It is interesting to be aware of the behavior when the treatment of strings with characters using different encodings.

<?php
# Works like expected. There is no accent
var_dump(strpos("Fabio", 'b'));
#int(2)

# The "á" letter is occupying two positions
var_dump(strpos("Fábio", 'b')) ;
#int(3)

# Now, encoding the string "Fábio" to utf8, we get some "unexpected" outputs. Every letter that is no in regular ASCII table, will use 4 positions(bytes). The starting point remains like before.
# We cant find the characted, because the haystack string is now encoded.
var_dump(strpos(utf8_encode("Fábio"), 'á'));
#bool(false)

# To get the expected result, we need to encode the needle too
var_dump(strpos(utf8_encode("Fábio"), utf8_encode('á')));
#int(1) 

# And, like said before, "á" occupies 4 positions(bytes)
var_dump(strpos(utf8_encode("Fábio"), 'b'));
#int(5)