docling-ocr-onnxtr 0.1.2__py3-none-any.whl → 0.2.0__py3-none-any.whl

docling_ocr_onnxtr/onnxtr_model.py

@@ -12,7 +12,7 @@ import numpy
 import numpy as np
 from docling.datamodel.base_models import Page
 from docling.datamodel.document import ConversionResult
-from docling.datamodel.pipeline_options import (
+from docling.datamodel.pipeline_options import (  # type: ignore[attr-defined]
     AcceleratorOptions,
     OcrOptions,
 )
@@ -91,11 +91,13 @@ class OnnxtrOcrModel(BaseOcrModel):
 
             self.reader = ocr_predictor(
                 det_arch=(
-                    from_hub(self.options.det_arch) if self.options.det_arch.count("/") == 1 else self.options.det_arch
+                    from_hub(self.options.det_arch)
+                    if isinstance(self.options.det_arch, str) and self.options.det_arch.count("/") == 1
+                    else self.options.det_arch
                 ),
                 reco_arch=(
                     from_hub(self.options.reco_arch)
-                    if self.options.reco_arch.count("/") == 1
+                    if isinstance(self.options.reco_arch, str) and self.options.reco_arch.count("/") == 1
                     else self.options.reco_arch
                 ),
                 det_bs=1,  # NOTE: Should be always 1, because docling handles batching
@@ -195,7 +197,7 @@ class OnnxtrOcrModel(BaseOcrModel):
                                     )
 
                 # Post-process the cells
-                page.cells = self.post_process_cells(all_ocr_cells, page.cells)
+                self.post_process_cells(all_ocr_cells, page)
 
             # DEBUG code:
             if settings.debug.visualize_ocr:

docling_ocr_onnxtr/options.py

@@ -22,9 +22,9 @@ class OnnxtrOcrOptions(OcrOptions):
     # detection model objectness score threshold 'fast algorithm'
     objectness_score: float = 0.3
 
-    # NOTE: This can be also a hf hub model
-    det_arch: str = "fast_base"
-    reco_arch: str = "crnn_vgg16_bn"
+    # NOTE: This can be also a hf hub model or an instance of a model class.
+    det_arch: Any = "fast_base"
+    reco_arch: Any = "crnn_vgg16_bn"
     reco_bs: int = 512
     auto_correct_orientation: bool = False
     preserve_aspect_ratio: bool = True

docling_ocr_onnxtr/version.py

@@ -1 +1 @@
-__version__ = 'v0.1.2'
+__version__ = 'v0.2.0'

{docling_ocr_onnxtr-0.1.2.dist-info → docling_ocr_onnxtr-0.2.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: docling-ocr-onnxtr
-Version: 0.1.2
+Version: 0.2.0
 Summary: Onnx Text Recognition (OnnxTR) OCR plugin for docling
 Author-email: Felix Dittrich <felixdittrich92@gmail.com>
 Maintainer: Felix Dittrich
@@ -262,11 +262,11 @@ Dynamic: license-file
 </p>
 
 [![License](https://img.shields.io/badge/License-Apache%202.0-blue.svg)](LICENSE)
-[![Test Status](https://github.com/felixdittrich92/docling-OCR-OnnxTR/actions/workflows/main.yml/badge.svg)](https://github.com/felixdittrich92/docling-OCR-OnnxTR/actions/workflows/main.yml)
+[![Build Status](https://github.com/felixdittrich92/docling-OCR-OnnxTR/actions/workflows/builds.yml/badge.svg)](https://github.com/felixdittrich92/docling-OCR-OnnxTR/actions/workflows/builds.yml)
 [![codecov](https://codecov.io/gh/felixdittrich92/docling-OCR-OnnxTR/graph/badge.svg?token=L3AHXKV86A)](https://codecov.io/gh/felixdittrich92/docling-OCR-OnnxTR)
 [![Codacy Badge](https://app.codacy.com/project/badge/Grade/0d250447650240ee9ca573950fea8b99)](https://app.codacy.com/gh/felixdittrich92/docling-OCR-OnnxTR/dashboard?utm_source=gh&utm_medium=referral&utm_content=&utm_campaign=Badge_grade)
 [![CodeFactor](https://www.codefactor.io/repository/github/felixdittrich92/docling-ocr-onnxtr/badge)](https://www.codefactor.io/repository/github/felixdittrich92/docling-ocr-onnxtr)
-[![Pypi](https://img.shields.io/badge/pypi-v0.1.2-blue.svg)](https://pypi.org/project//)
+[![Pypi](https://img.shields.io/badge/pypi-v0.2.0-blue.svg)](https://pypi.org/project/docling-ocr-onnxtr/)
 ![PyPI - Downloads](https://img.shields.io/pypi/dm/docling-ocr-onnxtr)
 
 The `docling-OCR-OnnxTR` repository provides a plugin that integrates the [OnnxTR OCR engine](https://github.com/felixdittrich92/OnnxTR) into the [Docling framework](https://github.com/docling-project/docling), enhancing document processing capabilities with efficient and accurate text recognition.
@@ -289,19 +289,19 @@ For GPU support please take a look at: [ONNX Runtime](https://onnxruntime.ai/get
 
 ```bash
 # For CPU
-pip install docling-ocr-onnxtr[cpu]
+pip install "docling-ocr-onnxtr[cpu]"
 # For Nvidia GPU
-pip install docling-ocr-onnxtr[gpu]
+pip install "docling-ocr-onnxtr[gpu]"
 # For Intel GPU / Integrated Graphics
-pip install docling-ocr-onnxtr[openvino]
+pip install "docling-ocr-onnxtr[openvino]"
 
 # Headless mode (no GUI)
 # For CPU
-pip install docling-ocr-onnxtr[cpu-headless]
+pip install "docling-ocr-onnxtr[cpu-headless]"
 # For Nvidia GPU
-pip install docling-ocr-onnxtr[gpu-headless]
+pip install "docling-ocr-onnxtr[gpu-headless]"
 # For Intel GPU / Integrated Graphics
-pip install docling-ocr-onnxtr[openvino-headless]
+pip install "docling-ocr-onnxtr[openvino-headless]"
 ```
 
 By integrating OnnxTR with Docling, users can achieve more efficient and accurate OCR results, enhancing the overall document processing experience.
@@ -358,6 +358,62 @@ def main():
     print(md)
 
 
+if __name__ == "__main__":
+    main()
+```
+
+It is also possible to load the models from local files instead of using the Hugging Face Hub or downloading them from the repo:
+
+```python
+from docling.datamodel.pipeline_options import PdfPipelineOptions
+from docling.document_converter import (
+    ConversionResult,
+    DocumentConverter,
+    InputFormat,
+    PdfFormatOption,
+)
+from docling_ocr_onnxtr import OnnxtrOcrOptions
+from onnxtr.models import db_mobilenet_v3_large, parseq
+
+
+def main():
+    # Source document to convert
+    source = "https://arxiv.org/pdf/2408.09869v4"
+
+    # Load models from local files
+    # NOTE: You need to download the models first and then adjust the paths accordingly.
+    det_model = db_mobilenet_v3_large("/home/felix/.cache/onnxtr/models/db_mobilenet_v3_large-1866973f.onnx")
+    reco_model = parseq("/home/felix/.cache/onnxtr/models/parseq-00b40714.onnx")
+
+    ocr_options = OnnxtrOcrOptions(
+        # Text detection model
+        det_arch=det_model,
+        # Text recognition model
+        reco_arch=reco_model,
+        # This can be set to `True` to auto-correct the orientation of the pages
+        auto_correct_orientation=False,
+    )
+
+    pipeline_options = PdfPipelineOptions(
+        ocr_options=ocr_options,
+    )
+    pipeline_options.allow_external_plugins = True  # <-- enabled the external plugins
+
+    # Convert the document
+    converter = DocumentConverter(
+        format_options={
+            InputFormat.PDF: PdfFormatOption(
+                pipeline_options=pipeline_options,
+            ),
+        },
+    )
+
+    conversion_result: ConversionResult = converter.convert(source=source)
+    doc = conversion_result.document
+    md = doc.export_to_markdown()
+    print(md)
+
+
 if __name__ == "__main__":
     main()
 ```

docling_ocr_onnxtr-0.2.0.dist-info/RECORD

@@ -0,0 +1,13 @@
+docling_ocr_onnxtr/__init__.py,sha256=w8M8ctJDcDcxxrRJ_GT-G7JP3H7E1O50AuoK4l2djf0,85
+docling_ocr_onnxtr/onnxtr_model.py,sha256=4CWDtT2B4PVxEb01glcgmiTGd6bNj5ig0tvTK4UmXCk,8628
+docling_ocr_onnxtr/options.py,sha256=TBzwwDFmmkdhoXmmImKrV3tSEGX2FD8wtey5Mh1Z3EY,1299
+docling_ocr_onnxtr/plugin.py,sha256=qtPPjpsiXyExvK7Vu8ssZhqxDV2gqFTHIIBBv1qz1lA,317
+docling_ocr_onnxtr/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+docling_ocr_onnxtr/version.py,sha256=F-dLDdnsmwBrwqQCEb3nX0LORU_q1xg9pmFGVU_3j8o,23
+docling_ocr_onnxtr-0.2.0.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
+docling_ocr_onnxtr-0.2.0.dist-info/METADATA,sha256=58vZ_WCZUCZ0k6fLC2X7_WAxhjxgyMKCXgwIq9Nq-0I,23822
+docling_ocr_onnxtr-0.2.0.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+docling_ocr_onnxtr-0.2.0.dist-info/entry_points.txt,sha256=4bUh15gjHVjgG1_vEtZgZ0Kyo7ygkWeAugc_W-EtSfo,57
+docling_ocr_onnxtr-0.2.0.dist-info/top_level.txt,sha256=vmOD5JYg8FKSI-iXyYH73_9AwJdxPOP1TsSY1nDp5iE,24
+docling_ocr_onnxtr-0.2.0.dist-info/zip-safe,sha256=AbpHGcgLb-kRsJGnwFEktk7uzpZOCcBY74-YBdrKVGs,1
+docling_ocr_onnxtr-0.2.0.dist-info/RECORD,,

{docling_ocr_onnxtr-0.1.2.dist-info → docling_ocr_onnxtr-0.2.0.dist-info}/WHEEL

@@ -1,5 +1,5 @@
 Wheel-Version: 1.0
-Generator: setuptools (78.1.0)
+Generator: setuptools (80.9.0)
 Root-Is-Purelib: true
 Tag: py3-none-any
 

docling_ocr_onnxtr-0.1.2.dist-info/RECORD

@@ -1,13 +0,0 @@
-docling_ocr_onnxtr/__init__.py,sha256=w8M8ctJDcDcxxrRJ_GT-G7JP3H7E1O50AuoK4l2djf0,85
-docling_ocr_onnxtr/onnxtr_model.py,sha256=FEEyFbRttBh4ckz5QgfT_n5zgtmfCNLr_bLhoobnl68,8490
-docling_ocr_onnxtr/options.py,sha256=gOzKHo86zwAVI_5-YxOr4NaRkNFwlv1RXeag7ZKWxtM,1266
-docling_ocr_onnxtr/plugin.py,sha256=qtPPjpsiXyExvK7Vu8ssZhqxDV2gqFTHIIBBv1qz1lA,317
-docling_ocr_onnxtr/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-docling_ocr_onnxtr/version.py,sha256=t2ZLy6eqIRREzP9MNv-0UaNqpTXYFxlbMzz0JwDxS9I,23
-docling_ocr_onnxtr-0.1.2.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
-docling_ocr_onnxtr-0.1.2.dist-info/METADATA,sha256=2uPk15A60jvJM8tXk5mOpKTdeyZi5jh6QcLFl1F1P98,22059
-docling_ocr_onnxtr-0.1.2.dist-info/WHEEL,sha256=CmyFI0kx5cdEMTLiONQRbGQwjIoR1aIYB7eCAQ4KPJ0,91
-docling_ocr_onnxtr-0.1.2.dist-info/entry_points.txt,sha256=4bUh15gjHVjgG1_vEtZgZ0Kyo7ygkWeAugc_W-EtSfo,57
-docling_ocr_onnxtr-0.1.2.dist-info/top_level.txt,sha256=vmOD5JYg8FKSI-iXyYH73_9AwJdxPOP1TsSY1nDp5iE,24
-docling_ocr_onnxtr-0.1.2.dist-info/zip-safe,sha256=AbpHGcgLb-kRsJGnwFEktk7uzpZOCcBY74-YBdrKVGs,1
-docling_ocr_onnxtr-0.1.2.dist-info/RECORD,,