diff --git a/avro/main.py b/avro/main.py
index 9cc3319..9a8b4f9 100755
--- a/avro/main.py
+++ b/avro/main.py
@@ -5,7 +5,8 @@
 import re
 from concurrent.futures import ThreadPoolExecutor, as_completed
 from functools import lru_cache
-from typing import Any, Callable, Dict, List, Optional, Tuple, Union
+from itertools import chain
+from typing import Any, Callable, Dict, Generator, List, Optional, Tuple, Union
 
 # Import local modules.
 from . import config
@@ -50,59 +51,54 @@ def parse(*texts: str) -> Union[str, List[str]]:
     ```
     """
 
-    # Internal function for multiple parses.
+    # Compiled regular expression for UTF-8 validation
+    utf8_regex = re.compile(r'\A[\x00-\x7F]*\Z')
+
     @lru_cache
     def _parse_backend(text: str) -> str:
         fixed_text = validate.fix_string_case(
             text
-        )  # Sanitize text case to meet phonetic comparison standards.
-        output = []  # The output list of strings.
+        )  # Sanitize input text to meet phonetic comparison standards.
         cur_end = 0  # Cursor end point.
 
-        # Iterate through input text.
-        for cur, i in enumerate(fixed_text):
-            try:
-                i.encode('utf-8')
-            except UnicodeDecodeError:
-                uni_pass = False
-            else:
-                uni_pass = True
-
-            match = {'matched': False}
-
-            if not uni_pass:
-                cur_end = cur + 1
-                output.append(i)
-
-            elif cur >= cur_end and uni_pass:
-                match = match_patterns(fixed_text, cur, rule=False)
+        def output_generator() -> Generator[str, None, None]:
+            nonlocal cur_end
 
-                if match['matched']:
-                    output.append(match['replaced'])
-                    cur_end = cur + len(match['found'])
+            # Iterate through input text.
+            for cur, i in enumerate(fixed_text):
+                uni_pass = utf8_regex.match(i) is not None
 
-                else:
-                    match = match_patterns(fixed_text, cur, rule=True)
+                if not uni_pass:
+                    cur_end = cur + 1
+                    yield i
+                elif cur >= cur_end and uni_pass:
+                    match = match_patterns(fixed_text, cur, rule=False)
+                    matched = match['matched']
 
-                    if match['matched']:
+                    if matched:
+                        yield match['replaced']
                         cur_end = cur + len(match['found'])
-                        replaced = process_rules(
-                            rules=match['rules'], fixed_text=fixed_text, cur=cur, cur_end=cur_end
-                        )
+                    else:
+                        match = match_patterns(fixed_text, cur, rule=True)
+                        matched = match['matched']
 
-                        if replaced is not None:
-                            output.append(replaced)
+                        if matched:
+                            cur_end = cur + len(match['found'])
+                            replaced = process_rules(
+                                rules=match['rules'], fixed_text=fixed_text, cur=cur, cur_end=cur_end
+                            )
 
-                        else:
-                            output.append(match['replaced'])
+                            if replaced:
+                                yield replaced
+                            else:
+                                yield match['replaced']
 
-                if not match['matched']:
-                    cur_end = cur + 1
-                    output.append(i)
+                    if not matched:
+                        cur_end = cur + 1
+                        yield i
 
-        return ''.join(output)
+        return ''.join(chain.from_iterable(output_generator()))
 
-    # Do the final output.
     output = _concurrency_helper(_parse_backend, texts)
     return output[0] if len(output) == 1 else output
 
@@ -135,28 +131,16 @@ def _reverse_backend(text: str) -> str:
         for cur, i in enumerate(text):
             try:
                 i.encode('utf-8')
-            except UnicodeDecodeError:
-                uni_pass = False
-            else:
-                uni_pass = True
-
-            match = {'matched': False}
-
-            if not uni_pass:
-                output.append(i)
-
-            elif uni_pass:
                 match = match_patterns(text, cur, rule=False, reversed=True)
 
                 if match['matched']:
-                    if match['reversed']:
-                        output.append(match['reversed'])
-                    else:
-                        output.append(match['found'])
-
-                if not match['matched']:
+                    output.append(match['reversed'] if match['reversed'] else match['found'])
+                else:
                     output.append(i)
 
+            except UnicodeDecodeError:
+                output.append(i)
+
         return ''.join(output)
 
     # Split using regex to remove noise.
@@ -165,17 +149,12 @@ def _reverse_backend(text: str) -> str:
     # Extension for the _reverse_backend() function.
     @lru_cache
     def _reverse_backend_ext(text: str) -> str:
-        text_segments = []
         exceptions = config.EXCEPTIONS.get(text, None)
 
         if not exceptions:
             separated_texts = compiled_regex.split(text)
-
-            for separated_text in separated_texts:
-                text_segments.append(_reverse_backend(separated_text))
-
+            text_segments = [_reverse_backend(separated_text) for separated_text in separated_texts]
             return ''.join(text_segments)
-
         else:
             return exceptions