landingai-ade 0.11.0__py3-none-any.whl → 0.21.1__py3-none-any.whl

landingai_ade/_client.py

@@ -41,6 +41,7 @@ from ._response import (
     async_to_raw_response_wrapper,
     async_to_streamed_response_wrapper,
 )
+from .resources import parse_jobs
 from ._streaming import Stream as Stream, AsyncStream as AsyncStream
 from ._exceptions import APIStatusError, LandingAiadeError
 from ._base_client import (
@@ -49,6 +50,7 @@ from ._base_client import (
     AsyncAPIClient,
     make_request_options,
 )
+from .lib.url_utils import convert_url_to_file_if_local
 from .types.parse_response import ParseResponse
 from .types.extract_response import ExtractResponse
 
@@ -73,6 +75,7 @@ ENVIRONMENTS: Dict[str, str] = {
 
 
 class LandingAIADE(SyncAPIClient):
+    parse_jobs: parse_jobs.ParseJobsResource
     with_raw_response: LandingAIADEWithRawResponse
     with_streaming_response: LandingAIADEWithStreamedResponse
 
@@ -154,6 +157,7 @@ class LandingAIADE(SyncAPIClient):
             _strict_response_validation=_strict_response_validation,
         )
 
+        self.parse_jobs = parse_jobs.ParseJobsResource(self)
         self.with_raw_response = LandingAIADEWithRawResponse(self)
         self.with_streaming_response = LandingAIADEWithStreamedResponse(self)
 
@@ -259,11 +263,12 @@ class LandingAIADE(SyncAPIClient):
               are extracted from the Markdown. The schema must be a valid JSON object and will
               be validated before processing the document.
 
-          markdown: The Markdown file to extract data from.
+          markdown: The Markdown file or Markdown content to extract data from.
 
           markdown_url: The URL to the Markdown file to extract data from.
 
-          model: The version of the model to use for extraction.
+          model: The version of the model to use for extraction. Use `extract-latest` to use the
+              latest version.
 
           extra_headers: Send extra headers
 
@@ -273,6 +278,9 @@ class LandingAIADE(SyncAPIClient):
 
           timeout: Override the client-level default timeout for this request, in seconds
         """
+        # Convert local file paths to file parameters
+        markdown, markdown_url = convert_url_to_file_if_local(markdown, markdown_url)
+
         body = deepcopy_minimal(
             {
                 "schema": schema,
@@ -318,24 +326,23 @@ class LandingAIADE(SyncAPIClient):
         timeout: float | httpx.Timeout | None | NotGiven = not_given,
     ) -> ParseResponse:
         """
-        Parse a document.
+        Parse a document or spreadsheet.
 
-        This endpoint parses documents and structured Markdown, chunks, and metadata.
+        This endpoint parses documents (PDF, images) and spreadsheets (XLSX, CSV) into
+        structured Markdown, chunks, and metadata.
 
         For EU users, use this endpoint:
 
             `https://api.va.eu-west-1.landing.ai/v1/ade/parse`.
 
         Args:
-          document: A file to be parsed. The file can be a PDF (50 pages max) or an image (50MB).
-              See the list of supported file types here
-              (https://docs.landing.ai/ade/ade-file-types). Either this parameter or the
-              document_url parameter must be provided.
+          document: A file to be parsed. The file can be a PDF or an image. See the list of
+              supported file types here: https://docs.landing.ai/ade/ade-file-types. Either
+              this parameter or the `document_url` parameter must be provided.
 
-          document_url: The URL to the file to be parsed. The file can be a PDF (50 pages max) or an
-              image (50MB). See the list of supported file types here
-              (https://docs.landing.ai/ade/ade-file-types). Either this parameter or the
-              document parameter must be provided.
+          document_url: The URL to the file to be parsed. The file can be a PDF or an image. See the
+              list of supported file types here: https://docs.landing.ai/ade/ade-file-types.
+              Either this parameter or the `document` parameter must be provided.
 
           model: The version of the model to use for parsing.
 
@@ -351,6 +358,9 @@ class LandingAIADE(SyncAPIClient):
 
           timeout: Override the client-level default timeout for this request, in seconds
         """
+        # Convert local file paths to file parameters
+        document, document_url = convert_url_to_file_if_local(document, document_url)
+
         body = deepcopy_minimal(
             {
                 "document": document,
@@ -416,6 +426,7 @@ class LandingAIADE(SyncAPIClient):
 
 
 class AsyncLandingAIADE(AsyncAPIClient):
+    parse_jobs: parse_jobs.AsyncParseJobsResource
     with_raw_response: AsyncLandingAIADEWithRawResponse
     with_streaming_response: AsyncLandingAIADEWithStreamedResponse
 
@@ -497,6 +508,7 @@ class AsyncLandingAIADE(AsyncAPIClient):
             _strict_response_validation=_strict_response_validation,
         )
 
+        self.parse_jobs = parse_jobs.AsyncParseJobsResource(self)
         self.with_raw_response = AsyncLandingAIADEWithRawResponse(self)
         self.with_streaming_response = AsyncLandingAIADEWithStreamedResponse(self)
 
@@ -602,11 +614,12 @@ class AsyncLandingAIADE(AsyncAPIClient):
               are extracted from the Markdown. The schema must be a valid JSON object and will
               be validated before processing the document.
 
-          markdown: The Markdown file to extract data from.
+          markdown: The Markdown file or Markdown content to extract data from.
 
           markdown_url: The URL to the Markdown file to extract data from.
 
-          model: The version of the model to use for extraction.
+          model: The version of the model to use for extraction. Use `extract-latest` to use the
+              latest version.
 
           extra_headers: Send extra headers
 
@@ -616,6 +629,9 @@ class AsyncLandingAIADE(AsyncAPIClient):
 
           timeout: Override the client-level default timeout for this request, in seconds
         """
+        # Convert local file paths to file parameters
+        markdown, markdown_url = convert_url_to_file_if_local(markdown, markdown_url)
+
         body = deepcopy_minimal(
             {
                 "schema": schema,
@@ -661,24 +677,23 @@ class AsyncLandingAIADE(AsyncAPIClient):
         timeout: float | httpx.Timeout | None | NotGiven = not_given,
     ) -> ParseResponse:
         """
-        Parse a document.
+        Parse a document or spreadsheet.
 
-        This endpoint parses documents and structured Markdown, chunks, and metadata.
+        This endpoint parses documents (PDF, images) and spreadsheets (XLSX, CSV) into
+        structured Markdown, chunks, and metadata.
 
         For EU users, use this endpoint:
 
             `https://api.va.eu-west-1.landing.ai/v1/ade/parse`.
 
         Args:
-          document: A file to be parsed. The file can be a PDF (50 pages max) or an image (50MB).
-              See the list of supported file types here
-              (https://docs.landing.ai/ade/ade-file-types). Either this parameter or the
-              document_url parameter must be provided.
+          document: A file to be parsed. The file can be a PDF or an image. See the list of
+              supported file types here: https://docs.landing.ai/ade/ade-file-types. Either
+              this parameter or the `document_url` parameter must be provided.
 
-          document_url: The URL to the file to be parsed. The file can be a PDF (50 pages max) or an
-              image (50MB). See the list of supported file types here
-              (https://docs.landing.ai/ade/ade-file-types). Either this parameter or the
-              document parameter must be provided.
+          document_url: The URL to the file to be parsed. The file can be a PDF or an image. See the
+              list of supported file types here: https://docs.landing.ai/ade/ade-file-types.
+              Either this parameter or the `document` parameter must be provided.
 
           model: The version of the model to use for parsing.
 
@@ -694,6 +709,9 @@ class AsyncLandingAIADE(AsyncAPIClient):
 
           timeout: Override the client-level default timeout for this request, in seconds
         """
+        # Convert local file paths to file parameters
+        document, document_url = convert_url_to_file_if_local(document, document_url)
+
         body = deepcopy_minimal(
             {
                 "document": document,
@@ -760,6 +778,8 @@ class AsyncLandingAIADE(AsyncAPIClient):
 
 class LandingAIADEWithRawResponse:
     def __init__(self, client: LandingAIADE) -> None:
+        self.parse_jobs = parse_jobs.ParseJobsResourceWithRawResponse(client.parse_jobs)
+
         self.extract = to_raw_response_wrapper(
             client.extract,
         )
@@ -770,6 +790,8 @@ class LandingAIADEWithRawResponse:
 
 class AsyncLandingAIADEWithRawResponse:
     def __init__(self, client: AsyncLandingAIADE) -> None:
+        self.parse_jobs = parse_jobs.AsyncParseJobsResourceWithRawResponse(client.parse_jobs)
+
         self.extract = async_to_raw_response_wrapper(
             client.extract,
         )
@@ -780,6 +802,8 @@ class AsyncLandingAIADEWithRawResponse:
 
 class LandingAIADEWithStreamedResponse:
     def __init__(self, client: LandingAIADE) -> None:
+        self.parse_jobs = parse_jobs.ParseJobsResourceWithStreamingResponse(client.parse_jobs)
+
         self.extract = to_streamed_response_wrapper(
             client.extract,
         )
@@ -790,6 +814,8 @@ class LandingAIADEWithStreamedResponse:
 
 class AsyncLandingAIADEWithStreamedResponse:
     def __init__(self, client: AsyncLandingAIADE) -> None:
+        self.parse_jobs = parse_jobs.AsyncParseJobsResourceWithStreamingResponse(client.parse_jobs)
+
         self.extract = async_to_streamed_response_wrapper(
             client.extract,
         )

landingai_ade/_constants.py

@@ -5,8 +5,8 @@ import httpx
 RAW_RESPONSE_HEADER = "X-Stainless-Raw-Response"
 OVERRIDE_CAST_TO_HEADER = "____stainless_override_cast_to"
 
-# default timeout is 1 minute
-DEFAULT_TIMEOUT = httpx.Timeout(timeout=60, connect=5.0)
+# default timeout is 8 minutes
+DEFAULT_TIMEOUT = httpx.Timeout(timeout=480, connect=5.0)
 DEFAULT_MAX_RETRIES = 2
 DEFAULT_CONNECTION_LIMITS = httpx.Limits(max_connections=100, max_keepalive_connections=20)
 

landingai_ade/_files.py

@@ -26,7 +26,7 @@ def is_base64_file_input(obj: object) -> TypeGuard[Base64FileInput]:
 
 def is_file_content(obj: object) -> TypeGuard[FileContent]:
     return (
-        isinstance(obj, bytes) or isinstance(obj, tuple) or isinstance(obj, io.IOBase) or isinstance(obj, os.PathLike)
+        isinstance(obj, bytes) or isinstance(obj, str) or isinstance(obj, tuple) or isinstance(obj, io.IOBase) or isinstance(obj, os.PathLike)
     )
 
 
@@ -66,6 +66,9 @@ def _transform_file(file: FileTypes) -> HttpxFileTypes:
             path = pathlib.Path(file)
             return (path.name, path.read_bytes())
 
+        if isinstance(file, str):
+            return file.encode('utf-8')
+
         return file
 
     if is_tuple_t(file):
@@ -77,6 +80,8 @@ def _transform_file(file: FileTypes) -> HttpxFileTypes:
 def read_file_content(file: FileContent) -> HttpxFileContent:
     if isinstance(file, os.PathLike):
         return pathlib.Path(file).read_bytes()
+    if isinstance(file, str):
+        return file.encode('utf-8')
     return file
 
 
@@ -108,6 +113,9 @@ async def _async_transform_file(file: FileTypes) -> HttpxFileTypes:
             path = anyio.Path(file)
             return (path.name, await path.read_bytes())
 
+        if isinstance(file, str):
+            return file.encode('utf-8')
+
         return file
 
     if is_tuple_t(file):
@@ -120,4 +128,7 @@ async def async_read_file_content(file: FileContent) -> HttpxFileContent:
     if isinstance(file, os.PathLike):
         return await anyio.Path(file).read_bytes()
 
+    if isinstance(file, str):
+        return file.encode('utf-8')
+
     return file

landingai_ade/_models.py

@@ -2,6 +2,7 @@ from __future__ import annotations
 
 import os
 import inspect
+import weakref
 from typing import TYPE_CHECKING, Any, Type, Union, Generic, TypeVar, Callable, Optional, cast
 from datetime import date, datetime
 from typing_extensions import (
@@ -573,6 +574,9 @@ class CachedDiscriminatorType(Protocol):
     __discriminator__: DiscriminatorDetails
 
 
+DISCRIMINATOR_CACHE: weakref.WeakKeyDictionary[type, DiscriminatorDetails] = weakref.WeakKeyDictionary()
+
+
 class DiscriminatorDetails:
     field_name: str
     """The name of the discriminator field in the variant class, e.g.
@@ -615,8 +619,9 @@ class DiscriminatorDetails:
 
 
 def _build_discriminated_union_meta(*, union: type, meta_annotations: tuple[Any, ...]) -> DiscriminatorDetails | None:
-    if isinstance(union, CachedDiscriminatorType):
-        return union.__discriminator__
+    cached = DISCRIMINATOR_CACHE.get(union)
+    if cached is not None:
+        return cached
 
     discriminator_field_name: str | None = None
 
@@ -669,7 +674,7 @@ def _build_discriminated_union_meta(*, union: type, meta_annotations: tuple[Any,
         discriminator_field=discriminator_field_name,
         discriminator_alias=discriminator_alias,
     )
-    cast(CachedDiscriminatorType, union).__discriminator__ = details
+    DISCRIMINATOR_CACHE.setdefault(union, details)
     return details
 
 

landingai_ade/_streaming.py

@@ -57,9 +57,8 @@ class Stream(Generic[_T]):
         for sse in iterator:
             yield process_data(data=sse.json(), cast_to=cast_to, response=response)
 
-        # Ensure the entire stream is consumed
-        for _sse in iterator:
-            ...
+        # As we might not fully consume the response stream, we need to close it explicitly
+        response.close()
 
     def __enter__(self) -> Self:
         return self
@@ -121,9 +120,8 @@ class AsyncStream(Generic[_T]):
         async for sse in iterator:
             yield process_data(data=sse.json(), cast_to=cast_to, response=response)
 
-        # Ensure the entire stream is consumed
-        async for _sse in iterator:
-            ...
+        # As we might not fully consume the response stream, we need to close it explicitly
+        await response.aclose()
 
     async def __aenter__(self) -> Self:
         return self

landingai_ade/_types.py

@@ -52,10 +52,10 @@ ProxiesDict = Dict["str | URL", Union[None, str, URL, Proxy]]
 ProxiesTypes = Union[str, Proxy, ProxiesDict]
 if TYPE_CHECKING:
     Base64FileInput = Union[IO[bytes], PathLike[str]]
-    FileContent = Union[IO[bytes], bytes, PathLike[str]]
+    FileContent = Union[IO[bytes], bytes, str, PathLike[str]]
 else:
     Base64FileInput = Union[IO[bytes], PathLike]
-    FileContent = Union[IO[bytes], bytes, PathLike]  # PathLike is not subscriptable in Python 3.8.
+    FileContent = Union[IO[bytes], bytes, str, PathLike]  # PathLike is not subscriptable in Python 3.8.
 FileTypes = Union[
     # file (or bytes)
     FileContent,

landingai_ade/_utils/_sync.py

@@ -1,10 +1,8 @@
 from __future__ import annotations
 
-import sys
 import asyncio
 import functools
-import contextvars
-from typing import Any, TypeVar, Callable, Awaitable
+from typing import TypeVar, Callable, Awaitable
 from typing_extensions import ParamSpec
 
 import anyio
@@ -15,34 +13,11 @@ T_Retval = TypeVar("T_Retval")
 T_ParamSpec = ParamSpec("T_ParamSpec")
 
 
-if sys.version_info >= (3, 9):
-    _asyncio_to_thread = asyncio.to_thread
-else:
-    # backport of https://docs.python.org/3/library/asyncio-task.html#asyncio.to_thread
-    # for Python 3.8 support
-    async def _asyncio_to_thread(
-        func: Callable[T_ParamSpec, T_Retval], /, *args: T_ParamSpec.args, **kwargs: T_ParamSpec.kwargs
-    ) -> Any:
-        """Asynchronously run function *func* in a separate thread.
-
-        Any *args and **kwargs supplied for this function are directly passed
-        to *func*. Also, the current :class:`contextvars.Context` is propagated,
-        allowing context variables from the main thread to be accessed in the
-        separate thread.
-
-        Returns a coroutine that can be awaited to get the eventual result of *func*.
-        """
-        loop = asyncio.events.get_running_loop()
-        ctx = contextvars.copy_context()
-        func_call = functools.partial(ctx.run, func, *args, **kwargs)
-        return await loop.run_in_executor(None, func_call)
-
-
 async def to_thread(
     func: Callable[T_ParamSpec, T_Retval], /, *args: T_ParamSpec.args, **kwargs: T_ParamSpec.kwargs
 ) -> T_Retval:
     if sniffio.current_async_library() == "asyncio":
-        return await _asyncio_to_thread(func, *args, **kwargs)
+        return await asyncio.to_thread(func, *args, **kwargs)
 
     return await anyio.to_thread.run_sync(
         functools.partial(func, *args, **kwargs),
@@ -53,10 +28,7 @@ async def to_thread(
 def asyncify(function: Callable[T_ParamSpec, T_Retval]) -> Callable[T_ParamSpec, Awaitable[T_Retval]]:
     """
     Take a blocking function and create an async one that receives the same
-    positional and keyword arguments. For python version 3.9 and above, it uses
-    asyncio.to_thread to run the function in a separate thread. For python version
-    3.8, it uses locally defined copy of the asyncio.to_thread function which was
-    introduced in python 3.9.
+    positional and keyword arguments.
 
     Usage:
 

landingai_ade/_utils/_utils.py

@@ -133,7 +133,7 @@ def is_given(obj: _T | NotGiven | Omit) -> TypeGuard[_T]:
 # Type safe methods for narrowing types with TypeVars.
 # The default narrowing for isinstance(obj, dict) is dict[unknown, unknown],
 # however this cause Pyright to rightfully report errors. As we know we don't
-# care about the contained types we can safely use `object` in it's place.
+# care about the contained types we can safely use `object` in its place.
 #
 # There are two separate functions defined, `is_*` and `is_*_t` for different use cases.
 # `is_*` is for when you're dealing with an unknown input

landingai_ade/_version.py

@@ -1,4 +1,4 @@
 # File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
 
 __title__ = "landingai_ade"
-__version__ = "0.11.0"  # x-release-please-version
+__version__ = "0.21.1"  # x-release-please-version

landingai_ade/lib/url_utils.py

@@ -0,0 +1,51 @@
+"""Utility functions for handling URL and file path conversions."""
+
+from typing import Tuple, Union, Optional, cast
+from pathlib import Path
+from urllib.parse import urlparse
+
+from .._types import Omit, FileTypes, omit
+
+
+def convert_url_to_file_if_local(
+    file: Union[Optional[FileTypes], Omit],
+    file_url: Union[Optional[str], Omit],
+) -> Tuple[Union[Optional[FileTypes], Omit], Union[Optional[str], Omit]]:
+    """
+    Convert a URL parameter to a file parameter if it's a local file path.
+
+    If the file_url is a local file path that exists, it will be converted to a Path object
+    and returned as the file parameter, with the file_url parameter set to omit.
+
+    If the file_url is a remote URL (http/https) or doesn't exist as a local file,
+    it will be returned unchanged.
+
+    Args:
+        file: The existing file parameter
+        file_url: The URL parameter that might be a local file path
+
+    Returns:
+        A tuple of (file, file_url) where one will be set and the other omit
+    """
+    # If file_url is omit or None, return unchanged
+    if file_url is omit or file_url is None:
+        return file, file_url
+
+    # At this point, file_url is guaranteed to be a string, use cast for type narrowing
+    url_str = cast(str, file_url)
+
+    # Check if it's a remote URL (http/https)
+    parsed = urlparse(url_str)
+    if parsed.scheme in ("http", "https", "ftp", "ftps"):
+        # It's a remote URL, keep it as is
+        return file, file_url
+
+    # Check if it's a local file path
+    path = Path(url_str)
+    if path.exists() and path.is_file():
+        # It's a local file, convert to file parameter
+        return path, omit
+
+    # Path doesn't exist or is not a file, treat as URL
+    # (could be a URL with a different scheme or a typo)
+    return file, file_url

landingai_ade/resources/__init__.py

@@ -1 +1,19 @@
 # File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
+
+from .parse_jobs import (
+    ParseJobsResource,
+    AsyncParseJobsResource,
+    ParseJobsResourceWithRawResponse,
+    AsyncParseJobsResourceWithRawResponse,
+    ParseJobsResourceWithStreamingResponse,
+    AsyncParseJobsResourceWithStreamingResponse,
+)
+
+__all__ = [
+    "ParseJobsResource",
+    "AsyncParseJobsResource",
+    "ParseJobsResourceWithRawResponse",
+    "AsyncParseJobsResourceWithRawResponse",
+    "ParseJobsResourceWithStreamingResponse",
+    "AsyncParseJobsResourceWithStreamingResponse",
+]