nv-ingest-api 2025.4.17.dev20250417__py3-none-any.whl → 2025.4.19.dev20250419__py3-none-any.whl

nv_ingest_api/util/service_clients/rest/rest_client.py

@@ -1,398 +0,0 @@
-# SPDX-FileCopyrightText: Copyright (c) 2024-25, NVIDIA CORPORATION & AFFILIATES.
-# All rights reserved.
-# SPDX-License-Identifier: Apache-2.0
-
-# pylint: skip-file
-
-import logging
-import re
-import time
-from typing import Any
-
-import httpx
-import requests
-
-from nv_ingest_api.internal.schemas.message_brokers.response_schema import ResponseSchema
-from nv_ingest_api.util.service_clients.client_base import MessageBrokerClientBase
-
-logger = logging.getLogger(__name__)
-
-# HTTP Response Statuses that result in marking submission as failed
-# 4XX - Any 4XX status is considered a client derived error and will result in failure
-# 5XX - Not all 500's are terminal but most are. Those which are listed below
-_TERMINAL_RESPONSE_STATUSES = [
-    400,
-    401,
-    402,
-    403,
-    404,
-    405,
-    406,
-    407,
-    408,
-    409,
-    410,
-    411,
-    412,
-    413,
-    414,
-    415,
-    416,
-    417,
-    418,
-    421,
-    422,
-    423,
-    424,
-    425,
-    426,
-    428,
-    429,
-    431,
-    451,
-    500,
-    501,
-    503,
-    505,
-    506,
-    507,
-    508,
-    510,
-    511,
-]
-
-
-class RestClient(MessageBrokerClientBase):
-    """
-    A client for interfacing with the nv-ingest HTTP endpoint, providing mechanisms for sending and receiving messages
-    with retry logic and connection management.
-
-    Parameters
-    ----------
-    host : str
-        The hostname of the HTTP server.
-    port : int
-        The port number of the HTTP server.
-    max_retries : int, optional
-        The maximum number of retry attempts for operations. Default is 0 (no retries).
-    max_backoff : int, optional
-        The maximum backoff delay between retries in seconds. Default is 32 seconds.
-    connection_timeout : int, optional
-        The timeout in seconds for connecting to the HTTP server. Default is 300 seconds.
-    http_allocator : Any, optional
-        The HTTP client allocator.
-
-    Attributes
-    ----------
-    client : Any
-        The HTTP client instance used for operations.
-    """
-
-    def __init__(
-        self,
-        host: str,
-        port: int,
-        max_retries: int = 0,
-        max_backoff: int = 32,
-        connection_timeout: int = 300,
-        http_allocator: Any = httpx.AsyncClient,
-        **kwargs,
-    ):
-        self._host = host
-        self._port = port
-        self._max_retries = max_retries
-        self._max_backoff = max_backoff
-        self._connection_timeout = connection_timeout
-        self._http_allocator = http_allocator
-        self._client = self._http_allocator()
-        self._retries = 0
-
-        self._submit_endpoint = "/v1/submit_job"
-        self._fetch_endpoint = "/v1/fetch_job"
-
-        if "base_url" in kwargs:
-            logger.debug("Using custom base_url; ignoring host and port")
-
-        self._base_url = kwargs.get("base_url") or self.generate_url(self._host, self._port)
-        self._headers = kwargs.get("headers", {})
-        self._auth = kwargs.get("auth", None)
-
-    def _connect(self) -> None:
-        """
-        Attempts to reconnect to the HTTP server if the current connection is not responsive.
-        """
-        ping_result = self.ping()
-
-        if ping_result.response_code != 0:
-            logger.debug("Reconnecting to HTTP server")
-            self._client = self._http_allocator()
-
-    @property
-    def max_retries(self) -> int:
-        return self._max_retries
-
-    @max_retries.setter
-    def max_retries(self, value: int) -> None:
-        self._max_retries = value
-
-    def get_client(self) -> Any:
-        """
-        Returns a HTTP client instance, reconnecting if necessary.
-
-        Returns
-        -------
-        Any
-            The HTTP client instance.
-        """
-        if self._client is None:
-            self._connect()
-        return self._client
-
-    def ping(self) -> ResponseSchema:
-        """
-        Checks if the HTTP server is responsive.
-
-        Returns
-        -------
-        bool
-            True if the server responds to a ping, False otherwise.
-        """
-        try:
-            # Implement a simple GET request to a health endpoint or root
-            self._client.ping()
-            return ResponseSchema(response_code=0)
-        except (httpx.HTTPError, AttributeError):
-            return ResponseSchema(response_code=1, response_reason="Failed to ping HTTP server")
-
-    @staticmethod
-    def generate_url(user_provided_url, user_provided_port) -> str:
-        """Examines the user defined URL for http*://. If that
-        pattern is detected the URL is used as provided by the user.
-        If that pattern does not exist then the assumption is made that
-        the endpoint is simply `http://` and that is prepended
-        to the user supplied endpoint.
-
-        Args:
-            user_provided_url str: Endpoint where the Rest service is running
-
-        Returns:
-            str: Fully validated URL
-        """
-        if not re.match(r"^https?://", user_provided_url):
-            # Add the default `http://` if it's not already present in the URL
-            user_provided_url = f"http://{user_provided_url}:{user_provided_port}"
-        else:
-            user_provided_url = f"{user_provided_url}:{user_provided_port}"
-        return user_provided_url
-
-    def fetch_message(self, job_id: str, timeout: float = (10, 600)) -> ResponseSchema:
-        """
-        Fetches a message from the specified queue with retries on failure, handling streaming HTTP responses.
-
-        Parameters
-        ----------
-        job_id : str
-            The server-side job identifier.
-        timeout : float
-            The timeout in seconds for blocking until a message is available.
-
-        Returns
-        -------
-        ResponseSchema
-            The fetched message wrapped in a ResponseSchema object.
-        """
-        retries = 0
-        url = f"{self._base_url}{self._fetch_endpoint}/{job_id}"
-
-        # Ensure headers are included
-        headers = {"Content-Type": "application/json"}
-        headers.update(self._headers)
-
-        while True:
-            try:
-                logger.debug(f"Invoking fetch_message http endpoint @ '{url}'")
-
-                # Fetch using streaming response
-                with requests.get(
-                    url,
-                    timeout=(30, 600),
-                    stream=True,
-                    headers=headers,
-                    auth=self._auth,
-                ) as result:
-                    response_code = result.status_code
-
-                    if response_code in _TERMINAL_RESPONSE_STATUSES:
-                        # Terminal response code; return error ResponseSchema
-                        return ResponseSchema(
-                            response_code=1,
-                            response_reason=(
-                                f"Terminal response code {response_code} received when fetching JobSpec: {job_id}"
-                            ),
-                            response=result.text,
-                        )
-
-                    if response_code == 200:
-                        # Handle streaming response, reconstructing payload incrementally
-                        response_chunks = []
-                        for chunk in result.iter_content(chunk_size=1024 * 1024):  # 1MB chunks
-                            if chunk:
-                                response_chunks.append(chunk)
-                        full_response = b"".join(response_chunks).decode("utf-8")
-
-                        return ResponseSchema(
-                            response_code=0,
-                            response_reason="OK",
-                            response=full_response,
-                        )
-
-                    elif response_code == 202:
-                        # Job is not ready yet
-                        return ResponseSchema(
-                            response_code=1,
-                            response_reason="Job is not ready yet. Retry later.",
-                        )
-
-                    else:
-                        try:
-                            # Retry the operation
-                            retries = self.perform_retry_backoff(retries)
-                        except RuntimeError as rte:
-                            raise rte
-
-            except (ConnectionError, requests.HTTPError, requests.exceptions.ConnectionError) as err:
-                logger.error(f"Error during fetching, retrying... Error: {err}")
-                self._client = None  # Invalidate client to force reconnection
-                if "Connection refused" in str(err):
-                    logger.debug(
-                        "Connection refused encountered during fetch; sleeping for 10 seconds before retrying."
-                    )
-                    time.sleep(10)
-                try:
-                    retries = self.perform_retry_backoff(retries)
-                except RuntimeError as rte:
-                    # Max retries reached
-                    return ResponseSchema(response_code=1, response_reason=str(rte), response=str(err))
-                except TimeoutError:
-                    raise
-            except Exception as e:
-                # Handle non-http specific exceptions
-                logger.error(f"Unexpected error during fetch from {url}: {e}")
-                return ResponseSchema(
-                    response_code=1, response_reason=f"Unexpected error during fetch: {e}", response=None
-                )
-
-    def submit_message(self, channel_name: str, message: str, for_nv_ingest: bool = False) -> ResponseSchema:
-        """
-        Submits a JobSpec to a specified HTTP endpoint with retries on failure.
-
-        Parameters
-        ----------
-        channel_name : str
-            Not used as part of RestClient but defined in MessageClientBase.
-        message : str
-            The message to submit.
-        for_nv_ingest : bool
-            Not used as part of RestClient but defined in MessageClientBase.
-
-        Returns
-        -------
-        ResponseSchema
-            The response from the server wrapped in a ResponseSchema object.
-        """
-        retries = 0
-        url = f"{self._base_url}{self._submit_endpoint}"
-
-        # Ensure content-type is present
-        headers = {"Content-Type": "application/json"}
-        headers.update(self._headers)
-
-        while True:
-            try:
-                # Submit via HTTP
-                result = requests.post(
-                    url,
-                    json={"payload": message},
-                    headers=headers,
-                    auth=self._auth,
-                    timeout=self._connection_timeout,
-                )
-
-                response_code = result.status_code
-                if response_code in _TERMINAL_RESPONSE_STATUSES:
-                    # Terminal response code; return error ResponseSchema
-                    return ResponseSchema(
-                        response_code=1,
-                        response_reason=f"Terminal response code {response_code} received when submitting JobSpec",
-                        trace_id=result.headers.get("x-trace-id"),
-                    )
-                else:
-                    # If 200 we are good, otherwise let's try again
-                    if response_code == 200:
-                        logger.debug(f"JobSpec successfully submitted to http endpoint {self._submit_endpoint}")
-                        # The REST interface returns a JobId, so we capture that here
-                        x_trace_id = result.headers.get("x-trace-id")
-                        return ResponseSchema(
-                            response_code=0,
-                            response_reason="OK",
-                            response=result.text,
-                            transaction_id=result.text,
-                            trace_id=x_trace_id,
-                        )
-                    else:
-                        # Retry the operation
-                        retries = self.perform_retry_backoff(retries)
-            except requests.RequestException as e:
-                logger.error(f"Failed to submit job, retrying... Error: {e}")
-                self._client = None  # Invalidate client to force reconnection
-                if "Connection refused" in str(e):
-                    logger.debug(
-                        "Connection refused encountered during submission; sleeping for 10 seconds before retrying."
-                    )
-                    time.sleep(10)
-                try:
-                    retries = self.perform_retry_backoff(retries)
-                except RuntimeError as rte:
-                    # Max retries reached
-                    return ResponseSchema(response_code=1, response_reason=str(rte), response=str(e))
-            except Exception as e:
-                # Handle non-http specific exceptions
-                logger.error(f"Unexpected error during submission of JobSpec to {url}: {e}")
-                return ResponseSchema(
-                    response_code=1, response_reason=f"Unexpected error during JobSpec submission: {e}", response=None
-                )
-
-    def perform_retry_backoff(self, existing_retries) -> int:
-        """
-        Attempts to perform a backoff retry delay. This function accepts the
-        current number of retries that have been attempted and compares
-        that with the maximum number of retries allowed. If the current
-        number of retries exceeds the max then a RuntimeError is raised.
-
-        Parameters
-        ----------
-        existing_retries : int
-            The number of retries that have been attempted for this operation thus far
-
-        Returns
-        -------
-        int
-            The updated number of retry attempts that have been made for this operation
-
-        Raises
-        ------
-        RuntimeError
-            Raised if the maximum number of retry attempts has been reached.
-        """
-        backoff_delay = min(2**existing_retries, self._max_backoff)
-        logger.debug(
-            f"Retry #: {existing_retries} of max_retries: {self.max_retries} | "
-            f"current backoff_delay: {backoff_delay}s of max_backoff: {self._max_backoff}s"
-        )
-
-        if self.max_retries > 0 and existing_retries < self.max_retries:
-            logger.error(f"Operation failed, retrying in {backoff_delay}s...")
-            time.sleep(backoff_delay)
-            return existing_retries + 1
-        else:
-            raise RuntimeError(f"Max retry attempts of {self.max_retries} reached")

nv_ingest_api/util/string_processing/__init__.py

@@ -1,51 +0,0 @@
-# SPDX-FileCopyrightText: Copyright (c) 2024, NVIDIA CORPORATION & AFFILIATES.
-# All rights reserved.
-# SPDX-License-Identifier: Apache-2.0
-
-import logging
-import re
-
-logger = logging.getLogger(__name__)
-
-DEPLOT_MAX_TOKENS = 128
-DEPLOT_TEMPERATURE = 1.0
-DEPLOT_TOP_P = 1.0
-
-
-def remove_url_endpoints(url) -> str:
-    """Some configurations provide the full endpoint in the URL.
-    Ex: http://deplot:8000/v1/chat/completions. For hitting the
-    health endpoint we need to get just the hostname:port combo
-    that we can append the health/ready endpoint to so we attempt
-    to parse that information here.
-
-    Args:
-        url str: Incoming URL
-
-    Returns:
-        str: URL with just the hostname:port portion remaining
-    """
-    if "/v1" in url:
-        url = url.split("/v1")[0]
-
-    return url
-
-
-def generate_url(url) -> str:
-    """Examines the user defined URL for http*://. If that
-    pattern is detected the URL is used as provided by the user.
-    If that pattern does not exist then the assumption is made that
-    the endpoint is simply `http://` and that is prepended
-    to the user supplied endpoint.
-
-    Args:
-        url str: Endpoint where the Rest service is running
-
-    Returns:
-        str: Fully validated URL
-    """
-    if not re.match(r"^https?://", url):
-        # Add the default `http://` if it's not already present in the URL
-        url = f"http://{url}"
-
-    return url

nv_ingest_api-2025.4.17.dev20250417.dist-info/RECORD

@@ -1,152 +0,0 @@
-nv_ingest_api/__init__.py,sha256=wQSlVx3T14ZgQAt-EPzEczQusXVW0W8yynnUaFFGE3s,143
-nv_ingest_api/interface/__init__.py,sha256=ltWlfmtCewHSRK4B7DF__QvlSUPuliz58JEcEIeIgI0,10134
-nv_ingest_api/interface/extract.py,sha256=GyBfXKKTGwSb-y0k0nMiTf4HcCT2E-lxLY4aMYAPeOI,38815
-nv_ingest_api/interface/mutate.py,sha256=eZkd3sbHEJQiEPJyMbhewlPxQNMnL_Xur15icclnb-U,5934
-nv_ingest_api/interface/store.py,sha256=aR3Cf19lq9Yo9AHlAy1VVcrOP2dgyN01yYhwxyTprkQ,8207
-nv_ingest_api/interface/transform.py,sha256=g6YnFR7TpEU0xNtzCvv6kqnFbuCwQ6vRMjjBxz3G4n4,15815
-nv_ingest_api/interface/utility.py,sha256=oXHV2Miz2BKviQg5vOVfiGSvPs2fKJsPDmnxe3fJL9c,7857
-nv_ingest_api/internal/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-nv_ingest_api/internal/enums/__init__.py,sha256=wQSlVx3T14ZgQAt-EPzEczQusXVW0W8yynnUaFFGE3s,143
-nv_ingest_api/internal/enums/common.py,sha256=HSj7qqNr6KXu_FIyK_Wvel24R-r8lV7dLA173z5XFBc,12321
-nv_ingest_api/internal/extract/__init__.py,sha256=uLsBITo_XfgbwpzqXUm1IYX6XlZrTfx6T1cIhdILwG8,140
-nv_ingest_api/internal/extract/audio/__init__.py,sha256=wQSlVx3T14ZgQAt-EPzEczQusXVW0W8yynnUaFFGE3s,143
-nv_ingest_api/internal/extract/audio/audio_extraction.py,sha256=L8cK7xB6QTaSx8gsrdyaYHYh0HpW6lycGfduCk7XSMg,5364
-nv_ingest_api/internal/extract/docx/__init__.py,sha256=HIHfzSig66GT0Uk8qsGBm_f13fKYcPtItBicRUWOOVA,183
-nv_ingest_api/internal/extract/docx/docx_extractor.py,sha256=lzZPSa-oHBmLk7ynop5aOLM2rVbYAbzSkPqVlAuT8RE,8319
-nv_ingest_api/internal/extract/docx/engines/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-nv_ingest_api/internal/extract/docx/engines/docxreader_helpers/__init__.py,sha256=uLsBITo_XfgbwpzqXUm1IYX6XlZrTfx6T1cIhdILwG8,140
-nv_ingest_api/internal/extract/docx/engines/docxreader_helpers/docx_helper.py,sha256=1wkciAxu8lz9WuPuoleJFy2s09ieSzXl1S71F9r0BWA,4385
-nv_ingest_api/internal/extract/docx/engines/docxreader_helpers/docxreader.py,sha256=CM2yV8lfEw1F1ORAjupD4gyIKX0PDDJrL3nsZ5Mnrgg,31539
-nv_ingest_api/internal/extract/image/__init__.py,sha256=wQSlVx3T14ZgQAt-EPzEczQusXVW0W8yynnUaFFGE3s,143
-nv_ingest_api/internal/extract/image/chart_extractor.py,sha256=Jy_fNmDbZcdni55Fq7vT6NdbYnCyGoyw0J7QjpK-KPc,13315
-nv_ingest_api/internal/extract/image/image_extractor.py,sha256=ocLvlVMzO9CQvduxbjupOeKxnt2aq1_CzJCqcdD-loo,8783
-nv_ingest_api/internal/extract/image/infographic_extractor.py,sha256=k4Z6JwsoNKsyfmpaQkN_dxJpAv9-RVsRL1BfSWUtXTM,8908
-nv_ingest_api/internal/extract/image/table_extractor.py,sha256=80FQef4Dsn6__MNIRCQzFf32s4wUyTOzBFgmA84JZJk,13133
-nv_ingest_api/internal/extract/image/image_helpers/__init__.py,sha256=wQSlVx3T14ZgQAt-EPzEczQusXVW0W8yynnUaFFGE3s,143
-nv_ingest_api/internal/extract/image/image_helpers/common.py,sha256=NU8TEU9p2aIL_KppyhtTgRUPqD4MsanxATG19rKhGjw,15032
-nv_ingest_api/internal/extract/pdf/__init__.py,sha256=wQSlVx3T14ZgQAt-EPzEczQusXVW0W8yynnUaFFGE3s,143
-nv_ingest_api/internal/extract/pdf/pdf_extractor.py,sha256=CxtWaD6mql9MEqSdk2CfSQ9T-Bn87beBkCOuGGjxGt8,2934
-nv_ingest_api/internal/extract/pdf/engines/__init__.py,sha256=u4GnAZmDKRl0RwYGIRiozIRw70Kybw3A72-lcKFeoTI,582
-nv_ingest_api/internal/extract/pdf/engines/adobe.py,sha256=VT0dEqkU-y2uGkaCqxtKYov_Q8R1028UQVBchgMLca4,17466
-nv_ingest_api/internal/extract/pdf/engines/llama.py,sha256=PpKTqS8jGHBV6mKLGZWwjpfT8ga6Fy8ffrvL-gPAf2c,8182
-nv_ingest_api/internal/extract/pdf/engines/nemoretriever.py,sha256=F8hZdqYRr0CTNeIRJIG6H__CCh_3GWQ4_ySCM0WPLPU,22913
-nv_ingest_api/internal/extract/pdf/engines/pdfium.py,sha256=jUcquCWbyQPNCHZLaV-XnVqUFsajX4YxVFCiWWwD4QQ,22367
-nv_ingest_api/internal/extract/pdf/engines/tika.py,sha256=6GyR2l6EsgNZl9jnYDXLeKNK9Fj2Mw9y2UWDq-eSkOc,3169
-nv_ingest_api/internal/extract/pdf/engines/unstructured_io.py,sha256=jrv2B4VZAH4PevAQrFz965qz8UyXq3rViiOTbGLejec,14908
-nv_ingest_api/internal/extract/pdf/engines/pdf_helpers/__init__.py,sha256=Jk3wrQ2CZs167juvEZ-uV6qXWQjR08hhIu8otk2MWj4,4931
-nv_ingest_api/internal/extract/pptx/__init__.py,sha256=HIHfzSig66GT0Uk8qsGBm_f13fKYcPtItBicRUWOOVA,183
-nv_ingest_api/internal/extract/pptx/pptx_extractor.py,sha256=15gU7NtTmTwr1ml679gABQABXI463ZKoqPOh31EK98s,7867
-nv_ingest_api/internal/extract/pptx/engines/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-nv_ingest_api/internal/extract/pptx/engines/pptx_helper.py,sha256=tmUXw4H35o6dMcsS73Q6L_zd-qDqwCshTGfCv_V610c,28435
-nv_ingest_api/internal/mutate/__init__.py,sha256=wQSlVx3T14ZgQAt-EPzEczQusXVW0W8yynnUaFFGE3s,143
-nv_ingest_api/internal/mutate/deduplicate.py,sha256=hmvTTGevpCtlkM_wVZSoc8-Exr6rUJwqLjoEnbPcPzY,3849
-nv_ingest_api/internal/mutate/filter.py,sha256=H-hOTBVP-zLpvQr-FoGIJKxkhtj4l_sZ9V2Fgu3rTEM,5183
-nv_ingest_api/internal/primitives/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-nv_ingest_api/internal/primitives/control_message_task.py,sha256=nWVB3QsP6p8BKwHThd-SNv_zwJAEA1mKCRharuju1mc,439
-nv_ingest_api/internal/primitives/ingest_control_message.py,sha256=rvipBiiUaHuRhupFCFDCG8rv0PylSJibCiJ7rDeb98A,8514
-nv_ingest_api/internal/primitives/nim/__init__.py,sha256=i_i_fBR2EcRCh2Y19DF6GM3s_Q0VPgo_thPnhEIJUyg,266
-nv_ingest_api/internal/primitives/nim/default_values.py,sha256=W92XjfyeC6uuVxut6J7p00x1kpNsnXIDb97gSVytZJk,380
-nv_ingest_api/internal/primitives/nim/nim_client.py,sha256=7C_t3BnYz_hL2H8RmvOShLCKlfYmwIREC6vnOnzOHWA,14483
-nv_ingest_api/internal/primitives/nim/nim_model_interface.py,sha256=wMEgoi79YQn_4338MVemkeZgM1J-vnz0aZWpvqDhib4,2392
-nv_ingest_api/internal/primitives/nim/model_interface/__init__.py,sha256=wQSlVx3T14ZgQAt-EPzEczQusXVW0W8yynnUaFFGE3s,143
-nv_ingest_api/internal/primitives/nim/model_interface/cached.py,sha256=b1HX-PY1ExW5V6pXC1ZiHdobeG_BmbPr3rBbVJef13s,11003
-nv_ingest_api/internal/primitives/nim/model_interface/decorators.py,sha256=qwubkHs4WjnexM6rI0wkjWCsrVNEbA4Wjk2oKL9OYCU,1499
-nv_ingest_api/internal/primitives/nim/model_interface/deplot.py,sha256=TvKdk6PTuI1WNhRmNNrvygaI_DIutkJkDL-XdtLZQac,10787
-nv_ingest_api/internal/primitives/nim/model_interface/helpers.py,sha256=x35a9AyTYxpESQflLo_YnhVOKblQKVen6vGGFaXmNiE,9927
-nv_ingest_api/internal/primitives/nim/model_interface/nemoretriever_parse.py,sha256=MFWPqMTXs_MZG3ripRR21o7f_mVeoE46Q10yvJ8KNr0,7023
-nv_ingest_api/internal/primitives/nim/model_interface/paddle.py,sha256=rSUPwl5XOrqneoS6aKhatVjrNBg_LhP3nwUWS_aTwz0,17950
-nv_ingest_api/internal/primitives/nim/model_interface/parakeet.py,sha256=OYg4AGki_wm--Np9VlSm0eZC-r54GbDOISbe9v0B9fw,12967
-nv_ingest_api/internal/primitives/nim/model_interface/text_embedding.py,sha256=oai0-moKDJOnOMfTaGQf-vo6qMRD6pbcf7_XRIt-oJ8,4934
-nv_ingest_api/internal/primitives/nim/model_interface/vlm.py,sha256=qJ382PU1ZrIM-SR3cqIhtY_W2rmHec2HIa2aUB2SvaU,6031
-nv_ingest_api/internal/primitives/nim/model_interface/yolox.py,sha256=exN0pKTBXd3pb5kKP96jinTYisgz1Y7EyWmWUuDNnCY,49312
-nv_ingest_api/internal/primitives/tracing/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-nv_ingest_api/internal/primitives/tracing/latency.py,sha256=5kVTeYRbRdTlT_aI4MeS20N_S7mqCcLqZR6YHtxhXkY,2215
-nv_ingest_api/internal/primitives/tracing/logging.py,sha256=SSzIgS7afLH-e1C7VagYDmkkA6rTXmQ-bmtLjoEguhg,3851
-nv_ingest_api/internal/primitives/tracing/tagging.py,sha256=RC-sF6-w8YBb74nZ2JjEnkCBAxBIamhIHdn1ksxDtRI,7577
-nv_ingest_api/internal/schemas/__init__.py,sha256=wQSlVx3T14ZgQAt-EPzEczQusXVW0W8yynnUaFFGE3s,143
-nv_ingest_api/internal/schemas/extract/__init__.py,sha256=wQSlVx3T14ZgQAt-EPzEczQusXVW0W8yynnUaFFGE3s,143
-nv_ingest_api/internal/schemas/extract/extract_audio_schema.py,sha256=VVppZgV1lnyJCTfADexzoj3V0lOSq3t6Dw_6VhIxZ7k,3771
-nv_ingest_api/internal/schemas/extract/extract_chart_schema.py,sha256=mNsv628oslNieU6KPUHw_Iwr4WohtK2dIHoVo2HnaEs,4302
-nv_ingest_api/internal/schemas/extract/extract_docx_schema.py,sha256=M2N7WjMNvSemHcJHWeNUD_kFG0wC5VE2W3K6SVrJqvA,3761
-nv_ingest_api/internal/schemas/extract/extract_image_schema.py,sha256=GC4xV8Z9TPLOuxlEtf2fbklSSp8ETGMrDpZgMQ02UwA,3766
-nv_ingest_api/internal/schemas/extract/extract_infographic_schema.py,sha256=_ptTrxN74tpasJ0aQZgaXEUYFe298PJGbGNk6gyeM94,3992
-nv_ingest_api/internal/schemas/extract/extract_pdf_schema.py,sha256=ZBCppSNmnr4jrPl2-R_j0RBw2L4ej_r0hVdFn02AG18,6569
-nv_ingest_api/internal/schemas/extract/extract_pptx_schema.py,sha256=5dT0kv-Mmpe5KW-BZc1JOW3rUlgzVZI0rpB79NWytmw,3761
-nv_ingest_api/internal/schemas/extract/extract_table_schema.py,sha256=SXBYDU3V97-pPOLfhFmXQveP_awARXP7k1aGcMMEJtU,3951
-nv_ingest_api/internal/schemas/message_brokers/__init__.py,sha256=uLsBITo_XfgbwpzqXUm1IYX6XlZrTfx6T1cIhdILwG8,140
-nv_ingest_api/internal/schemas/message_brokers/message_broker_client_schema.py,sha256=nbnNzCQCCduoFw4k8XPfkpn3jyyMRpDLROTwEosaSG8,766
-nv_ingest_api/internal/schemas/message_brokers/request_schema.py,sha256=LZX_wXDxTamVFqTQs2Yd8uvWyPE5mddHAWSU4PtfEIQ,966
-nv_ingest_api/internal/schemas/message_brokers/response_schema.py,sha256=4b275HlzBSzpmuE2wdoeaGKPCdKki3wuWldtRIfrj8w,727
-nv_ingest_api/internal/schemas/meta/__init__.py,sha256=wQSlVx3T14ZgQAt-EPzEczQusXVW0W8yynnUaFFGE3s,143
-nv_ingest_api/internal/schemas/meta/base_model_noext.py,sha256=8hXU1uuiqZ6t8EsoZ8vlC5EFf2zSZrKEX133FcfZMwI,316
-nv_ingest_api/internal/schemas/meta/ingest_job_schema.py,sha256=4Ylcz5CDJXYUKd79-CnyrG7mI463jLd4Uachy7uTRVE,7735
-nv_ingest_api/internal/schemas/meta/metadata_schema.py,sha256=_FAE-yeb01hxq05SXrV3NLM4DPUPSfnIbH6ZMliWsEg,6625
-nv_ingest_api/internal/schemas/mutate/__init__.py,sha256=wQSlVx3T14ZgQAt-EPzEczQusXVW0W8yynnUaFFGE3s,143
-nv_ingest_api/internal/schemas/mutate/mutate_image_dedup_schema.py,sha256=k1JOdlPPpsipc0XhHf-9YxJ_-W0HvpVE1ZhYmr7fzj0,395
-nv_ingest_api/internal/schemas/store/__init__.py,sha256=wQSlVx3T14ZgQAt-EPzEczQusXVW0W8yynnUaFFGE3s,143
-nv_ingest_api/internal/schemas/store/store_embedding_schema.py,sha256=tdKeiraim9CDL9htgp4oUSCoPMoO5PrHBnlXqDyCpMw,956
-nv_ingest_api/internal/schemas/store/store_image_schema.py,sha256=p2LGij9i6sG6RYmsfdiQOiWIc2j-POjxYrNuMrp3ELU,1010
-nv_ingest_api/internal/schemas/transform/__init__.py,sha256=wQSlVx3T14ZgQAt-EPzEczQusXVW0W8yynnUaFFGE3s,143
-nv_ingest_api/internal/schemas/transform/transform_image_caption_schema.py,sha256=ORXAowdjxBUyfkw95eg2F82DRFqEsuV9PwNKVBulcmY,568
-nv_ingest_api/internal/schemas/transform/transform_image_filter_schema.py,sha256=31ThI5fr0yyENeJeE1xMAA-pxk1QVJLwM842zMate_k,429
-nv_ingest_api/internal/schemas/transform/transform_text_embedding_schema.py,sha256=vlTjAj1T78QkQXYkC83vZQKTW04x7PeoukEzmkam7sY,732
-nv_ingest_api/internal/schemas/transform/transform_text_splitter_schema.py,sha256=iM1sUklcZVA6fdeEWRsMqV_ls-E4UcUsGwewv0JJRi4,759
-nv_ingest_api/internal/store/__init__.py,sha256=wQSlVx3T14ZgQAt-EPzEczQusXVW0W8yynnUaFFGE3s,143
-nv_ingest_api/internal/store/embed_text_upload.py,sha256=maxb4FPsBvWgvlrjAPEBlRZEFdJX5NxPG-p8kUbzV7I,9898
-nv_ingest_api/internal/store/image_upload.py,sha256=J5EHNng7Z5I6M4f3UcbniKQB29Scr3Qe05wsBpaVXds,9653
-nv_ingest_api/internal/transform/__init__.py,sha256=wQSlVx3T14ZgQAt-EPzEczQusXVW0W8yynnUaFFGE3s,143
-nv_ingest_api/internal/transform/caption_image.py,sha256=0ILCG2F8ESqKtZiPUM-6F1BHUflFZ76Dzi2GNzkE-lU,8517
-nv_ingest_api/internal/transform/embed_text.py,sha256=MACFgVHUxK3aVlEmymF7F4pT_aKoCcOKxrmmHalk1f0,15622
-nv_ingest_api/internal/transform/split_text.py,sha256=y6NYRkCEVpVsDu-AqrKx2D6JPp1vwxclw9obNZNJIIs,6561
-nv_ingest_api/util/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-nv_ingest_api/util/control_message/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-nv_ingest_api/util/control_message/validators.py,sha256=KvvbyheJ5rbzvJbH9JKpMR9VfoI0b0uM6eTAZte1p44,1315
-nv_ingest_api/util/converters/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-nv_ingest_api/util/converters/bytetools.py,sha256=xt57nCLcBgDOQ5S3q4HZ7L6dYcNFUzYuuQ_-e6zR-wo,1418
-nv_ingest_api/util/converters/containers.py,sha256=RV1ooujhq6dVujAzC0MIdZvpOZyQqUeGyQONSZLT2RA,2188
-nv_ingest_api/util/converters/datetools.py,sha256=9tskk4BkdLOMLI9ejXvRmri-otE1_Ast3oyX3HQoJZc,2579
-nv_ingest_api/util/converters/dftools.py,sha256=FjHjazIeiUd1LdFwWuummJmraqZe1a90YrWzSjZKzB4,3284
-nv_ingest_api/util/converters/formats.py,sha256=L11FtormO2SeHSebbwsGE_uuCv6Jk0D3VvVW2avU0vI,2258
-nv_ingest_api/util/converters/type_mappings.py,sha256=VFVK5IXfnXJjG4ijDTzKSEZQvJ7xK6iO8snsLgG2vv8,1108
-nv_ingest_api/util/detectors/__init__.py,sha256=HIHfzSig66GT0Uk8qsGBm_f13fKYcPtItBicRUWOOVA,183
-nv_ingest_api/util/detectors/language.py,sha256=TvzcESYY0bn0U4aLY6GjB4VaCWA6XrXxAGZbVzHTMuE,965
-nv_ingest_api/util/exception_handlers/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-nv_ingest_api/util/exception_handlers/converters.py,sha256=MYl7b-_V1g661EVDgaep9-nO3A6ka9uyV6pKoN0pEDA,2223
-nv_ingest_api/util/exception_handlers/decorators.py,sha256=F1E_xIpNY5S9TvBEZcsbli88iI5vHrcT1PGmPu5KwcE,8865
-nv_ingest_api/util/exception_handlers/detectors.py,sha256=Q1O-QOzsShPpNian2lawXVAOCstIE7nSytNw516hTg8,2288
-nv_ingest_api/util/exception_handlers/pdf.py,sha256=FUC41QJKDCfiTv-1c1_8Isxwt1xMxDZw9BN8JLEJKBw,3654
-nv_ingest_api/util/exception_handlers/schemas.py,sha256=NJngVNf9sk5Uz6CFFfkNO_LBAMt2QZUcMYGxX64oYRk,2179
-nv_ingest_api/util/image_processing/__init__.py,sha256=Jiy8C1ZuSrNb_eBM1ZTV9IKFIsnjhZi6Ku3JJhVLimA,104
-nv_ingest_api/util/image_processing/clustering.py,sha256=sUGlZI4cx1q8h4Pns1N9JVpdfSM2BOH8zRmn9QFCtzI,9236
-nv_ingest_api/util/image_processing/processing.py,sha256=dHyoxoI2btKT04ODJK0ChB8MR6eCnZ0ZLpbEQowCb5A,6561
-nv_ingest_api/util/image_processing/table_and_chart.py,sha256=bxOu9PZYkG_WFCDGw_JLaO60S2pDSN8EOWK3xkIwr2A,14376
-nv_ingest_api/util/image_processing/transforms.py,sha256=Kz9hrizV314Hy7cRCYK9ZmhmBbVUOZ_z0HEpzZYcslQ,14081
-nv_ingest_api/util/logging/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-nv_ingest_api/util/logging/configuration.py,sha256=GFO7Fofco00O3DbovXDNqInVKpFuvS0i_-WT-GvxKlE,814
-nv_ingest_api/util/message_brokers/__init__.py,sha256=wQSlVx3T14ZgQAt-EPzEczQusXVW0W8yynnUaFFGE3s,143
-nv_ingest_api/util/message_brokers/simple_message_broker/__init__.py,sha256=WaQ3CWIpIKWEivT5kL-bkmzcSQKLGFNFHdXHUJjqZFs,325
-nv_ingest_api/util/message_brokers/simple_message_broker/broker.py,sha256=h9Q4q_alXGxCLNlJUZPan46q8fJ7B72sQy2eBfHdk6I,17265
-nv_ingest_api/util/message_brokers/simple_message_broker/ordered_message_queue.py,sha256=3p-LRqG8qLnsfEhBNf73_DG22C08JKahTqUvPLS2Apg,2554
-nv_ingest_api/util/message_brokers/simple_message_broker/simple_client.py,sha256=roNuqWjfNSoi-Dus-RDeF4aBIcSjUpyq8eUDnWrRRBg,15338
-nv_ingest_api/util/metadata/__init__.py,sha256=HIHfzSig66GT0Uk8qsGBm_f13fKYcPtItBicRUWOOVA,183
-nv_ingest_api/util/metadata/aggregators.py,sha256=Y5JSKuLhhk_ldpzT3eRIcVg7QM7cTNhfQZn4g5bcbq4,15884
-nv_ingest_api/util/multi_processing/__init__.py,sha256=4fojP8Rp_5Hu1YAkqGylqTyEZ-HBVVEunn5Z9I99swA,242
-nv_ingest_api/util/multi_processing/mp_pool_singleton.py,sha256=dTfP82DgGPaXEJH3jywTO8rNlLZUniD4FFzwv84_giE,7372
-nv_ingest_api/util/nim/__init__.py,sha256=UqbiXFCqjWcjNvoduXd_0gOUOGBT8JvppiYHOmMyneA,1775
-nv_ingest_api/util/pdf/__init__.py,sha256=uLsBITo_XfgbwpzqXUm1IYX6XlZrTfx6T1cIhdILwG8,140
-nv_ingest_api/util/pdf/pdfium.py,sha256=puqw9lYloNJwjdx9X63TQ9u6vA6FRJqmmPiaquUm3HM,15767
-nv_ingest_api/util/schema/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-nv_ingest_api/util/schema/schema_validator.py,sha256=H0yZ_i_HZaiBRUCGmTBfRB9-hURhVqyd10aS_ynM1_0,321
-nv_ingest_api/util/service_clients/__init__.py,sha256=wQSlVx3T14ZgQAt-EPzEczQusXVW0W8yynnUaFFGE3s,143
-nv_ingest_api/util/service_clients/client_base.py,sha256=TUe67EtHKNCx7kLJYRF62AsOJyf3KZ24CFdo0ujRx0M,2035
-nv_ingest_api/util/service_clients/kafka/__init__.py,sha256=uLsBITo_XfgbwpzqXUm1IYX6XlZrTfx6T1cIhdILwG8,140
-nv_ingest_api/util/service_clients/redis/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-nv_ingest_api/util/service_clients/redis/redis_client.py,sha256=cT4kC2ys7c12lrWPinDuPPgs9gwr3Cw75QqJXQCmtCc,12842
-nv_ingest_api/util/service_clients/rest/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-nv_ingest_api/util/service_clients/rest/rest_client.py,sha256=F9V0YxhRwRfaoyiGsybnmxXieR-1ZkiBmJl7QLcZrl0,14413
-nv_ingest_api/util/string_processing/__init__.py,sha256=mkwHthyS-IILcLcL1tJYeF6mpqX3pxEw5aUzDGjTSeU,1411
-nv_ingest_api-2025.4.17.dev20250417.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
-nv_ingest_api-2025.4.17.dev20250417.dist-info/METADATA,sha256=-eK-AA8NgqIwYd1oDZjNaVZNXlVOAwtLRuogarWvnj0,13889
-nv_ingest_api-2025.4.17.dev20250417.dist-info/WHEEL,sha256=CmyFI0kx5cdEMTLiONQRbGQwjIoR1aIYB7eCAQ4KPJ0,91
-nv_ingest_api-2025.4.17.dev20250417.dist-info/top_level.txt,sha256=abjYMlTJGoG5tOdfIB-IWvLyKclw6HLaRSc8MxX4X6I,14
-nv_ingest_api-2025.4.17.dev20250417.dist-info/RECORD,,