0-validate_utf8.py

0x04-utf8_validation/0-main.py

@@ -0,0 +1,15 @@
+#!/usr/bin/python3
+"""
+Main file for testing
+"""
+
+validUTF8 = __import__('0-validate_utf8').validUTF8
+
+data = [65]
+print(validUTF8(data))
+
+data = [80, 121, 116, 104, 111, 110, 32, 105, 115, 32, 99, 111, 111, 108, 33]
+print(validUTF8(data))
+
+data = [229, 65, 127, 256]
+print(validUTF8(data))

0x04-utf8_validation/0-validate_utf8.py

@@ -0,0 +1,33 @@
+#!/usr/bin/python3
+"""
+Defines a UTF-8 Validation function
+"""
+
+
+def validUTF8(data):
+    """
+    UTF-8 Validation
+    Args:
+        data (list[int]): an array of characters represented as 1byte int
+    Returns:
+        (True): if all characters in data are valid UTF-8 code point
+        (False): if one or more characters in data are invalid code point
+    """
+    tesla = 1 << 7
+    spaceX = 1 << 6
+    byteCount = 0
+    for codePoint in data:
+        elon = 1 << 7
+        if byteCount == 0:
+            while elon & codePoint:
+                byteCount += 1
+                elon = elon >> 1
+            if byteCount == 0:
+                continue
+            if byteCount == 1 or byteCount > 4:
+                return False
+        else:
+            if not (codePoint & tesla and not (codePoint & spaceX)):
+                return False
+        byteCount -= 1
+    return not byteCount