prize18-2  Exercise 3: Truncate to a Byte Limit Without Breaking a Character
=============================================================================

Idea: a continuation byte always looks like 10xxxxxx, so the byte at the cut position tells you whether you are mid-character.
If data[limit] is a continuation byte, the character containing it started earlier, so step back until data[end] is a lead byte or ASCII.
You never step back more than three bytes.

ex3.py:

  def safe_truncate(data, limit):
      if len(data) <= limit:
          return data
      end = limit
      while end > 0 and data[end] & 0xC0 == 0x80:
          end -= 1
      return data[:end]

  data = "caf\u00e9 \u20ac \U0001F600!".encode("utf-8")
  print(len(data))
  for limit in range(4, 9):
      cut = safe_truncate(data, limit)
      print(limit, len(cut), cut.decode("utf-8"))
  naive = data[:8]
  print(naive.decode("utf-8", errors="replace"))

Real output (Python 3.14):

  15
  4 3 caf
  5 5 café
  6 6 café
  7 6 café
  8 6 café
  café �

Limits 7 and 8 both fall inside the 3-byte euro sign (bytes 6-8), so the safe version keeps only 6 bytes. The naive slice data[:8] ends mid-character
and decodes to a replacement character. (Trailing space in the lines for limits 6-8 is the real space character.)