Created
October 11, 2020 22:03
-
-
Save Egor-Skriptunoff/8f5caf39b4611a297dc82bc43c70a158 to your computer and use it in GitHub Desktop.
Function wcwidth implemented in pure Lua
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
local wcwidth -- function wcwidth(codepoint) returns 0, 1 or 2 for codepoints from the range 0..0x10FFFF | |
do -- implementation from the C library "utf8proc" was rewritten in Lua, compressed and minified | |
local d = [[r6rur7!&B!!(!"my!"arCssststssrNx|}ssrDr+r&sr{zsxttsvr7sssr4r1rq}rP{{sr.vs{suswrAu!!/rGrLusr(sy|tr | |
3urNssytttu{s|tr0sturNsswvttuusr4tus}urNsszsusur*tr,xsurNssytttuyu|tr4srQwuusv{sr>wrOysusvyt}tr3urNssysusvyt}tr2vrMt | |
sysusv{s|tr3ur\svxssszr(trSsty~zrxst{}xr`tr1sssssvtrGr*stw}sr:{srzr*r-vvusutyuvr#~ss|vrxrv!%=u!)zur3ur3tr4trVr6{srCv | |
!!.tr8s!!.~v~!"5wrO|sr3tsrFr'rUwrEr'r<{~ur4r#rNr$rFr*!!Pusr+vsxstu!!~rPsw!%Kwr.yrGws|rvr7!%itr#t!!vvusts!%Itr+trH~rA | |
sr)sr#sztr'twtzswsr+sytssvstsysvtr2sr9sssvussrSur.sr$s!)#trIsvs!$7u!!Esrvr6!!8r0sro~!"0r0~vr@vr'trlttr{wrAsrtsrj~rEs | |
r>z!jZrV#C1u!,;urM!$Fvs|r6trft!"jsusvsr-wvsritrHr(r0r(r#sr<zr/r#~r3uvrEr$r:srYr$~sztrCurHssuttwtssr?wwt!"Fzstr("8frr | |
!5`!c|!#jt!!"rZs!'er&|xr&r9sr)sv!!Kssrv!!7yr(u!%As!"<s!!Mw!1^ustwvr>uvs!!]t!%}v!$!t!!Q}!!gurKr%rNvrC}tsr%srHur:r$r&t | |
rBs~urFr$zvstrr~xs!!X~r+vrMtsytttu{s|ttyuw!!xr(r-srgr*!"Eyt{r1trhr'!!"r#r{r%!"Zr%!"Oxsttvssst!!Eytyusr2|r>ysvzs{}rDr | |
&!$3zszrhr,sr$!!2xusstsyssrXwstsw!#Wv!Ya{"VuwrQy!+$ssrMyvrcvs}tr$!a0z!-&r@{"!+!"yrGur'vz!$*!:+tsv!XGwur,tyr4v!!Lu!4y | |
rMvrHzsr$sr,wsr%!.Bysr'tystsw!"_y!$Sv!/ty!!%y!23s!"$s!!vst|r{ur#rBv{ytr$x!!Rr7~{sr\sr,~rAvw~r'usur]sss!!strUr#vsr.r( | |
sr0tr#srlrkrFr\xsuutur)ty{!"=~!"zrEs|srHsrhsrI!!(wuuwy{r/yy{ur#y!-y%n@r8!L#}!"8t!]2r$!oE!@u!%^!/v!TIr"Lsr4rv!!8!"J]] | |
local i, a = 1, {2, 72, 68, 6, 9, 0, 3, 1, 19, 30, 2, 32, 2, 2, 4, 3, 1, 1, 5, 2, 4, 164, 1, 1, 16, 44, 106, 6} | |
local w, s, e, h, z, c, W, b = 0, 1, 239, d:gmatch"%S", 1, 0, {0}, d.byte; local function g() if e == 0 then | |
e, i, s = a[i], i+1, 4-s; if not e then return elseif w ~= 1 then w = 2-w; return w end end; e, w = e-1, s-w | |
return w end; repeat local pr, w = c, g(); if w then c = b(h())-114; if c < 1 then c = c*94+b(h())-22 | |
if c < 13 then c = c*94+b(h())+714754 end end; c, z = pr+c, z+2; W[z], W[z-1] = w, c end until not w | |
function wcwidth(codepoint) local m, l, r = 1, 1, z; repeat local s = (r-l)/4; if s >= m then s, m = m, m*4 else | |
s = s-s%1 end; local k = l+s*2; if codepoint < W[k+1] then r = k else l = k+2 end until l == r; return W[l] end | |
end | |
local function utf8_string_width(str) | |
local width = 0 | |
for u in str:gmatch".[\128-\191]*" do | |
local c = u:byte()%2^(8-#u) | |
for j = 2, #u do c = (c-2)*64+u:byte(j) end | |
width = width + wcwidth(c) | |
end | |
return width | |
end | |
-- Examples: | |
print(utf8_string_width("Hello")) --> 5 | |
print(utf8_string_width("Привет")) --> 6 | |
print(utf8_string_width("你好")) --> 4 |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment