docling 2.42.1__tar.gz → 2.42.2__tar.gz

{docling-2.42.1 → docling-2.42.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: docling
-Version: 2.42.1
+Version: 2.42.2
 Summary: SDK and CLI for parsing PDF, DOCX, HTML, and more, to a unified document representation for powering downstream workflows such as gen AI applications.
 Author-email: Christoph Auer <cau@zurich.ibm.com>, Michele Dolfi <dol@zurich.ibm.com>, Maxim Lysak <mly@zurich.ibm.com>, Nikos Livathinos <nli@zurich.ibm.com>, Ahmed Nassar <ahn@zurich.ibm.com>, Panos Vagenas <pva@zurich.ibm.com>, Peter Staar <taa@zurich.ibm.com>
 License-Expression: MIT
@@ -89,6 +89,7 @@ Dynamic: license-file
 [![License MIT](https://img.shields.io/github/license/docling-project/docling)](https://opensource.org/licenses/MIT)
 [![PyPI Downloads](https://static.pepy.tech/badge/docling/month)](https://pepy.tech/projects/docling)
 [![Docling Actor](https://apify.com/actor-badge?actor=vancura/docling?fpr=docling)](https://apify.com/vancura/docling)
+[![Chat with Dosu](https://dosu.dev/dosu-chat-badge.svg)](https://app.dosu.dev/097760a8-135e-4789-8234-90c8837d7f1c/ask?utm_source=github)
 [![OpenSSF Best Practices](https://www.bestpractices.dev/projects/10101/badge)](https://www.bestpractices.dev/projects/10101)
 [![LF AI & Data](https://img.shields.io/badge/LF%20AI%20%26%20Data-003778?logo=linuxfoundation&logoColor=fff&color=0094ff&labelColor=003778)](https://lfaidata.foundation/projects/)
 

{docling-2.42.1 → docling-2.42.2}/README.md

@@ -21,6 +21,7 @@
 [![License MIT](https://img.shields.io/github/license/docling-project/docling)](https://opensource.org/licenses/MIT)
 [![PyPI Downloads](https://static.pepy.tech/badge/docling/month)](https://pepy.tech/projects/docling)
 [![Docling Actor](https://apify.com/actor-badge?actor=vancura/docling?fpr=docling)](https://apify.com/vancura/docling)
+[![Chat with Dosu](https://dosu.dev/dosu-chat-badge.svg)](https://app.dosu.dev/097760a8-135e-4789-8234-90c8837d7f1c/ask?utm_source=github)
 [![OpenSSF Best Practices](https://www.bestpractices.dev/projects/10101/badge)](https://www.bestpractices.dev/projects/10101)
 [![LF AI & Data](https://img.shields.io/badge/LF%20AI%20%26%20Data-003778?logo=linuxfoundation&logoColor=fff&color=0094ff&labelColor=003778)](https://lfaidata.foundation/projects/)
 

{docling-2.42.1 → docling-2.42.2}/docling/backend/html_backend.py

@@ -5,7 +5,7 @@ from io import BytesIO
 from pathlib import Path
 from typing import Final, Optional, Union, cast
 
-from bs4 import BeautifulSoup, NavigableString, Tag
+from bs4 import BeautifulSoup, NavigableString, PageElement, Tag
 from bs4.element import PreformattedString
 from docling_core.types.doc import (
     DocItem,
@@ -297,7 +297,7 @@ class HTMLDocumentBackend(DeclarativeDocumentBackend):
                     ):
                         parts.append(child)
                     elif isinstance(child, Tag) and child.name not in ("ul", "ol"):
-                        text_part = child.get_text()
+                        text_part = HTMLDocumentBackend.get_text(child)
                         if text_part:
                             parts.append(text_part)
                 li_text = re.sub(r"\s+|\n+", " ", "".join(parts)).strip()
@@ -417,6 +417,36 @@ class HTMLDocumentBackend(DeclarativeDocumentBackend):
             content_layer=self.content_layer,
         )
 
+    @staticmethod
+    def get_text(item: PageElement) -> str:
+        """Concatenate all child strings of a PageElement.
+
+        This method is equivalent to `PageElement.get_text()` but also considers
+        certain tags. When called on a <p> or <li> tags, it returns the text with a
+        trailing space, otherwise the text is concatenated without separators.
+        """
+
+        def _extract_text_recursively(item: PageElement) -> list[str]:
+            """Recursively extract text from all child nodes."""
+            result: list[str] = []
+
+            if isinstance(item, NavigableString):
+                result = [item]
+            elif isinstance(item, Tag):
+                tag = cast(Tag, item)
+                parts: list[str] = []
+                for child in tag:
+                    parts.extend(_extract_text_recursively(child))
+                result.append(
+                    "".join(parts) + " " if tag.name in {"p", "li"} else "".join(parts)
+                )
+
+            return result
+
+        parts: list[str] = _extract_text_recursively(item)
+
+        return "".join(parts)
+
     @staticmethod
     def _get_cell_spans(cell: Tag) -> tuple[int, int]:
         """Extract colspan and rowspan values from a table cell tag.
@@ -510,9 +540,7 @@ class HTMLDocumentBackend(DeclarativeDocumentBackend):
                         formula.replace_with(NavigableString(math_formula))
 
                 # TODO: extract content correctly from table-cells with lists
-                text = html_cell.text
-
-                # label = html_cell.name
+                text = HTMLDocumentBackend.get_text(html_cell).strip()
                 col_span, row_span = HTMLDocumentBackend._get_cell_spans(html_cell)
                 if row_header:
                     row_span -= 1

{docling-2.42.1 → docling-2.42.2}/docling/backend/msword_backend.py

@@ -1104,8 +1104,17 @@ class MsWordDocumentBackend(DeclarativeDocumentBackend):
                     )
                 _log.debug(f"  spanned before row {spanned_idx}")
 
+                # Detect equations in cell text
+                text, equations = self._handle_equations_in_text(
+                    element=cell._element, text=cell.text
+                )
+                if len(equations) == 0:
+                    text = cell.text
+                else:
+                    text = text.replace("<eq>", "$").replace("</eq>", "$")
+
                 table_cell = TableCell(
-                    text=cell.text,
+                    text=text,
                     row_span=spanned_idx - row_idx,
                     col_span=cell.grid_span,
                     start_row_offset_idx=row.grid_cols_before + row_idx,

{docling-2.42.1 → docling-2.42.2}/docling/backend/pdf_backend.py

@@ -57,7 +57,31 @@ class PdfDocumentBackend(PaginatedDocumentBackend):
             if self.input_format is InputFormat.IMAGE:
                 buf = BytesIO()
                 img = Image.open(self.path_or_stream)
-                img.save(buf, "PDF")
+
+                # Handle multi-page TIFF images
+                if hasattr(img, "n_frames") and img.n_frames > 1:
+                    # Extract all frames from multi-page image
+                    frames = []
+                    try:
+                        for i in range(img.n_frames):
+                            img.seek(i)
+                            frame = img.copy().convert("RGB")
+                            frames.append(frame)
+                    except EOFError:
+                        pass
+
+                    # Save as multi-page PDF
+                    if frames:
+                        frames[0].save(
+                            buf, "PDF", save_all=True, append_images=frames[1:]
+                        )
+                    else:
+                        # Fallback to single page if frame extraction fails
+                        img.convert("RGB").save(buf, "PDF")
+                else:
+                    # Single page image - convert to RGB and save
+                    img.convert("RGB").save(buf, "PDF")
+
                 buf.seek(0)
                 self.path_or_stream = buf
             else:

{docling-2.42.1 → docling-2.42.2}/docling/pipeline/base_pipeline.py

@@ -217,7 +217,13 @@ class PaginatedPipeline(BasePipeline):  # TODO this is a bad name.
         return conv_res
 
     def _determine_status(self, conv_res: ConversionResult) -> ConversionStatus:
-        status = ConversionStatus.SUCCESS
+        status = conv_res.status
+        if status in [
+            ConversionStatus.PENDING,
+            ConversionStatus.STARTED,
+        ]:  # preserves ConversionStatus.PARTIAL_SUCCESS
+            status = ConversionStatus.SUCCESS
+
         for page in conv_res.pages:
             if page._backend is None or not page._backend.is_valid():
                 conv_res.errors.append(

{docling-2.42.1 → docling-2.42.2}/docling.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: docling
-Version: 2.42.1
+Version: 2.42.2
 Summary: SDK and CLI for parsing PDF, DOCX, HTML, and more, to a unified document representation for powering downstream workflows such as gen AI applications.
 Author-email: Christoph Auer <cau@zurich.ibm.com>, Michele Dolfi <dol@zurich.ibm.com>, Maxim Lysak <mly@zurich.ibm.com>, Nikos Livathinos <nli@zurich.ibm.com>, Ahmed Nassar <ahn@zurich.ibm.com>, Panos Vagenas <pva@zurich.ibm.com>, Peter Staar <taa@zurich.ibm.com>
 License-Expression: MIT
@@ -89,6 +89,7 @@ Dynamic: license-file
 [![License MIT](https://img.shields.io/github/license/docling-project/docling)](https://opensource.org/licenses/MIT)
 [![PyPI Downloads](https://static.pepy.tech/badge/docling/month)](https://pepy.tech/projects/docling)
 [![Docling Actor](https://apify.com/actor-badge?actor=vancura/docling?fpr=docling)](https://apify.com/vancura/docling)
+[![Chat with Dosu](https://dosu.dev/dosu-chat-badge.svg)](https://app.dosu.dev/097760a8-135e-4789-8234-90c8837d7f1c/ask?utm_source=github)
 [![OpenSSF Best Practices](https://www.bestpractices.dev/projects/10101/badge)](https://www.bestpractices.dev/projects/10101)
 [![LF AI & Data](https://img.shields.io/badge/LF%20AI%20%26%20Data-003778?logo=linuxfoundation&logoColor=fff&color=0094ff&labelColor=003778)](https://lfaidata.foundation/projects/)
 

{docling-2.42.1 → docling-2.42.2}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "docling"
-version = "2.42.1"  # DO NOT EDIT, updated automatically
+version = "2.42.2"  # DO NOT EDIT, updated automatically
 description = "SDK and CLI for parsing PDF, DOCX, HTML, and more, to a unified document representation for powering downstream workflows such as gen AI applications."
 license = "MIT"
 keywords = [

{docling-2.42.1 → docling-2.42.2}/tests/test_input_doc.py

@@ -243,3 +243,26 @@ def _make_input_doc_from_stream(doc_stream):
         backend=PdfFormatOption().backend,  # use default
     )
     return in_doc
+
+
+def test_tiff_two_pages():
+    tiff_path = Path("./tests/data/tiff/2206.01062.tif")
+    doc = InputDocument(
+        path_or_stream=tiff_path,
+        format=InputFormat.IMAGE,
+        backend=PdfFormatOption().backend,  # use default backend
+    )
+    assert doc.valid is True
+    assert doc.page_count == 2
+
+    # Expect two full-page rectangles
+    rects_page1 = doc._backend.load_page(0).get_bitmap_rects()
+    rects_page2 = doc._backend.load_page(1).get_bitmap_rects()
+
+    page1_rect = next(rects_page1)
+    page2_rect = next(rects_page2)
+
+    assert page1_rect.t == page2_rect.t == 0
+    assert page1_rect.l == page2_rect.l == 0
+    assert page1_rect.r == page2_rect.r == 612.0
+    assert page1_rect.b == page2_rect.b == 792.0