fix(markdown): ensure correct parsing of nested lists (#1995)

* fix(markdown): ensure correct parsing of nested lists

Signed-off-by: Cesar Berrospi Ramis <75900930+ceberam@users.noreply.github.com>

* chore: update dependencies in uv.lock file

Signed-off-by: Cesar Berrospi Ramis <75900930+ceberam@users.noreply.github.com>

---------

Signed-off-by: Cesar Berrospi Ramis <75900930+ceberam@users.noreply.github.com>

tests/data/groundtruth/docling_v2/mixed_without_h1.md.md

@@ -3,6 +3,6 @@
 - A. first
     - subitem
 - B. second
-    1. strange
+- 2 . strange
 
 The end!

tests/data/groundtruth/docling_v2/mixed_without_h1.md.yaml

@@ -0,0 +1,139 @@
+body:
+  children:
+  - $ref: '#/texts/0'
+  - $ref: '#/texts/1'
+  - $ref: '#/groups/0'
+  content_layer: body
+  label: unspecified
+  name: _root_
+  self_ref: '#/body'
+form_items: []
+furniture:
+  children: []
+  content_layer: furniture
+  label: unspecified
+  name: _root_
+  self_ref: '#/furniture'
+groups:
+- children:
+  - $ref: '#/texts/2'
+  content_layer: body
+  label: section
+  name: header-1
+  parent:
+    $ref: '#/body'
+  self_ref: '#/groups/0'
+- children:
+  - $ref: '#/texts/3'
+  - $ref: '#/texts/5'
+  - $ref: '#/texts/6'
+  content_layer: body
+  label: list
+  name: list
+  parent:
+    $ref: '#/texts/2'
+  self_ref: '#/groups/1'
+- children:
+  - $ref: '#/texts/4'
+  content_layer: body
+  label: list
+  name: list
+  parent:
+    $ref: '#/texts/3'
+  self_ref: '#/groups/2'
+key_value_items: []
+name: mixed_without_h1
+origin:
+  binary_hash: 7394721163373597328
+  filename: mixed_without_h1.md
+  mimetype: text/html
+pages: {}
+pictures: []
+schema_name: DoclingDocument
+tables: []
+texts:
+- children: []
+  content_layer: furniture
+  label: title
+  orig: mixed_without_h1
+  parent:
+    $ref: '#/body'
+  prov: []
+  self_ref: '#/texts/0'
+  text: mixed_without_h1
+- children: []
+  content_layer: furniture
+  label: text
+  orig: Content before first heading
+  parent:
+    $ref: '#/body'
+  prov: []
+  self_ref: '#/texts/1'
+  text: Content before first heading
+- children:
+  - $ref: '#/groups/1'
+  - $ref: '#/texts/7'
+  content_layer: body
+  label: section_header
+  level: 1
+  orig: Some heading
+  parent:
+    $ref: '#/groups/0'
+  prov: []
+  self_ref: '#/texts/2'
+  text: Some heading
+- children:
+  - $ref: '#/groups/2'
+  content_layer: body
+  enumerated: false
+  label: list_item
+  marker: ''
+  orig: A. first
+  parent:
+    $ref: '#/groups/1'
+  prov: []
+  self_ref: '#/texts/3'
+  text: A. first
+- children: []
+  content_layer: body
+  enumerated: false
+  label: list_item
+  marker: ''
+  orig: subitem
+  parent:
+    $ref: '#/groups/2'
+  prov: []
+  self_ref: '#/texts/4'
+  text: subitem
+- children: []
+  content_layer: body
+  enumerated: false
+  label: list_item
+  marker: ''
+  orig: B. second
+  parent:
+    $ref: '#/groups/1'
+  prov: []
+  self_ref: '#/texts/5'
+  text: B. second
+- children: []
+  content_layer: body
+  enumerated: false
+  label: list_item
+  marker: ''
+  orig: 2 . strange
+  parent:
+    $ref: '#/groups/1'
+  prov: []
+  self_ref: '#/texts/6'
+  text: 2 . strange
+- children: []
+  content_layer: body
+  label: text
+  orig: The end!
+  parent:
+    $ref: '#/texts/2'
+  prov: []
+  self_ref: '#/texts/7'
+  text: The end!
+version: 1.5.0

tests/data/md/mixed_without_h1.md

@@ -7,6 +7,6 @@ Content before first heading
 - A. first
     - subitem
 - B. second
-- 2. strange
+- 2\. strange
 
 The end!

tests/test_backend_markdown.py

@@ -16,7 +16,7 @@ def test_convert_valid():
     relevant_paths = sorted((root_path / "md").rglob("*.md"))
     assert len(relevant_paths) > 0
 
-    yaml_filter = ["inline_and_formatting"]
+    yaml_filter = ["inline_and_formatting", "mixed_without_h1"]
 
     for in_path in relevant_paths:
         md_gt_path = root_path / "groundtruth" / "docling_v2" / f"{in_path.name}.md"
@@ -41,12 +41,11 @@ def test_convert_valid():
                 f.write(f"{act_data}\n")
 
             if in_path.stem in yaml_filter:
-                with open(yaml_gt_path, mode="w", encoding="utf-8") as f:
-                    act_doc.save_as_yaml(
-                        yaml_gt_path,
-                        coord_precision=COORD_PREC,
-                        confid_precision=CONFID_PREC,
-                    )
+                act_doc.save_as_yaml(
+                    yaml_gt_path,
+                    coord_precision=COORD_PREC,
+                    confid_precision=CONFID_PREC,
+                )
         else:
             with open(md_gt_path, encoding="utf-8") as f:
                 exp_data = f.read().rstrip()
@@ -54,4 +53,4 @@ def test_convert_valid():
 
             if in_path.stem in yaml_filter:
                 exp_doc = DoclingDocument.load_from_yaml(yaml_gt_path)
-                assert act_doc == exp_doc
+                assert act_doc == exp_doc, f"export to yaml failed on {in_path}"