added latest version

tts_processor.py

@@ -108,10 +108,30 @@ def replace_abbreviations(string):
             words[i] = ''.join([alphabet_map.get(char, char) for char in word])
     return ' '.join(words)
 
-# Clean up whitespace in the text
 def clean_whitespace(string):
+    # Remove spaces before punctuation
     string = re.sub(r'\s+([.,?!])', r'\1', string)
-    return ' '.join(string.split())
+    # Collapse multiple spaces into one, but don’t touch inside tokens like "test.com"
+    string = re.sub(r'\s{2,}', ' ', string)
+    return string.strip()
+
+def make_dots_tts_friendly(text):
+    # Handle IP addresses (force "dot")
+    ipv4_pattern = r'\b\d{1,3}(\.\d{1,3}){3}\b'
+    text = re.sub(ipv4_pattern, lambda m: m.group(0).replace('.', ' dot '), text)
+
+    # Handle domain-like endings (force "dot")
+    domain_pattern = r'\b([\w-]+)\.(com|net|org|io|gov|edu|exe|dll|local)\b'
+    text = re.sub(domain_pattern, lambda m: m.group(0).replace('.', ' dot '), text)
+
+    # Handle decimals (use "point")
+    decimal_pattern = r'\b\d+\.\d+\b'
+    text = re.sub(decimal_pattern, lambda m: m.group(0).replace('.', ' point '), text)
+
+    # Handle leading dot words (.Net → dot Net)
+    text = re.sub(r'\.(?=\w)', 'dot ', text)
+
+    return text
 
 # Main preprocessing pipeline
 def preprocess_all(string):
@@ -119,6 +139,7 @@ def preprocess_all(string):
     string = replace_invalid_chars(string)
     string = replace_numbers(string)
     string = replace_abbreviations(string)
+    string = make_dots_tts_friendly(string)
     string = clean_whitespace(string)
     return string