Update app.py

app.py

@@ -174,31 +174,44 @@ def on_select(instruction1, instruction2, instruction3, evt: gr.SelectData):
 import re
 
 def preprocess_text(text):
-    # Remove special characters and extra spaces
-    text = re.sub(r'[^\w\s]', '', text)
+    # Remove extra spaces and special characters except punctuation
     text = re.sub(r'\s+', ' ', text).strip()
     return text
 
+def count_sentences(text):
+    # Split sentences using punctuation followed by a space or end of string
+    sentences = re.split(r'(?<=[.!?]) +', text)
+    return len(sentences), sentences
+
+def count_syllables(word):
+    return len(re.findall(r'[aeiouyAEIOUY]', word))
+
 def flesch_kincaid_grade_level(text):
-    sentences = len(re.split(r'[.!?]', text))
+    sentences_count, sentences = count_sentences(text)
     words = len(text.split())
-    syllables = sum([len(re.findall(r'[aeiouyAEIOUY]', word)) for word in text.split()])
-    
-    return 0.39 * (words / sentences) + 11.8 * (syllables / words) - 15.59
+    syllables = sum([count_syllables(word) for word in text.split()])
+
+    if sentences_count == 0 or words == 0:
+        return float('nan')  # Return NaN to indicate an error
+    return 0.39 * (words / sentences_count) + 11.8 * (syllables / words) - 15.59
 
 def flesch_reading_ease(text):
-    sentences = len(re.split(r'[.!?]', text))
+    sentences_count, sentences = count_sentences(text)
     words = len(text.split())
-    syllables = sum([len(re.findall(r'[aeiouyAEIOUY]', word)) for word in text.split()])
-    
-    return 206.835 - 1.015 * (words / sentences) - 84.6 * (syllables / words)
+    syllables = sum([count_syllables(word) for word in text.split()])
+
+    if sentences_count == 0 or words == 0:
+        return float('nan')  # Return NaN to indicate an error
+    return 206.835 - 1.015 * (words / sentences_count) - 84.6 * (syllables / words)
 
 def gunning_fog_index(text):
-    sentences = len(re.split(r'[.!?]', text))
+    sentences_count, sentences = count_sentences(text)
     words = len(text.split())
-    complex_words = len([word for word in text.split() if len(re.findall(r'[aeiouyAEIOUY]', word)) >= 3])
-    
-    return 0.4 * ((words / sentences) + 100 * (complex_words / words))
+    complex_words = len([word for word in text.split() if count_syllables(word) >= 3])
+
+    if sentences_count == 0 or words == 0:
+        return float('nan')  # Return NaN to indicate an error
+    return 0.4 * ((words / sentences_count) + 100 * (complex_words / words))
 
 def calculate_readability_metrics(text):
     preprocessed_text = preprocess_text(text)