docling-core 2.38.0__tar.gz → 2.38.2__tar.gz

{docling_core-2.38.0 → docling_core-2.38.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: docling-core
-Version: 2.38.0
+Version: 2.38.2
 Summary: A python library to define and validate data types in Docling.
 Author-email: Cesar Berrospi Ramis <ceb@zurich.ibm.com>, Panos Vagenas <pva@zurich.ibm.com>, Michele Dolfi <dol@zurich.ibm.com>, Christoph Auer <cau@zurich.ibm.com>, Peter Staar <taa@zurich.ibm.com>
 Maintainer-email: Panos Vagenas <pva@zurich.ibm.com>, Michele Dolfi <dol@zurich.ibm.com>, Christoph Auer <cau@zurich.ibm.com>, Peter Staar <taa@zurich.ibm.com>, Cesar Berrospi Ramis <ceb@zurich.ibm.com>

{docling_core-2.38.0 → docling_core-2.38.2}/docling_core/transforms/serializer/common.py

@@ -349,6 +349,7 @@ class DocSerializer(BaseModel, BaseDocSerializer):
                         doc_serializer=self,
                         doc=self.doc,
                         is_inline_scope=is_inline_scope,
+                        visited=my_visited,
                         **my_kwargs,
                     )
                     if item.self_ref not in self.get_excluded_refs(**kwargs)

{docling_core-2.38.0 → docling_core-2.38.2}/docling_core/transforms/serializer/markdown.py

@@ -106,26 +106,49 @@ class MarkdownTextSerializer(BaseModel, BaseTextSerializer):
         doc_serializer: BaseDocSerializer,
         doc: DoclingDocument,
         is_inline_scope: bool = False,
+        visited: Optional[set[str]] = None,  # refs of visited items
         **kwargs: Any,
     ) -> SerializationResult:
         """Serializes the passed item."""
+        my_visited = visited if visited is not None else set()
         params = MarkdownParams(**kwargs)
         res_parts: list[SerializationResult] = []
+        text = item.text
         escape_html = True
         escape_underscores = True
-        if isinstance(item, TitleItem):
-            text_part = f"# {item.text}"
-        elif isinstance(item, SectionHeaderItem):
-            text_part = f"{(item.level + 1) * '#'} {item.text}"
+        processing_pending = True
+        if isinstance(item, (TitleItem, SectionHeaderItem)):
+            # case where processing/formatting should be applied first (in inner scope)
+            processing_pending = False
+            if (
+                text == ""
+                and len(item.children) == 1
+                and isinstance(
+                    (child_group := item.children[0].resolve(doc)), InlineGroup
+                )
+            ):
+                # case of heading with inline
+                ser_res = doc_serializer.serialize(item=child_group)
+                text = ser_res.text
+                for span in ser_res.spans:
+                    my_visited.add(span.item.self_ref)
+            else:
+                text = doc_serializer.post_process(
+                    text=text,
+                    escape_html=escape_html,
+                    escape_underscores=escape_underscores,
+                    formatting=item.formatting,
+                    hyperlink=item.hyperlink,
+                )
+            num_hashes = 1 if isinstance(item, TitleItem) else item.level + 1
+            text_part = f"{num_hashes * '#'} {text}"
         elif isinstance(item, CodeItem):
-            text_part = (
-                f"`{item.text}`" if is_inline_scope else f"```\n{item.text}\n```"
-            )
+            text_part = f"`{text}`" if is_inline_scope else f"```\n{text}\n```"
             escape_html = False
             escape_underscores = False
         elif isinstance(item, FormulaItem):
-            if item.text:
-                text_part = f"${item.text}$" if is_inline_scope else f"$${item.text}$$"
+            if text:
+                text_part = f"${text}$" if is_inline_scope else f"$${text}$$"
             elif item.orig:
                 text_part = "<!-- formula-not-decoded -->"
             else:
@@ -133,9 +156,10 @@ class MarkdownTextSerializer(BaseModel, BaseTextSerializer):
             escape_html = False
             escape_underscores = False
         elif params.wrap_width:
-            text_part = textwrap.fill(item.text, width=params.wrap_width)
+            # although wrapping is not guaranteed if post-processing makes changes
+            text_part = textwrap.fill(text, width=params.wrap_width)
         else:
-            text_part = item.text
+            text_part = text
 
         if text_part:
             text_res = create_ser_result(text=text_part, span_source=item)
@@ -147,13 +171,14 @@ class MarkdownTextSerializer(BaseModel, BaseTextSerializer):
                 res_parts.append(cap_res)
 
         text = (" " if is_inline_scope else "\n\n").join([r.text for r in res_parts])
-        text = doc_serializer.post_process(
-            text=text,
-            escape_html=escape_html,
-            escape_underscores=escape_underscores,
-            formatting=item.formatting,
-            hyperlink=item.hyperlink,
-        )
+        if processing_pending:
+            text = doc_serializer.post_process(
+                text=text,
+                escape_html=escape_html,
+                escape_underscores=escape_underscores,
+                formatting=item.formatting,
+                hyperlink=item.hyperlink,
+            )
         return create_ser_result(text=text, span_source=res_parts)
 
 

{docling_core-2.38.0 → docling_core-2.38.2}/docling_core/types/doc/document.py

@@ -540,6 +540,9 @@ class DocumentOrigin(BaseModel):
         "text/asciidoc",
         "text/markdown",
         "text/csv",
+        "audio/x-wav",
+        "audio/wav",
+        "audio/mp3",
     ]
 
     @field_validator("binary_hash", mode="before")

{docling_core-2.38.0 → docling_core-2.38.2}/docling_core/types/doc/page.py

@@ -649,6 +649,7 @@ class SegmentedPdfPage(SegmentedPage):
         add_location: bool = True,
         add_fontkey: bool = False,
         add_fontname: bool = True,
+        add_text_direction: bool = True,
     ) -> List[str]:
         """Export text cells as formatted text lines.
 
@@ -676,6 +677,9 @@ class SegmentedPdfPage(SegmentedPage):
             if add_fontname and isinstance(cell, PdfTextCell):
                 line += f"{cell.font_name:>10} "
 
+            if add_text_direction and isinstance(cell, PdfTextCell):
+                line += f"{cell.text_direction} "
+
             line += f"{cell.text}"
             lines.append(line)
 

{docling_core-2.38.0 → docling_core-2.38.2}/docling_core.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: docling-core
-Version: 2.38.0
+Version: 2.38.2
 Summary: A python library to define and validate data types in Docling.
 Author-email: Cesar Berrospi Ramis <ceb@zurich.ibm.com>, Panos Vagenas <pva@zurich.ibm.com>, Michele Dolfi <dol@zurich.ibm.com>, Christoph Auer <cau@zurich.ibm.com>, Peter Staar <taa@zurich.ibm.com>
 Maintainer-email: Panos Vagenas <pva@zurich.ibm.com>, Michele Dolfi <dol@zurich.ibm.com>, Christoph Auer <cau@zurich.ibm.com>, Peter Staar <taa@zurich.ibm.com>, Cesar Berrospi Ramis <ceb@zurich.ibm.com>

{docling_core-2.38.0 → docling_core-2.38.2}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "docling-core"
-version = "2.38.0"  # DO NOT EDIT, updated automatically
+version = "2.38.2"  # DO NOT EDIT, updated automatically
 description = "A python library to define and validate data types in Docling."
 license = "MIT"
 license-files = ["LICENSE"]

{docling_core-2.38.0 → docling_core-2.38.2}/test/test_serialization.py

@@ -196,6 +196,20 @@ def test_md_charts():
     verify(exp_file=src.parent / f"{src.stem}.gt.md", actual=actual)
 
 
+def test_md_inline_and_formatting():
+    src = Path("./test/data/doc/inline_and_formatting.yaml")
+    doc = DoclingDocument.load_from_yaml(src)
+
+    ser = MarkdownDocSerializer(
+        doc=doc,
+        params=MarkdownParams(
+            image_mode=ImageRefMode.PLACEHOLDER,
+        ),
+    )
+    actual = ser.serialize().text
+    verify(exp_file=src.parent / f"{src.stem}.md", actual=actual)
+
+
 def test_html_cross_page_list_page_break():
     src = Path("./test/data/doc/activities.json")
     doc = DoclingDocument.load_from_json(src)