Update app.py

app.py

@@ -6,22 +6,25 @@ import nltk
 import re
 from nltk.tokenize import sent_tokenize
 
-# Download punkt for sentence tokenization
+# Download punkt tokenizer
 nltk.download("punkt")
 
-# Use local LanguageTool server (runs via setup.sh)
-grammar_tool = language_tool_python.LanguageTool('en-US')
+# Connect to local LanguageTool server (started via setup.sh)
+grammar_tool = language_tool_python.LanguageTool('en-US', host='localhost', port=8081)
 
-# Load summarizer and hate-speech detection model
+# Load transformers pipelines
 summarizer = pipeline("summarization", model="sshleifer/distilbart-cnn-6-6")
 toxicity_classifier = pipeline("text-classification", model="Hate-speech-CNERG/bert-base-uncased-hatexplain")
 
 def extract_text(input_type, text_input, url_input):
     if input_type == "URL" and url_input:
-        article = Article(url_input)
-        article.download()
-        article.parse()
-        return article.text
+        try:
+            article = Article(url_input)
+            article.download()
+            article.parse()
+            return article.text
+        except Exception as e:
+            return f"Error fetching URL: {e}"
     return text_input
 
 def check_grammar(text):
@@ -37,8 +40,7 @@ def check_grammar(text):
             } for match in matches
         ]
     except Exception as e:
-        print(f"[Grammar Check Error] {e}")
-        return [{"text": "", "error": "Grammar check skipped (server error)", "suggestions": [], "offset": 0, "length": 0}]
+        return [{"text": "", "error": f"Grammar check failed: {str(e)}", "suggestions": [], "offset": 0, "length": 0}]
 
 def detect_sensitive_content(text):
     sentences = sent_tokenize(text)
@@ -87,6 +89,9 @@ def review_blog(input_type, text_input, url_input):
         return "Please provide text or a URL.", "", []
 
     text = extract_text(input_type, text_input, url_input)
+    if text.startswith("Error"):
+        return text, "", []
+
     grammar_issues = check_grammar(text)
     sensitive_issues = detect_sensitive_content(text)
     suggestions = generate_suggestions(text, grammar_issues, sensitive_issues)
@@ -119,7 +124,7 @@ def apply_changes(text, suggestions, approved_indices):
             continue
     return text
 
-# Gradio Interface
+# Gradio UI
 with gr.Blocks(theme=gr.themes.Soft()) as demo:
     gr.Markdown("## 🧠 AI Blog Reviewer with Grammar & Bias Detection")
     gr.Markdown("Enter blog content or a URL. Detect grammar issues and sensitive (toxic, biased) content.")