Editing Punycode (section)

===Encoding the non-ASCII characters===

The non-ASCII characters are sorted by Unicode value, lowest first (if a character occurs more than once, they are sorted by position). Each is then encoded as a single number. This single number defines both the location to insert the character at and which character to insert.

* An {{mvar|index}} into the result to insert the code at, starting at 0 (for insertion at the start).{{cn|date=March 2024}}
* The number of {{mvar|insertionPoints}} (current length of the result plus one).
* The {{mvar|reducedCodepoint}} is the Unicode code point to insert minus 127.{{cn|date=March 2024}}

The encoded number is {{math|''insertionPoints'' × ''reducedCodepoint'' + ''index''}}. By dividing by {{mvar|insertionPoints}} and also getting the remainder, a decoder can determine {{mvar|reducedCodepoint}} and {{mvar|index}}.

There are 6 possible insertion points for a character in the string "bcher" (including before the first character and after the last one). {{char|ü}} is Unicode code point {{tt|0xFC}} or 252 (see [[Latin-1_Supplement_(Unicode_block)|Latin-1 Supplement]]), and the reduced code point is {{math|252 − 127}}, or {{math|124}}. The {{char|ü}} is inserted at position 1, after the {{char|b}}. Thus the encoder will add the number {{math|1=6 × 124 + 1 = 745}}, and the decoder can retrieve these by {{math|1=&lfloor;745 / 6&rfloor; = 124}} and {{math|1=745 mod 6 = 1}}.

These numbers are strictly increasing. For the second and subsequent inserted characters, the difference between the number and the previous one is written.