premmukund
commited on
Commit
•
4f49ea4
1
Parent(s):
282546b
Create app.py
Browse files
app.py
ADDED
@@ -0,0 +1,70 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import nltk
|
2 |
+
from nltk.corpus import wordnet
|
3 |
+
import streamlit as st
|
4 |
+
|
5 |
+
# Download required NLTK resources if you haven't already
|
6 |
+
nltk.download('punkt')
|
7 |
+
nltk.download('wordnet')
|
8 |
+
nltk.download('averaged_perceptron_tagger')
|
9 |
+
|
10 |
+
def rewrite_content(text):
|
11 |
+
# Tokenize the text
|
12 |
+
tokens = nltk.word_tokenize(text)
|
13 |
+
|
14 |
+
# Tag the tokens with their part-of-speech
|
15 |
+
tagged_tokens = nltk.pos_tag(tokens)
|
16 |
+
|
17 |
+
# Iterate over the tagged tokens and replace words with synonyms
|
18 |
+
rewritten_text = []
|
19 |
+
for token, tag in tagged_tokens:
|
20 |
+
# Fetch synonyms for the word
|
21 |
+
synonyms = wordnet.synsets(token)
|
22 |
+
if synonyms:
|
23 |
+
# Choose the first synonym
|
24 |
+
synonym = synonyms[0].lemmas()[0].name()
|
25 |
+
# Ensure the synonym is not the same as the original word
|
26 |
+
if synonym.lower() != token.lower():
|
27 |
+
rewritten_text.append(synonym.replace('_', ' '))
|
28 |
+
else:
|
29 |
+
rewritten_text.append(token)
|
30 |
+
else:
|
31 |
+
rewritten_text.append(token)
|
32 |
+
|
33 |
+
# Join the rewritten tokens back into a string
|
34 |
+
rewritten_content = ' '.join(rewritten_text)
|
35 |
+
|
36 |
+
# Maintain paragraph structure
|
37 |
+
paragraphs = text.split('\n')
|
38 |
+
rewritten_paragraphs = []
|
39 |
+
token_index = 0
|
40 |
+
|
41 |
+
for paragraph in paragraphs:
|
42 |
+
paragraph_tokens = nltk.word_tokenize(paragraph)
|
43 |
+
rewritten_paragraph = ' '.join(rewritten_text[token_index:token_index + len(paragraph_tokens)])
|
44 |
+
rewritten_paragraphs.append(rewritten_paragraph)
|
45 |
+
token_index += len(paragraph_tokens)
|
46 |
+
|
47 |
+
return '\n'.join(rewritten_paragraphs)
|
48 |
+
|
49 |
+
# Create a Streamlit interface
|
50 |
+
st.title("Text Rewriter")
|
51 |
+
st.write("Note: The rewrite content limit is 200000 words.")
|
52 |
+
|
53 |
+
input_text = st.text_area("Enter text to rewrite", height=300)
|
54 |
+
|
55 |
+
# Count words in the input text
|
56 |
+
word_count = len(nltk.word_tokenize(input_text))
|
57 |
+
|
58 |
+
# Display the current word count
|
59 |
+
st.write(f"Word count: {word_count} / 200000")
|
60 |
+
|
61 |
+
if st.button("Rewrite"):
|
62 |
+
if input_text:
|
63 |
+
if word_count <= 200000:
|
64 |
+
rewritten_text = rewrite_content(input_text)
|
65 |
+
st.subheader("Rewritten Text")
|
66 |
+
st.text_area("", rewritten_text, height=300)
|
67 |
+
else:
|
68 |
+
st.error("The text exceeds 200000 words. Please enter fewer words.")
|
69 |
+
else:
|
70 |
+
st.error("Please enter some text.")
|