NeoAnd/Docling
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

fix: Handling of single-cell tables in DOCX backend (#314)

Browse Source 
* Handling of single-cell tables in DOCX backend

Signed-off-by: Maksym Lysak <mly@zurich.ibm.com>

* returned try-catch on tables handling

Signed-off-by: Maksym Lysak <mly@zurich.ibm.com>

* cleaned

Signed-off-by: Maksym Lysak <mly@zurich.ibm.com>

* proceed processing the content of single cell table as if its just part of the body

Signed-off-by: Maksym Lysak <mly@zurich.ibm.com>

* Added example of trickly 1 cell table docx

Signed-off-by: Maksym Lysak <mly@zurich.ibm.com>

---------

Signed-off-by: Maksym Lysak <mly@zurich.ibm.com>
Co-authored-by: Maksym Lysak <mly@zurich.ibm.com>
This commit is contained in:
Maxim Lysak 2024-11-12 15:20:55 +01:00 committed by GitHub
parent 7f5d35ea3c
commit fb8ba861e2
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
2 changed files with 8 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

16
docling/backend/msword_backend.py
View File

@ -130,7 +130,6 @@ class MsWordDocumentBackend(DeclarativeDocumentBackend):
def walk_linear(self, body, docx_obj, doc) -> DoclingDocument: def walk_linear(self, body, docx_obj, doc) -> DoclingDocument:
for element in body: for element in body:
tag_name = etree.QName(element).localname tag_name = etree.QName(element).localname
# Check for Inline Images (drawings or blip elements) # Check for Inline Images (drawings or blip elements)
found_drawing = etree.ElementBase.xpath( found_drawing = etree.ElementBase.xpath(
element, ".//w:drawing", namespaces=self.xml_namespaces element, ".//w:drawing", namespaces=self.xml_namespaces
@ -201,7 +200,6 @@ class MsWordDocumentBackend(DeclarativeDocumentBackend):
label_str = "" label_str = ""
label_level = 0 label_level = 0
if parts[0] == "Heading": if parts[0] == "Heading":
# print("{} - {}".format(parts[0], parts[1]))
label_str = parts[0] label_str = parts[0]
label_level = self.str_to_int(parts[1], default=None) label_level = self.str_to_int(parts[1], default=None)
if parts[1] == "Heading": if parts[1] == "Heading":
@ -217,19 +215,16 @@ class MsWordDocumentBackend(DeclarativeDocumentBackend):
if paragraph.text is None: if paragraph.text is None:
# _log.warn(f"paragraph has text==None") # _log.warn(f"paragraph has text==None")
return return
text = paragraph.text.strip() text = paragraph.text.strip()
# if len(text)==0 # keep empty paragraphs, they seperate adjacent lists! # if len(text)==0 # keep empty paragraphs, they seperate adjacent lists!
# Common styles for bullet and numbered lists. # Common styles for bullet and numbered lists.
# "List Bullet", "List Number", "List Paragraph" # "List Bullet", "List Number", "List Paragraph"
# TODO: reliably identify wether list is a numbered list or not # Identify wether list is a numbered list or not
# is_numbered = "List Bullet" not in paragraph.style.name # is_numbered = "List Bullet" not in paragraph.style.name
is_numbered = False is_numbered = False
p_style_name, p_level = self.get_label_and_level(paragraph) p_style_name, p_level = self.get_label_and_level(paragraph)
numid, ilevel = self.get_numId_and_ilvl(paragraph) numid, ilevel = self.get_numId_and_ilvl(paragraph)
# print("numid: {}, ilevel: {}, text: {}".format(numid, ilevel, text))
if numid == 0: if numid == 0:
numid = None numid = None
@ -450,8 +445,13 @@ class MsWordDocumentBackend(DeclarativeDocumentBackend):
for row in table.rows: for row in table.rows:
# Calculate the max number of columns # Calculate the max number of columns
num_cols = max(num_cols, sum(get_colspan(cell) for cell in row.cells)) num_cols = max(num_cols, sum(get_colspan(cell) for cell in row.cells))
# if row.cells:
# num_cols = max(num_cols, len(row.cells)) if num_rows == 1 and num_cols == 1:
cell_element = table.rows[0].cells[0]
# In case we have a table of only 1 cell, we consider it furniture
# And proceed processing the content of the cell as though it's in the document body
self.walk_linear(cell_element._element, docx_obj, doc)
return
# Initialize the table grid # Initialize the table grid
table_grid = [[None for _ in range(num_cols)] for _ in range(num_rows)] table_grid = [[None for _ in range(num_cols)] for _ in range(num_rows)]

BIN
tests/data/docx/tablecell.docx Normal file
View File

Binary file not shown.