docling 1.2.1__py3-none-any.whl → 1.4.0__py3-none-any.whl

docling/backend/docling_parse_backend.py

@@ -150,10 +150,11 @@ 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))
+        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))
 
     def page_count(self) -> int:
         return len(self._parser_doc["pages"])

docling/datamodel/base_models.py

@@ -265,3 +265,9 @@ class PipelineOptions(BaseModel):
     do_ocr: bool = False  # True: perform OCR, replace programmatic PDF text
 
     table_structure_options: TableStructureOptions = TableStructureOptions()
+
+
+class AssembleOptions(BaseModel):
+    keep_page_images: bool = (
+        False  # False: page images are removed in the assemble step
+    )

docling/datamodel/document.py

@@ -14,7 +14,7 @@ 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,
@@ -64,7 +64,7 @@ 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__()
 
@@ -308,7 +308,7 @@ 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

@@ -14,6 +14,7 @@ from pydantic import AnyHttpUrl, TypeAdapter, ValidationError
 from docling.backend.abstract_backend import PdfDocumentBackend
 from docling.datamodel.base_models import (
     AssembledUnit,
+    AssembleOptions,
     ConversionStatus,
     Page,
     PipelineOptions,
@@ -44,6 +45,7 @@ class DocumentConverter:
         pipeline_options: PipelineOptions = PipelineOptions(),
         pdf_backend: Type[PdfDocumentBackend] = DocumentConversionInput.DEFAULT_BACKEND,
         pipeline_cls: Type[BaseModelPipeline] = StandardModelPipeline,
+        assemble_options: AssembleOptions = AssembleOptions(),
     ):
         if not artifacts_path:
             artifacts_path = self.download_models_hf()
@@ -57,6 +59,7 @@ class DocumentConverter:
         self.page_assemble_model = PageAssembleModel(config={})
         self.glm_model = GlmModel(config={})
         self.pdf_backend = pdf_backend
+        self.assemble_options = assemble_options
 
     @staticmethod
     def download_models_hf(
@@ -174,17 +177,23 @@ class DocumentConverter:
                     pages_with_images,
                 )
 
+                # 4. Run pipeline stages
                 pipeline_pages = self.model_pipeline.apply(pages_with_cells)
 
-                # 7. Assemble page elements (per page)
+                # 5. Assemble page elements (per page)
                 assembled_pages = self.page_assemble_model(pipeline_pages)
 
                 # exhaust assembled_pages
                 for assembled_page in assembled_pages:
                     # Free up mem resources before moving on with next batch
-                    assembled_page.image = (
-                        None  # Comment this if you want to visualize page images
-                    )
+
+                    # 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
+                        )
+
+                    # Unload backend
                     assembled_page._backend.unload()
 
                     all_assembled_pages.append(assembled_page)

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

{docling-1.2.1.dist-info → docling-1.4.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: docling
-Version: 1.2.1
+Version: 1.4.0
 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)

{docling-1.2.1.dist-info → docling-1.4.0.dist-info}/RECORD

@@ -1,17 +1,17 @@
 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/docling_parse_backend.py,sha256=bgsmnwDmroBwuOwkEUzlN9KMEIFJ1xUaCZW6rsr5G-c,5924
 docling/backend/pypdfium2_backend.py,sha256=tv6JxyTkTdT2qr2ghsQgYA2zgpCDxKYSdHVBTAR7FSk,7411
 docling/datamodel/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-docling/datamodel/base_models.py,sha256=k7gLFPnq3ArEMAFz6qUcp5qemlYzVhOmR9qtBTkAiX4,6862
-docling/datamodel/document.py,sha256=FG_ntDFRBWj-MhV52D0sC8XaZOwN3yryyXahsVHGnyI,12517
+docling/datamodel/base_models.py,sha256=irZLAHdsROOOwRbywKIA0mk3H8GrLwtGjOgTV6G0QoU,7004
+docling/datamodel/document.py,sha256=lZHXINmPWvpzrV3PTilgJs1blqTMCnJdLEww_qfcqdE,12533
 docling/datamodel/settings.py,sha256=t5g6wrEJnPa9gBzMMl8ppgBRUYz-8xgopEtfMS0ZH28,733
-docling/document_converter.py,sha256=I9vjTLCLahsMrcs9ozM3C5r_CtBN-9qHk7-ANma7fkc,9895
+docling/document_converter.py,sha256=dMucsq6M_nwPsC1ChogVwJgNDv8sJuFklQWWinDZaug,10246
 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/layout_model.py,sha256=3mOgNvCYPh99_oLxJy-ZaIqGOFgG5bcIQ0tTubW656Q,11204
 docling/models/page_assemble_model.py,sha256=8eoG2WiFxPxq9TPvM-wkngb2gkr0tdtCRVXg1JcTETo,5550
 docling/models/table_structure_model.py,sha256=xUmfunZNYC30P0fRdESdztqy1FVlMzlhJjLBp-xcn4A,5638
 docling/pipeline/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
@@ -20,7 +20,7 @@ docling/pipeline/standard_model_pipeline.py,sha256=UTwodKUKrisLoVcntbNUBDhjzRyFv
 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-1.2.1.dist-info/LICENSE,sha256=ACwmltkrXIz5VsEQcrqljq-fat6ZXAMepjXGoe40KtE,1069
-docling-1.2.1.dist-info/METADATA,sha256=EwclgZsLCgm5qOT5na5QRxSwSKYZCIfSeotTlwMRyYk,7042
-docling-1.2.1.dist-info/WHEEL,sha256=sP946D7jFCHeNz5Iq4fL4Lu-PrWrFsgfLXbbkciIZwg,88
-docling-1.2.1.dist-info/RECORD,,
+docling-1.4.0.dist-info/LICENSE,sha256=ACwmltkrXIz5VsEQcrqljq-fat6ZXAMepjXGoe40KtE,1069
+docling-1.4.0.dist-info/METADATA,sha256=Hu8pvrxpc0b1qzQvvzI_ijRAQWjOfcfNl4_1Zb7oyoc,7042
+docling-1.4.0.dist-info/WHEEL,sha256=sP946D7jFCHeNz5Iq4fL4Lu-PrWrFsgfLXbbkciIZwg,88
+docling-1.4.0.dist-info/RECORD,,