docling 2.29.0__py3-none-any.whl → 2.31.0__py3-none-any.whl

docling/models/readingorder_model.py

@@ -1,12 +1,7 @@
-import copy
-import random
 from pathlib import Path
 from typing import Dict, List
 
 from docling_core.types.doc import (
-    BoundingBox,
-    CoordOrigin,
-    DocItem,
     DocItemLabel,
     DoclingDocument,
     DocumentOrigin,
@@ -17,13 +12,10 @@ from docling_core.types.doc import (
     TableData,
 )
 from docling_core.types.doc.document import ContentLayer
-from docling_core.types.legacy_doc.base import Ref
-from docling_core.types.legacy_doc.document import BaseText
 from docling_ibm_models.reading_order.reading_order_rb import (
     PageElement as ReadingOrderPageElement,
+    ReadingOrderPredictor,
 )
-from docling_ibm_models.reading_order.reading_order_rb import ReadingOrderPredictor
-from PIL import ImageDraw
 from pydantic import BaseModel, ConfigDict
 
 from docling.datamodel.base_models import (
@@ -35,7 +27,6 @@ from docling.datamodel.base_models import (
     TextElement,
 )
 from docling.datamodel.document import ConversionResult
-from docling.datamodel.settings import settings
 from docling.utils.profiling import ProfilingScope, TimeRecorder
 
 
@@ -53,12 +44,10 @@ class ReadingOrderModel:
     def _assembled_to_readingorder_elements(
         self, conv_res: ConversionResult
     ) -> List[ReadingOrderPageElement]:
-
         elements: List[ReadingOrderPageElement] = []
         page_no_to_pages = {p.page_no: p for p in conv_res.pages}
 
         for element in conv_res.assembled.elements:
-
             page_height = page_no_to_pages[element.page_no].size.height  # type: ignore
             bbox = element.cluster.bbox.to_bottom_left_origin(page_height)
             text = element.text or ""
@@ -84,7 +73,6 @@ class ReadingOrderModel:
     def _add_child_elements(
         self, element: BasePageElement, doc_item: NodeItem, doc: DoclingDocument
     ):
-
         child: Cluster
         for child in element.cluster.children:
             c_label = child.label
@@ -110,7 +98,7 @@ class ReadingOrderModel:
             else:
                 doc.add_text(parent=doc_item, label=c_label, text=c_text, prov=c_prov)
 
-    def _readingorder_elements_to_docling_doc(
+    def _readingorder_elements_to_docling_doc(  # noqa: C901
         self,
         conv_res: ConversionResult,
         ro_elements: List[ReadingOrderPageElement],
@@ -118,7 +106,6 @@ class ReadingOrderModel:
         el_to_footnotes_mapping: Dict[int, List[int]],
         el_merges_mapping: Dict[int, List[int]],
     ) -> DoclingDocument:
-
         id_to_elem = {
             RefItem(cref=f"#/{elem.page_no}/{elem.cluster.id}").cref: elem
             for elem in conv_res.assembled.elements
@@ -192,7 +179,6 @@ class ReadingOrderModel:
 
                             code_item.footnotes.append(new_footnote_item.get_ref())
                 else:
-
                     new_item, current_list = self._handle_text_element(
                         element, out_doc, current_list, page_height
                     )
@@ -206,7 +192,6 @@ class ReadingOrderModel:
                             )
 
             elif isinstance(element, Table):
-
                 tbl_data = TableData(
                     num_rows=element.num_rows,
                     num_cols=element.num_cols,
@@ -342,12 +327,12 @@ class ReadingOrderModel:
         return new_item, current_list
 
     def _merge_elements(self, element, merged_elem, new_item, page_height):
-        assert isinstance(
-            merged_elem, type(element)
-        ), "Merged element must be of same type as element."
-        assert (
-            merged_elem.label == new_item.label
-        ), "Labels of merged elements must match."
+        assert isinstance(merged_elem, type(element)), (
+            "Merged element must be of same type as element."
+        )
+        assert merged_elem.label == new_item.label, (
+            "Labels of merged elements must match."
+        )
         prov = ProvenanceItem(
             page_no=element.page_no + 1,
             charspan=(

docling/models/table_structure_model.py

@@ -1,13 +1,13 @@
 import copy
 import warnings
+from collections.abc import Iterable
 from pathlib import Path
-from typing import Iterable, Optional, Union
+from typing import Optional
 
 import numpy
 from docling_core.types.doc import BoundingBox, DocItemLabel, TableCell
 from docling_core.types.doc.page import (
     BoundingRectangle,
-    SegmentedPdfPage,
     TextCellUnit,
 )
 from docling_ibm_models.tableformer.data_management.tf_predictor import TFPredictor
@@ -44,7 +44,6 @@ class TableStructureModel(BasePageModel):
 
         self.enabled = enabled
         if self.enabled:
-
             if artifacts_path is None:
                 artifacts_path = self.download_models() / self._model_path
             else:
@@ -175,7 +174,6 @@ class TableStructureModel(BasePageModel):
     def __call__(
         self, conv_res: ConversionResult, page_batch: Iterable[Page]
     ) -> Iterable[Page]:
-
         if not self.enabled:
             yield from page_batch
             return
@@ -186,7 +184,6 @@ class TableStructureModel(BasePageModel):
                 yield page
             else:
                 with TimeRecorder(conv_res, "table_structure"):
-
                     assert page.predictions.layout is not None
                     assert page.size is not None
 
@@ -260,7 +257,6 @@ class TableStructureModel(BasePageModel):
                             table_out = tf_output[0]
                             table_cells = []
                             for element in table_out["tf_responses"]:
-
                                 if not self.do_cell_matching:
                                     the_bbox = BoundingBox.model_validate(
                                         element["bbox"]

docling/models/tesseract_ocr_cli_model.py

@@ -3,9 +3,10 @@ import io
 import logging
 import os
 import tempfile
+from collections.abc import Iterable
 from pathlib import Path
 from subprocess import DEVNULL, PIPE, Popen
-from typing import Iterable, List, Optional, Tuple, Type
+from typing import List, Optional, Tuple, Type
 
 import pandas as pd
 from docling_core.types.doc import BoundingBox, CoordOrigin
@@ -63,8 +64,7 @@ class TesseractOcrCliModel(BaseOcrModel):
                 )
 
     def _get_name_and_version(self) -> Tuple[str, str]:
-
-        if self._name != None and self._version != None:
+        if self._name is not None and self._version is not None:
             return self._name, self._version  # type: ignore
 
         cmd = [self.options.tesseract_cmd, "--version"]
@@ -125,14 +125,16 @@ class TesseractOcrCliModel(BaseOcrModel):
         # _log.info(decoded_data)
 
         # Read the TSV file generated by Tesseract
-        df = pd.read_csv(io.StringIO(decoded_data), quoting=csv.QUOTE_NONE, sep="\t")
+        df_result = pd.read_csv(
+            io.StringIO(decoded_data), quoting=csv.QUOTE_NONE, sep="\t"
+        )
 
         # Display the dataframe (optional)
         # _log.info("df: ", df.head())
 
         # Filter rows that contain actual text (ignore header or empty rows)
-        df_filtered = df[
-            df["text"].notnull() & (df["text"].apply(str).str.strip() != "")
+        df_filtered = df_result[
+            df_result["text"].notna() & (df_result["text"].apply(str).str.strip() != "")
         ]
 
         return df_filtered
@@ -149,10 +151,10 @@ class TesseractOcrCliModel(BaseOcrModel):
         proc = Popen(cmd, stdout=PIPE, stderr=DEVNULL)
         output, _ = proc.communicate()
         decoded_data = output.decode("utf-8")
-        df = pd.read_csv(
+        df_detected = pd.read_csv(
             io.StringIO(decoded_data), sep=":", header=None, names=["key", "value"]
         )
-        scripts = df.loc[df["key"] == "Script"].value.tolist()
+        scripts = df_detected.loc[df_detected["key"] == "Script"].value.tolist()
         if len(scripts) == 0:
             _log.warning("Tesseract cannot detect the script of the page")
             return None
@@ -183,11 +185,11 @@ class TesseractOcrCliModel(BaseOcrModel):
         proc = Popen(cmd, stdout=PIPE, stderr=DEVNULL)
         output, _ = proc.communicate()
         decoded_data = output.decode("utf-8")
-        df = pd.read_csv(io.StringIO(decoded_data), header=None)
-        self._tesseract_languages = df[0].tolist()[1:]
+        df_list = pd.read_csv(io.StringIO(decoded_data), header=None)
+        self._tesseract_languages = df_list[0].tolist()[1:]
 
         # Decide the script prefix
-        if any([l.startswith("script/") for l in self._tesseract_languages]):
+        if any(lang.startswith("script/") for lang in self._tesseract_languages):
             script_prefix = "script/"
         else:
             script_prefix = ""
@@ -197,7 +199,6 @@ class TesseractOcrCliModel(BaseOcrModel):
     def __call__(
         self, conv_res: ConversionResult, page_batch: Iterable[Page]
     ) -> Iterable[Page]:
-
         if not self.enabled:
             yield from page_batch
             return
@@ -225,19 +226,19 @@ class TesseractOcrCliModel(BaseOcrModel):
                                 fname = image_file.name
                                 high_res_image.save(image_file)
 
-                            df = self._run_tesseract(fname)
+                            df_result = self._run_tesseract(fname)
                         finally:
                             if os.path.exists(fname):
                                 os.remove(fname)
 
-                        # _log.info(df)
+                        # _log.info(df_result)
 
                         # Print relevant columns (bounding box and text)
-                        for ix, row in df.iterrows():
+                        for ix, row in df_result.iterrows():
                             text = row["text"]
                             conf = row["conf"]
 
-                            l = float(row["left"])
+                            l = float(row["left"])  # noqa: E741
                             b = float(row["top"])
                             w = float(row["width"])
                             h = float(row["height"])

docling/models/tesseract_ocr_model.py

@@ -1,6 +1,7 @@
 import logging
+from collections.abc import Iterable
 from pathlib import Path
-from typing import Iterable, Optional, Type
+from typing import Optional, Type
 
 from docling_core.types.doc import BoundingBox, CoordOrigin
 from docling_core.types.doc.page import BoundingRectangle, TextCell
@@ -37,9 +38,6 @@ class TesseractOcrModel(BaseOcrModel):
         self.options: TesseractOcrOptions
 
         self.scale = 3  # multiplier for 72 dpi == 216 dpi.
-        self.reader = None
-        self.osd_reader = None
-        self.script_readers: dict[str, tesserocr.PyTessBaseAPI] = {}
 
         if self.enabled:
             install_errmsg = (
@@ -64,7 +62,7 @@ class TesseractOcrModel(BaseOcrModel):
                 raise ImportError(install_errmsg)
             try:
                 tesseract_version = tesserocr.tesseract_version()
-            except:
+            except Exception:
                 raise ImportError(install_errmsg)
 
             _, self._tesserocr_languages = tesserocr.get_languages()
@@ -75,7 +73,7 @@ class TesseractOcrModel(BaseOcrModel):
             _log.debug("Initializing TesserOCR: %s", tesseract_version)
             lang = "+".join(self.options.lang)
 
-            if any([l.startswith("script/") for l in self._tesserocr_languages]):
+            if any(lang.startswith("script/") for lang in self._tesserocr_languages):
                 self.script_prefix = "script/"
             else:
                 self.script_prefix = ""
@@ -86,6 +84,10 @@ class TesseractOcrModel(BaseOcrModel):
                 "oem": tesserocr.OEM.DEFAULT,
             }
 
+            self.reader = None
+            self.osd_reader = None
+            self.script_readers: dict[str, tesserocr.PyTessBaseAPI] = {}
+
             if self.options.path is not None:
                 tesserocr_kwargs["path"] = self.options.path
 

docling/pipeline/base_pipeline.py

@@ -3,9 +3,10 @@ import logging
 import time
 import traceback
 from abc import ABC, abstractmethod
-from typing import Any, Callable, Iterable, List
+from collections.abc import Iterable
+from typing import Any, Callable, List
 
-from docling_core.types.doc import DoclingDocument, NodeItem
+from docling_core.types.doc import NodeItem
 
 from docling.backend.abstract_backend import AbstractDocumentBackend
 from docling.backend.pdf_backend import PdfDocumentBackend
@@ -64,7 +65,6 @@ class BasePipeline(ABC):
         return conv_res
 
     def _enrich_document(self, conv_res: ConversionResult) -> ConversionResult:
-
         def _prepare_elements(
             conv_res: ConversionResult, model: GenericEnrichmentModel[Any]
         ) -> Iterable[NodeItem]:
@@ -113,7 +113,6 @@ class BasePipeline(ABC):
 
 
 class PaginatedPipeline(BasePipeline):  # TODO this is a bad name.
-
     def __init__(self, pipeline_options: PipelineOptions):
         super().__init__(pipeline_options)
         self.keep_backend = False
@@ -127,7 +126,6 @@ class PaginatedPipeline(BasePipeline):  # TODO this is a bad name.
         yield from page_batch
 
     def _build_document(self, conv_res: ConversionResult) -> ConversionResult:
-
         if not isinstance(conv_res.input._backend, PdfDocumentBackend):
             raise RuntimeError(
                 f"The selected backend {type(conv_res.input._backend).__name__} for {conv_res.input.file} is not a PDF backend. "
@@ -139,8 +137,7 @@ class PaginatedPipeline(BasePipeline):  # TODO this is a bad name.
 
         total_elapsed_time = 0.0
         with TimeRecorder(conv_res, "doc_build", scope=ProfilingScope.DOCUMENT):
-
-            for i in range(0, conv_res.input.page_count):
+            for i in range(conv_res.input.page_count):
                 start_page, end_page = conv_res.input.limits.page_range
                 if (start_page - 1) <= i <= (end_page - 1):
                     conv_res.pages.append(Page(page_no=i))
@@ -161,7 +158,6 @@ class PaginatedPipeline(BasePipeline):  # TODO this is a bad name.
                     pipeline_pages = self._apply_on_pages(conv_res, init_pages)
 
                     for p in pipeline_pages:  # Must exhaust!
-
                         # Cleanup cached images
                         if not self.keep_images:
                             p._image_cache = {}

docling/pipeline/simple_pipeline.py

@@ -24,7 +24,6 @@ class SimplePipeline(BasePipeline):
         super().__init__(pipeline_options)
 
     def _build_document(self, conv_res: ConversionResult) -> ConversionResult:
-
         if not isinstance(conv_res.input._backend, DeclarativeDocumentBackend):
             raise RuntimeError(
                 f"The selected backend {type(conv_res.input._backend).__name__} for {conv_res.input.file} is not a declarative backend. "

docling/pipeline/standard_pdf_pipeline.py

@@ -1,8 +1,7 @@
 import logging
-import sys
 import warnings
 from pathlib import Path
-from typing import Optional
+from typing import Optional, cast
 
 from docling_core.types.doc import DocItem, ImageRef, PictureItem, TableItem
 
@@ -226,7 +225,11 @@ class StandardPdfPipeline(PaginatedPipeline):
                         and self.pipeline_options.generate_table_images
                     ):
                         page_ix = element.prov[0].page_no - 1
-                        page = conv_res.pages[page_ix]
+                        page = next(
+                            (p for p in conv_res.pages if p.page_no == page_ix),
+                            cast("Page", None),
+                        )
+                        assert page is not None
                         assert page.size is not None
                         assert page.image is not None
 

docling/pipeline/vlm_pipeline.py

@@ -1,5 +1,4 @@
 import logging
-import warnings
 from io import BytesIO
 from pathlib import Path
 from typing import List, Optional, Union, cast
@@ -15,11 +14,14 @@ from docling.backend.pdf_backend import PdfDocumentBackend
 from docling.datamodel.base_models import InputFormat, Page
 from docling.datamodel.document import ConversionResult, InputDocument
 from docling.datamodel.pipeline_options import (
+    ApiVlmOptions,
+    HuggingFaceVlmOptions,
     InferenceFramework,
     ResponseFormat,
     VlmPipelineOptions,
 )
 from docling.datamodel.settings import settings
+from docling.models.api_vlm_model import ApiVlmModel
 from docling.models.hf_mlx_model import HuggingFaceMlxModel
 from docling.models.hf_vlm_model import HuggingFaceVlmModel
 from docling.pipeline.base_pipeline import PaginatedPipeline
@@ -29,7 +31,6 @@ _log = logging.getLogger(__name__)
 
 
 class VlmPipeline(PaginatedPipeline):
-
     def __init__(self, pipeline_options: VlmPipelineOptions):
         super().__init__(pipeline_options)
         self.keep_backend = True
@@ -57,27 +58,34 @@ class VlmPipeline(PaginatedPipeline):
 
         self.keep_images = self.pipeline_options.generate_page_images
 
-        if (
-            self.pipeline_options.vlm_options.inference_framework
-            == InferenceFramework.MLX
-        ):
+        if isinstance(pipeline_options.vlm_options, ApiVlmOptions):
             self.build_pipe = [
-                HuggingFaceMlxModel(
+                ApiVlmModel(
                     enabled=True,  # must be always enabled for this pipeline to make sense.
-                    artifacts_path=artifacts_path,
-                    accelerator_options=pipeline_options.accelerator_options,
-                    vlm_options=self.pipeline_options.vlm_options,
-                ),
-            ]
-        else:
-            self.build_pipe = [
-                HuggingFaceVlmModel(
-                    enabled=True,  # must be always enabled for this pipeline to make sense.
-                    artifacts_path=artifacts_path,
-                    accelerator_options=pipeline_options.accelerator_options,
-                    vlm_options=self.pipeline_options.vlm_options,
+                    enable_remote_services=self.pipeline_options.enable_remote_services,
+                    vlm_options=cast(ApiVlmOptions, self.pipeline_options.vlm_options),
                 ),
             ]
+        elif isinstance(self.pipeline_options.vlm_options, HuggingFaceVlmOptions):
+            vlm_options = cast(HuggingFaceVlmOptions, self.pipeline_options.vlm_options)
+            if vlm_options.inference_framework == InferenceFramework.MLX:
+                self.build_pipe = [
+                    HuggingFaceMlxModel(
+                        enabled=True,  # must be always enabled for this pipeline to make sense.
+                        artifacts_path=artifacts_path,
+                        accelerator_options=pipeline_options.accelerator_options,
+                        vlm_options=vlm_options,
+                    ),
+                ]
+            else:
+                self.build_pipe = [
+                    HuggingFaceVlmModel(
+                        enabled=True,  # must be always enabled for this pipeline to make sense.
+                        artifacts_path=artifacts_path,
+                        accelerator_options=pipeline_options.accelerator_options,
+                        vlm_options=vlm_options,
+                    ),
+                ]
 
         self.enrichment_pipe = [
             # Other models working on `NodeItem` elements in the DoclingDocument
@@ -104,7 +112,6 @@ class VlmPipeline(PaginatedPipeline):
 
     def _assemble_document(self, conv_res: ConversionResult) -> ConversionResult:
         with TimeRecorder(conv_res, "doc_assemble", scope=ProfilingScope.DOCUMENT):
-
             if (
                 self.pipeline_options.vlm_options.response_format
                 == ResponseFormat.DOCTAGS

docling/utils/api_image_request.py

@@ -0,0 +1,61 @@
+import base64
+import logging
+from io import BytesIO
+from typing import Dict, Optional
+
+import requests
+from PIL import Image
+from pydantic import AnyUrl
+
+from docling.datamodel.base_models import OpenAiApiResponse
+
+_log = logging.getLogger(__name__)
+
+
+def api_image_request(
+    image: Image.Image,
+    prompt: str,
+    url: AnyUrl,
+    timeout: float = 20,
+    headers: Optional[Dict[str, str]] = None,
+    **params,
+) -> str:
+    img_io = BytesIO()
+    image.save(img_io, "PNG")
+    image_base64 = base64.b64encode(img_io.getvalue()).decode("utf-8")
+    messages = [
+        {
+            "role": "user",
+            "content": [
+                {
+                    "type": "image_url",
+                    "image_url": {"url": f"data:image/png;base64,{image_base64}"},
+                },
+                {
+                    "type": "text",
+                    "text": prompt,
+                },
+            ],
+        }
+    ]
+
+    payload = {
+        "messages": messages,
+        **params,
+    }
+
+    headers = headers or {}
+
+    r = requests.post(
+        str(url),
+        headers=headers,
+        json=payload,
+        timeout=timeout,
+    )
+    if not r.ok:
+        _log.error(f"Error calling the API. Response was {r.text}")
+    r.raise_for_status()
+
+    api_resp = OpenAiApiResponse.model_validate_json(r.text)
+    generated_text = api_resp.choices[0].message.content.strip()
+    return generated_text

docling/utils/export.py

@@ -1,8 +1,8 @@
 import logging
-from typing import Any, Dict, Iterable, List, Tuple, Union
+from collections.abc import Iterable
+from typing import Any, Dict, List, Tuple, Union
 
 from docling_core.types.doc import BoundingBox, CoordOrigin
-from docling_core.types.doc.page import TextCell
 from docling_core.types.legacy_doc.base import BaseCell, BaseText, Ref, Table
 
 from docling.datamodel.document import ConversionResult, Page
@@ -13,7 +13,6 @@ _log = logging.getLogger(__name__)
 def generate_multimodal_pages(
     doc_result: ConversionResult,
 ) -> Iterable[Tuple[str, str, List[Dict[str, Any]], List[Dict[str, Any]], Page]]:
-
     label_to_doclaynet = {
         "title": "title",
         "table-of-contents": "document_index",
@@ -122,7 +121,6 @@ def generate_multimodal_pages(
     if doc.main_text is None:
         return
     for ix, orig_item in enumerate(doc.main_text):
-
         item = doc._resolve_ref(orig_item) if isinstance(orig_item, Ref) else orig_item
         if item is None or item.prov is None or len(item.prov) == 0:
             _log.debug(f"Skipping item {orig_item}")

docling/utils/glm_utils.py

@@ -29,7 +29,7 @@ def resolve_item(paths, obj):
 
     try:
         key = int(paths[0])
-    except:
+    except Exception:
         key = paths[0]
 
     if len(paths) == 1:
@@ -67,7 +67,7 @@ def _flatten_table_grid(grid: List[List[dict]]) -> List[dict]:
     return unique_objects
 
 
-def to_docling_document(doc_glm, update_name_label=False) -> DoclingDocument:
+def to_docling_document(doc_glm, update_name_label=False) -> DoclingDocument:  # noqa: C901
     origin = DocumentOrigin(
         mimetype="application/pdf",
         filename=doc_glm["file-info"]["filename"],

docling/utils/layout_postprocessor.py

@@ -18,7 +18,7 @@ class UnionFind:
 
     def __init__(self, elements):
         self.parent = {elem: elem for elem in elements}
-        self.rank = {elem: 0 for elem in elements}
+        self.rank = dict.fromkeys(elements, 0)
 
     def find(self, x):
         if self.parent[x] != x:
@@ -484,7 +484,9 @@ class LayoutPostprocessor:
         spatial_index = (
             self.regular_index
             if cluster_type == "regular"
-            else self.picture_index if cluster_type == "picture" else self.wrapper_index
+            else self.picture_index
+            if cluster_type == "picture"
+            else self.wrapper_index
         )
 
         # Map of currently valid clusters

docling/utils/model_downloader.py

@@ -37,7 +37,7 @@ def download_models(
     output_dir.mkdir(exist_ok=True, parents=True)
 
     if with_layout:
-        _log.info(f"Downloading layout model...")
+        _log.info("Downloading layout model...")
         LayoutModel.download_models(
             local_dir=output_dir / LayoutModel._model_repo_folder,
             force=force,
@@ -45,7 +45,7 @@ def download_models(
         )
 
     if with_tableformer:
-        _log.info(f"Downloading tableformer model...")
+        _log.info("Downloading tableformer model...")
         TableStructureModel.download_models(
             local_dir=output_dir / TableStructureModel._model_repo_folder,
             force=force,
@@ -53,7 +53,7 @@ def download_models(
         )
 
     if with_picture_classifier:
-        _log.info(f"Downloading picture classifier model...")
+        _log.info("Downloading picture classifier model...")
         DocumentPictureClassifier.download_models(
             local_dir=output_dir / DocumentPictureClassifier._model_repo_folder,
             force=force,
@@ -61,7 +61,7 @@ def download_models(
         )
 
     if with_code_formula:
-        _log.info(f"Downloading code formula model...")
+        _log.info("Downloading code formula model...")
         CodeFormulaModel.download_models(
             local_dir=output_dir / CodeFormulaModel._model_repo_folder,
             force=force,
@@ -69,7 +69,7 @@ def download_models(
         )
 
     if with_smolvlm:
-        _log.info(f"Downloading SmolVlm model...")
+        _log.info("Downloading SmolVlm model...")
         PictureDescriptionVlmModel.download_models(
             repo_id=smolvlm_picture_description.repo_id,
             local_dir=output_dir / smolvlm_picture_description.repo_cache_folder,
@@ -78,7 +78,7 @@ def download_models(
         )
 
     if with_granite_vision:
-        _log.info(f"Downloading Granite Vision model...")
+        _log.info("Downloading Granite Vision model...")
         PictureDescriptionVlmModel.download_models(
             repo_id=granite_picture_description.repo_id,
             local_dir=output_dir / granite_picture_description.repo_cache_folder,
@@ -87,7 +87,7 @@ def download_models(
         )
 
     if with_easyocr:
-        _log.info(f"Downloading easyocr models...")
+        _log.info("Downloading easyocr models...")
         EasyOcrModel.download_models(
             local_dir=output_dir / EasyOcrModel._model_repo_folder,
             force=force,

docling/utils/utils.py

@@ -13,7 +13,7 @@ def chunkify(iterator, chunk_size):
     if isinstance(iterator, List):
         iterator = iter(iterator)
     for first in iterator:  # Take the first element from the iterator
-        yield [first] + list(islice(iterator, chunk_size - 1))
+        yield [first, *list(islice(iterator, chunk_size - 1))]
 
 
 def create_file_hash(path_or_stream: Union[BytesIO, Path]) -> str: