Structure):
+    _fields_ = [
+        ("item_map", ctypes.POINTER(CapabilityItemMapInner)),
+        ("capability_type", ctypes.c_int)
+    ]
+
+    def get(self):
+        return Capabilities(
+            item_map=self.item_map.contents.get() if self.item_map else None,
+            capability_type=self.capability_type
+        )
+
+    @staticmethod
+    def create(capabilities: Capabilities) -> 'CapabilitiesInner':
+        item_map_ptr = None
+        if capabilities.item_map:
+            item_map_inner = CapabilityItemMapInner.create(capabilities.item_map)
+            item_map_ptr = ctypes.pointer(item_map_inner)
+        
+        return CapabilitiesInner(
+            item_map_ptr,
+            capabilities.capability_type
+        )
+
+
+
+class CapabilitiesItemMapInner(ctypes.Structure):
+    _fields_ = [
+        ("item", ctypes.POINTER(CapabilityItemInner)),
+        ("size", ctypes.c_size_t)
+    ]
+
+    def get(self):
+        return CapabilitiesItemMap(
+            item=self.item,
+            size=self.size
+        )