extract-python 0.10.4__tar.gz → 0.11.1__tar.gz

{extract_python-0.10.4 → extract_python-0.11.1}/PKG-INFO

@@ -1,13 +1,13 @@
 Metadata-Version: 2.5
 Name: extract-python
-Version: 0.10.4
+Version: 0.11.1
 Summary: Structured content extraction
 Project-URL: Homepage, https://github.com/ICIJ/extract-python
 Project-URL: Repository, https://github.com/ICIJ/extract-python
 Project-URL: Issues, https://github.com/ICIJ/extract-python/issues
 Author-email: Clément Doumouro <cdoumouro@icij.org>
 Requires-Python: <3.15,>=3.13
-Requires-Dist: extract-core~=0.10.4
+Requires-Dist: extract-core~=0.11.0
 Requires-Dist: icij-common~=0.8.2
 Requires-Dist: pympler~=1.1
 Provides-Extra: benches
@@ -15,6 +15,8 @@ Requires-Dist: html2image~=2.0.7; extra == 'benches'
 Requires-Dist: markdown2>=2.5.4; extra == 'benches'
 Requires-Dist: notebook>=7.4.5; extra == 'benches'
 Requires-Dist: pypdfium2>=4.30.0; extra == 'benches'
+Provides-Extra: cli
+Requires-Dist: typer~=0.21.2; extra == 'cli'
 Provides-Extra: docling
 Requires-Dist: docling-slim[feat-ocr-easyocr,feat-ocr-mac,feat-ocr-rapidocr,feat-ocr-tesserocr,standard]~=2.127; extra == 'docling'
 Provides-Extra: docling-gpu

{extract_python-0.10.4 → extract_python-0.11.1}/extract_python/__init__.py

@@ -5,20 +5,20 @@ logger = logging.getLogger(__name__)
 try:
     from .docling_ import DOCLING_DEFAULT_ARTIFACTS_PATH, DoclingPipeline
 except ModuleNotFoundError:
-    logger.exception("docling is not available, might be optional...")
+    logger.warning("docling is not available, might be optional...")
     DOCKING_DEFAULT_ARTIFACTS_PATH, DoclingPipeline = None, None
 
 try:
     from .marker_ import MarkerPipeline
 except ModuleNotFoundError:
-    logger.exception("marker is not available, might be optional...")
+    logger.warning("marker is not available, might be optional...")
     MarkerPipeline = None
 
 
 try:
     from .miner_u import MinerUPipeline
 except ModuleNotFoundError:
-    logger.exception("mineru is not available, might be optional...")
+    logger.warning("mineru is not available, might be optional...")
     MinerUPipeline = None
 
 

extract_python-0.11.1/extract_python/__main__.py

@@ -0,0 +1,9 @@
+from .cli import cli_app
+
+
+def main() -> None:
+    cli_app()
+
+
+if __name__ == "__main__":
+    main()

extract_python-0.11.1/extract_python/cli/__init__.py

@@ -0,0 +1,50 @@
+import importlib
+import logging
+import os
+from typing import Annotated
+
+import typer
+from icij_common.logging_utils import setup_loggers
+
+import extract_python
+
+from .doc import configs_app
+from .utils import AsyncTyper
+
+cli_app = AsyncTyper(
+    context_settings={"help_option_names": ["-h", "--help"]},
+    pretty_exceptions_enable=False,
+)
+cli_app.add_typer(configs_app)
+
+
+def version_callback(value: bool) -> None:  # noqa: FBT001
+    if value:
+        package_version = importlib.metadata.version(extract_python.__name__)
+        print(package_version)
+        raise typer.Exit()
+
+
+def pretty_exc_callback(value: bool) -> None:  # noqa: FBT001
+    if not value:
+        os.environ["TYPER_STANDARD_TRACEBACK"] = "1"
+
+
+@cli_app.callback()
+def main(
+    version: Annotated[  # noqa: ARG001
+        bool | None,
+        typer.Option("--version", callback=version_callback, is_eager=True),
+    ] = None,
+    *,
+    pretty_exceptions: Annotated[  # noqa: ARG001
+        bool,
+        typer.Option(
+            "--pretty-exceptions", callback=pretty_exc_callback, is_eager=True
+        ),
+    ] = False,
+) -> None:
+    setup_loggers(
+        ["__main__", extract_python.__name__, "docling", "mineru", "marker"],
+        level=logging.DEBUG,
+    )

extract_python-0.11.1/extract_python/cli/doc.py

@@ -0,0 +1,78 @@
+import logging
+import os
+from pathlib import Path
+from typing import Annotated, Never
+
+import typer
+from extract_core import (
+    Device,
+    InputDoc,
+    OutputFormat,
+    Pipeline,
+    PipelineConfig,
+    PipelineSize,
+    PipelineType,
+)
+from extract_core.default_config import default_config
+from pydantic import TypeAdapter
+
+from .utils import AsyncTyper
+
+logger = logging.getLogger(__name__)
+
+_DOC = "doc"
+
+_DOC_CONVERT_HELP = "Converts documents to the provided output format"
+_DOC_CONVERT_DOCS_HELP = "input documents (can be file or directories)"
+_DOC_CONVERT_OUTPUT_PATH_HELP = "directory where the converted documents will be saved"
+_DOC_CONVERT_DEVICE_HELP = "device used to run inference"
+_DOC_CONVERT_OUTPUT_FMT_HELP = "output format"
+_DOC_CONVERT_PIPELINE_CONFIG = "path to a JSON pipeline configuration"
+
+
+configs_app = AsyncTyper(name=_DOC)
+
+
+@configs_app.async_command(help=_DOC_CONVERT_HELP)
+async def convert(
+    docs: Annotated[list[Path], typer.Argument(help=_DOC_CONVERT_DOCS_HELP)],
+    output_path: Annotated[Path, typer.Argument(help=_DOC_CONVERT_OUTPUT_PATH_HELP)],
+    output_format: Annotated[
+        OutputFormat, typer.Option(help=_DOC_CONVERT_OUTPUT_FMT_HELP)
+    ] = OutputFormat.MARKDOWN,
+    config: Annotated[
+        Path | None,
+        typer.Option("-c", "--config", help=_DOC_CONVERT_PIPELINE_CONFIG),
+    ] = None,
+) -> None:
+    if config is not None:
+        config = TypeAdapter(PipelineConfig).validate_json(config.read_text())
+    else:
+        config = default_config(
+            pipeline_type=PipelineType.DOCLING,
+            device=Device.CPU,
+            size=PipelineSize.SMALL,
+        )
+    config = config.to_device()
+    docs = [InputDoc.from_path(d, n_pages=1) for p in docs for d in _walkdir(p)]
+    pipeline = Pipeline.from_config(config)
+    output_paths = []
+    async for res in pipeline.extract_content(docs, output_format, output_path):
+        output_path = res.output.path
+        logger.info("converted %s to %s", res.input.path, output_path / output_path)
+        output_paths.append(output_path / output_path)
+    print("\n".join(output_paths))
+
+
+def _raise(exc: OSError) -> Never:
+    raise exc
+
+
+def _walkdir(path: Path) -> list[Path]:
+    if path.is_file():
+        return [path]
+    paths = []
+    for root, _, files in os.walk(path, onerror=_raise):
+        root = Path(root)  # noqa: PLW2901
+        paths.extend(root / f for f in files)
+    return paths

extract_python-0.11.1/extract_python/cli/utils.py

@@ -0,0 +1,25 @@
+import asyncio
+import sys
+from collections.abc import Callable
+from functools import wraps
+from typing import Any
+
+import typer
+
+
+class AsyncTyper(typer.Typer):
+    def async_command(self, *args, **kwargs) -> Callable[[Callable], Callable]:
+        def decorator(async_func: Callable) -> Callable:
+            @wraps(async_func)
+            def sync_func(*_args, **_kwargs) -> Any:
+                res = asyncio.run(async_func(*_args, **_kwargs))
+                return res
+
+            self.command(*args, **kwargs)(sync_func)
+            return async_func
+
+        return decorator
+
+
+def eprint(*args, **kwargs) -> None:
+    print(*args, file=sys.stderr, **kwargs)

{extract_python-0.10.4 → extract_python-0.11.1}/extract_python/docling_.py

@@ -59,8 +59,7 @@ class DoclingPipeline(Pipeline):
     def __init__(self, config: DoclingPipelineConfig):
         super().__init__(config)
         format_options = {
-            k: v.to_docling(self._device)
-            for k, v in self._config.format_options.items()
+            k: v.to_docling() for k, v in self._config.format_options.items()
         }
         logger.info(
             "resolved format options to: %s",

{extract_python-0.10.4 → extract_python-0.11.1}/pyproject.toml

@@ -8,12 +8,16 @@ authors = [
 readme = "README.md"
 requires-python = ">=3.13,<3.15"
 dependencies = [
-  "extract-core~=0.10.4",
+  "extract-core~=0.11.0",
   "icij-common~=0.8.2",
   "pympler~=1.1",
 ]
 
 [project.optional-dependencies]
+cli = [
+  "typer~=0.21.2",
+]
+
 benches = [
   "html2image~=2.0.7",
   "markdown2>=2.5.4",
@@ -58,6 +62,9 @@ override-dependencies = [
   "pillow==11.3.0",
 ]
 
+[project.scripts]
+extract_python = "extract_python.__main__:main"
+
 [[tool.uv.index]]
 name = "pytorch-cpu"
 url = "https://download.pytorch.org/whl/cpu"

{extract_python-0.10.4 → extract_python-0.11.1}/uv.lock

@@ -691,7 +691,7 @@ wheels = [
 
 [[package]]
 name = "docling-core"
-version = "2.98.0"
+version = "2.100.0"
 source = { registry = "https://pypi.org/simple" }
 dependencies = [
     { name = "defusedxml" },
@@ -709,9 +709,9 @@ dependencies = [
     { name = "typer" },
     { name = "typing-extensions" },
 ]
-sdist = { url = "https://files.pythonhosted.org/packages/6f/1f/03528148da2a97c35fcf0a0a9bc7614b7585cf1f63f624fce85aafaa60c9/docling_core-2.98.0.tar.gz", hash = "sha256:a5dd76d747d23b6e1c83b434e29285798a2547d6e23a0ddfc487ed09ff5663ea", size = 383348, upload-time = "2026-09-22T09:42:51.273Z" }
+sdist = { url = "https://files.pythonhosted.org/packages/4d/f7/5fe0635e3a3251d08eb4a30dffd02414b6614986431a671c5a4352fcf90f/docling_core-2.100.0.tar.gz", hash = "sha256:87a4f7318d774194364a3ac6c1e6b1df0517f28c7ac395d92aa3535eb3e5c446", size = 404111, upload-time = "2026-10-06T12:14:49.988Z" }
 wheels = [
-    { url = "https://files.pythonhosted.org/packages/23/c6/479371d25ad9fd0137c3036bb21d0b1f96075b85958a81ce868aae4bc515/docling_core-2.98.0-py3-none-any.whl", hash = "sha256:655903b45f4c66c759842df8172237bda3ef35055197bf371fa5c855d3df0bd3", size = 305641, upload-time = "2026-09-22T09:42:49.591Z" },
+    { url = "https://files.pythonhosted.org/packages/19/4c/9d319f71922af600271cb6e36b15bae0969444182d55361c39b566e29ab7/docling_core-2.100.0-py3-none-any.whl", hash = "sha256:4cb57c4ccb5740e83e5e63d2215201711125e3603115bc720b099bcc49d09029", size = 316004, upload-time = "2026-10-06T12:14:48.161Z" },
 ]
 
 [package.optional-dependencies]
@@ -752,7 +752,7 @@ wheels = [
 
 [[package]]
 name = "docling-parse"
-version = "7.21.0"
+version = "7.22.2"
 source = { registry = "https://pypi.org/simple" }
 dependencies = [
     { name = "docling-core" },
@@ -760,23 +760,23 @@ dependencies = [
     { name = "pydantic" },
     { name = "pywin32", marker = "sys_platform == 'win32' or (extra == 'extra-14-extract-python-marker' and extra == 'extra-14-extract-python-mineru')" },
 ]
-sdist = { url = "https://files.pythonhosted.org/packages/dc/d3/c0b03384b7735261662a5298837f5da49727ba0bf0d63baaeedc97236955/docling_parse-7.21.0.tar.gz", hash = "sha256:babbd8b43deb69706963ebc9b54283b458395a53519f68fd4bd0e7c5f793943b", size = 7035738, upload-time = "2026-09-22T09:53:15.861Z" }
+sdist = { url = "https://files.pythonhosted.org/packages/bc/19/4036a3e844a6385b720b40f9c44d6ab4b6733707dd698dcfec3beda0a48d/docling_parse-7.22.2.tar.gz", hash = "sha256:368d076d2b887fcb44a90c37c219a190b975f01b257b61bd83ba75dc06861536", size = 7045499, upload-time = "2026-10-05T07:23:55.9Z" }
 wheels = [
-    { url = "https://files.pythonhosted.org/packages/53/94/836d8fe9d079bc1d1d92498386e9ab7000bc7d1ae5154134e9677d8ec27a/docling_parse-7.21.0-cp313-cp313-macosx_14_0_arm64.whl", hash = "sha256:da68374d71ad50fde66318db0034e56159fb74bf2f73602084b986471f2faee6", size = 9878142, upload-time = "2026-09-22T09:52:54.478Z" },
-    { url = "https://files.pythonhosted.org/packages/94/9e/dcd2704f99ac831bc736c0c2bf30cedeb168cba908536f825946bec9f729/docling_parse-7.21.0-cp313-cp313-manylinux_2_26_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:3d554d9f12b5d8d3d3b239afdcfad53996b4912d6d8e61dfd7635ecb4b15663a", size = 10437326, upload-time = "2026-09-22T09:52:57.327Z" },
-    { url = "https://files.pythonhosted.org/packages/1d/39/51f01a03c39e40939f0e8761bb399462ddc2cb9d95920f59f84615468d9e/docling_parse-7.21.0-cp313-cp313-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl", hash = "sha256:40201d5ffbdae0d21d52ecfd5c5045167caec27330a66706c35abc128fd33f40", size = 10845161, upload-time = "2026-09-22T09:52:59.175Z" },
-    { url = "https://files.pythonhosted.org/packages/91/f4/9b900d863146dcd42c9eb984aaa41b5bcaeffe51f8743ca13f9ae3f0ee1d/docling_parse-7.21.0-cp313-cp313-win_amd64.whl", hash = "sha256:10b612c08c537221d5e1ab927a2226602479801e4132d6ebeea131fee9a09c72", size = 11886411, upload-time = "2026-09-22T09:53:01.888Z" },
-    { url = "https://files.pythonhosted.org/packages/bc/ff/9ed39e6bdc256df9a4f69b3e0c33dcd7b7cdc9a491ee5da6be9645fe4a47/docling_parse-7.21.0-cp313-cp313-win_arm64.whl", hash = "sha256:e3f8740bad16bbc1d264708153751787529a06777a0cb51e29242e5ad7f4fd30", size = 9160130, upload-time = "2026-09-22T09:53:04.334Z" },
-    { url = "https://files.pythonhosted.org/packages/97/87/fa8ac21dddf36ee89531e0c8de316e50003885551a99ec0e6eccd76efb12/docling_parse-7.21.0-cp314-cp314-macosx_14_0_arm64.whl", hash = "sha256:fe2a63c0b92ec73fcc0db674ee57ed9972ea2661c08eaf40c099976acd58e708", size = 9878785, upload-time = "2026-09-22T09:53:06.226Z" },
-    { url = "https://files.pythonhosted.org/packages/7b/f7/a29021e0953b4f562c673bb8adf9737a35fc34500680553a05cbfaf6c085/docling_parse-7.21.0-cp314-cp314-manylinux_2_26_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:40bfdf4e7ede1557b58c1680aea1fba1c22bf512e786d2889973fe330178d8df", size = 10437964, upload-time = "2026-09-22T09:53:08.473Z" },
-    { url = "https://files.pythonhosted.org/packages/43/1b/8058f7a4a96ea97ee4a6fe53525c3505b5ca8e95361dce82223c40595c58/docling_parse-7.21.0-cp314-cp314-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl", hash = "sha256:8f79205a3ae29104720b6136ed581b0574b3f365984eea5239d9e89ba4196032", size = 10845654, upload-time = "2026-09-22T09:53:10.309Z" },
-    { url = "https://files.pythonhosted.org/packages/48/f1/82049b4c466df0c87864481b80d874cf4c1880a4b7a70c587f608400aced/docling_parse-7.21.0-cp314-cp314-win_amd64.whl", hash = "sha256:d3c57efa2ab1759373c440a2d869d08ff1079a9a3244231c7167b1e85440de51", size = 12322863, upload-time = "2026-09-22T09:53:12.31Z" },
-    { url = "https://files.pythonhosted.org/packages/f4/3f/a5d44d7f51a5f3e2e96620c59770de7a98f3299d98959c1ec9b289937f61/docling_parse-7.21.0-cp314-cp314-win_arm64.whl", hash = "sha256:8cda2b5a6e8e9933ef68a114f283dfa83f6a2876ae70468ed055dbb0cdc8914b", size = 9528726, upload-time = "2026-09-22T09:53:14.311Z" },
+    { url = "https://files.pythonhosted.org/packages/d2/61/05b5d2fbc5dcb3023aa980daaebfaa03dd2447e21271668cd6b3058552c2/docling_parse-7.22.2-cp313-cp313-macosx_14_0_arm64.whl", hash = "sha256:53156ac7c1e389857f23b118f8cd59331d3a215be94f9f9ce9ae3145e4125759", size = 9878343, upload-time = "2026-10-05T07:23:28.238Z" },
+    { url = "https://files.pythonhosted.org/packages/b9/d1/538efc18c43a0c6e678e6dea7c14d0599a14a0d6c8fc3b192bd6d6a92f6d/docling_parse-7.22.2-cp313-cp313-manylinux_2_26_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:f8eb209efd837d7ba055bc582ae0aa978e32b6fa9851fb39632a1c303f2b0693", size = 10437835, upload-time = "2026-10-05T07:23:30.77Z" },
+    { url = "https://files.pythonhosted.org/packages/5a/3f/c004e1eef07e24059130e9fe97ac5870c11a187162f1f7eb93e9c20c3a75/docling_parse-7.22.2-cp313-cp313-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl", hash = "sha256:73265a552c3f28bee591a8ca1e0dc38b51ee17f5051a626aef2fb625ee81f305", size = 10842974, upload-time = "2026-10-05T07:23:33.225Z" },
+    { url = "https://files.pythonhosted.org/packages/65/a0/8943af36a45726fe5b59c0ed213ef260ed0111309547d5b01b9820dcd5bf/docling_parse-7.22.2-cp313-cp313-win_amd64.whl", hash = "sha256:5a1433aceca0b1be834a79b6a7bddd89681426b8c1b06725aebe7740583d38a3", size = 11886875, upload-time = "2026-10-05T07:23:35.722Z" },
+    { url = "https://files.pythonhosted.org/packages/08/44/5d44f7ca3801fbcc5e0f6e51d92cb85a6952873203ecf56227503fb53f6a/docling_parse-7.22.2-cp313-cp313-win_arm64.whl", hash = "sha256:e23c55d856d0124634b42d8b63a879ecb1056769a3b38cb6ea186e03d20c0421", size = 9274980, upload-time = "2026-10-05T07:23:38.454Z" },
+    { url = "https://files.pythonhosted.org/packages/1f/71/75b2492ea7de1dea7b23405d8afbbf6be44e194857fad328c400dbcc9ef4/docling_parse-7.22.2-cp314-cp314-macosx_14_0_arm64.whl", hash = "sha256:9a003356d71578a56718abd343baf5c6d745ea7d7e988ca524b3af5bbd35760d", size = 9878449, upload-time = "2026-10-05T07:23:41.387Z" },
+    { url = "https://files.pythonhosted.org/packages/65/e4/82e930b4f534f77dc70871d3b5f7cc4672fa6b89320d14bcad95a6ae3a72/docling_parse-7.22.2-cp314-cp314-manylinux_2_26_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:7699b216003ef5759ee4fa61bc7a403a6ce63afcbbf316cd17e4562d5b306e8b", size = 10439277, upload-time = "2026-10-05T07:23:44.53Z" },
+    { url = "https://files.pythonhosted.org/packages/f1/b2/6161ee6ab8e519177451620976abd1a002dbe1bdc98d4f8cddd7c49f68f4/docling_parse-7.22.2-cp314-cp314-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl", hash = "sha256:a6a5aa4e0c0f62c4e0032c24117410d5773dbc303da5b4a282ff5abf9f491391", size = 10843921, upload-time = "2026-10-05T07:23:47.72Z" },
+    { url = "https://files.pythonhosted.org/packages/28/52/3593b031a4319379f9850dd3e8bc76adbcc62d7533f0741e6e3257d77d9c/docling_parse-7.22.2-cp314-cp314-win_amd64.whl", hash = "sha256:4bca9aeefa92896a12eb96e8858adc6b3adbd34c457095e5be68db84c6bb8c56", size = 12324184, upload-time = "2026-10-05T07:23:50.939Z" },
+    { url = "https://files.pythonhosted.org/packages/79/ec/c439cced7ed74407b0fd70d48e4de19b7a8e0da88f786b95bfb0af10c77f/docling_parse-7.22.2-cp314-cp314-win_arm64.whl", hash = "sha256:6bfff00d2577f288e037069bbf5087bcf01305f6b78b9432d599c6ffe03557a2", size = 9652594, upload-time = "2026-10-05T07:23:53.627Z" },
 ]
 
 [[package]]
 name = "docling-slim"
-version = "2.130.0"
+version = "2.134.0"
 source = { registry = "https://pypi.org/simple" }
 dependencies = [
     { name = "certifi" },
@@ -789,9 +789,9 @@ dependencies = [
     { name = "requests" },
     { name = "tqdm" },
 ]
-sdist = { url = "https://files.pythonhosted.org/packages/0e/d5/4db4fd7e756d40db15b07ad072294548b0f508fb35fdd3496fa191223822/docling_slim-2.130.0.tar.gz", hash = "sha256:d45b467d322cd15f03fcc1e792bd4a52aa2dc075dfee229b13ae43a8603f82a6", size = 708768, upload-time = "2026-09-22T15:36:53.243Z" }
+sdist = { url = "https://files.pythonhosted.org/packages/7b/75/ee898b4b7a310c187d0d7ad89002b000efeead0af177f9e50f71554b194b/docling_slim-2.134.0.tar.gz", hash = "sha256:63f12c79bdd0bbe205f244f97f0a8640c590ea46de4d913d23e963ec57e531dd", size = 795345, upload-time = "2026-10-06T04:57:33.257Z" }
 wheels = [
-    { url = "https://files.pythonhosted.org/packages/db/57/9e21add9ec842b314d4d193db8c4186354b872529ef3a17f8aeedbd0c515/docling_slim-2.130.0-py3-none-any.whl", hash = "sha256:913043ee069eb4f107d47ebfca8bdbfbfd7ec84160a7193b7260777e3adbea19", size = 876695, upload-time = "2026-09-22T15:36:51.06Z" },
+    { url = "https://files.pythonhosted.org/packages/cd/58/526bc8f7c7f46dc22afa6f548c67e0bd685cf5f97faf7ee3d96c26355732/docling_slim-2.134.0-py3-none-any.whl", hash = "sha256:818b350e1e9878c0d54bd4ff972bd1918cc9827f193257162cc6f3a797175c29", size = 974718, upload-time = "2026-10-06T04:57:31.022Z" },
 ]
 
 [package.optional-dependencies]
@@ -964,7 +964,7 @@ dependencies = [
 
 [package.metadata]
 requires-dist = [
-    { name = "docling-slim", specifier = "~=2.127" },
+    { name = "docling-slim", specifier = "~=2.134" },
     { name = "icij-common", specifier = "~=0.8.2" },
     { name = "pydantic", specifier = "~=2.13" },
     { name = "pydantic-extra-types", extras = ["pycountry"], specifier = "~=2.11.1" },
@@ -1003,6 +1003,9 @@ benches = [
     { name = "pypdfium2", version = "4.30.0", source = { registry = "https://pypi.org/simple" }, marker = "extra == 'extra-14-extract-python-marker'" },
     { name = "pypdfium2", version = "5.13.0", source = { registry = "https://pypi.org/simple" }, marker = "extra == 'extra-14-extract-python-mineru' or extra != 'extra-14-extract-python-marker'" },
 ]
+cli = [
+    { name = "typer" },
+]
 docling = [
     { name = "docling-slim", extra = ["feat-ocr-easyocr", "feat-ocr-mac", "feat-ocr-rapidocr", "feat-ocr-tesserocr", "standard"] },
 ]
@@ -1046,8 +1049,9 @@ requires-dist = [
     { name = "pypdfium2", marker = "extra == 'benches'", specifier = ">=4.30.0" },
     { name = "python-pptx", marker = "extra == 'mineru'", specifier = "~=1.0" },
     { name = "six", marker = "extra == 'mineru'", specifier = "~=1.17" },
+    { name = "typer", marker = "extra == 'cli'", specifier = "~=0.21.2" },
 ]
-provides-extras = ["benches", "docling", "docling-gpu", "marker", "mineru"]
+provides-extras = ["benches", "cli", "docling", "docling-gpu", "marker", "mineru"]
 
 [package.metadata.requires-dev]
 dev = [
@@ -5186,7 +5190,7 @@ wheels = [
 
 [[package]]
 name = "typer"
-version = "0.24.2"
+version = "0.21.2"
 source = { registry = "https://pypi.org/simple" }
 dependencies = [
     { name = "annotated-doc" },
@@ -5194,9 +5198,9 @@ dependencies = [
     { name = "rich" },
     { name = "shellingham" },
 ]
-sdist = { url = "https://files.pythonhosted.org/packages/83/b8/9ebb531b6c2d377af08ac6746a5df3425b21853a5d2260876919b58a2a4a/typer-0.24.2.tar.gz", hash = "sha256:ec070dcfca1408e85ee203c6365001e818c3b7fffe686fd07ff2d68095ca0480", size = 119849, upload-time = "2026-04-22T17:45:34.413Z" }
+sdist = { url = "https://files.pythonhosted.org/packages/f2/1e/a27cc02a0cd715118c71fa2aef2c687fdefc3c28d90fd0dd789c5118154c/typer-0.21.2.tar.gz", hash = "sha256:1abd95a3b675e17ff61b0838ac637fe9478d446d62ad17fa4bb81ea57cc54028", size = 120426, upload-time = "2026-02-10T19:33:46.182Z" }
 wheels = [
-    { url = "https://files.pythonhosted.org/packages/39/d1/9484b497e0a0410b901c12b8251c3e746e1e863f7d28419ffe06f7892fda/typer-0.24.2-py3-none-any.whl", hash = "sha256:b618bc3d721f9a8d30f3e05565be26416d06e9bcc29d49bc491dc26aba674fa8", size = 55977, upload-time = "2026-04-22T17:45:33.055Z" },
+    { url = "https://files.pythonhosted.org/packages/b8/cc/d59f893fbdfb5f58770c05febfc4086a46875f1084453621c35605cec946/typer-0.21.2-py3-none-any.whl", hash = "sha256:c3d8de54d00347ef90b82131ca946274f017cffb46683ae3883c360fa958f55c", size = 56728, upload-time = "2026-02-10T19:33:48.01Z" },
 ]
 
 [[package]]