Spaces:
Running
on
Zero
Running
on
Zero
update load_from_doctags usage
Browse files
app.py
CHANGED
@@ -124,13 +124,12 @@ def model_inference(
|
|
124 |
yield cleaned_output
|
125 |
|
126 |
if any(tag in doctag_output for tag in ["<doctag>", "<otsl>", "<code>", "<chart>", "<formula>"]):
|
127 |
-
doc = DoclingDocument(name="Document")
|
128 |
if "<chart>" in doctag_output:
|
129 |
doctag_output = doctag_output.replace("<chart>", "<otsl>").replace("</chart>", "</otsl>")
|
130 |
doctag_output = re.sub(r'(<loc_500>)(?!.*<loc_500>)<[^>]+>', r'\1', doctag_output)
|
131 |
|
132 |
doctags_doc = DocTagsDocument.from_doctags_and_image_pairs([doctag_output], images)
|
133 |
-
doc.load_from_doctags(doctags_doc)
|
134 |
yield f"**MD Output:**\n\n{doc.export_to_markdown()}"
|
135 |
|
136 |
examples=[[{"text": "Convert this page to docling.", "files": ["example_images/2d0fbcc50e88065a040a537b717620e964fb4453314b71d83f3ed3425addcef6.png"]}],
|
|
|
124 |
yield cleaned_output
|
125 |
|
126 |
if any(tag in doctag_output for tag in ["<doctag>", "<otsl>", "<code>", "<chart>", "<formula>"]):
|
|
|
127 |
if "<chart>" in doctag_output:
|
128 |
doctag_output = doctag_output.replace("<chart>", "<otsl>").replace("</chart>", "</otsl>")
|
129 |
doctag_output = re.sub(r'(<loc_500>)(?!.*<loc_500>)<[^>]+>', r'\1', doctag_output)
|
130 |
|
131 |
doctags_doc = DocTagsDocument.from_doctags_and_image_pairs([doctag_output], images)
|
132 |
+
doc = DoclingDocument.load_from_doctags(doctags_doc, document_name="Document")
|
133 |
yield f"**MD Output:**\n\n{doc.export_to_markdown()}"
|
134 |
|
135 |
examples=[[{"text": "Convert this page to docling.", "files": ["example_images/2d0fbcc50e88065a040a537b717620e964fb4453314b71d83f3ed3425addcef6.png"]}],
|