Zeitpunkt Nutzer Delta Tröts TNR Titel Version maxTL Sa 18.05.2024 00:00:03 192.112 -5 8.720.153 45,4 Mastodon 4.3.0... 500 Fr 17.05.2024 00:00:03 192.117 -1 8.710.099 45,3 Mastodon 4.3.0... 500 Do 16.05.2024 00:00:02 192.118 -1 8.707.027 45,3 Mastodon 4.3.0... 500 Mi 15.05.2024 00:00:02 192.119 -2 8.697.532 45,3 Mastodon 4.3.0... 500 Di 14.05.2024 00:00:03 192.121 -4 8.689.581 45,2 Mastodon 4.3.0... 500 Mo 13.05.2024 00:00:06 192.125 -4 8.681.677 45,2 Mastodon 4.3.0... 500 So 12.05.2024 00:00:03 192.129 0 8.674.347 45,1 Mastodon 4.3.0... 500 Sa 11.05.2024 00:00:03 192.129 -2 8.665.940 45,1 Mastodon 4.3.0... 500 Fr 10.05.2024 00:00:03 192.131 -2 8.657.514 45,1 Mastodon 4.3.0... 500 Do 09.05.2024 00:00:03 192.133 0 8.648.316 45,0 Mastodon 4.3.0... 500
weilawei (@weilawei) · 04/2023 · Tröts: 1.200 · Folger: 852
Sa 18.05.2024 08:09
Alright, here's the fast explanation of why n > -65 works to tell UTF-8 starts apart from continuations.
1) Continuations are all 0b10xxxxxx = [-128..-65].
2) ASCII is 0b0xxxxxxx = [0..127].
3) Multi-byte starts are 0b11xxxxxx = [-64..-1].
4) Because the continuation range is all the way at one end of the representational range, you can do it in the one check.
5) It's in hardware. Conversion is just reinterpretation, which is free. In from a string, then it's 1 instruction to the output bool.
[Öffentlich] Antw.: 0 Wtrl.: 0 Fav.: 0 · via Web