S15-unicode, change .chars to .codes where this was actually intended

`.chars` was used be all the examples output's specified the number of
codes. Use `.codes` in the example because of this

S15-unicode.pod

@@ -57,7 +57,7 @@ Perl 6 by default operates on graphemes, so counting by graphemes involves:
 To count by codepoints, conversion of a string to one of NFC, NFD, NFKC, or NFKD
 is needed:
 
-    "string".NFC.chars
+    "string".NFC.codes
     "string".NFKD.codes
 
 To count by code units, you can convert to the appropriate buffer type.
@@ -126,11 +126,11 @@ software.
 All four of NFC, NFD, NFKC, and NFKD can be considered valid "codepoint views",
 though each differ in their exact formulation of the contents of a string:
 
-    say "ẛ̣".chars;               # OUTPUT: 1 (NFG, ẛ̣)
-    say "ẛ̣".NFC.chars;           # OUTPUT: 2 (NFC, ẛ + ̣)
-    say "ẛ̣".NFD.chars;           # OUTPUT: 3 (NFD, ſ + ̣+ ̇)
-    say "ẛ̣".NFKC.chars;          # OUTPUT: 1 (NFKC, ṩ)
-    say "ẛ̣".NFKD.chars;          # OUTPUT: 3 (NFKD, s + ̣+ ̇)
+    say "ẛ̣".codes;               # OUTPUT: 2 (NFG, ẛ̣)
+    say "ẛ̣".NFC.codes;           # OUTPUT: 2 (NFC, ẛ + ̣)
+    say "ẛ̣".NFD.codes;           # OUTPUT: 3 (NFD, ſ + ̣+ ̇)
+    say "ẛ̣".NFKC.codes;          # OUTPUT: 1 (NFKC, ṩ)
+    say "ẛ̣".NFKD.codes;          # OUTPUT: 3 (NFKD, s + ̣+ ̇)
 
 Those who wish to operate with strings on the codepoint level may wish to use
 NFC, as it is the least different from NFG, as well as Perl 6's default form for