Efficiently replace all accented characters in a string?

Question

For a poor man\'s implementation of near-collation-correct sorting on the client side I need a JavaScript function that does efficient single character rep

Answer 1

Basing on existing answers and some suggestions, I've created this one:

String.prototype.removeAccents = function() {

    var removalMap = {
        'A'  : /[AⒶＡÀÁÂẦẤẪẨÃĀĂẰẮẴẲȦǠÄǞẢÅǺǍȀȂẠẬẶḀĄ]/g,
        'AA' : /[Ꜳ]/g,
        'AE' : /[ÆǼǢ]/g,
        'AO' : /[Ꜵ]/g,
        'AU' : /[Ꜷ]/g,
        'AV' : /[ꜸꜺ]/g,
        'AY' : /[Ꜽ]/g,
        'B'  : /[BⒷＢḂḄḆɃƂƁ]/g,
        'C'  : /[CⒸＣĆĈĊČÇḈƇȻꜾ]/g,
        'D'  : /[DⒹＤḊĎḌḐḒḎĐƋƊƉꝹ]/g,
        'DZ' : /[DZDŽ]/g,
        'Dz' : /[DzDž]/g,
        'E'  : /[EⒺＥÈÉÊỀẾỄỂẼĒḔḖĔĖËẺĚȄȆẸỆȨḜĘḘḚƐƎ]/g,
        'F'  : /[FⒻＦḞƑꝻ]/g,
        'G'  : /[GⒼＧǴĜḠĞĠǦĢǤƓꞠꝽꝾ]/g,
        'H'  : /[HⒽＨĤḢḦȞḤḨḪĦⱧⱵꞍ]/g,
        'I'  : /[IⒾＩÌÍÎĨĪĬİÏḮỈǏȈȊỊĮḬƗ]/g,
        'J'  : /[JⒿＪĴɈ]/g,
        'K'  : /[KⓀＫḰǨḲĶḴƘⱩꝀꝂꝄꞢ]/g,
        'L'  : /[LⓁＬĿĹĽḶḸĻḼḺŁȽⱢⱠꝈꝆꞀ]/g,
        'LJ' : /[LJ]/g,
        'Lj' : /[Lj]/g,
        'M'  : /[MⓂＭḾṀṂⱮƜ]/g,
        'N'  : /[NⓃＮǸŃÑṄŇṆŅṊṈȠƝꞐꞤ]/g,
        'NJ' : /[NJ]/g,
        'Nj' : /[Nj]/g,
        'O'  : /[OⓄＯÒÓÔỒỐỖỔÕṌȬṎŌṐṒŎȮȰÖȪỎŐǑȌȎƠỜỚỠỞỢỌỘǪǬØǾƆƟꝊꝌ]/g,
        'OI' : /[Ƣ]/g,
        'OO' : /[Ꝏ]/g,
        'OU' : /[Ȣ]/g,
        'P'  : /[PⓅＰṔṖƤⱣꝐꝒꝔ]/g,
        'Q'  : /[QⓆＱꝖꝘɊ]/g,
        'R'  : /[RⓇＲŔṘŘȐȒṚṜŖṞɌⱤꝚꞦꞂ]/g,
        'S'  : /[SⓈＳẞŚṤŜṠŠṦṢṨȘŞⱾꞨꞄ]/g,
        'T'  : /[TⓉＴṪŤṬȚŢṰṮŦƬƮȾꞆ]/g,
        'TZ' : /[Ꜩ]/g,
        'U'  : /[UⓊＵÙÚÛŨṸŪṺŬÜǛǗǕǙỦŮŰǓȔȖƯỪỨỮỬỰỤṲŲṶṴɄ]/g,
        'V'  : /[VⓋＶṼṾƲꝞɅ]/g,
        'VY' : /[Ꝡ]/g,
        'W'  : /[WⓌＷẀẂŴẆẄẈⱲ]/g,
        'X'  : /[XⓍＸẊẌ]/g,
        'Y'  : /[YⓎＹỲÝŶỸȲẎŸỶỴƳɎỾ]/g,
        'Z'  : /[ZⓏＺŹẐŻŽẒẔƵȤⱿⱫꝢ]/g,
        'a'  : /[aⓐａẚàáâầấẫẩãāăằắẵẳȧǡäǟảåǻǎȁȃạậặḁąⱥɐ]/g,
        'aa' : /[ꜳ]/g,
        'ae' : /[æǽǣ]/g,
        'ao' : /[ꜵ]/g,
        'au' : /[ꜷ]/g,
        'av' : /[ꜹꜻ]/g,
        'ay' : /[ꜽ]/g,
        'b'  : /[bⓑｂḃḅḇƀƃɓ]/g,
        'c'  : /[cⓒｃćĉċčçḉƈȼꜿↄ]/g,
        'd'  : /[dⓓｄḋďḍḑḓḏđƌɖɗꝺ]/g,
        'dz' : /[dzdž]/g,
        'e'  : /[eⓔｅèéêềếễểẽēḕḗĕėëẻěȅȇẹệȩḝęḙḛɇɛǝ]/g,
        'f'  : /[fⓕｆḟƒꝼ]/g,
        'g'  : /[gⓖｇǵĝḡğġǧģǥɠꞡᵹꝿ]/g,
        'h'  : /[hⓗｈĥḣḧȟḥḩḫẖħⱨⱶɥ]/g,
        'hv' : /[ƕ]/g,
        'i'  : /[iⓘｉìíîĩīĭïḯỉǐȉȋịįḭɨı]/g,
        'j'  : /[jⓙｊĵǰɉ]/g,
        'k'  : /[kⓚｋḱǩḳķḵƙⱪꝁꝃꝅꞣ]/g,
        'l'  : /[lⓛｌŀĺľḷḹļḽḻſłƚɫⱡꝉꞁꝇ]/g,
        'lj' : /[lj]/g,
        'm'  : /[mⓜｍḿṁṃɱɯ]/g,
        'n'  : /[nⓝｎǹńñṅňṇņṋṉƞɲʼnꞑꞥ]/g,
        'nj' : /[nj]/g,
        'o'  : /[oⓞｏòóôồốỗổõṍȭṏōṑṓŏȯȱöȫỏőǒȍȏơờớỡởợọộǫǭøǿɔꝋꝍɵ]/g,
        'oi' : /[ƣ]/g,
        'ou' : /[ȣ]/g,
        'oo' : /[ꝏ]/g,
        'p'  : /[pⓟｐṕṗƥᵽꝑꝓꝕ]/g,
        'q'  : /[qⓠｑɋꝗꝙ]/g,
        'r'  : /[rⓡｒŕṙřȑȓṛṝŗṟɍɽꝛꞧꞃ]/g,
        's'  : /[sⓢｓßśṥŝṡšṧṣṩșşȿꞩꞅẛ]/g,
        't'  : /[tⓣｔṫẗťṭțţṱṯŧƭʈⱦꞇ]/g,
        'tz' : /[ꜩ]/g,
        'u'  : /[uⓤｕùúûũṹūṻŭüǜǘǖǚủůűǔȕȗưừứữửựụṳųṷṵʉ]/g,
        'v'  : /[vⓥｖṽṿʋꝟʌ]/g,
        'vy' : /[ꝡ]/g,
        'w'  : /[wⓦｗẁẃŵẇẅẘẉⱳ]/g,
        'x'  : /[xⓧｘẋẍ]/g,
        'y'  : /[yⓨｙỳýŷỹȳẏÿỷẙỵƴɏỿ]/g,
        'z'  : /[zⓩｚźẑżžẓẕƶȥɀⱬꝣ]/g,
    };

    var str = this;

    for(var latin in removalMap) {
      var nonLatin = removalMap[latin];
      str = str.replace(nonLatin , latin);
    }

    return str;
}

It uses real chars instead of unicode list and works well.

You can use it like

"ąąą".removeAccents(); // returns "aaa"

You can easily convert this function to not be string prototype. However, as I'm fan of using string prototype in such cases, you'll have to do it yourself.