Fix encoding detection on PHP 8.1

[come-nc]

Use mb_check_encoding to detect encoding as mb_detect_encoding is misbehaving under PHP 8.1.
Also use mb_convert_encoding instead of utf8_encode as it’s getting deprecated in PHP 8.2.
Fixes #181

[staabm]

do we need additional test-coverage for this change?

[codecov]

Codecov Report

Merging #182 (42e1411) into master (315f592) will decrease coverage by 0.15%.
The diff coverage is 100.00%.

@@             Coverage Diff              @@
##             master     #182      +/-   ##
============================================
- Coverage     89.75%   89.60%   -0.16%     
  Complexity      262      262              
============================================
  Files            15       15              
  Lines           898      885      -13     
============================================
- Hits            806      793      -13     
  Misses           92       92              

Impacted Files	Coverage Δ
lib/functions.php	95.65% <100.00%> (-0.07%)	⬇️
lib/Client.php	84.50% <0.00%> (-0.67%)	⬇️
lib/Response.php	96.77% <0.00%> (-0.20%)	⬇️

📣 Codecov can now indicate which changes are the most critical in Pull Requests. Learn more

[come-nc]

do we need additional test-coverage for this change?

If the current coverage did not catch the bug, yes.

Here are some example strings that could be used in the test: https://3v4l.org/RrjlE
You can see that both 'Dušan' and 'Živko' (stolen from https://en.wikipedia.org/wiki/Slavic_names) are messed up by the function.

[staabm]

please investigate whether we need such a test and add it, if required.

[come-nc]

please investigate whether we need such a test and add it, if required.

Test added.

[staabm]

since this method is deprecated as of php 8.2, this is a step into a good direction

👍

[PVince81]

@DeepDiver1975 @phil-davis any objections to merging this ? thanks 😄

[come-nc]

@DeepDiver1975 There is the same problem in https://github.com/sabre-io/dav/blob/master/lib/DAV/StringUtil.php#L78 and https://github.com/sabre-io/vobject/blob/master/lib/StringUtil.php#L41

Should I open PR on these repos as well?

[DeepDiver1975]

Should I open PR on these repos as well?

yes please

[phil-davis]

Just a note for "information". There are some impossibilities for this kind of automated "educated guess" detection of encoding.
For example, hex C2A3 is the UTF-8 for the UK pound symbol.
But in ISO-8859-1 that is 2 code-points - C2 is  and A3 happens to be the UK pound symbol £

So if any software is presented with C2A3 as an "encoded string" and no other meta-data about how to interpret it, then there is no way to know if it is meant to represent £ or just £.

There will be plenty of other examples. Have a play at https://dencode.com/en/string/hex and try putting in hex for https://en.wikipedia.org/wiki/UTF-8 code points, and find ones that match sets of ISO-8859-1 code points that represent sequences of characters that could also be a valid, reasonable combination that might occur in a file name, for example.

CEA9 is UFT-8 Greek Omega Ω and ISO-8859-1 Ω

Any automated algorithm has to have a heuristic that "guesses" that Greek Omega is more common than the combination Ω