docling 1.3.0__py3-none-any.whl → 1.5.0__py3-none-any.whl

docling/backend/docling_parse_backend.py

@@ -1,4 +1,6 @@
+import logging
 import random
+import time
 from io import BytesIO
 from pathlib import Path
 from typing import Iterable, List, Optional, Union
@@ -11,6 +13,8 @@ from pypdfium2 import PdfPage
 from docling.backend.abstract_backend import PdfDocumentBackend, PdfPageBackend
 from docling.datamodel.base_models import BoundingBox, Cell, CoordOrigin, PageSize
 
+_log = logging.getLogger(__name__)
+
 
 class DoclingParsePageBackend(PdfPageBackend):
     def __init__(self, page_obj: PdfPage, docling_page_obj):
@@ -80,7 +84,9 @@ class DoclingParsePageBackend(PdfPageBackend):
             cell_counter += 1
 
         def draw_clusters_and_cells():
-            image = self.get_page_image()
+            image = (
+                self.get_page_image()
+            )  # make new image to avoid drawing on the saved ones
             draw = ImageDraw.Draw(image)
             for c in cells:
                 x0, y0, x1, y1 = c.bbox.as_tuple()
@@ -150,10 +156,19 @@ class DoclingParseDocumentBackend(PdfDocumentBackend):
         super().__init__(path_or_stream)
         self._pdoc = pdfium.PdfDocument(path_or_stream)
         # Parsing cells with docling_parser call
-        if isinstance(path_or_stream, BytesIO):
-            raise NotImplemented("This backend does not support byte streams yet.")
         parser = pdf_parser()
-        self._parser_doc = parser.find_cells(str(path_or_stream))
+
+        start_pb_time = time.time()
+
+        if isinstance(path_or_stream, BytesIO):
+            self._parser_doc = parser.find_cells_from_bytesio(path_or_stream)
+        else:
+            self._parser_doc = parser.find_cells(str(path_or_stream))
+
+        end_pb_time = time.time() - start_pb_time
+        _log.info(
+            f"Time to parse {path_or_stream.name} with docling-parse: time={end_pb_time:.3f}"
+        )
 
     def page_count(self) -> int:
         return len(self._parser_doc["pages"])

docling/backend/pypdfium2_backend.py

@@ -134,7 +134,9 @@ class PyPdfiumPageBackend(PdfPageBackend):
             return merged_cells
 
         def draw_clusters_and_cells():
-            image = self.get_page_image()
+            image = (
+                self.get_page_image()
+            )  # make new image to avoid drawing on the saved ones
             draw = ImageDraw.Draw(image)
             for c in cells:
                 x0, y0, x1, y1 = c.bbox.as_tuple()

docling/datamodel/base_models.py

@@ -1,10 +1,12 @@
 import copy
+import warnings
 from enum import Enum, auto
 from io import BytesIO
-from typing import Any, Dict, List, Optional, Tuple, Union
+from typing import Annotated, Any, Dict, List, Optional, Tuple, Union
 
 from PIL.Image import Image
-from pydantic import BaseModel, ConfigDict, model_validator
+from pydantic import BaseModel, ConfigDict, Field, model_validator
+from typing_extensions import Self
 
 from docling.backend.abstract_backend import PdfPageBackend
 
@@ -234,14 +236,30 @@ class Page(BaseModel):
     model_config = ConfigDict(arbitrary_types_allowed=True)
 
     page_no: int
-    page_hash: str = None
-    size: PageSize = None
-    image: Image = None
+    page_hash: Optional[str] = None
+    size: Optional[PageSize] = None
     cells: List[Cell] = None
     predictions: PagePredictions = PagePredictions()
-    assembled: AssembledUnit = None
+    assembled: Optional[AssembledUnit] = None
 
-    _backend: PdfPageBackend = None  # Internal PDF backend
+    _backend: Optional[PdfPageBackend] = (
+        None  # Internal PDF backend. By default it is cleared during assembling.
+    )
+    _default_image_scale: float = 1.0  # Default image scale for external usage.
+    _image_cache: Dict[float, Image] = (
+        {}
+    )  # Cache of images in different scales. By default it is cleared during assembling.
+
+    def get_image(self, scale: float = 1.0) -> Optional[Image]:
+        if self._backend is None:
+            return self._image_cache.get(scale, None)
+        if not scale in self._image_cache:
+            self._image_cache[scale] = self._backend.get_page_image(scale=scale)
+        return self._image_cache[scale]
+
+    @property
+    def image(self) -> Optional[Image]:
+        return self.get_image(scale=self._default_image_scale)
 
 
 class DocumentStream(BaseModel):
@@ -268,6 +286,19 @@ class PipelineOptions(BaseModel):
 
 
 class AssembleOptions(BaseModel):
-    keep_page_images: bool = (
-        False  # False: page images are removed in the assemble step
-    )
+    keep_page_images: Annotated[
+        bool,
+        Field(
+            deprecated="`keep_page_images` is depreacted, set the value of `images_scale` instead"
+        ),
+    ] = False  # False: page images are removed in the assemble step
+    images_scale: Optional[float] = None  # if set, the scale for generated images
+
+    @model_validator(mode="after")
+    def set_page_images_from_deprecated(self) -> Self:
+        with warnings.catch_warnings():
+            warnings.simplefilter("ignore", DeprecationWarning)
+            default_scale = 1.0
+            if self.keep_page_images and self.images_scale is None:
+                self.images_scale = default_scale
+        return self

docling/datamodel/document.py

@@ -1,7 +1,7 @@
 import logging
 from io import BytesIO
 from pathlib import Path, PurePath
-from typing import ClassVar, Dict, Iterable, List, Optional, Type, Union
+from typing import ClassVar, Dict, Iterable, List, Optional, Tuple, Type, Union
 
 from docling_core.types import BaseCell, BaseText
 from docling_core.types import BoundingBox as DsBoundingBox
@@ -14,13 +14,14 @@ from docling_core.types import TableCell
 from pydantic import BaseModel
 
 from docling.backend.abstract_backend import PdfDocumentBackend
-from docling.backend.pypdfium2_backend import PyPdfiumDocumentBackend
+from docling.backend.docling_parse_backend import DoclingParseDocumentBackend
 from docling.datamodel.base_models import (
     AssembledUnit,
     ConversionStatus,
     DocumentStream,
     FigureElement,
     Page,
+    PageElement,
     TableElement,
     TextElement,
 )
@@ -64,7 +65,7 @@ class InputDocument(BaseModel):
         path_or_stream: Union[BytesIO, Path],
         filename: Optional[str] = None,
         limits: Optional[DocumentLimits] = None,
-        pdf_backend=PyPdfiumDocumentBackend,
+        pdf_backend=DoclingParseDocumentBackend,
     ):
         super().__init__()
 
@@ -302,13 +303,27 @@ class ConvertedDocument(BaseModel):
         else:
             return ""
 
+    def render_element_images(
+        self, element_types: Tuple[PageElement] = (FigureElement,)
+    ):
+        for element in self.assembled.elements:
+            if isinstance(element, element_types):
+                page_ix = element.page_no
+                scale = self.pages[page_ix]._default_image_scale
+                crop_bbox = element.cluster.bbox.scaled(scale=scale).to_top_left_origin(
+                    page_height=self.pages[page_ix].size.height * scale
+                )
+
+                cropped_im = self.pages[page_ix].image.crop(crop_bbox.as_tuple())
+                yield element, cropped_im
+
 
 class DocumentConversionInput(BaseModel):
 
     _path_or_stream_iterator: Iterable[Union[Path, DocumentStream]] = None
     limits: Optional[DocumentLimits] = DocumentLimits()
 
-    DEFAULT_BACKEND: ClassVar = PyPdfiumDocumentBackend
+    DEFAULT_BACKEND: ClassVar = DoclingParseDocumentBackend
 
     def docs(
         self, pdf_backend: Optional[Type[PdfDocumentBackend]] = None

docling/document_converter.py

@@ -188,10 +188,8 @@ class DocumentConverter:
                     # Free up mem resources before moving on with next batch
 
                     # Remove page images (can be disabled)
-                    if not self.assemble_options.keep_page_images:
-                        assembled_page.image = (
-                            None  # Comment this if you want to visualize page images
-                        )
+                    if self.assemble_options.images_scale is None:
+                        assembled_page._image_cache = {}
 
                     # Unload backend
                     assembled_page._backend.unload()
@@ -231,7 +229,15 @@ class DocumentConverter:
 
     # Generate the page image and store it in the page object
     def populate_page_images(self, doc: InputDocument, page: Page) -> Page:
-        page.image = page._backend.get_page_image()
+        # default scale
+        page.get_image(scale=1.0)
+
+        # user requested scales
+        if self.assemble_options.images_scale is not None:
+            page._default_image_scale = self.assemble_options.images_scale
+            page.get_image(
+                scale=self.assemble_options.images_scale
+            )  # this will trigger storing the image in the internal cache
 
         return page
 
@@ -247,7 +253,7 @@ class DocumentConverter:
                 draw.rectangle([(x0, y0), (x1, y1)], outline="red")
             image.show()
 
-        # draw_text_boxes(page.image, cells)
+        # draw_text_boxes(page.get_image(scale=1.0), cells)
 
         return page
 

docling/models/easyocr_model.py

@@ -30,7 +30,7 @@ class EasyOcrModel:
 
         for page in page_batch:
             # rects = page._fpage.
-            high_res_image = page._backend.get_page_image(scale=self.scale)
+            high_res_image = page.get_image(scale=self.scale)
             im = numpy.array(high_res_image)
             result = self.reader.readtext(im)
 

docling/models/layout_model.py

@@ -69,6 +69,10 @@ class LayoutModel:
             "Key-Value Region": 0.45,
         }
 
+        CLASS_REMAPPINGS = {
+            "Document Index": "Table",
+        }
+
         _log.debug("================= Start postprocess function ====================")
         start_time = time.time()
         # Apply Confidence Threshold to cluster predictions
@@ -79,6 +83,10 @@ class LayoutModel:
             confidence = CLASS_THRESHOLDS[cluster.label]
             if cluster.confidence >= confidence:
                 # annotation["created_by"] = "high_conf_pred"
+
+                # Remap class labels where needed.
+                if cluster.label in CLASS_REMAPPINGS.keys():
+                    cluster.label = CLASS_REMAPPINGS[cluster.label]
                 clusters_out.append(cluster)
 
         # map to dictionary clusters and cells, with bottom left origin
@@ -259,7 +267,9 @@ class LayoutModel:
     def __call__(self, page_batch: Iterable[Page]) -> Iterable[Page]:
         for page in page_batch:
             clusters = []
-            for ix, pred_item in enumerate(self.layout_predictor.predict(page.image)):
+            for ix, pred_item in enumerate(
+                self.layout_predictor.predict(page.get_image(scale=1.0))
+            ):
                 cluster = Cluster(
                     id=ix,
                     label=pred_item["label"],

docling/models/table_structure_model.py

@@ -34,7 +34,9 @@ class TableStructureModel:
             self.scale = 2.0  # Scale up table input images to 144 dpi
 
     def draw_table_and_cells(self, page: Page, tbl_list: List[TableElement]):
-        image = page._backend.get_page_image()
+        image = (
+            page._backend.get_page_image()
+        )  # make new image to avoid drawing on the saved ones
         draw = ImageDraw.Draw(image)
 
         for table_element in tbl_list:
@@ -94,13 +96,7 @@ class TableStructureModel:
                 "width": page.size.width * self.scale,
                 "height": page.size.height * self.scale,
             }
-            # add image to page input.
-            if self.scale == 1.0:
-                page_input["image"] = numpy.asarray(page.image)
-            else:  # render new page image on the fly at desired scale
-                page_input["image"] = numpy.asarray(
-                    page._backend.get_page_image(scale=self.scale)
-                )
+            page_input["image"] = numpy.asarray(page.get_image(scale=self.scale))
 
             table_clusters, table_bboxes = zip(*in_tables)
 

{docling-1.3.0.dist-info → docling-1.5.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: docling
-Version: 1.3.0
+Version: 1.5.0
 Summary: Docling PDF conversion package
 Home-page: https://github.com/DS4SD/docling
 License: MIT
@@ -24,8 +24,8 @@ Provides-Extra: ocr
 Requires-Dist: certifi (>=2024.7.4)
 Requires-Dist: deepsearch-glm (>=0.19.0,<1)
 Requires-Dist: docling-core (>=1.1.2,<2.0.0)
-Requires-Dist: docling-ibm-models (>=1.1.0,<2.0.0)
-Requires-Dist: docling-parse (>=0.0.1,<0.0.2)
+Requires-Dist: docling-ibm-models (>=1.1.1,<2.0.0)
+Requires-Dist: docling-parse (>=0.2.0,<0.3.0)
 Requires-Dist: easyocr (>=1.7,<2.0) ; extra == "easyocr" or extra == "ocr"
 Requires-Dist: filetype (>=1.2.0,<2.0.0)
 Requires-Dist: huggingface_hub (>=0.23,<1)
@@ -44,6 +44,7 @@ Description-Content-Type: text/markdown
 
 # Docling
 
+[![arXiv](https://img.shields.io/badge/arXiv-2408.09869-b31b1b.svg)](https://arxiv.org/abs/2408.09869)
 [![PyPI version](https://img.shields.io/pypi/v/docling)](https://pypi.org/project/docling/)
 ![Python](https://img.shields.io/badge/python-3.10%20%7C%203.11%20%7C%203.12-blue)
 [![Poetry](https://img.shields.io/endpoint?url=https://python-poetry.org/badge/v0.json)](https://python-poetry.org/)
@@ -172,13 +173,15 @@ Please read [Contributing to Docling](https://github.com/DS4SD/docling/blob/main
 If you use Docling in your projects, please consider citing the following:
 
 ```bib
-@software{Docling,
-author = {Deep Search Team},
-month = {7},
-title = {{Docling}},
-url = {https://github.com/DS4SD/docling},
-version = {main},
-year = {2024}
+@techreport{Docling,
+  author = {Deep Search Team},
+  month = {8},
+  title = {{Docling Technical Report}},
+  url={https://arxiv.org/abs/2408.09869},
+  eprint={2408.09869},
+  doi = "10.48550/arXiv.2408.09869",
+  version = {1.0.0},
+  year = {2024}
 }
 ```
 

{docling-1.3.0.dist-info → docling-1.5.0.dist-info}/RECORD

@@ -1,26 +1,26 @@
 docling/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 docling/backend/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 docling/backend/abstract_backend.py,sha256=swwmXzNueZSHqEOvw4j-IFhP2OUJhBeB--gV7NtzKgo,1112
-docling/backend/docling_parse_backend.py,sha256=mGuJCpMVqyrZK-cXKRWrELPz0Wt1h6uydx4QwWI1rew,5912
-docling/backend/pypdfium2_backend.py,sha256=tv6JxyTkTdT2qr2ghsQgYA2zgpCDxKYSdHVBTAR7FSk,7411
+docling/backend/docling_parse_backend.py,sha256=-bIjYJ-80R2SArAEw_lAyzgW5_BFEoX83n1oBMmUGF4,6284
+docling/backend/pypdfium2_backend.py,sha256=3Qeeal8z6DunUe4S10Z2TXrdeucanCpa8evt6SQtpKQ,7496
 docling/datamodel/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-docling/datamodel/base_models.py,sha256=irZLAHdsROOOwRbywKIA0mk3H8GrLwtGjOgTV6G0QoU,7004
-docling/datamodel/document.py,sha256=FG_ntDFRBWj-MhV52D0sC8XaZOwN3yryyXahsVHGnyI,12517
+docling/datamodel/base_models.py,sha256=uOq0zjUS60aIkROREiypp3Jn1yqQTlWEf34jXTT43ls,8391
+docling/datamodel/document.py,sha256=Dgi9pSwXCgIoR26MKiRDiVMyMaFKdvGSKq2Fm5Lef9M,13173
 docling/datamodel/settings.py,sha256=t5g6wrEJnPa9gBzMMl8ppgBRUYz-8xgopEtfMS0ZH28,733
-docling/document_converter.py,sha256=dMucsq6M_nwPsC1ChogVwJgNDv8sJuFklQWWinDZaug,10246
+docling/document_converter.py,sha256=r9z48VjL_hkq-rbAgyZ135njzUGBJ5AnhEH6-1zfyCA,10490
 docling/models/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 docling/models/ds_glm_model.py,sha256=wmb--2JKFQby-kvidw6PyM8wURPXYPQ_Z_eKKCBAdYQ,3192
-docling/models/easyocr_model.py,sha256=NaHVs8IN0eW9KB076E2Kae1s-bq74_4IMWueze9QqtE,2290
-docling/models/layout_model.py,sha256=4AfPFiu6pXc8wIQ1sQlEZnHRt7SnBmfzDdctiRveOWw,10944
+docling/models/easyocr_model.py,sha256=Y-RWolIFE3By6gk8dnb2qFy7Cr9qcHs6eo65fWPT0Nc,2276
+docling/models/layout_model.py,sha256=ZFmaLXlRWUfsT1pJCiYVxhQFrBBsiz6Aw0m9GM3UvVM,11249
 docling/models/page_assemble_model.py,sha256=8eoG2WiFxPxq9TPvM-wkngb2gkr0tdtCRVXg1JcTETo,5550
-docling/models/table_structure_model.py,sha256=xUmfunZNYC30P0fRdESdztqy1FVlMzlhJjLBp-xcn4A,5638
+docling/models/table_structure_model.py,sha256=lKsodvfZaGwxOHp-CbRW5nzCKZYMwf770h0Ka6Bdbgw,5451
 docling/pipeline/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 docling/pipeline/base_model_pipeline.py,sha256=ozHdJak0yQAxQf7pQN_C480vI35A2e5KL5Qq1xSkq5c,560
 docling/pipeline/standard_model_pipeline.py,sha256=UTwodKUKrisLoVcntbNUBDhjzRyFvpdUvyVw-gNmBlM,1541
 docling/utils/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 docling/utils/layout_utils.py,sha256=FOFbL0hKzUoWXdZaeUvEtFqKv0IkPifIr4sdGW4suKs,31804
 docling/utils/utils.py,sha256=llhXSbIDNZ1MHOwBEfLHBAoJIAYI7QlPIonlI1jLUJ0,1208
-docling-1.3.0.dist-info/LICENSE,sha256=ACwmltkrXIz5VsEQcrqljq-fat6ZXAMepjXGoe40KtE,1069
-docling-1.3.0.dist-info/METADATA,sha256=wi2DOn77z_BIMSLsrmzebYZUgpjHYWbNTOIVEY3A4-o,7042
-docling-1.3.0.dist-info/WHEEL,sha256=sP946D7jFCHeNz5Iq4fL4Lu-PrWrFsgfLXbbkciIZwg,88
-docling-1.3.0.dist-info/RECORD,,
+docling-1.5.0.dist-info/LICENSE,sha256=ACwmltkrXIz5VsEQcrqljq-fat6ZXAMepjXGoe40KtE,1069
+docling-1.5.0.dist-info/METADATA,sha256=jWcjsrdfYcpeYFCRQ1h5C1b8MyaKsJWyUhGheXQEGvY,7235
+docling-1.5.0.dist-info/WHEEL,sha256=sP946D7jFCHeNz5Iq4fL4Lu-PrWrFsgfLXbbkciIZwg,88
+docling-1.5.0.dist-info/RECORD,,