Spaces:
Sleeping
Sleeping
Create app.py
Browse files
app.py
ADDED
@@ -0,0 +1,64 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import fitz # PyMuPDF
|
2 |
+
import gradio as gr
|
3 |
+
from transformers import pipeline
|
4 |
+
import re
|
5 |
+
|
6 |
+
# μμ½μ μν λͺ¨λΈ λ‘λ
|
7 |
+
summarizer = pipeline("summarization", model="facebook/bart-large-cnn")
|
8 |
+
|
9 |
+
def extract_text_from_pdf(pdf_path):
|
10 |
+
doc = fitz.open(pdf_path)
|
11 |
+
text = ""
|
12 |
+
for page_num in range(doc.page_count):
|
13 |
+
page = doc.load_page(page_num)
|
14 |
+
text += page.get_text("text") + "\n"
|
15 |
+
return text
|
16 |
+
|
17 |
+
def find_section(text, section_title):
|
18 |
+
# μ κ· ννμμ μ¬μ©νμ¬ μΉμ
μ λͺ©μ μ°Ύμ΅λλ€.
|
19 |
+
pattern = re.compile(r'(?i)^.*{}.*$'.format(section_title), re.MULTILINE)
|
20 |
+
matches = list(pattern.finditer(text))
|
21 |
+
if not matches:
|
22 |
+
return None
|
23 |
+
start_idx = matches[0].start()
|
24 |
+
end_idx = text.find('\n\n', start_idx)
|
25 |
+
if end_idx == -1:
|
26 |
+
end_idx = len(text)
|
27 |
+
section_text = text[start_idx:end_idx].strip()
|
28 |
+
return section_text
|
29 |
+
|
30 |
+
def summarize_section(text, section_title, max_length=150):
|
31 |
+
try:
|
32 |
+
section_text = find_section(text, section_title)
|
33 |
+
if section_text:
|
34 |
+
summary = summarizer(section_text, max_length=max_length, min_length=30, do_sample=False)
|
35 |
+
return summary[0]['summary_text']
|
36 |
+
return f"Section '{section_title}' not found."
|
37 |
+
except Exception as e:
|
38 |
+
return f"Error processing section '{section_title}': {str(e)}"
|
39 |
+
|
40 |
+
def process_pdf(file):
|
41 |
+
try:
|
42 |
+
text = extract_text_from_pdf(file.name)
|
43 |
+
except Exception as e:
|
44 |
+
return [f"Error extracting text from PDF: {str(e)}"] * 3
|
45 |
+
|
46 |
+
abstract_summary = summarize_section(text, "abstract")
|
47 |
+
research_question_summary = summarize_section(text, "research question")
|
48 |
+
results_summary = summarize_section(text, "results")
|
49 |
+
|
50 |
+
return [abstract_summary, research_question_summary, results_summary]
|
51 |
+
|
52 |
+
# Gradio μΈν°νμ΄μ€ μ€μ
|
53 |
+
interface = gr.Interface(
|
54 |
+
fn=process_pdf,
|
55 |
+
inputs=gr.File(label="Upload PDF"),
|
56 |
+
outputs=[
|
57 |
+
gr.Textbox(label="Abstract Summary"),
|
58 |
+
gr.Textbox(label="Research Question Summary"),
|
59 |
+
gr.Textbox(label="Results Summary")
|
60 |
+
]
|
61 |
+
)
|
62 |
+
|
63 |
+
# μΈν°νμ΄μ€ μ€ν
|
64 |
+
interface.launch()
|