fix end of string

This commit is contained in:
Maurizio Dipierro 2025-03-19 16:34:02 +01:00
parent f0c802eb3d
commit 5b9eafdf6a
1 changed files with 15 additions and 2 deletions

17
app.py
View File

@ -116,11 +116,24 @@ def model_inference(
buffer += html.escape(new_text) buffer += html.escape(new_text)
yield buffer yield buffer
# After finishing the streamer loop:
cleaned_output = full_output.replace("<end_of_utterance>", "").strip() cleaned_output = full_output.replace("<end_of_utterance>", "").strip()
final_output = cleaned_output
if cleaned_output: # If markdown conversion is needed, combine it with cleaned_output
if any(tag in cleaned_output for tag in ["<doctag>", "<otsl>", "<code>", "<chart>", "<formula>"]):
doctag_output = cleaned_output doctag_output = cleaned_output
yield cleaned_output if "<chart>" in doctag_output:
doctag_output = doctag_output.replace("<chart>", "<otsl>").replace("</chart>", "</otsl>")
doctag_output = re.sub(r'(<loc_500>)(?!.*<loc_500>)<[^>]+>', r'\1', doctag_output)
doc = DoclingDocument(name="Document")
doctags_doc = DocTagsDocument.from_doctags_and_image_pairs([doctag_output], images)
doc.load_from_doctags(doctags_doc)
final_output += "\n\n**MD Output:**\n\n" + doc.export_to_markdown()
# Yield the final combined output only once
yield final_output
if any(tag in doctag_output for tag in ["<doctag>", "<otsl>", "<code>", "<chart>", "<formula>"]): if any(tag in doctag_output for tag in ["<doctag>", "<otsl>", "<code>", "<chart>", "<formula>"]):
doc = DoclingDocument(name="Document") doc = DoclingDocument(name="Document")