docling 1.19.0__py3-none-any.whl → 2.1.0__py3-none-any.whl

docling/cli/main.py

@@ -5,22 +5,27 @@ import time
 import warnings
 from enum import Enum
 from pathlib import Path
-from typing import Annotated, Iterable, List, Optional
+from typing import Annotated, Dict, Iterable, List, Optional
 
 import typer
 from docling_core.utils.file import resolve_file_source
 
 from docling.backend.docling_parse_backend import DoclingParseDocumentBackend
-from docling.backend.pypdfium2_backend import PyPdfiumDocumentBackend
-from docling.datamodel.base_models import ConversionStatus
-from docling.datamodel.document import ConversionResult, DocumentConversionInput
+from docling.datamodel.base_models import (
+    ConversionStatus,
+    FormatToExtensions,
+    InputFormat,
+    OutputFormat,
+)
+from docling.datamodel.document import ConversionResult
 from docling.datamodel.pipeline_options import (
     EasyOcrOptions,
-    PipelineOptions,
+    OcrOptions,
+    PdfPipelineOptions,
     TesseractCliOcrOptions,
     TesseractOcrOptions,
 )
-from docling.document_converter import DocumentConverter
+from docling.document_converter import DocumentConverter, FormatOption, PdfFormatOption
 
 warnings.filterwarnings(action="ignore", category=UserWarning, module="pydantic|torch")
 warnings.filterwarnings(action="ignore", category=FutureWarning, module="easyocr")
@@ -87,28 +92,28 @@ def export_documents(
                 fname = output_dir / f"{doc_filename}.json"
                 with fname.open("w") as fp:
                     _log.info(f"writing JSON output to {fname}")
-                    fp.write(json.dumps(conv_res.render_as_dict()))
+                    fp.write(json.dumps(conv_res.document.export_to_dict()))
 
             # Export Text format:
             if export_txt:
                 fname = output_dir / f"{doc_filename}.txt"
                 with fname.open("w") as fp:
                     _log.info(f"writing Text output to {fname}")
-                    fp.write(conv_res.render_as_text())
+                    fp.write(conv_res.document.export_to_markdown(strict_text=True))
 
             # Export Markdown format:
             if export_md:
                 fname = output_dir / f"{doc_filename}.md"
                 with fname.open("w") as fp:
                     _log.info(f"writing Markdown output to {fname}")
-                    fp.write(conv_res.render_as_markdown())
+                    fp.write(conv_res.document.export_to_markdown())
 
             # Export Document Tags format:
             if export_doctags:
                 fname = output_dir / f"{doc_filename}.doctags"
                 with fname.open("w") as fp:
                     _log.info(f"writing Doc Tags output to {fname}")
-                    fp.write(conv_res.render_as_doctags())
+                    fp.write(conv_res.document.export_to_document_tokens())
 
         else:
             _log.warning(f"Document {conv_res.input.file} failed to convert.")
@@ -129,44 +134,31 @@ def convert(
             help="PDF files to convert. Can be local file / directory paths or URL.",
         ),
     ],
-    export_json: Annotated[
-        bool,
-        typer.Option(
-            ..., "--json/--no-json", help="If enabled the document is exported as JSON."
-        ),
-    ] = False,
-    export_md: Annotated[
-        bool,
-        typer.Option(
-            ..., "--md/--no-md", help="If enabled the document is exported as Markdown."
-        ),
-    ] = True,
-    export_txt: Annotated[
-        bool,
-        typer.Option(
-            ..., "--txt/--no-txt", help="If enabled the document is exported as Text."
-        ),
-    ] = False,
-    export_doctags: Annotated[
-        bool,
-        typer.Option(
-            ...,
-            "--doctags/--no-doctags",
-            help="If enabled the document is exported as Doc Tags.",
-        ),
-    ] = False,
+    from_formats: List[InputFormat] = typer.Option(
+        None,
+        "--from",
+        help="Specify input formats to convert from. Defaults to all formats.",
+    ),
+    to_formats: List[OutputFormat] = typer.Option(
+        None, "--to", help="Specify output formats. Defaults to Markdown."
+    ),
     ocr: Annotated[
         bool,
         typer.Option(
             ..., help="If enabled, the bitmap content will be processed using OCR."
         ),
     ] = True,
-    backend: Annotated[
-        Backend, typer.Option(..., help="The PDF backend to use.")
-    ] = Backend.DOCLING,
     ocr_engine: Annotated[
         OcrEngine, typer.Option(..., help="The OCR engine to use.")
     ] = OcrEngine.EASYOCR,
+    abort_on_error: Annotated[
+        bool,
+        typer.Option(
+            ...,
+            "--abort-on-error/--no-abort-on-error",
+            help="If enabled, the bitmap content will be processed using OCR.",
+        ),
+    ] = False,
     output: Annotated[
         Path, typer.Option(..., help="Output directory where results are saved.")
     ] = Path("."),
@@ -182,6 +174,9 @@ def convert(
 ):
     logging.basicConfig(level=logging.INFO)
 
+    if from_formats is None:
+        from_formats = [e for e in InputFormat]
+
     input_doc_paths: List[Path] = []
     for src in input_sources:
         source = resolve_file_source(source=src)
@@ -191,48 +186,54 @@ def convert(
             )
             raise typer.Abort()
         elif source.is_dir():
-            input_doc_paths.extend(list(source.glob("**/*.pdf")))
-            input_doc_paths.extend(list(source.glob("**/*.PDF")))
+            for fmt in from_formats:
+                for ext in FormatToExtensions[fmt]:
+                    input_doc_paths.extend(list(source.glob(f"**/*.{ext}")))
+                    input_doc_paths.extend(list(source.glob(f"**/*.{ext.upper()}")))
         else:
             input_doc_paths.append(source)
 
-    match backend:
-        case Backend.PYPDFIUM2:
-            do_cell_matching = ocr  # only do cell matching when OCR enabled
-            pdf_backend = PyPdfiumDocumentBackend
-        case Backend.DOCLING:
-            do_cell_matching = True
-            pdf_backend = DoclingParseDocumentBackend
-        case _:
-            raise RuntimeError(f"Unexpected backend type {backend}")
+    if to_formats is None:
+        to_formats = [OutputFormat.MARKDOWN]
+
+    export_json = OutputFormat.JSON in to_formats
+    export_md = OutputFormat.MARKDOWN in to_formats
+    export_txt = OutputFormat.TEXT in to_formats
+    export_doctags = OutputFormat.DOCTAGS in to_formats
 
     match ocr_engine:
         case OcrEngine.EASYOCR:
-            ocr_options = EasyOcrOptions()
+            ocr_options: OcrOptions = EasyOcrOptions()
         case OcrEngine.TESSERACT_CLI:
             ocr_options = TesseractCliOcrOptions()
         case OcrEngine.TESSERACT:
             ocr_options = TesseractOcrOptions()
         case _:
-            raise RuntimeError(f"Unexpected backend type {backend}")
+            raise RuntimeError(f"Unexpected OCR engine type {ocr_engine}")
 
-    pipeline_options = PipelineOptions(
+    pipeline_options = PdfPipelineOptions(
         do_ocr=ocr,
         ocr_options=ocr_options,
         do_table_structure=True,
     )
-    pipeline_options.table_structure_options.do_cell_matching = do_cell_matching
+    pipeline_options.table_structure_options.do_cell_matching = True  # do_cell_matching
+
+    format_options: Dict[InputFormat, FormatOption] = {
+        InputFormat.PDF: PdfFormatOption(
+            pipeline_options=pipeline_options,
+            backend=DoclingParseDocumentBackend,  # pdf_backend
+        )
+    }
     doc_converter = DocumentConverter(
-        pipeline_options=pipeline_options,
-        pdf_backend=pdf_backend,
+        allowed_formats=from_formats,
+        format_options=format_options,
     )
 
-    # Define input files
-    input = DocumentConversionInput.from_paths(input_doc_paths)
-
     start_time = time.time()
 
-    conv_results = doc_converter.convert(input)
+    conv_results = doc_converter.convert_all(
+        input_doc_paths, raises_on_error=abort_on_error
+    )
 
     output.mkdir(parents=True, exist_ok=True)
     export_documents(

docling/datamodel/base_models.py

@@ -1,18 +1,19 @@
-import copy
-import warnings
 from enum import Enum, auto
 from io import BytesIO
-from typing import Annotated, Any, Dict, List, Optional, Tuple, Union
-
+from typing import TYPE_CHECKING, Dict, List, Optional, Set, Union
+
+from docling_core.types.doc import (
+    BoundingBox,
+    DocItemLabel,
+    PictureDataType,
+    Size,
+    TableCell,
+)
 from PIL.Image import Image
-from pydantic import BaseModel, ConfigDict, Field, model_validator
-from typing_extensions import Self
+from pydantic import BaseModel, ConfigDict
 
-from docling.backend.abstract_backend import PdfPageBackend
-from docling.datamodel.pipeline_options import (  # Must be imported here for backward compatibility.
-    PipelineOptions,
-    TableStructureOptions,
-)
+if TYPE_CHECKING:
+    from docling.backend.pdf_backend import PdfPageBackend
 
 
 class ConversionStatus(str, Enum):
@@ -23,18 +24,61 @@ class ConversionStatus(str, Enum):
     PARTIAL_SUCCESS = auto()
 
 
+class InputFormat(str, Enum):
+    DOCX = "docx"
+    PPTX = "pptx"
+    HTML = "html"
+    IMAGE = "image"
+    PDF = "pdf"
+
+
+class OutputFormat(str, Enum):
+    MARKDOWN = "md"
+    JSON = "json"
+    TEXT = "text"
+    DOCTAGS = "doctags"
+
+
+FormatToExtensions: Dict[InputFormat, List[str]] = {
+    InputFormat.DOCX: ["docx", "dotx", "docm", "dotm"],
+    InputFormat.PPTX: ["pptx", "potx", "ppsx", "pptm", "potm", "ppsm"],
+    InputFormat.PDF: ["pdf"],
+    InputFormat.HTML: ["html", "htm", "xhtml"],
+    InputFormat.IMAGE: ["jpg", "jpeg", "png", "tif", "tiff", "bmp"],
+}
+
+FormatToMimeType: Dict[InputFormat, Set[str]] = {
+    InputFormat.DOCX: {
+        "application/vnd.openxmlformats-officedocument.wordprocessingml.document",
+        "application/vnd.openxmlformats-officedocument.wordprocessingml.template",
+    },
+    InputFormat.PPTX: {
+        "application/vnd.openxmlformats-officedocument.presentationml.template",
+        "application/vnd.openxmlformats-officedocument.presentationml.slideshow",
+        "application/vnd.openxmlformats-officedocument.presentationml.presentation",
+    },
+    InputFormat.HTML: {"text/html", "application/xhtml+xml"},
+    InputFormat.IMAGE: {
+        "image/png",
+        "image/jpeg",
+        "image/tiff",
+        "image/gif",
+        "image/bmp",
+    },
+    InputFormat.PDF: {"application/pdf"},
+}
+MimeTypeToFormat = {
+    mime: fmt for fmt, mimes in FormatToMimeType.items() for mime in mimes
+}
+
+
 class DocInputType(str, Enum):
     PATH = auto()
     STREAM = auto()
 
 
-class CoordOrigin(str, Enum):
-    TOPLEFT = auto()
-    BOTTOMLEFT = auto()
-
-
 class DoclingComponentType(str, Enum):
-    PDF_BACKEND = auto()
+    DOCUMENT_BACKEND = auto()
     MODEL = auto()
     DOC_ASSEMBLER = auto()
 
@@ -45,118 +89,6 @@ class ErrorItem(BaseModel):
     error_message: str
 
 
-class PageSize(BaseModel):
-    width: float = 0.0
-    height: float = 0.0
-
-
-class BoundingBox(BaseModel):
-    l: float  # left
-    t: float  # top
-    r: float  # right
-    b: float  # bottom
-
-    coord_origin: CoordOrigin = CoordOrigin.TOPLEFT
-
-    @property
-    def width(self):
-        return self.r - self.l
-
-    @property
-    def height(self):
-        return abs(self.t - self.b)
-
-    def scaled(self, scale: float) -> "BoundingBox":
-        out_bbox = copy.deepcopy(self)
-        out_bbox.l *= scale
-        out_bbox.r *= scale
-        out_bbox.t *= scale
-        out_bbox.b *= scale
-
-        return out_bbox
-
-    def normalized(self, page_size: PageSize) -> "BoundingBox":
-        out_bbox = copy.deepcopy(self)
-        out_bbox.l /= page_size.width
-        out_bbox.r /= page_size.width
-        out_bbox.t /= page_size.height
-        out_bbox.b /= page_size.height
-
-        return out_bbox
-
-    def as_tuple(self):
-        if self.coord_origin == CoordOrigin.TOPLEFT:
-            return (self.l, self.t, self.r, self.b)
-        elif self.coord_origin == CoordOrigin.BOTTOMLEFT:
-            return (self.l, self.b, self.r, self.t)
-
-    @classmethod
-    def from_tuple(cls, coord: Tuple[float, ...], origin: CoordOrigin):
-        if origin == CoordOrigin.TOPLEFT:
-            l, t, r, b = coord[0], coord[1], coord[2], coord[3]
-            if r < l:
-                l, r = r, l
-            if b < t:
-                b, t = t, b
-
-            return BoundingBox(l=l, t=t, r=r, b=b, coord_origin=origin)
-        elif origin == CoordOrigin.BOTTOMLEFT:
-            l, b, r, t = coord[0], coord[1], coord[2], coord[3]
-            if r < l:
-                l, r = r, l
-            if b > t:
-                b, t = t, b
-
-            return BoundingBox(l=l, t=t, r=r, b=b, coord_origin=origin)
-
-    def area(self) -> float:
-        area = (self.r - self.l) * (self.b - self.t)
-        if self.coord_origin == CoordOrigin.BOTTOMLEFT:
-            area = -area
-        return area
-
-    def intersection_area_with(self, other: "BoundingBox") -> float:
-        # Calculate intersection coordinates
-        left = max(self.l, other.l)
-        top = max(self.t, other.t)
-        right = min(self.r, other.r)
-        bottom = min(self.b, other.b)
-
-        # Calculate intersection dimensions
-        width = right - left
-        height = bottom - top
-
-        # If the bounding boxes do not overlap, width or height will be negative
-        if width <= 0 or height <= 0:
-            return 0.0
-
-        return width * height
-
-    def to_bottom_left_origin(self, page_height) -> "BoundingBox":
-        if self.coord_origin == CoordOrigin.BOTTOMLEFT:
-            return self
-        elif self.coord_origin == CoordOrigin.TOPLEFT:
-            return BoundingBox(
-                l=self.l,
-                r=self.r,
-                t=page_height - self.t,
-                b=page_height - self.b,
-                coord_origin=CoordOrigin.BOTTOMLEFT,
-            )
-
-    def to_top_left_origin(self, page_height):
-        if self.coord_origin == CoordOrigin.TOPLEFT:
-            return self
-        elif self.coord_origin == CoordOrigin.BOTTOMLEFT:
-            return BoundingBox(
-                l=self.l,
-                r=self.r,
-                t=page_height - self.t,  # self.b
-                b=page_height - self.b,  # self.t
-                coord_origin=CoordOrigin.TOPLEFT,
-            )
-
-
 class Cell(BaseModel):
     id: int
     text: str
@@ -169,14 +101,14 @@ class OcrCell(Cell):
 
 class Cluster(BaseModel):
     id: int
-    label: str
+    label: DocItemLabel
     bbox: BoundingBox
     confidence: float = 1.0
     cells: List[Cell] = []
 
 
 class BasePageElement(BaseModel):
-    label: str
+    label: DocItemLabel
     id: int
     page_no: int
     cluster: Cluster
@@ -187,37 +119,7 @@ class LayoutPrediction(BaseModel):
     clusters: List[Cluster] = []
 
 
-class TableCell(BaseModel):
-    bbox: BoundingBox
-    row_span: int
-    col_span: int
-    start_row_offset_idx: int
-    end_row_offset_idx: int
-    start_col_offset_idx: int
-    end_col_offset_idx: int
-    text: str
-    column_header: bool = False
-    row_header: bool = False
-    row_section: bool = False
-
-    @model_validator(mode="before")
-    @classmethod
-    def from_dict_format(cls, data: Any) -> Any:
-        if isinstance(data, Dict):
-            text = data["bbox"].get("token", "")
-            if not len(text):
-                text_cells = data.pop("text_cell_bboxes", None)
-                if text_cells:
-                    for el in text_cells:
-                        text += el["token"] + " "
-
-                text = text.strip()
-            data["text"] = text
-
-        return data
-
-
-class TableElement(BasePageElement):
+class Table(BasePageElement):
     otsl_seq: List[str]
     num_rows: int = 0
     num_cols: int = 0
@@ -225,18 +127,15 @@ class TableElement(BasePageElement):
 
 
 class TableStructurePrediction(BaseModel):
-    table_map: Dict[int, TableElement] = {}
-
-
-class TextElement(BasePageElement): ...
+    table_map: Dict[int, Table] = {}
 
 
-class FigureData(BaseModel):
-    pass
+class TextElement(BasePageElement):
+    text: str
 
 
 class FigureElement(BasePageElement):
-    data: Optional[FigureData] = None
+    annotations: List[PictureDataType] = []
     provenance: Optional[str] = None
     predicted_class: Optional[str] = None
     confidence: Optional[float] = None
@@ -259,7 +158,7 @@ class PagePredictions(BaseModel):
     equations_prediction: Optional[EquationPrediction] = None
 
 
-PageElement = Union[TextElement, TableElement, FigureElement]
+PageElement = Union[TextElement, Table, FigureElement]
 
 
 class AssembledUnit(BaseModel):
@@ -272,13 +171,13 @@ class Page(BaseModel):
     model_config = ConfigDict(arbitrary_types_allowed=True)
 
     page_no: int
-    page_hash: Optional[str] = None
-    size: Optional[PageSize] = None
+    # page_hash: Optional[str] = None
+    size: Optional[Size] = None
     cells: List[Cell] = []
     predictions: PagePredictions = PagePredictions()
     assembled: Optional[AssembledUnit] = None
 
-    _backend: Optional[PdfPageBackend] = (
+    _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.
@@ -301,24 +200,5 @@ class Page(BaseModel):
 class DocumentStream(BaseModel):
     model_config = ConfigDict(arbitrary_types_allowed=True)
 
-    filename: str
+    name: str
     stream: BytesIO
-
-
-class AssembleOptions(BaseModel):
-    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