mirror of
https://github.com/DS4SD/docling.git
synced 2025-12-08 20:58:11 +00:00
feat(xlsx): create a page for each worksheet in XLSX backend (#1332)
* sytle(xlsx): enforce type hints in XLSX backend Signed-off-by: Cesar Berrospi Ramis <75900930+ceberam@users.noreply.github.com> * feat(xlsx): create a page for each worksheet in XLSX backend Signed-off-by: Cesar Berrospi Ramis <75900930+ceberam@users.noreply.github.com> * docs(xlsx): add docstrings to XLSX backend module. Signed-off-by: Cesar Berrospi Ramis <75900930+ceberam@users.noreply.github.com> * docling(xlsx): add bounding boxes and page size information in cell units Signed-off-by: Cesar Berrospi Ramis <75900930+ceberam@users.noreply.github.com> --------- Signed-off-by: Cesar Berrospi Ramis <75900930+ceberam@users.noreply.github.com>
This commit is contained in:
Cesar Berrospi Ramis
committed by
GitHub
GitHub
parent
c605edd8e9
commit
eef2bdea77
@@ -97,7 +97,22 @@
|
||||
"children": [],
|
||||
"content_layer": "body",
|
||||
"label": "picture",
|
||||
"prov": [],
|
||||
"prov": [
|
||||
{
|
||||
"page_no": 3,
|
||||
"bbox": {
|
||||
"l": 8.0,
|
||||
"t": 18.0,
|
||||
"r": 13.0,
|
||||
"b": 36.0,
|
||||
"coord_origin": "TOPLEFT"
|
||||
},
|
||||
"charspan": [
|
||||
0,
|
||||
0
|
||||
]
|
||||
}
|
||||
],
|
||||
"captions": [],
|
||||
"references": [],
|
||||
"footnotes": [],
|
||||
@@ -122,7 +137,22 @@
|
||||
"children": [],
|
||||
"content_layer": "body",
|
||||
"label": "table",
|
||||
"prov": [],
|
||||
"prov": [
|
||||
{
|
||||
"page_no": 1,
|
||||
"bbox": {
|
||||
"l": 0.0,
|
||||
"t": 0.0,
|
||||
"r": 3.0,
|
||||
"b": 7.0,
|
||||
"coord_origin": "TOPLEFT"
|
||||
},
|
||||
"charspan": [
|
||||
0,
|
||||
0
|
||||
]
|
||||
}
|
||||
],
|
||||
"captions": [],
|
||||
"references": [],
|
||||
"footnotes": [],
|
||||
@@ -661,7 +691,22 @@
|
||||
"children": [],
|
||||
"content_layer": "body",
|
||||
"label": "table",
|
||||
"prov": [],
|
||||
"prov": [
|
||||
{
|
||||
"page_no": 2,
|
||||
"bbox": {
|
||||
"l": 0.0,
|
||||
"t": 0.0,
|
||||
"r": 4.0,
|
||||
"b": 9.0,
|
||||
"coord_origin": "TOPLEFT"
|
||||
},
|
||||
"charspan": [
|
||||
0,
|
||||
0
|
||||
]
|
||||
}
|
||||
],
|
||||
"captions": [],
|
||||
"references": [],
|
||||
"footnotes": [],
|
||||
@@ -1564,7 +1609,22 @@
|
||||
"children": [],
|
||||
"content_layer": "body",
|
||||
"label": "table",
|
||||
"prov": [],
|
||||
"prov": [
|
||||
{
|
||||
"page_no": 2,
|
||||
"bbox": {
|
||||
"l": 6.0,
|
||||
"t": 4.0,
|
||||
"r": 9.0,
|
||||
"b": 9.0,
|
||||
"coord_origin": "TOPLEFT"
|
||||
},
|
||||
"charspan": [
|
||||
0,
|
||||
0
|
||||
]
|
||||
}
|
||||
],
|
||||
"captions": [],
|
||||
"references": [],
|
||||
"footnotes": [],
|
||||
@@ -1955,7 +2015,22 @@
|
||||
"children": [],
|
||||
"content_layer": "body",
|
||||
"label": "table",
|
||||
"prov": [],
|
||||
"prov": [
|
||||
{
|
||||
"page_no": 2,
|
||||
"bbox": {
|
||||
"l": 2.0,
|
||||
"t": 13.0,
|
||||
"r": 5.0,
|
||||
"b": 18.0,
|
||||
"coord_origin": "TOPLEFT"
|
||||
},
|
||||
"charspan": [
|
||||
0,
|
||||
0
|
||||
]
|
||||
}
|
||||
],
|
||||
"captions": [],
|
||||
"references": [],
|
||||
"footnotes": [],
|
||||
@@ -2346,7 +2421,22 @@
|
||||
"children": [],
|
||||
"content_layer": "body",
|
||||
"label": "table",
|
||||
"prov": [],
|
||||
"prov": [
|
||||
{
|
||||
"page_no": 3,
|
||||
"bbox": {
|
||||
"l": 0.0,
|
||||
"t": 0.0,
|
||||
"r": 3.0,
|
||||
"b": 7.0,
|
||||
"coord_origin": "TOPLEFT"
|
||||
},
|
||||
"charspan": [
|
||||
0,
|
||||
0
|
||||
]
|
||||
}
|
||||
],
|
||||
"captions": [],
|
||||
"references": [],
|
||||
"footnotes": [],
|
||||
@@ -2813,7 +2903,22 @@
|
||||
"children": [],
|
||||
"content_layer": "body",
|
||||
"label": "table",
|
||||
"prov": [],
|
||||
"prov": [
|
||||
{
|
||||
"page_no": 3,
|
||||
"bbox": {
|
||||
"l": 4.0,
|
||||
"t": 6.0,
|
||||
"r": 7.0,
|
||||
"b": 13.0,
|
||||
"coord_origin": "TOPLEFT"
|
||||
},
|
||||
"charspan": [
|
||||
0,
|
||||
0
|
||||
]
|
||||
}
|
||||
],
|
||||
"captions": [],
|
||||
"references": [],
|
||||
"footnotes": [],
|
||||
@@ -3275,5 +3380,27 @@
|
||||
],
|
||||
"key_value_items": [],
|
||||
"form_items": [],
|
||||
"pages": {}
|
||||
"pages": {
|
||||
"1": {
|
||||
"size": {
|
||||
"width": 3.0,
|
||||
"height": 7.0
|
||||
},
|
||||
"page_no": 1
|
||||
},
|
||||
"2": {
|
||||
"size": {
|
||||
"width": 9.0,
|
||||
"height": 18.0
|
||||
},
|
||||
"page_no": 2
|
||||
},
|
||||
"3": {
|
||||
"size": {
|
||||
"width": 13.0,
|
||||
"height": 36.0
|
||||
},
|
||||
"page_no": 3
|
||||
}
|
||||
}
|
||||
}
|
||||
@@ -1,13 +1,18 @@
|
||||
import os
|
||||
import logging
|
||||
from pathlib import Path
|
||||
|
||||
import pytest
|
||||
|
||||
from docling.backend.msexcel_backend import MsExcelDocumentBackend
|
||||
from docling.datamodel.base_models import InputFormat
|
||||
from docling.datamodel.document import ConversionResult, DoclingDocument
|
||||
from docling.datamodel.document import ConversionResult, DoclingDocument, InputDocument
|
||||
from docling.document_converter import DocumentConverter
|
||||
|
||||
from .test_data_gen_flag import GEN_TEST_DATA
|
||||
from .verify_utils import verify_document, verify_export
|
||||
|
||||
_log = logging.getLogger(__name__)
|
||||
|
||||
GENERATE = GEN_TEST_DATA
|
||||
|
||||
|
||||
@@ -28,13 +33,15 @@ def get_converter():
|
||||
return converter
|
||||
|
||||
|
||||
def test_e2e_xlsx_conversions():
|
||||
@pytest.fixture(scope="module")
|
||||
def documents() -> list[tuple[Path, DoclingDocument]]:
|
||||
documents: list[dict[Path, DoclingDocument]] = []
|
||||
|
||||
xlsx_paths = get_xlsx_paths()
|
||||
converter = get_converter()
|
||||
|
||||
for xlsx_path in xlsx_paths:
|
||||
print(f"converting {xlsx_path}")
|
||||
_log.debug(f"converting {xlsx_path}")
|
||||
|
||||
gt_path = (
|
||||
xlsx_path.parent.parent / "groundtruth" / "docling_v2" / xlsx_path.name
|
||||
@@ -44,6 +51,14 @@ def test_e2e_xlsx_conversions():
|
||||
|
||||
doc: DoclingDocument = conv_result.document
|
||||
|
||||
assert doc, f"Failed to convert document from file {gt_path}"
|
||||
documents.append((gt_path, doc))
|
||||
|
||||
return documents
|
||||
|
||||
|
||||
def test_e2e_xlsx_conversions(documents) -> None:
|
||||
for gt_path, doc in documents:
|
||||
pred_md: str = doc.export_to_markdown()
|
||||
assert verify_export(pred_md, str(gt_path) + ".md"), "export to md"
|
||||
|
||||
@@ -57,3 +72,30 @@ def test_e2e_xlsx_conversions():
|
||||
assert verify_document(
|
||||
doc, str(gt_path) + ".json", GENERATE
|
||||
), "document document"
|
||||
|
||||
|
||||
def test_pages(documents) -> None:
|
||||
"""Test the page count and page size of converted documents.
|
||||
|
||||
Args:
|
||||
documents: The paths and converted documents.
|
||||
"""
|
||||
# number of pages from the backend method
|
||||
path = [item for item in get_xlsx_paths() if item.stem == "test-01"][0]
|
||||
in_doc = InputDocument(
|
||||
path_or_stream=path,
|
||||
format=InputFormat.XLSX,
|
||||
filename=path.stem,
|
||||
backend=MsExcelDocumentBackend,
|
||||
)
|
||||
backend = MsExcelDocumentBackend(in_doc=in_doc, path_or_stream=path)
|
||||
assert backend.page_count() == 3
|
||||
|
||||
# number of pages from the converted document
|
||||
doc = [item for path, item in documents if path.stem == "test-01"][0]
|
||||
assert len(doc.pages) == 3
|
||||
|
||||
# page sizes as number of cells
|
||||
assert doc.pages.get(1).size.as_tuple() == (3.0, 7.0)
|
||||
assert doc.pages.get(2).size.as_tuple() == (9.0, 18.0)
|
||||
assert doc.pages.get(3).size.as_tuple() == (13.0, 36.0)
|
||||
|
||||
Reference in New Issue
Block a user