Add dotdump_utf8 helper for UTF-8-aware file viewer rendering

[C-Oliver]

Summary

Adds dotdump_utf8() to assemblyline.common.str_utils — a UTF-8-aware
counterpart to the existing dotdump() helper. Valid UTF-8 multi-byte
sequences (accented Latin, CJK, emoji, etc.) are preserved as their decoded
characters; only invalid or non-printable bytes are replaced with ..

This is the supporting library change for an upcoming
assemblyline-ui PR that switches the File Viewer ASCII tab to render
UTF-8 instead of dot-replacing every non-ASCII byte.

Why

The current File Viewer applies bytes.translate(FILTER_ASCII), which maps
every byte outside printable ASCII (32–126, plus tab/LF/CR) to .. Because
every byte of a UTF-8 multi-byte sequence is in the 0x80–0xFF range,
files containing café, 日本語, 🎉, etc. are rendered as solid dots.

Implementation

• Reuses the existing _valid_utf8 regex (RFC 3629, already used by
  escape_str_strict) so the definition of "valid UTF-8" stays consistent
  across the codebase.
• _valid_utf8.split(s) returns alternating non-matching / captured
  segments. Captured (valid UTF-8) segments are decoded to str;
  non-matching bytes are replaced one-for-one with ., preserving the
  byte length of invalid sections.

Behaviour

Input	Output
b'hello\nworld'	'hello\nworld'
'café'.encode()	'café'
'日本語'.encode()	'日本語'
'🎉ok'.encode()	'🎉ok'
b'A\x00B\x07C'	'A.B.C'
b'\xc3(' (bad 2-byte start)	'.('
b'\xed\xa0\x80' (UTF-16 surrogate)	'...'
b'\t\r\n abc'	'\t\r\n abc'

Risk

• New, additive helper. No existing call sites are changed in this PR.
• No new dependencies; reuses the existing compiled _valid_utf8 regex.