Better fallbacks for UnicodeEncodeError

[Erotemic]

In 3.13, CPython added utf8 characters to a module that was profiled, which exposed an issue on the windows CI. Currently this is being worked around by simply not printing the offending text, but this is a hacky solution and we should probably just replace the utf character with a placeholder.

This workaround was introduced here: #287

On this line:

line_profiler/line_profiler/line_profiler.py

Line 439 in 4c6d2b0

except UnicodeEncodeError:

I think this should be easy, but I don't currently have the cycles to test it. Looking for some help on this.