prize18-1  Exercise 3: Make Garbled Text, Then Repair It
=========================================================

Idea: garbled text ("mojibake") appears when bytes written with one encoding are read with another. UTF-8 bytes read as Latin-1 give
two odd characters where one accented letter should be. Because Latin-1 maps every byte to a character, the damage can be reversed
by encoding back to Latin-1 (recovering the original bytes) and decoding as UTF-8.

ex3.py:

  original = "na\u00efve caf\u00e9"
  raw = original.encode("utf-8")
  print("bytes:", raw)
  garbled = raw.decode("latin-1")
  print("decoded as latin-1:", garbled)
  repaired = garbled.encode("latin-1").decode("utf-8")
  print("repaired:", repaired, repaired == original)
  try:
      raw.decode("ascii")
  except UnicodeDecodeError as e:
      print("ascii error:", e)

Real output (Python 3.14):

  bytes: b'na\xc3\xafve caf\xc3\xa9'
  decoded as latin-1: naïve café
  repaired: naïve café True
  ascii error: 'ascii' codec can't decode byte 0xc3 in position 2: ordinal not in range(128)

Caveat: this repair only works when no information was lost. If the garbled text was saved through a step that replaced or dropped
bytes, no code can fully recover it.