(MD export) add inline formulas, placeholders for KeyValueItem, `Fo…

…rmItem` Signed-off-by: Panos Vagenas <[email protected]>
DS4SD · Feb 24, 2025 · d28c100 · d28c100
1 parent 1953b73
commit d28c100
Show file tree

Hide file tree

Showing 18 changed files with 841 additions and 56 deletions.
diff --git a/docling_core/types/doc/document.py b/docling_core/types/doc/document.py
@@ -1397,6 +1397,10 @@ class KeyValueItem(FloatingItem):
 
     graph: GraphData
 
+    def _export_to_markdown(self) -> str:
+        # TODO add actual implementation
+        return "<!-- missing-key-value-item -->"
+
 
 class FormItem(FloatingItem):
     """FormItem."""
@@ -1405,6 +1409,10 @@ class FormItem(FloatingItem):
 
     graph: GraphData
 
+    def _export_to_markdown(self) -> str:
+        # TODO add actual implementation
+        return "<!-- missing-form-item -->"
+
 
 ContentItem = Annotated[
     Union[
@@ -2537,14 +2545,14 @@ def _ingest_text(text: str, do_escape_html=True, do_escape_underscores=True):
                 text = f"{marker} {item.text}"
                 _ingest_text(text.strip())
 
-            elif isinstance(item, CodeItem) and item.label in labels:
+            elif isinstance(item, CodeItem):
                 text = f"`{item.text}`" if is_inline_scope else f"```\n{item.text}\n```"
                 _ingest_text(text, do_escape_underscores=False, do_escape_html=False)
 
             elif isinstance(item, TextItem) and item.label in [DocItemLabel.FORMULA]:
                 if item.text != "":
                     _ingest_text(
-                        f"$${item.text}$$",
+                        f"${item.text}$" if is_inline_scope else f"$${item.text}$$",
                         do_escape_underscores=False,
                         do_escape_html=False,
                     )
@@ -2555,7 +2563,7 @@ def _ingest_text(text: str, do_escape_html=True, do_escape_underscores=True):
                         do_escape_html=False,
                     )
 
-            elif isinstance(item, TextItem) and item.label in labels:
+            elif isinstance(item, TextItem):
                 if len(item.text) and text_width > 0:
                     text = item.text
                     wrapped_text = textwrap.fill(text, width=text_width)
@@ -2580,7 +2588,11 @@ def _ingest_text(text: str, do_escape_html=True, do_escape_underscores=True):
 
                 _ingest_text(line, do_escape_html=False, do_escape_underscores=False)
 
-            elif isinstance(item, DocItem) and item.label in labels:
+            elif isinstance(item, (KeyValueItem, FormItem)):
+                text = item._export_to_markdown()
+                _ingest_text(text, do_escape_html=False, do_escape_underscores=False)
+
+            elif isinstance(item, DocItem):
                 text = "<!-- missing-text -->"
                 _ingest_text(text, do_escape_html=False, do_escape_underscores=False)
 

diff --git a/test/data/doc/constructed_doc.dt b/test/data/doc/constructed_doc.dt
@@ -29,7 +29,13 @@ Affiliation 2</text>
 <code<_unknown_>print("Hello world")</code
 <paragraph>(to be displayed inline)</paragraph>
 </unordered_list>
+<paragraph>Here a formula:</paragraph>
+<formula>E=mc^2</formula>
+<paragraph>(to be displayed inline)</paragraph>
 </unordered_list>
-<paragraph>And a code block:</paragraph>
+<paragraph>Here a code block:</paragraph>
 <code<_unknown_>print("Hello world")</code
+<paragraph>Here a formula block:</paragraph>
+<formula>E=mc^2</formula>
+<paragraph>The end.</paragraph>
 </doctag>
diff --git a/test/data/doc/constructed_doc.dt.gt b/test/data/doc/constructed_doc.dt.gt
@@ -29,7 +29,13 @@ Affiliation 2</text>
 <code<_unknown_>print("Hello world")</code
 <paragraph>(to be displayed inline)</paragraph>
 </unordered_list>
+<paragraph>Here a formula:</paragraph>
+<formula>E=mc^2</formula>
+<paragraph>(to be displayed inline)</paragraph>
 </unordered_list>
-<paragraph>And a code block:</paragraph>
+<paragraph>Here a code block:</paragraph>
 <code<_unknown_>print("Hello world")</code
+<paragraph>Here a formula block:</paragraph>
+<formula>E=mc^2</formula>
+<paragraph>The end.</paragraph>
 </doctag>
diff --git a/test/data/doc/constructed_doc.embedded.html.gt b/test/data/doc/constructed_doc.embedded.html.gt
@@ -104,7 +104,13 @@
 <pre><code>print("Hello world")</code></pre>
 <p>(to be displayed inline)</p>
 </ul>
+<p>Here a formula:</p>
+<div><math xmlns="http://www.w3.org/1998/Math/MathML" display="block"><mrow><mi>E</mi><mo>&#x0003D;</mo><mi>m</mi><msup><mi>c</mi><mn>2</mn></msup></mrow><annotation encoding="TeX">E=mc^2</annotation></math></div>
+<p>(to be displayed inline)</p>
 </ul>
-<p>And a code block:</p>
+<p>Here a code block:</p>
 <pre><code>print("Hello world")</code></pre>
+<p>Here a formula block:</p>
+<div><math xmlns="http://www.w3.org/1998/Math/MathML" display="block"><mrow><mi>E</mi><mo>&#x0003D;</mo><mi>m</mi><msup><mi>c</mi><mn>2</mn></msup></mrow><annotation encoding="TeX">E=mc^2</annotation></math></div>
+<p>The end.</p>
 </html>
diff --git a/test/data/doc/constructed_doc.embedded.json.gt b/test/data/doc/constructed_doc.embedded.json.gt
@@ -43,10 +43,25 @@
         "$ref": "#/groups/5"
       },
       {
-        "$ref": "#/texts/24"
+        "$ref": "#/texts/27"
       },
       {
-        "$ref": "#/texts/25"
+        "$ref": "#/texts/28"
+      },
+      {
+        "$ref": "#/texts/29"
+      },
+      {
+        "$ref": "#/texts/30"
+      },
+      {
+        "$ref": "#/key_value_items/0"
+      },
+      {
+        "$ref": "#/form_items/0"
+      },
+      {
+        "$ref": "#/texts/31"
       }
     ],
     "content_layer": "body",
@@ -185,6 +200,9 @@
         },
         {
           "$ref": "#/groups/7"
+        },
+        {
+          "$ref": "#/groups/8"
         }
       ],
       "content_layer": "body",
@@ -210,6 +228,26 @@
       "content_layer": "body",
       "name": "group",
       "label": "inline"
+    },
+    {
+      "self_ref": "#/groups/8",
+      "parent": {
+        "$ref": "#/groups/6"
+      },
+      "children": [
+        {
+          "$ref": "#/texts/24"
+        },
+        {
+          "$ref": "#/texts/25"
+        },
+        {
+          "$ref": "#/texts/26"
+        }
+      ],
+      "content_layer": "body",
+      "name": "group",
+      "label": "inline"
     }
   ],
   "texts": [
@@ -542,17 +580,53 @@
     {
       "self_ref": "#/texts/24",
       "parent": {
-        "$ref": "#/body"
+        "$ref": "#/groups/8"
       },
       "children": [],
       "content_layer": "body",
       "label": "paragraph",
       "prov": [],
-      "orig": "And a code block:",
-      "text": "And a code block:"
+      "orig": "Here a formula:",
+      "text": "Here a formula:"
     },
     {
       "self_ref": "#/texts/25",
+      "parent": {
+        "$ref": "#/groups/8"
+      },
+      "children": [],
+      "content_layer": "body",
+      "label": "formula",
+      "prov": [],
+      "orig": "E=mc^2",
+      "text": "E=mc^2"
+    },
+    {
+      "self_ref": "#/texts/26",
+      "parent": {
+        "$ref": "#/groups/8"
+      },
+      "children": [],
+      "content_layer": "body",
+      "label": "paragraph",
+      "prov": [],
+      "orig": "(to be displayed inline)",
+      "text": "(to be displayed inline)"
+    },
+    {
+      "self_ref": "#/texts/27",
+      "parent": {
+        "$ref": "#/body"
+      },
+      "children": [],
+      "content_layer": "body",
+      "label": "paragraph",
+      "prov": [],
+      "orig": "Here a code block:",
+      "text": "Here a code block:"
+    },
+    {
+      "self_ref": "#/texts/28",
       "parent": {
         "$ref": "#/body"
       },
@@ -566,6 +640,42 @@
       "references": [],
       "footnotes": [],
       "code_language": "unknown"
+    },
+    {
+      "self_ref": "#/texts/29",
+      "parent": {
+        "$ref": "#/body"
+      },
+      "children": [],
+      "content_layer": "body",
+      "label": "paragraph",
+      "prov": [],
+      "orig": "Here a formula block:",
+      "text": "Here a formula block:"
+    },
+    {
+      "self_ref": "#/texts/30",
+      "parent": {
+        "$ref": "#/body"
+      },
+      "children": [],
+      "content_layer": "body",
+      "label": "formula",
+      "prov": [],
+      "orig": "E=mc^2",
+      "text": "E=mc^2"
+    },
+    {
+      "self_ref": "#/texts/31",
+      "parent": {
+        "$ref": "#/body"
+      },
+      "children": [],
+      "content_layer": "body",
+      "label": "paragraph",
+      "prov": [],
+      "orig": "The end.",
+      "text": "The end."
     }
   ],
   "pictures": [
@@ -840,7 +950,91 @@
       }
     }
   ],
-  "key_value_items": [],
-  "form_items": [],
+  "key_value_items": [
+    {
+      "self_ref": "#/key_value_items/0",
+      "parent": {
+        "$ref": "#/body"
+      },
+      "children": [],
+      "content_layer": "body",
+      "label": "key_value_region",
+      "prov": [],
+      "captions": [],
+      "references": [],
+      "footnotes": [],
+      "graph": {
+        "cells": [
+          {
+            "label": "key",
+            "cell_id": 0,
+            "text": "number",
+            "orig": "#"
+          },
+          {
+            "label": "value",
+            "cell_id": 1,
+            "text": "1",
+            "orig": "1"
+          }
+        ],
+        "links": [
+          {
+            "label": "to_value",
+            "source_cell_id": 0,
+            "target_cell_id": 1
+          },
+          {
+            "label": "to_key",
+            "source_cell_id": 1,
+            "target_cell_id": 0
+          }
+        ]
+      }
+    }
+  ],
+  "form_items": [
+    {
+      "self_ref": "#/form_items/0",
+      "parent": {
+        "$ref": "#/body"
+      },
+      "children": [],
+      "content_layer": "body",
+      "label": "form",
+      "prov": [],
+      "captions": [],
+      "references": [],
+      "footnotes": [],
+      "graph": {
+        "cells": [
+          {
+            "label": "key",
+            "cell_id": 0,
+            "text": "number",
+            "orig": "#"
+          },
+          {
+            "label": "value",
+            "cell_id": 1,
+            "text": "1",
+            "orig": "1"
+          }
+        ],
+        "links": [
+          {
+            "label": "to_value",
+            "source_cell_id": 0,
+            "target_cell_id": 1
+          },
+          {
+            "label": "to_key",
+            "source_cell_id": 1,
+            "target_cell_id": 0
+          }
+        ]
+      }
+    }
+  ],
   "pages": {}
 }
diff --git a/test/data/doc/constructed_doc.embedded.md.gt b/test/data/doc/constructed_doc.embedded.md.gt
@@ -41,9 +41,16 @@ This is the caption of figure 2.
 - item 2 of neighboring list
     - item 1 of sub list
     - Here a code snippet: `print("Hello world")` (to be displayed inline)
+    - Here a formula: $E=mc^2$ (to be displayed inline)
 
-And a code block:
+Here a code block:
 
 ```
 print("Hello world")
-```
+```
+
+Here a formula block:
+
+$$E=mc^2$$
+
+The end.