deepdoctection 0.39.1__py3-none-any.whl → 0.39.2__py3-none-any.whl

deepdoctection/__init__.py

@@ -25,7 +25,7 @@ from .utils.logger import LoggingRecord, logger
 
 # pylint: enable=wrong-import-position
 
-__version__ = "0.39.1"
+__version__ = "0.39.2"
 
 _IMPORT_STRUCTURE = {
     "analyzer": ["config_sanity_checks", "get_dd_analyzer", "ServiceFactory"],

deepdoctection/analyzer/_config.py

@@ -40,6 +40,7 @@ cfg.TF.CELL.FILTER = None
 cfg.TF.ITEM.WEIGHTS = "item/model-1620000_inf_only.data-00000-of-00001"
 cfg.TF.ITEM.FILTER = None
 
+cfg.PT.ENFORCE_WEIGHTS = False
 cfg.PT.LAYOUT.WEIGHTS = "layout/d2_model_0829999_layout_inf_only.pt"
 cfg.PT.LAYOUT.WEIGHTS_TS = "layout/d2_model_0829999_layout_inf_only.ts"
 cfg.PT.LAYOUT.FILTER = None

deepdoctection/analyzer/factory.py

@@ -98,7 +98,11 @@ class ServiceFactory:
         weights = (
             getattr(config.TF, mode).WEIGHTS
             if config.LIB == "TF"
-            else (getattr(config.PT, mode).WEIGHTS if detectron2_available() else getattr(config.PT, mode).WEIGHTS_TS)
+            else (
+                getattr(config.PT, mode).WEIGHTS
+                if detectron2_available() or config.PT.ENFORCE_WEIGHTS
+                else getattr(config.PT, mode).WEIGHTS_TS
+            )
         )
         filter_categories = (
             getattr(getattr(config.TF, mode), "FILTER")

deepdoctection/datapoint/view.py

@@ -407,6 +407,35 @@ class Table(Layout):
         col_anns = self.base_page.get_annotation(annotation_ids=all_relation_ids, category_names=[LayoutType.COLUMN])
         return col_anns
 
+    def row(self, row_number: int) -> list[ImageAnnotationBaseView]:
+        """
+        Get a list of cells in a row.
+        """
+        all_relation_ids = self.get_relationship(Relationships.CHILD)
+        all_cells = self.base_page.get_annotation(
+            category_names=[LayoutType.CELL, CellType.SPANNING], annotation_ids=all_relation_ids
+        )
+        row_cells = list(
+            filter(lambda c: row_number in (c.row_number, c.row_number + c.row_span - 1), all_cells)  # type: ignore
+        )
+        row_cells.sort(key=lambda c: c.column_number)  # type: ignore
+        return row_cells  # type: ignore
+
+    def column(self, column_number: int) -> list[ImageAnnotationBaseView]:
+        """
+        Get a list of cells in a column.
+        """
+        all_relation_ids = self.get_relationship(Relationships.CHILD)
+        all_cells = self.base_page.get_annotation(
+            category_names=[LayoutType.CELL, CellType.SPANNING], annotation_ids=all_relation_ids
+        )
+        column_cells = list(
+            filter(lambda c: column_number in  # type: ignore
+                             (c.column_number, c.column_number + c.column_span - 1), all_cells)  # type: ignore
+        )
+        column_cells.sort(key=lambda c: c.row_number)  # type: ignore
+        return column_cells  # type: ignore
+
     @property
     def html(self) -> HTML:
         """

deepdoctection/pipe/base.py

@@ -24,7 +24,7 @@ from __future__ import annotations
 from abc import ABC, abstractmethod
 from collections import defaultdict
 from dataclasses import dataclass, field
-from typing import Any, Mapping, Optional, Union, Callable
+from typing import Any, Callable, Mapping, Optional, Union
 from uuid import uuid1
 
 from ..dataflow import DataFlow, MapData
@@ -100,7 +100,7 @@ class PipelineComponent(ABC):
 
         :param filter_func: A function that takes an image datapoint and returns a boolean value
         """
-        self.filter_func = filter_func # type: ignore
+        self.filter_func = filter_func  # type: ignore
 
     @abstractmethod
     def serve(self, dp: Image) -> None:
@@ -122,7 +122,6 @@ class PipelineComponent(ABC):
         if not self.filter_func(dp):
             self.serve(dp)
 
-
     def pass_datapoint(self, dp: Image) -> Image:
         """
         Acceptance, handover to dp_manager, transformation and forwarding of dp. To measure the time, use

deepdoctection/pipe/common.py

@@ -362,7 +362,7 @@ class AnnotationNmsService(PipelineComponent):
             self.threshold = [thresholds for _ in self.nms_pairs]
         else:
             assert len(self.nms_pairs) == len(thresholds), "Sequences of nms_pairs and thresholds must have same length"
-            self.threshold = thresholds # type: ignore
+            self.threshold = thresholds  # type: ignore
         if priority:
             assert len(self.nms_pairs) == len(priority), "Sequences of nms_pairs and priority must have same length"
 

deepdoctection/pipe/lm.py

@@ -265,7 +265,7 @@ class LMSequenceClassifierService(PipelineComponent):
         padding: Literal["max_length", "do_not_pad", "longest"] = "max_length",
         truncation: bool = True,
         return_overflowing_tokens: bool = False,
-        use_other_as_default_category: bool = False
+        use_other_as_default_category: bool = False,
     ) -> None:
         """
         :param tokenizer: Tokenizer, typing allows currently anything. This will be changed in the future
@@ -309,11 +309,10 @@ class LMSequenceClassifierService(PipelineComponent):
         lm_output = None
         if lm_input is None:
             if self.use_other_as_default_category:
-                class_id = self.language_model.categories.get_categories(as_dict=True,
-                                                                         name_as_key=True).get(TokenClasses.OTHER, 1)
-                lm_output = SequenceClassResult(class_name=TokenClasses.OTHER,
-                                                class_id = class_id,
-                                                score=-1.)
+                class_id = self.language_model.categories.get_categories(as_dict=True, name_as_key=True).get(
+                    TokenClasses.OTHER, 1
+                )
+                lm_output = SequenceClassResult(class_name=TokenClasses.OTHER, class_id=class_id, score=-1.0)
         else:
             lm_output = self.language_model.predict(**lm_input)
         if lm_output:

deepdoctection/train/hf_layoutlm_train.py

@@ -499,9 +499,7 @@ def train_hf_layoutlm(
         )
         pipeline_component_cls = pipeline_component_registry.get(pipeline_component_name)
         if dataset_type == DatasetType.SEQUENCE_CLASSIFICATION:
-            pipeline_component = pipeline_component_cls(tokenizer_fast,
-                                                        dd_model,
-                                                        use_other_as_default_category=True)
+            pipeline_component = pipeline_component_cls(tokenizer_fast, dd_model, use_other_as_default_category=True)
         else:
             pipeline_component = pipeline_component_cls(
                 tokenizer_fast,

{deepdoctection-0.39.1.dist-info → deepdoctection-0.39.2.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: deepdoctection
-Version: 0.39.1
+Version: 0.39.2
 Summary: Repository for Document AI
 Home-page: https://github.com/deepdoctection/deepdoctection
 Author: Dr. Janis Meyer

{deepdoctection-0.39.1.dist-info → deepdoctection-0.39.2.dist-info}/RECORD

@@ -1,9 +1,9 @@
-deepdoctection/__init__.py,sha256=uDowNayqaYZGYaqnGzPSz6pVuHQhtDVRAN_bvPq85Ko,12754
+deepdoctection/__init__.py,sha256=KpFnC7nCOVQOgeXDWt0fIjKoikD3MRj-_iSAbARIWeQ,12754
 deepdoctection/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 deepdoctection/analyzer/__init__.py,sha256=icClxrd20XutD6LxLgEPIWceSs4j_QfI3szCE-9BL2w,729
-deepdoctection/analyzer/_config.py,sha256=OZMOPlyFv4gcyabPG6KO08EYx-0tUH82Ehs9YDv2B1Q,5027
+deepdoctection/analyzer/_config.py,sha256=1rfvVrp7cI2YLzpahD77aa1tZ_KFAIQ21DM1NWhxYiI,5058
 deepdoctection/analyzer/dd.py,sha256=bfR7e1JV7BwUNDRLu0jYZU7qQXnyA_vbRAJl2Ylrq5o,5905
-deepdoctection/analyzer/factory.py,sha256=Kf3Ztv5FEcF5yJf6i4I557aOIUHybuxIP0moHryguTQ,32344
+deepdoctection/analyzer/factory.py,sha256=7L-bJ9957TBn_C6OGWJFmZobrh8MPq4Q-Espx5faEiY,32435
 deepdoctection/configs/__init__.py,sha256=TX_P6tqDOF1LK1mi9ruAl7x0mtv1Asm8cYWCz3Pe2dk,646
 deepdoctection/configs/conf_dd_one.yaml,sha256=qnrDAST1PHBtdIKE_hdkZexW22FqVvNTI-PEo9wvinM,3025
 deepdoctection/configs/conf_tesseract.yaml,sha256=oF6szDyoi15FHvq7yFUNIEjfA_jNLhGxoowiRsz_zY4,35
@@ -20,7 +20,7 @@ deepdoctection/datapoint/annotation.py,sha256=FEgz4COxVDfjic0gG7kS6iHnWLBIgFnquQ
 deepdoctection/datapoint/box.py,sha256=UAdSnLexvFyg4KK1u9kXdJxhaWTwRxTU-cnQcvl37Q8,23410
 deepdoctection/datapoint/convert.py,sha256=gJbHY2V8nlMht1N5VdxTmWSsOeydpFPTJsaJHp6XGgE,7516
 deepdoctection/datapoint/image.py,sha256=S6yfsIRQgMCl6HYAcHYJSBcbfdYKKtebtkEkkkrXsMQ,33619
-deepdoctection/datapoint/view.py,sha256=srMyPQGsK4OSiorxkyG6UAIgpViM6Ks1CI3b5k97cjY,49452
+deepdoctection/datapoint/view.py,sha256=XPyhbBr2cGIKdAiISBVZWxNxlSvN8kmGsD9P0mfpEEE,50772
 deepdoctection/datasets/__init__.py,sha256=-A3aR90aDsHPmVM35JavfnQ2itYSCn3ujl4krRni1QU,1076
 deepdoctection/datasets/adapter.py,sha256=Ly_vbOAgVI73V41FUccnSX1ECTOyesW_qsuvQuvOZbw,7796
 deepdoctection/datasets/base.py,sha256=DT4i-d74sIEiUNC6UspIHNJuHSK0t1dBv7qwadg4rLw,22341
@@ -103,13 +103,13 @@ deepdoctection/mapper/tpstruct.py,sha256=YNABRibvcISD5Lavg3jouoE4FMdqXEJoM-hNoB_
 deepdoctection/mapper/xfundstruct.py,sha256=_3r3c0K82fnF2h1HxA85h-9ETYrHwcERa6MNc6Ko6Z8,8807
 deepdoctection/pipe/__init__.py,sha256=ywTVoetftdL6plXg2YlBzMfmqBZupq7yXblSVyvvkcQ,1127
 deepdoctection/pipe/anngen.py,sha256=3319l4aaXzcY4w6ItVBNPX8LGS5fHFDVtyVY9KMefac,16393
-deepdoctection/pipe/base.py,sha256=F4NusbZ-xYc6wuO-XAngmC8uzahT2ubsu2g9NO8PpVw,15390
-deepdoctection/pipe/common.py,sha256=vlWzvwn8wl7baPbK-917HUWujEGJEkHur_-ilkweKjk,17751
+deepdoctection/pipe/base.py,sha256=wlza9aDOKnHKrXmaz8MLyLz0nMqqcIWQ-6Lu944aicE,15390
+deepdoctection/pipe/common.py,sha256=C1KxEfJFSPeDqlnkiJ1ZYPuA36P8BU_4jVhdsszW_V8,17752
 deepdoctection/pipe/concurrency.py,sha256=AAKRsVgaBEYNluntbDa46SBF1JZ_XqnWLDSWrNvAzEo,9657
 deepdoctection/pipe/doctectionpipe.py,sha256=bGW3ugky-fb-nEe-3bvO6Oc_4_6w82cQboGM_6p2eIo,12530
 deepdoctection/pipe/language.py,sha256=5zI0UQC6Fh12_r2pfVL42HoCGz2hpHrOhpXAn5m-rYw,5451
 deepdoctection/pipe/layout.py,sha256=xIhnJpyUSbvLbhTXyAKXY1hmG9352jihGYFSclTH_1g,5567
-deepdoctection/pipe/lm.py,sha256=Ygj6MmBvBZ7l4RGCwBuhmMsOM0Ep3LWteNg7bzh-UmI,17703
+deepdoctection/pipe/lm.py,sha256=x9NoYpivdjQF1r76a7PPrUuBEmuHP7ZukuXFDkXhXBc,17572
 deepdoctection/pipe/order.py,sha256=PnJZiCnxFluJiECXLTZT0c1Rr66vIRBFraa_G41UA2k,40121
 deepdoctection/pipe/refine.py,sha256=dTfI396xydPdbzpfo4yqFcuxl3UAB1y-WbSQn1o76ec,22367
 deepdoctection/pipe/registry.py,sha256=aFx-Tn0xhVA5l5H18duNW5QoTNKQltybsEUEzsMgUfg,902
@@ -120,7 +120,7 @@ deepdoctection/pipe/transform.py,sha256=9Om7X7hJeL4jgUwHM1CHa4sb5v7Qo1PtVG0ls_3n
 deepdoctection/train/__init__.py,sha256=YFTRAZF1F7cEAKTdAIi1BLyYb6rSRcwq09Ui5Lu8d6E,1071
 deepdoctection/train/d2_frcnn_train.py,sha256=sFc_G-mEpaM8d1CCE0_6Gl4nBh11X2RYRBA3p_ylFJQ,16000
 deepdoctection/train/hf_detr_train.py,sha256=eHSdI11U8oGy93noZxAISfukhRBElj4dBerJ4Xcercw,10785
-deepdoctection/train/hf_layoutlm_train.py,sha256=irSg-IpbVoSlaw1-vZCej2mCZcctONtXr5Z2NQAc_a4,22680
+deepdoctection/train/hf_layoutlm_train.py,sha256=DTPJZYKeDdRtDpcObYh93uh5D4sgT4c0ckHiAknCroY,22568
 deepdoctection/train/tp_frcnn_train.py,sha256=pEpXokSVGveqo82pRnhnAmHPmjQ_8wQWpqM4ZyNHJgs,13049
 deepdoctection/utils/__init__.py,sha256=brBceRWeov9WXMiJTjyJOF2rHMP8trGGRRjhMdZ61nI,2371
 deepdoctection/utils/concurrency.py,sha256=nIhpkSncmv0LBB8PtcOLY-BsRGlfcDpz7foVdgzZd20,4598
@@ -141,8 +141,8 @@ deepdoctection/utils/transform.py,sha256=3kCgsEeRkG1efCdkfvj7tUFMs-e2jbjbflq826F
 deepdoctection/utils/types.py,sha256=_3dmPdCIZNLbgU5QP5k_c5phDf18xLe1kYL6t2nM45s,2953
 deepdoctection/utils/utils.py,sha256=csVs_VvCq4QBETPoE2JdTTL4MFYnD4xh-Js5vRb612g,6492
 deepdoctection/utils/viz.py,sha256=Jf8ePNYWlpuyaS6SeTYQ4OyA3eNhtgjvAQZnGNdgHC0,27051
-deepdoctection-0.39.1.dist-info/LICENSE,sha256=GQ0rUvuGdrMNEI3iHK5UQx6dIMU1QwAuyXsxUHn5MEQ,11351
-deepdoctection-0.39.1.dist-info/METADATA,sha256=NBN2dqFMUiXkcJ28xJDwyN6eNP-MmFw64F7dm3kUWTA,19741
-deepdoctection-0.39.1.dist-info/WHEEL,sha256=In9FTNxeP60KnTkGw7wk6mJPYd_dQSjEZmXdBdMCI-8,91
-deepdoctection-0.39.1.dist-info/top_level.txt,sha256=hs2DdoOL9h4mnHhmO82BT4pz4QATIoOZ20PZmlnxFI8,15
-deepdoctection-0.39.1.dist-info/RECORD,,
+deepdoctection-0.39.2.dist-info/LICENSE,sha256=GQ0rUvuGdrMNEI3iHK5UQx6dIMU1QwAuyXsxUHn5MEQ,11351
+deepdoctection-0.39.2.dist-info/METADATA,sha256=Z6b0DeMKoXZOR6O7Yxu7Qt5_MLgCxxvyNuWsPAhFo0M,19741
+deepdoctection-0.39.2.dist-info/WHEEL,sha256=In9FTNxeP60KnTkGw7wk6mJPYd_dQSjEZmXdBdMCI-8,91
+deepdoctection-0.39.2.dist-info/top_level.txt,sha256=hs2DdoOL9h4mnHhmO82BT4pz4QATIoOZ20PZmlnxFI8,15
+deepdoctection-0.39.2.dist-info/RECORD,,