From ed20124544a1b10f068b11bbdf12e1bfc7567195 Mon Sep 17 00:00:00 2001
From: Cesar Berrospi Ramis <75900930+ceberam@users.noreply.github.com>
Date: Wed, 23 Apr 2025 09:30:59 +0200
Subject: [PATCH] fix(html): handle address, details, and summary tags (#1436)
* fix(html): handle 'address' tag
Signed-off-by: Cesar Berrospi Ramis <75900930+ceberam@users.noreply.github.com>
* fix(html): handle 'details' tag
Signed-off-by: Cesar Berrospi Ramis <75900930+ceberam@users.noreply.github.com>
---------
Signed-off-by: Cesar Berrospi Ramis <75900930+ceberam@users.noreply.github.com>
---
docling/backend/html_backend.py | 24 ++++++++-
.../docling_v2/example_06.html.itxt | 5 +-
.../docling_v2/example_06.html.json | 51 +++++++++++++++++--
.../groundtruth/docling_v2/example_06.html.md | 4 ++
tests/data/html/example_06.html | 4 ++
5 files changed, 82 insertions(+), 6 deletions(-)
diff --git a/docling/backend/html_backend.py b/docling/backend/html_backend.py
index aa2637f..88a315d 100644
--- a/docling/backend/html_backend.py
+++ b/docling/backend/html_backend.py
@@ -26,6 +26,8 @@ _log = logging.getLogger(__name__)
# tags that generate NodeItem elements
TAGS_FOR_NODE_ITEMS: Final = [
+ "address",
+ "details",
"h1",
"h2",
"h3",
@@ -38,6 +40,7 @@ TAGS_FOR_NODE_ITEMS: Final = [
"ul",
"ol",
"li",
+ "summary",
"table",
"figure",
"img",
@@ -163,7 +166,7 @@ class HTMLDocumentBackend(DeclarativeDocumentBackend):
def analyze_tag(self, tag: Tag, doc: DoclingDocument) -> None:
if tag.name in ["h1", "h2", "h3", "h4", "h5", "h6"]:
self.handle_header(tag, doc)
- elif tag.name in ["p"]:
+ elif tag.name in ["p", "address", "summary"]:
self.handle_paragraph(tag, doc)
elif tag.name in ["pre", "code"]:
self.handle_code(tag, doc)
@@ -177,6 +180,8 @@ class HTMLDocumentBackend(DeclarativeDocumentBackend):
self.handle_figure(tag, doc)
elif tag.name == "img":
self.handle_image(tag, doc)
+ elif tag.name == "details":
+ self.handle_details(tag, doc)
else:
self.walk(tag, doc)
@@ -201,6 +206,21 @@ class HTMLDocumentBackend(DeclarativeDocumentBackend):
return ["".join(result) + " "]
+ def handle_details(self, element: Tag, doc: DoclingDocument) -> None:
+ """Handle details tag (details) and its content."""
+
+ self.parents[self.level + 1] = doc.add_group(
+ name="details",
+ label=GroupLabel.SECTION,
+ parent=self.parents[self.level],
+ content_layer=self.content_layer,
+ )
+
+ self.level += 1
+ self.walk(element, doc)
+ self.parents[self.level + 1] = None
+ self.level -= 1
+
def handle_header(self, element: Tag, doc: DoclingDocument) -> None:
"""Handles header tags (h1, h2, etc.)."""
hlevel = int(element.name.replace("h", ""))
@@ -258,7 +278,7 @@ class HTMLDocumentBackend(DeclarativeDocumentBackend):
)
def handle_paragraph(self, element: Tag, doc: DoclingDocument) -> None:
- """Handles paragraph tags (p)."""
+ """Handles paragraph tags (p) or equivalent ones."""
if element.text is None:
return
text = element.text.strip()
diff --git a/tests/data/groundtruth/docling_v2/example_06.html.itxt b/tests/data/groundtruth/docling_v2/example_06.html.itxt
index 0cbdcf4..f9d5ef5 100644
--- a/tests/data/groundtruth/docling_v2/example_06.html.itxt
+++ b/tests/data/groundtruth/docling_v2/example_06.html.itxt
@@ -4,4 +4,7 @@ item-0 at level 0: unspecified: group _root_
item-3 at level 1: text: This is a regular paragraph.
item-4 at level 1: text: This is a third div
with a new line.
- item-5 at level 1: text: This is a fourth div with a bold paragraph.
\ No newline at end of file
+ item-5 at level 1: section: group details
+ item-6 at level 2: text: Heading for the details element
+ item-7 at level 2: text: Description of the details element.
+ item-8 at level 1: text: This is a fourth div with a bold paragraph.
\ No newline at end of file
diff --git a/tests/data/groundtruth/docling_v2/example_06.html.json b/tests/data/groundtruth/docling_v2/example_06.html.json
index 2f48245..f62a2cf 100644
--- a/tests/data/groundtruth/docling_v2/example_06.html.json
+++ b/tests/data/groundtruth/docling_v2/example_06.html.json
@@ -4,7 +4,7 @@
"name": "example_06",
"origin": {
"mimetype": "text/html",
- "binary_hash": 14574683870626799530,
+ "binary_hash": 10224930410364781672,
"filename": "example_06.html"
},
"furniture": {
@@ -30,14 +30,35 @@
"$ref": "#/texts/3"
},
{
- "$ref": "#/texts/4"
+ "$ref": "#/groups/0"
+ },
+ {
+ "$ref": "#/texts/6"
}
],
"content_layer": "body",
"name": "_root_",
"label": "unspecified"
},
- "groups": [],
+ "groups": [
+ {
+ "self_ref": "#/groups/0",
+ "parent": {
+ "$ref": "#/body"
+ },
+ "children": [
+ {
+ "$ref": "#/texts/4"
+ },
+ {
+ "$ref": "#/texts/5"
+ }
+ ],
+ "content_layer": "body",
+ "name": "details",
+ "label": "section"
+ }
+ ],
"texts": [
{
"self_ref": "#/texts/0",
@@ -89,6 +110,30 @@
},
{
"self_ref": "#/texts/4",
+ "parent": {
+ "$ref": "#/groups/0"
+ },
+ "children": [],
+ "content_layer": "body",
+ "label": "text",
+ "prov": [],
+ "orig": "Heading for the details element",
+ "text": "Heading for the details element"
+ },
+ {
+ "self_ref": "#/texts/5",
+ "parent": {
+ "$ref": "#/groups/0"
+ },
+ "children": [],
+ "content_layer": "body",
+ "label": "text",
+ "prov": [],
+ "orig": "Description of the details element.",
+ "text": "Description of the details element."
+ },
+ {
+ "self_ref": "#/texts/6",
"parent": {
"$ref": "#/body"
},
diff --git a/tests/data/groundtruth/docling_v2/example_06.html.md b/tests/data/groundtruth/docling_v2/example_06.html.md
index ed105e6..de672b8 100644
--- a/tests/data/groundtruth/docling_v2/example_06.html.md
+++ b/tests/data/groundtruth/docling_v2/example_06.html.md
@@ -7,4 +7,8 @@ This is a regular paragraph.
This is a third div
with a new line.
+Heading for the details element
+
+Description of the details element.
+
This is a fourth div with a bold paragraph.
\ No newline at end of file
diff --git a/tests/data/html/example_06.html b/tests/data/html/example_06.html
index efafd27..8c244a5 100644
--- a/tests/data/html/example_06.html
+++ b/tests/data/html/example_06.html
@@ -7,6 +7,10 @@
This is another div with text.
This is a regular paragraph.
This is a third div
with a new line.
+
+ Heading for the details element
+ Description of the details element.
+
This is a fourth div with a bold paragraph.