phonexia-enhanced-speech-to-text-built-on-whisper-client 1.2.1__tar.gz

phonexia_enhanced_speech_to_text_built_on_whisper_client-1.2.1/PKG-INFO

@@ -0,0 +1,33 @@
+Metadata-Version: 2.1
+Name: phonexia-enhanced-speech-to-text-built-on-whisper-client
+Version: 1.2.1
+Summary: Client for communication with Phonexia Enhanced Speech To Text Built On Whisper microservice.
+Keywords: grpc,transcription,STT,ASR,speech to text,speech,language,microservice
+Author: Phonexia
+Author-email: info@phonexia.com
+Requires-Python: >=3.8,<4.0
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.8
+Classifier: Programming Language :: Python :: 3.9
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Requires-Dist: grpcio (>=1.54.0,<2.0.0)
+Requires-Dist: phonexia-grpc (>=1.0.0,<2.0.0)
+Project-URL: Homepage, https://phonexia.com
+Project-URL: Issues, https://phonexia.atlassian.net/servicedesk/customer/portal/15/group/20/create/40
+Project-URL: protofiles, https://github.com/phonexia/protofiles
+Description-Content-Type: text/markdown
+
+
+![](https://www.phonexia.com/wp-content/uploads/PHX_logotype_basic_2016_positive_transparent_RGB.png)
+
+# Phonexia enhanced speech to text built on whisper client
+
+This module contains client for communication with [enhanced speech to text built on whisper](https://hub.docker.com/r/phonexia/enhanced-speech-to-text-built-on-whisper/) developed by [Phonexia](https://phonexia.com).
+
+To use this client you will first need a running instance of any *Phonexia enhanced speech to text built on whisper microservice*. If you don't yet have any running instance, don't hesitate to [contact our sales department](mailto:info@phonexia.com).
+
+You can learn more about the enhanced speech to text built on whisper technology [here](https://docs.cloud.phonexia.com/docs/category/enhanced-speech-to-text-built-on-whisper).
+
+On [this page](https://docs.cloud.phonexia.com/docs/products/speech-platform-4/grpc/api/phonexia/grpc/technologies/enhanced_speech_to_text_built_on_whisper/v1/enhanced_speech_to_text_built_on_whisper.proto), you will find a *gRPC API* reference for *enhanced speech to text built on whisper microservice*.

phonexia_enhanced_speech_to_text_built_on_whisper_client-1.2.1/phonexia_enhanced_speech_to_text_built_on_whisper_client.py

@@ -0,0 +1,252 @@
+import argparse
+import logging
+import os
+from enum import Enum
+from typing import Iterator, Optional
+
+import google.protobuf.duration_pb2
+import grpc
+import phonexia.grpc.technologies.enhanced_speech_to_text_built_on_whisper.v1.enhanced_speech_to_text_built_on_whisper_pb2_grpc as stt_grpc
+from phonexia.grpc.common.core_pb2 import Audio, TimeRange
+from phonexia.grpc.technologies.enhanced_speech_to_text_built_on_whisper.v1.enhanced_speech_to_text_built_on_whisper_pb2 import (
+    TranscribeConfig,
+    TranscribeRequest,
+    TranslateConfig,
+    TranslateRequest,
+)
+
+CHUNK_SIZE = 32000
+
+
+class Task(Enum):
+    transcribe = "transcribe"
+    translate = "translate"
+
+    def __str__(self):
+        return self.value
+
+
+def time_to_duration(time: float) -> google.protobuf.duration_pb2.Duration | None:
+    if time is None:
+        return None
+    duration = google.protobuf.duration_pb2.Duration()
+    duration.seconds = int(time)
+    duration.nanos = int((time - duration.seconds) * 1e9)
+    return duration
+
+
+def transcribe_request_iterator(
+    file: str,
+    specified_language: Optional[str],
+    start: Optional[float],
+    end: Optional[float],
+    enable_language_switching: bool = False,
+) -> Iterator[TranscribeRequest]:
+    time_range = TimeRange(start=time_to_duration(start), end=time_to_duration(end))
+    config = TranscribeConfig(
+        language=specified_language, enable_language_switching=enable_language_switching
+    )
+
+    with open(file, "rb") as f:
+        while chunk := f.read(CHUNK_SIZE):
+            yield TranscribeRequest(
+                audio=Audio(content=chunk, time_range=time_range), config=config
+            )
+            time_range = None
+            config = None
+
+
+def translate_request_iterator(
+    file: str,
+    specified_language: Optional[str],
+    start: Optional[float],
+    end: Optional[float],
+    enable_language_switching: bool = False,
+) -> Iterator[TranslateRequest]:
+    time_range = TimeRange(start=time_to_duration(start), end=time_to_duration(end))
+    config = TranslateConfig(
+        source_language=specified_language, enable_language_switching=enable_language_switching
+    )
+
+    with open(file, "rb") as f:
+        while chunk := f.read(CHUNK_SIZE):
+            yield TranslateRequest(audio=Audio(content=chunk, time_range=time_range), config=config)
+            time_range = None
+            config = None
+
+
+def transcribe(
+    channel: grpc.Channel,
+    file: str,
+    language: Optional[str],
+    start: Optional[float],
+    end: Optional[float],
+    metadata: Optional[list],
+    task: Task,
+    enable_language_switching: bool = False,
+):
+    stub = stt_grpc.SpeechToTextStub(channel)
+    if task == Task.transcribe:
+        response = stub.Transcribe(
+            transcribe_request_iterator(
+                file=file,
+                specified_language=language,
+                start=start,
+                end=end,
+                enable_language_switching=enable_language_switching,
+            ),
+            metadata=metadata,
+        )
+    elif task == Task.translate:
+        response = stub.Translate(
+            translate_request_iterator(
+                file=file,
+                specified_language=language,
+                start=start,
+                end=end,
+                enable_language_switching=enable_language_switching,
+            ),
+            metadata=metadata,
+        )
+    else:
+        raise RuntimeError("Unknown task")
+
+    warning_message = []
+    for _response in response:
+        for segment in _response.result.one_best.segments:
+            print(
+                f"[{segment.start_time.ToJsonString()} -> {segment.end_time.ToJsonString()} "
+                + (
+                    f"{segment.language}"
+                    if task == Task.transcribe
+                    else f"{segment.source_language} -> {segment.language}"
+                )
+                + f"] {segment.text}"
+            )
+            if segment.source_language != segment.detected_source_language:
+                warning_message.append(
+                    f"Language '{segment.detected_source_language}' was detected, but the license does not support this language. "
+                    f"Instead the segment was {'transcribed' if task == Task.transcribe else 'translated'} with the "
+                    + (
+                        f"closest available source language '{segment.source_language}'"
+                        if language is None
+                        else f"language '{language}' that was enforced by argument '--language'"
+                    )
+                    + ". For more info on this problem don't hesitate to contact Phonexia."
+                )
+        if _response.HasField("processed_audio_length"):
+            print(f"Processed audio length: {_response.processed_audio_length.ToJsonString()}")
+
+    warning_message = set(warning_message)
+    if len(warning_message) > 0:
+        print()
+        for warning in warning_message:
+            print(f"WARNING: {warning}")
+
+
+def main():
+    parser = argparse.ArgumentParser(
+        description=(
+            "Enhanced Speech to Text Built on Whisper gRPC client. Transcribes input audio into segments"
+            " with timestamps."
+        )
+    )
+
+    parser.add_argument(
+        "-H",
+        "--host",
+        type=str,
+        default="localhost:8080",
+        help="Server address, default: localhost:8080",
+    )
+    parser.add_argument(
+        "-l",
+        "--log_level",
+        type=str,
+        default="error",
+        choices=["critical", "error", "warning", "info", "debug"],
+    )
+    parser.add_argument(
+        "--metadata",
+        metavar="key=value",
+        nargs="+",
+        type=lambda x: tuple(x.split("=")),
+        help="Custom client metadata",
+    )
+    parser.add_argument("--use_ssl", action="store_true", help="Use SSL connection")
+    parser.add_argument("--start", type=float, help="Audio start time")
+    parser.add_argument("--end", type=float, help="Audio end time")
+
+    parser.add_argument(
+        "--language",
+        type=str,
+        default=None,
+        help=(
+            "Force transcription to specified language, if not set, language is detected"
+            " automatically"
+        ),
+    )
+    parser.add_argument(
+        "--task",
+        type=Task,
+        default=Task.transcribe,
+        choices=list(Task),
+        help="Select whether to transcribe or translate the recording",
+    )
+    parser.add_argument(
+        "--enable-language-switching",
+        action="store_true",
+        help="Enable dynamic language switching during transcription, with the language being detected approximately every 30 seconds",
+    )
+    parser.add_argument("file", type=str, help="Path to input file")
+
+    args = parser.parse_args()
+
+    if args.start is not None and args.start < 0:
+        raise ValueError("Parameter 'start' must be a non-negative float.")
+
+    if args.end is not None and args.end <= 0:
+        raise ValueError("Parameter 'end' must be a positive float.")
+
+    if args.start is not None and args.end is not None and args.start >= args.end:
+        raise ValueError("Parameter 'end' must be larger than 'start'.")
+
+    logging.basicConfig(
+        level=args.log_level.upper(),
+        format="[%(asctime)s.%(msecs)03d] [%(levelname)s] %(message)s",
+        datefmt="%Y-%m-%d %H:%M:%S",
+    )
+
+    if not os.path.isfile(args.file):
+        logging.error(f"no such file {args.file}")
+        exit(1)
+
+    try:
+        logging.info(f"Connecting to {args.host}")
+        channel = (
+            grpc.secure_channel(target=args.host, credentials=grpc.ssl_channel_credentials())
+            if args.use_ssl
+            else grpc.insecure_channel(target=args.host)
+        )
+
+        transcribe(
+            channel=channel,
+            file=args.file,
+            language=args.language,
+            start=args.start,
+            end=args.end,
+            metadata=args.metadata,
+            task=args.task,
+            enable_language_switching=args.enable_language_switching,
+        )
+
+    except grpc.RpcError:
+        logging.exception("RPC failed")
+        exit(1)
+    except Exception:
+        logging.exception("Unknown error")
+        exit(1)
+
+
+if __name__ == "__main__":
+    main()

phonexia_enhanced_speech_to_text_built_on_whisper_client-1.2.1/pypi-README.md

@@ -0,0 +1,12 @@
+
+![](https://www.phonexia.com/wp-content/uploads/PHX_logotype_basic_2016_positive_transparent_RGB.png)
+
+# Phonexia enhanced speech to text built on whisper client
+
+This module contains client for communication with [enhanced speech to text built on whisper](https://hub.docker.com/r/phonexia/enhanced-speech-to-text-built-on-whisper/) developed by [Phonexia](https://phonexia.com).
+
+To use this client you will first need a running instance of any *Phonexia enhanced speech to text built on whisper microservice*. If you don't yet have any running instance, don't hesitate to [contact our sales department](mailto:info@phonexia.com).
+
+You can learn more about the enhanced speech to text built on whisper technology [here](https://docs.cloud.phonexia.com/docs/category/enhanced-speech-to-text-built-on-whisper).
+
+On [this page](https://docs.cloud.phonexia.com/docs/products/speech-platform-4/grpc/api/phonexia/grpc/technologies/enhanced_speech_to_text_built_on_whisper/v1/enhanced_speech_to_text_built_on_whisper.proto), you will find a *gRPC API* reference for *enhanced speech to text built on whisper microservice*.

phonexia_enhanced_speech_to_text_built_on_whisper_client-1.2.1/pyproject.toml

@@ -0,0 +1,92 @@
+[tool.poetry]
+name = "phonexia-enhanced-speech-to-text-built-on-whisper-client"
+version = "1.2.1"
+description = "Client for communication with Phonexia Enhanced Speech To Text Built On Whisper microservice."
+readme = "pypi-README.md"
+keywords = ["grpc", "transcription", "STT", "ASR", "speech to text", "speech", "language", "microservice"]
+authors = ["Phonexia <info@phonexia.com>"]
+
+[tool.poetry.urls]
+Homepage = "https://phonexia.com"
+Issues = "https://phonexia.atlassian.net/servicedesk/customer/portal/15/group/20/create/40"
+protofiles = "https://github.com/phonexia/protofiles"
+
+[tool.poetry.scripts]
+enhanced_speech_to_text_built_on_whisper_client = 'phonexia_enhanced_speech_to_text_built_on_whisper_client:main'
+
+[tool.poetry.dependencies]
+python = ">=3.8,<4.0"
+grpcio = "^1.54.0"
+phonexia-grpc = {version="^1.0.0", source="pypi"}
+
+[tool.poetry.group.dev.dependencies]
+pytest = "^8.0.0"
+pytest-cov = "^5.0.0"
+pytest-env = "^1.0.0"
+pytest-random-order = "^1.1.0"
+black = "^24.0.0"
+ruff = "^0.4.0"
+
+[[tool.poetry.source]]
+name = "gitlab"
+url = "https://gitlab.cloud.phonexia.com/api/v4/groups/39/-/packages/pypi/simple"
+priority = "primary"
+
+
+[[tool.poetry.source]]
+name = "PyPI"
+priority = "default"
+
+[build-system]
+requires = ["poetry-core>=1.0.0"]
+build-backend = "poetry.core.masonry.api"
+
+[tool.black]
+line-length = 100
+target-version = ['py38']
+preview = true
+
+[tool.ruff]
+target-version = "py38"
+line-length = 100
+fix = true
+select = [
+    # flake8-2020
+    "YTT",
+    # flake8-bandit
+    "S",
+    # flake8-bugbear
+    "B",
+    # flake8-builtins
+    "A",
+    # flake8-comprehensions
+    "C4",
+    # flake8-debugger
+    "T10",
+    # flake8-simplify
+    "SIM",
+    # isort
+    "I",
+    # mccabe
+    "C90",
+    # pycodestyle
+    "E", "W",
+    # pyflakes
+    "F",
+    # pygrep-hooks
+    "PGH",
+    # pyupgrade
+    "UP",
+    # ruff
+    "RUF",
+    # tryceratops
+    "TRY",
+]
+ignore = [
+    # LineTooLong
+    "E501",
+    # DoNotAssignLambda
+    "E731",
+    # RaiseVanillaArgs aka Avoid specifying long messages outside the exception class
+    "TRY003",
+]