10
\$\begingroup\$

I saw this question recently, and I was wondering the title of my question. Essentially is stuff like this:

H̠̹͖

A single character in reference to questions that count characters? If so, I could theoretically create a language that does any challenge in a single "character" by programming it in the zalgo areas of the character.

\$\endgroup\$
  • 4
    \$\begingroup\$ I assume it's cases like this which motivated the community to score by bytes rather than by characters. \$\endgroup\$ – Alex A. Dec 21 '15 at 2:07
  • \$\begingroup\$ @AlexA. I think it's because UTF-8 can encode 4,294,967,296 characters, and golfing langauges would use all of them. \$\endgroup\$ – Dennis Dec 21 '15 at 2:24
  • 1
    \$\begingroup\$ @AlexA., I think it's more Sclipting, which was seen as deliberately abusing the previous tendency to score by characters. \$\endgroup\$ – Peter Taylor Dec 21 '15 at 8:40
  • 2
    \$\begingroup\$ The safest way to count characters is to use Twitter, because they implement the most advanced Unicode compliant character count algorithm. (btw, Twitter also counts the above as 4 chars). More info: dev.twitter.com/overview/api/counting-characters \$\endgroup\$ – mınxomaτ Jan 1 '16 at 19:39
24
\$\begingroup\$

Again, just as I answered on the linked question, according to a simple char/byte counting tool, that's 4 chars, 7 bytes.

Combining diacritical marks are counted as separate Unicode characters.

\$\endgroup\$
  • 4
    \$\begingroup\$ It's multiple characters, and a single grapheme cluster. We're not counting grapheme clusters. \$\endgroup\$ – Riking Dec 28 '15 at 3:37
9
\$\begingroup\$

No.

A simple search shows that combining diacritical marks are used. These are all their own character, so H̠̹͖ > 1 char. They use codepoints U+0300 to U+036F. If you have ̃., it has 2 Unicode codepoints -- U+002E and U+0303 -- that is definitely not 1 char.

\$\endgroup\$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .