A couple comments on utf-8 string length.

core/libstr.py

@@ -67,7 +67,16 @@ def _CheckContinuationByte(byte):


def NumOfUtf8Chars(bytes):
  """Returns the number of utf-8 characters in the byte string 's'."""
  """Returns the number of utf-8 characters in the byte string 's'.

  TODO: Raise exception rather than returning a string, so we can set the exit
  code of the command to 1 ?

  $ echo ${#bad}
  Invalid utf-8 at index 3 of string 'bad': 'ab\xffd'
  $ echo $?
  1
  """
  num_of_utf8_chars = 0

  num_bytes = len(bytes)

core/word_eval.py

@@ -298,6 +298,12 @@ def _ApplyPrefixOp(self, val, op_id):

    if op_id == Id.VSub_Pound:  # LENGTH
      if val.tag == value_e.Str:
        # NOTE: Whether bash counts bytes or chars is affected by LANG
        # environment variables.
        # Should we respect that, or another way to select?  set -o
        # count-bytes?

        # https://stackoverflow.com/questions/17368067/length-of-string-in-bash
        length = libstr.NumOfUtf8Chars(val.s)
      elif val.tag == value_e.StrArray:
        # There can be empty placeholder values in the array.