Lots of improvements

Signed-off-by: Christoph Auer <cau@zurich.ibm.com>

examples/batch_convert.py

@@ -6,8 +6,9 @@ from typing import Iterable
 
 import yaml
 
-from docling.datamodel.base_models import ConversionStatus, PdfPipelineOptions
+from docling.datamodel.base_models import ConversionStatus
 from docling.datamodel.document import ConversionResult, DocumentConversionInput
+from docling.document_converter import DocumentConverter
 
 _log = logging.getLogger(__name__)
 
@@ -110,11 +111,7 @@ def main():
     # docs = [DocumentStream(filename="my_doc.pdf", stream=buf)]
     # input = DocumentConversionInput.from_streams(docs)
 
-    doc_converter = PdfDocumentConverter(
-        pipeline_options=PdfPipelineOptions(),
-        pdf_backend=DocumentConversionInput.DEFAULT_BACKEND,
-        pipeline_cls=StandardModelPipeline,
-    )
+    doc_converter = DocumentConverter()
 
     input = DocumentConversionInput.from_paths(input_doc_paths)
 

examples/custom_convert.py

@@ -4,10 +4,11 @@ import time
 from pathlib import Path
 from typing import Iterable
 
-from docling.backend.docling_parse_backend import DoclingParseDocumentBackend
-from docling.datamodel.base_models import ConversionStatus, PdfPipelineOptions
+from docling.datamodel.base_models import ConversionStatus, InputFormat
 from docling.datamodel.document import ConversionResult, DocumentConversionInput
-from docling.pdf_document_converter import PdfDocumentConverter
+from docling.datamodel.pipeline_options import PdfPipelineOptions
+from docling.document_converter import DocumentConverter, FormatOption
+from docling.pipeline.standard_pdf_model_pipeline import StandardPdfModelPipeline
 
 _log = logging.getLogger(__name__)
 
@@ -101,9 +102,12 @@ def main():
     pipeline_options.do_table_structure = True
     pipeline_options.table_structure_options.do_cell_matching = True
 
-    doc_converter = PdfDocumentConverter(
-        pipeline_options=pipeline_options,
-        pdf_backend=DoclingParseDocumentBackend,
+    doc_converter = DocumentConverter(
+        format_options={
+            InputFormat.PDF: FormatOption(
+                pipeline_cls=StandardPdfModelPipeline, pipeline_options=pipeline_options
+            )
+        }
     )
 
     # Docling Parse with OCR

examples/export_figures.py

@@ -5,11 +5,12 @@ from pathlib import Path
 from docling.datamodel.base_models import (
     ConversionStatus,
     FigureElement,
-    PdfPipelineOptions,
+    InputFormat,
     Table,
 )
 from docling.datamodel.document import DocumentConversionInput
-from docling.pdf_document_converter import PdfDocumentConverter
+from docling.datamodel.pipeline_options import PdfPipelineOptions
+from docling.document_converter import DocumentConverter, PdfFormatOption
 
 _log = logging.getLogger(__name__)
 
@@ -28,12 +29,16 @@ def main():
 
     # Important: For operating with page images, we must keep them, otherwise the DocumentConverter
     # will destroy them for cleaning up memory.
-    # This is done by setting PipelineOptions.images_scale, which also defines the scale of images.
+    # This is done by setting AssembleOptions.images_scale, which also defines the scale of images.
     # scale=1 correspond of a standard 72 DPI image
     pipeline_options = PdfPipelineOptions()
     pipeline_options.images_scale = IMAGE_RESOLUTION_SCALE
 
-    doc_converter = PdfDocumentConverter(pipeline_options=pipeline_options)
+    doc_converter = DocumentConverter(
+        format_options={
+            InputFormat.PDF: PdfFormatOption(pipeline_options=pipeline_options)
+        }
+    )
 
     start_time = time.time()
 

examples/export_multimodal.py

@@ -5,9 +5,10 @@ from pathlib import Path
 
 import pandas as pd
 
-from docling.datamodel.base_models import ConversionStatus, PdfPipelineOptions
+from docling.datamodel.base_models import ConversionStatus, InputFormat
 from docling.datamodel.document import DocumentConversionInput
-from docling.pdf_document_converter import PdfDocumentConverter
+from docling.datamodel.pipeline_options import PdfPipelineOptions
+from docling.document_converter import DocumentConverter, PdfFormatOption
 from docling.utils.export import generate_multimodal_pages
 
 _log = logging.getLogger(__name__)
@@ -27,12 +28,16 @@ def main():
 
     # Important: For operating with page images, we must keep them, otherwise the DocumentConverter
     # will destroy them for cleaning up memory.
-    # This is done by setting PipelineOptions.images_scale, which also defines the scale of images.
+    # This is done by setting AssembleOptions.images_scale, which also defines the scale of images.
     # scale=1 correspond of a standard 72 DPI image
     pipeline_options = PdfPipelineOptions()
     pipeline_options.images_scale = IMAGE_RESOLUTION_SCALE
 
-    doc_converter = PdfDocumentConverter(pipeline_options=pipeline_options)
+    doc_converter = DocumentConverter(
+        format_options={
+            InputFormat.PDF: PdfFormatOption(pipeline_options=pipeline_options)
+        }
+    )
 
     start_time = time.time()
 

examples/export_tables.py

@@ -6,7 +6,7 @@ import pandas as pd
 
 from docling.datamodel.base_models import ConversionStatus
 from docling.datamodel.document import DocumentConversionInput
-from docling.pdf_document_converter import PdfDocumentConverter
+from docling.document_converter import DocumentConverter
 
 _log = logging.getLogger(__name__)
 
@@ -21,7 +21,7 @@ def main():
 
     input_files = DocumentConversionInput.from_paths(input_doc_paths)
 
-    doc_converter = PdfDocumentConverter()
+    doc_converter = DocumentConverter()
 
     start_time = time.time()
 

examples/minimal.py

@@ -1,6 +1,6 @@
-from docling.pdf_document_converter import PdfDocumentConverter
+from docling.document_converter import DocumentConverter
 
 source = "https://arxiv.org/pdf/2408.09869"  # PDF path or URL
-converter = PdfDocumentConverter()
+converter = DocumentConverter()
 doc = converter.convert_single(source)
 print(doc.render_as_markdown())  # output: ## Docling Technical Report [...]"

examples/run_with_formats.py

@@ -1,8 +1,13 @@
 import logging
 from pathlib import Path
 
+from docling.backend.docling_parse_backend import DoclingParseDocumentBackend
+from docling.backend.msword_backend import MsWordDocumentBackend
+from docling.datamodel.base_models import InputFormat
 from docling.datamodel.document import DocumentConversionInput
-from docling.document_converter import DocumentConverter
+from docling.document_converter import DocumentConverter, FormatOption, PdfFormatOption
+from docling.pipeline.simple_model_pipeline import SimpleModelPipeline
+from docling.pipeline.standard_pdf_model_pipeline import StandardPdfModelPipeline
 
 _log = logging.getLogger(__name__)
 
@@ -13,6 +18,7 @@ input_paths = [
     Path("tests/data/word_sample.docx"),
     Path("tests/data/lorem_ipsum.docx"),
     Path("tests/data/powerpoint_sample.pptx"),
+    Path("tests/data/powerpoint_sample.pptx"),
     Path("tests/data/2206.01062.pdf"),
 ]
 input = DocumentConversionInput.from_paths(input_paths)
@@ -21,22 +27,29 @@ input = DocumentConversionInput.from_paths(input_paths)
 doc_converter = DocumentConverter()
 
 # to customize use:
-# doc_converter = DocumentConverter( # all of the below is optional, has internal defaults.
-#     formats=[InputFormat.PDF, InputFormat.DOCX],
+# doc_converter = DocumentConverter(  # all of the below is optional, has internal defaults.
+#     formats=[
+#         InputFormat.PDF,
+#         InputFormat.DOCX,
+#     ],  # whitelist formats, other files are ignored.
 #     format_options={
-#         InputFormat.PDF: FormatOption(pipeline_cls=StandardPdfModelPipeline, backend=PyPdfiumDocumentBackend),
-#         InputFormat.DOCX: FormatOption(pipeline_cls=SimpleModelPipeline, backend=MsWordDocumentBackend)
-#     }
+#         InputFormat.PDF: PdfFormatOption(backend=DoclingParseDocumentBackend),
+#         InputFormat.DOCX: FormatOption(
+#             pipeline_cls=StandardPdfModelPipeline, backend=MsWordDocumentBackend
+#         ),
+#         # InputFormat.IMAGE: PdfFormatOption(),
+#     },
 # )
 
 conv_results = doc_converter.convert(input)
 
 for res in conv_results:
-    print("")
+    out_path = Path("./scratch") / f"{res.input.file.name}.experimental.md"
     print(
-        f"Document {res.input.file.name} converted with status {res.status}. Content:"
+        f"Document {res.input.file.name} converted with status {res.status}."
+        f"\nSaved markdown output to: {str(out_path)}"
     )
-    print(res.experimental.export_to_markdown())
+    # print(res.experimental.export_to_markdown())
     # Export Docling document format to markdown (experimental):
-    with (Path("./scratch") / f"{res.input.file.name}.experimental.md").open("w") as fp:
+    with out_path.open("w") as fp:
         fp.write(res.experimental.export_to_markdown())