Spaces:
Starting
Starting
Create app.py
Browse files
app.py
ADDED
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import requests
|
2 |
+
import gradio as gr
|
3 |
+
|
4 |
+
def extract_text_from_url(url):
|
5 |
+
try:
|
6 |
+
# Request data from the URL
|
7 |
+
response = requests.get(url)
|
8 |
+
response.raise_for_status() # Raise an error for bad responses
|
9 |
+
data = response.json()
|
10 |
+
|
11 |
+
# Extract 'text' from each row
|
12 |
+
rows = data.get("rows", [])
|
13 |
+
texts = [row["row"]["text"] for row in rows if "text" in row["row"]]
|
14 |
+
|
15 |
+
# Return as a single string with newlines
|
16 |
+
return "\n".join(texts)
|
17 |
+
except Exception as e:
|
18 |
+
return f"An error occurred: {e}"
|
19 |
+
|
20 |
+
# Gradio interface
|
21 |
+
interface = gr.Interface(
|
22 |
+
fn=extract_text_from_url,
|
23 |
+
inputs=gr.Textbox(label="Dataset URL", placeholder="Enter the dataset URL"),
|
24 |
+
outputs=gr.Textbox(label="Extracted Texts", lines=20, placeholder="Extracted texts will appear here"),
|
25 |
+
title="Extract Text from Hugging Face Dataset",
|
26 |
+
description="Enter the URL of a Hugging Face dataset to extract and display the 'text' fields."
|
27 |
+
)
|
28 |
+
|
29 |
+
if __name__ == "__main__":
|
30 |
+
interface.launch()
|