sarvamai 0.1.13a2__py3-none-any.whl → 0.1.15__py3-none-any.whl

sarvamai/speech_to_text_translate_job/job.py

@@ -0,0 +1,505 @@
+import asyncio
+import mimetypes
+import os
+import time
+import typing
+import httpx
+from http import HTTPStatus
+
+from ..types import JobStatusV1Response
+
+if typing.TYPE_CHECKING:
+    from .client import (
+        AsyncSpeechToTextTranslateJobClient,
+        SpeechToTextTranslateJobClient,
+    )
+
+
+class AsyncSpeechToTextTranslateJob:
+    def __init__(
+        self,
+        job_id: str,
+        client: "AsyncSpeechToTextTranslateJobClient",
+    ):
+        """
+        Initialise the asynchronous speech-to-text-translate-translate job.
+
+        Parameters
+        ----------
+        job_id : str
+            The unique job identifier returned from a previous job initialisation.
+
+        client : AsyncSpeechToTextTranslateJobClient
+            The async client instance used to create the job.
+
+            !!! important
+                This must be the **same client instance** that was used to initialise
+                the job originally, as it contains the subscription key and configuration
+                required to authenticate and manage the job.
+
+        """
+        self._job_id = job_id
+        self._client = client
+
+    @property
+    def job_id(self) -> str:
+        """
+        Returns the job ID associated with this job instance.
+
+        Returns
+        -------
+        str
+        """
+        return self._job_id
+
+    async def upload_files(
+        self, file_paths: typing.Sequence[str], timeout: float = 60.0
+    ) -> bool:
+        """
+        Upload input audio files for the speech-to-text-translate job.
+
+        Parameters
+        ----------
+        file_paths : Sequence[str]
+            List of full paths to local audio files.
+
+        timeout : float, optional
+            The maximum time to wait for the upload to complete (in seconds),
+            by default 60.0
+        Returns
+        -------
+        bool
+            True if all files are uploaded successfully.
+        """
+        upload_links = await self._client.get_upload_links(
+            job_id=self._job_id,
+            files=[os.path.basename(p) for p in file_paths],
+        )
+        client_timeout = httpx.Timeout(timeout=timeout)
+        async with httpx.AsyncClient(timeout=client_timeout) as session:
+            for path in file_paths:
+                file_name = os.path.basename(path)
+                url = upload_links.upload_urls[file_name].file_url
+                with open(path, "rb") as f:
+                    response = await session.put(
+                        url,
+                        content=f.read(),
+                        headers={
+                            "x-ms-blob-type": "BlockBlob",
+                            "Content-Type": "audio/wav",
+                        },
+                    )
+                if (
+                    response.status_code > HTTPStatus.IM_USED
+                    or response.status_code < HTTPStatus.OK
+                ):
+                    raise RuntimeError(
+                        f"Upload failed for {file_name}: {response.status_code}"
+                    )
+        return True
+
+    async def wait_until_complete(
+        self, poll_interval: int = 5, timeout: int = 600
+    ) -> JobStatusV1Response:
+        """
+        Polls job status until it completes or fails.
+
+        Parameters
+        ----------
+        poll_interval : int, optional
+            Time in seconds between polling attempts (default is 5).
+
+        timeout : int, optional
+            Maximum time to wait for completion in seconds (default is 600).
+
+        Returns
+        -------
+        JobStatusV1Response
+            Final job status.
+
+        Raises
+        ------
+        TimeoutError
+            If the job does not complete within the given timeout.
+        """
+        start = asyncio.get_event_loop().time()
+        while True:
+            status = await self.get_status()
+            state = status.job_state.lower()
+            if state in {"completed", "failed"}:
+                return status
+            if asyncio.get_event_loop().time() - start > timeout:
+                raise TimeoutError(
+                    f"Job {self._job_id} did not complete within {timeout} seconds."
+                )
+            await asyncio.sleep(poll_interval)
+
+    async def get_output_mappings(self) -> typing.List[typing.Dict[str, str]]:
+        """
+        Get the mapping of input files to their corresponding output files.
+
+        Returns
+        -------
+        List[Dict[str, str]]
+            List of mappings with keys 'input_file' and 'output_file'.
+        """
+        job_status = await self.get_status()
+        return [
+            {
+                "input_file": detail.inputs[0].file_name,
+                "output_file": detail.outputs[0].file_name,
+            }
+            for detail in (job_status.job_details or [])
+            if detail.inputs and detail.outputs
+        ]
+
+    async def download_outputs(self, output_dir: str) -> bool:
+        """
+        Download output files to the specified directory.
+
+        Parameters
+        ----------
+        output_dir : str
+            Local directory where outputs will be saved.
+
+        Returns
+        -------
+        bool
+            True if all files downloaded successfully.
+
+        Raises
+        ------
+        RuntimeError
+            If a file fails to download.
+        """
+        mappings = await self.get_output_mappings()
+        file_names = [m["output_file"] for m in mappings]
+        download_links = await self._client.get_download_links(
+            job_id=self._job_id, files=file_names
+        )
+
+        os.makedirs(output_dir, exist_ok=True)
+        async with httpx.AsyncClient() as session:
+            for m in mappings:
+                url = download_links.download_urls[m["output_file"]].file_url
+                response = await session.get(url)
+                if (
+                    response.status_code > HTTPStatus.IM_USED
+                    or response.status_code < HTTPStatus.OK
+                ):
+                    raise RuntimeError(
+                        f"Download failed for {m['output_file']}: {response.status_code}"
+                    )
+                output_path = os.path.join(output_dir, f"{m['input_file']}.json")
+                with open(output_path, "wb") as f:
+                    f.write(response.content)
+        return True
+
+    async def get_status(self) -> JobStatusV1Response:
+        """
+        Retrieve the current status of the job.
+
+        Returns
+        -------
+        JobStatusV1Response
+        """
+        return await self._client.get_status(self._job_id)
+
+    async def start(self) -> JobStatusV1Response:
+        """
+        Start the speech-to-text-translate job processing.
+
+        Returns
+        -------
+        JobStatusV1Response
+        """
+        return await self._client.start(job_id=self._job_id)
+
+    async def exists(self) -> bool:
+        """
+        Check if the job exists in the system.
+
+        Returns
+        -------
+        bool
+        """
+        try:
+            await self.get_status()
+            return True
+        except httpx.HTTPStatusError:
+            return False
+
+    async def is_complete(self) -> bool:
+        """
+        Check if the job is either completed or failed.
+
+        Returns
+        -------
+        bool
+        """
+        state = (await self.get_status()).job_state.lower()
+        return state in {"completed", "failed"}
+
+    async def is_successful(self) -> bool:
+        """
+        Check if the job completed successfully.
+
+        Returns
+        -------
+        bool
+        """
+        return (await self.get_status()).job_state.lower() == "completed"
+
+    async def is_failed(self) -> bool:
+        """
+        Check if the job has failed.
+
+        Returns
+        -------
+        bool
+        """
+        return (await self.get_status()).job_state.lower() == "failed"
+
+
+class SpeechToTextTranslateJob:
+    def __init__(self, job_id: str, client: "SpeechToTextTranslateJobClient"):
+        """
+        Initialise the synchronous speech-to-text-translate job.
+
+        Parameters
+        ----------
+        job_id : str
+            The unique job identifier returned from a previous job initialisation.
+
+        client : SpeechToTextTranslateJobClient
+            The client instance used to create the job.
+
+            !!! important
+                This must be the **same client instance** that was used to initialise
+                the job originally, as it contains the subscription key and configuration
+                required to authenticate and manage the job.
+
+        """
+        self._job_id = job_id
+        self._client = client
+
+    @property
+    def job_id(self) -> str:
+        """
+        Returns the job ID associated with this job instance.
+
+        Returns
+        -------
+        str
+        """
+        return self._job_id
+
+    def upload_files(
+        self, file_paths: typing.Sequence[str], timeout: float = 60.0
+    ) -> bool:
+        """
+        Upload input audio files for the speech-to-text-translate job.
+
+        Parameters
+        ----------
+        file_paths : Sequence[str]
+            List of full paths to local audio files.
+
+        timeout : float, optional
+            The maximum time to wait for the upload to complete (in seconds),
+            by default 60.0
+
+        Returns
+        -------
+        bool
+            True if all files are uploaded successfully.
+        """
+        upload_links = self._client.get_upload_links(
+            job_id=self._job_id, files=[os.path.basename(p) for p in file_paths]
+        )
+        client_timeout = httpx.Timeout(timeout=timeout)
+        with httpx.Client(timeout=client_timeout) as client:
+            for path in file_paths:
+                file_name = os.path.basename(path)
+                url = upload_links.upload_urls[file_name].file_url
+                content_type, _ = mimetypes.guess_type(path)
+                if content_type is None:
+                    content_type = "audio/wav"
+                with open(path, "rb") as f:
+                    response = client.put(
+                        url,
+                        content=f,
+                        headers={
+                            "x-ms-blob-type": "BlockBlob",
+                            "Content-Type": content_type,
+                        },
+                    )
+                if (
+                    response.status_code > HTTPStatus.IM_USED
+                    or response.status_code < HTTPStatus.OK
+                ):
+                    raise RuntimeError(
+                        f"Upload failed for {file_name}: {response.status_code}"
+                    )
+        return True
+
+    def wait_until_complete(
+        self, poll_interval: int = 5, timeout: int = 600
+    ) -> JobStatusV1Response:
+        """
+        Polls job status until it completes or fails.
+
+        Parameters
+        ----------
+        poll_interval : int, optional
+            Time in seconds between polling attempts (default is 5).
+
+        timeout : int, optional
+            Maximum time to wait for completion in seconds (default is 600).
+
+        Returns
+        -------
+        JobStatusV1Response
+            Final job status.
+
+        Raises
+        ------
+        TimeoutError
+            If the job does not complete within the given timeout.
+        """
+        start = time.monotonic()
+        while True:
+            status = self.get_status()
+            state = status.job_state.lower()
+            if state in {"completed", "failed"}:
+                return status
+            if time.monotonic() - start > timeout:
+                raise TimeoutError(
+                    f"Job {self._job_id} did not complete within {timeout} seconds."
+                )
+            time.sleep(poll_interval)
+
+    def get_output_mappings(self) -> typing.List[typing.Dict[str, str]]:
+        """
+        Get the mapping of input files to their corresponding output files.
+
+        Returns
+        -------
+        List[Dict[str, str]]
+            List of mappings with keys 'input_file' and 'output_file'.
+        """
+        job_status = self.get_status()
+        return [
+            {
+                "input_file": detail.inputs[0].file_name,
+                "output_file": detail.outputs[0].file_name,
+            }
+            for detail in (job_status.job_details or [])
+            if detail.inputs and detail.outputs
+        ]
+
+    def download_outputs(self, output_dir: str) -> bool:
+        """
+        Download output files to the specified directory.
+
+        Parameters
+        ----------
+        output_dir : str
+            Local directory where outputs will be saved.
+
+        Returns
+        -------
+        bool
+            True if all files downloaded successfully.
+
+        Raises
+        ------
+        RuntimeError
+            If a file fails to download.
+        """
+        mappings = self.get_output_mappings()
+        file_names = [m["output_file"] for m in mappings]
+        download_links = self._client.get_download_links(
+            job_id=self._job_id, files=file_names
+        )
+
+        os.makedirs(output_dir, exist_ok=True)
+        with httpx.Client() as client:
+            for m in mappings:
+                url = download_links.download_urls[m["output_file"]].file_url
+                response = client.get(url)
+                if (
+                    response.status_code > HTTPStatus.IM_USED
+                    or response.status_code < HTTPStatus.OK
+                ):
+                    raise RuntimeError(
+                        f"Download failed for {m['output_file']}: {response.status_code}"
+                    )
+                output_path = os.path.join(output_dir, f"{m['input_file']}.json")
+                with open(output_path, "wb") as f:
+                    f.write(response.content)
+        return True
+
+    def get_status(self) -> JobStatusV1Response:
+        """
+        Retrieve the current status of the job.
+
+        Returns
+        -------
+        JobStatusV1Response
+        """
+        return self._client.get_status(self._job_id)
+
+    def start(self) -> JobStatusV1Response:
+        """
+        Start the speech-to-text-translate job processing.
+
+        Returns
+        -------
+        JobStatusV1Response
+        """
+        return self._client.start(job_id=self._job_id)
+
+    def exists(self) -> bool:
+        """
+        Check if the job exists in the system.
+
+        Returns
+        -------
+        bool
+        """
+        try:
+            self.get_status()
+            return True
+        except httpx.HTTPStatusError:
+            return False
+
+    def is_complete(self) -> bool:
+        """
+        Check if the job is either completed or failed.
+
+        Returns
+        -------
+        bool
+        """
+        return self.get_status().job_state.lower() in {"completed", "failed"}
+
+    def is_successful(self) -> bool:
+        """
+        Check if the job completed successfully.
+
+        Returns
+        -------
+        bool
+        """
+        return self.get_status().job_state.lower() == "completed"
+
+    def is_failed(self) -> bool:
+        """
+        Check if the job has failed.
+
+        Returns
+        -------
+        bool
+        """
+        return self.get_status().job_state.lower() == "failed"

{sarvamai-0.1.13a2.dist-info → sarvamai-0.1.15.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: sarvamai
-Version: 0.1.13a2
+Version: 0.1.15
 Summary: 
 Requires-Python: >=3.8,<4.0
 Classifier: Intended Audience :: Developers

{sarvamai-0.1.13a2.dist-info → sarvamai-0.1.15.dist-info}/RECORD

@@ -5,7 +5,7 @@ sarvamai/chat/raw_client.py,sha256=A2kRuZcVWlJhyYCD7YKgqNkZEp3cYa1731KhRkhirU0,1
 sarvamai/client.py,sha256=J30X_os1lPf8Wml0KDFEf6p8VGHhgF_lf3nw1T2D3qo,8207
 sarvamai/core/__init__.py,sha256=YE2CtXeASe1RAbaI39twKWYKCuT4tW5is9HWHhJjR_g,1653
 sarvamai/core/api_error.py,sha256=44vPoTyWN59gonCIZMdzw7M1uspygiLnr3GNFOoVL2Q,614
-sarvamai/core/client_wrapper.py,sha256=FC_3Kep5NGzWOSiwBhHJ1vKfR3_Pqm6CZ7kWWh_r5AI,2570
+sarvamai/core/client_wrapper.py,sha256=PqRROSSVXjzMUxkf9c2QHUsL_UOhEa2AJ6nE9vcLb1o,2566
 sarvamai/core/datetime_utils.py,sha256=nBys2IsYrhPdszxGKCNRPSOCwa-5DWOHG95FB8G9PKo,1047
 sarvamai/core/events.py,sha256=HvKBdSoYcFetk7cgNXb7FxuY-FtY8NtUhZIN7mGVx8U,1159
 sarvamai/core/file.py,sha256=d4NNbX8XvXP32z8KpK2Xovv33nFfruIrpz0QWxlgpZk,2663
@@ -84,10 +84,11 @@ sarvamai/requests/transcription_metrics.py,sha256=FDclX2Z9Z3azrDXxtZW8xbkxxWMZQX
 sarvamai/requests/translation_response.py,sha256=8iwQeZB1purHY757bIQI-n9QeVRBItaAVcBJ_la-k1Y,414
 sarvamai/requests/transliteration_response.py,sha256=KqRkqnegLmt7LjdVxjRePX6RoqaLm64KFGZ6q7mXyfw,426
 sarvamai/speech_to_text/__init__.py,sha256=_VhToAyIt_5axN6CLJwtxg3-CO7THa_23pbUzqhXJa4,85
-sarvamai/speech_to_text/client.py,sha256=lp2G2fI9SUbeOBBE1S5tjcp-Xb8wIhAuVadLKwXveh8,11003
-sarvamai/speech_to_text/raw_client.py,sha256=A_56vEVeJdyttVJRiFxTMJ4n-s4l_PS8rI1DiLZlOmc,25331
+sarvamai/speech_to_text/client.py,sha256=v9p6bqRbA9h3f4wuylPTVDwetG7MCOlEtmrqW18KHCU,10857
+sarvamai/speech_to_text/raw_client.py,sha256=IhflpQX1P2kvZw8L2qbneMbRs5gtr9ye6ebvEK-hn3A,25185
 sarvamai/speech_to_text_job/__init__.py,sha256=_VhToAyIt_5axN6CLJwtxg3-CO7THa_23pbUzqhXJa4,85
-sarvamai/speech_to_text_job/client.py,sha256=MeX6V_34pbeWczBdsiT3VmAp8O4e4QmNwXNR1ue6s7A,12577
+sarvamai/speech_to_text_job/client.py,sha256=WSGBJxYcNxl77Zd1X6VVWjg4zshqecXf6WCyhfLXVlI,18007
+sarvamai/speech_to_text_job/job.py,sha256=K8HOmwrYd6l82-MZfWDBmNkZeeERyg9YOihnFfvl-Js,15021
 sarvamai/speech_to_text_job/raw_client.py,sha256=OZTPzMhAn-ckE_xKzfZ9QLsEX5EZVOJS0Pf-PBa19jM,48200
 sarvamai/speech_to_text_streaming/__init__.py,sha256=q7QygMmZCHJ-4FMhhL_6XNV_dsqlIFRCO1iSxoyxaaY,437
 sarvamai/speech_to_text_streaming/client.py,sha256=AzStfZDXhu2YAJEpnVbsy0WObub5ctlGBzqfeYOUlpA,8442
@@ -99,7 +100,8 @@ sarvamai/speech_to_text_streaming/types/speech_to_text_streaming_language_code.p
 sarvamai/speech_to_text_streaming/types/speech_to_text_streaming_model.py,sha256=b6F4ymgz4got6KVDqrweYvkET8itze63wUwWyjqDlO4,180
 sarvamai/speech_to_text_streaming/types/speech_to_text_streaming_vad_signals.py,sha256=8wiFOB7WDMbYCcMTYgNFJaIjEytYeXpJLwr_O_mH0TI,172
 sarvamai/speech_to_text_translate_job/__init__.py,sha256=_VhToAyIt_5axN6CLJwtxg3-CO7THa_23pbUzqhXJa4,85
-sarvamai/speech_to_text_translate_job/client.py,sha256=LxmP_T_KsPtI4Xg25RqAVN0Nn1c0V1VLZecZzngiCJ0,13674
+sarvamai/speech_to_text_translate_job/client.py,sha256=xu8kYtCESDB7LzL8YKBUq5qhTPMIl3_H3XD2L_7y4UU,18969
+sarvamai/speech_to_text_translate_job/job.py,sha256=DU4k3eB28V8N16M_QEchakVng4IOul6_Qrdn3FumgHA,15208
 sarvamai/speech_to_text_translate_job/raw_client.py,sha256=dAitbu2B9afPK6iT9zNjUJnE5BIr5-lrAlwrfwFxdkU,49507
 sarvamai/speech_to_text_translate_streaming/__init__.py,sha256=_hmlce1Zs1grylysZhBUdtKfkaUROwVydtwz6l-1qqg,411
 sarvamai/speech_to_text_translate_streaming/client.py,sha256=xPPg5_JgpH8tYDUte6FGtpzXO2LGBUSRADN-ICqqA6U,8286
@@ -207,6 +209,6 @@ sarvamai/types/transliterate_mode.py,sha256=1jSEMlGcoLkWuk12TgoOpSgwifa4rThGKZ1h
 sarvamai/types/transliterate_source_language.py,sha256=bSY9wJszF0sg-Cgg6F-YcWC8ly1mIlj9rqa15-jBtx8,283
 sarvamai/types/transliteration_response.py,sha256=yt-lzTbDeJ_ZL4I8kQa6oESxA9ebeJJY7LfFHpdEsmM,815
 sarvamai/version.py,sha256=Qkp3Ee9YH-O9RTix90e0i7iNrFAGN-QDt2AFwGA4n8k,75
-sarvamai-0.1.13a2.dist-info/METADATA,sha256=-i6GJmDknM_EKjKfdo2vWiwx0ipLvO5tmR6K8Sh9soY,26753
-sarvamai-0.1.13a2.dist-info/WHEEL,sha256=Zb28QaM1gQi8f4VCBhsUklF61CTlNYfs9YAZn-TOGFk,88
-sarvamai-0.1.13a2.dist-info/RECORD,,
+sarvamai-0.1.15.dist-info/METADATA,sha256=Y_dnyHQdRGnBxh2KGV6woOtFQDPOX4XeUIykiUSuYPg,26751
+sarvamai-0.1.15.dist-info/WHEEL,sha256=Zb28QaM1gQi8f4VCBhsUklF61CTlNYfs9YAZn-TOGFk,88
+sarvamai-0.1.15.dist-info/RECORD,,