User:Dinoguy1000/scripts/fullwidth2ascii.js

Code that you insert on this page could contain malicious content capable of compromising your account. If you import a script from another page with "importScript", "mw.loader.load", "iusc", or "lusc", take note that this causes you to dynamically load a remote script, which could be changed by others. Editors are responsible for all edits and actions they perform, including by scripts. User scripts are not centrally supported and may malfunction or become inoperable due to software changes. A guide to help you find broken scripts is available. If you are unsure whether code you are adding to this page is safe, you can ask at the appropriate village pump.
This code will be executed when previewing this page.
Documentation for this user script can be added at User:Dinoguy1000/scripts/fullwidth2ascii.
Note: After saving, you have to bypass your browser's cache to see the changes. Google Chrome, Firefox, Microsoft Edge and Safari: Hold down the ⇧ Shift key and click the Reload toolbar button. For details and instructions about other browsers, see Wikipedia:Bypass your cache.
/***********************************************************************
 *                                                                     *
 * [[en:User:Dinoguy1000/scripts/fullwidth2ascii.js]]                  *
 *                                                                     *
 * Replaces all instances of ASCII fullwidth characters with           *
 * their original ASCII counterparts, according to the list            *
 * at [[en:User:Dinoguy1000/Cheatsheet#Fullwidth character table]].    *
 *                                                                     *
 * Framework shamelessly ripped from [[en:User:Drilnoth/codefixer.js]] *
 *                                                                     *
 * Spinoff stuff (try to keep everything in sync):                     *
 *  - [[Wikipedia:AutoEd/fullwidth.js]]                                *
 *  - [[User:TheFarix/AWB/scripts/FullwidthReplacer]]                  *
 *                                                                     *
 ***********************************************************************/
 
function fullwidthReplacer() { //MAIN FUNCTION describes list of fixes
    var txt = document.editform.wpTextbox1;
 
    //replace
    //Common punctuation
    txt.value = txt.value.replace(/　/g, ' '); //fullwidth space
    //txt.value = txt.value.replace(/・/g, '·'); //fullwidth interpunct
    txt.value = txt.value.replace(/。/g, '.'); //replace ideographic period with a regular period (from TheFarix's AWB cleanup module)
    //txt.value = txt.value.replace(/・・・/g, '...'); //three interpuncts normally mean an ellipsis
    //txt.value = txt.value.replace(/···/g, '...'); //catch already converted triple interpuncts - it's after the interpunct rule so we only need one go-through to catch partial replacements
    txt.value = txt.value.replace(/…/g, '...'); //it's a little ballsy, I guess, but we're already doing lots of replacements, so... (see also [[WP:ELLIPSES]])
    txt.value = txt.value.replace(/！/g, '!');
    txt.value = txt.value.replace(/（/g, '(');
    txt.value = txt.value.replace(/）/g, ')');
    txt.value = txt.value.replace(/？/g, '?');
 
    //Numbers
    txt.value = txt.value.replace(/０/g, '0');
    txt.value = txt.value.replace(/１/g, '1');
    txt.value = txt.value.replace(/２/g, '2');
    txt.value = txt.value.replace(/３/g, '3');
    txt.value = txt.value.replace(/４/g, '4');
    txt.value = txt.value.replace(/５/g, '5');
    txt.value = txt.value.replace(/６/g, '6');
    txt.value = txt.value.replace(/７/g, '7');
    txt.value = txt.value.replace(/８/g, '8');
    txt.value = txt.value.replace(/９/g, '9');
 
    //Latin letters, uppercase
    txt.value = txt.value.replace(/Ａ/g, 'Α');
    txt.value = txt.value.replace(/Ｂ/g, 'B');
    txt.value = txt.value.replace(/Ｃ/g, 'C');
    txt.value = txt.value.replace(/Ｄ/g, 'D');
    txt.value = txt.value.replace(/Ｅ/g, 'E');
    txt.value = txt.value.replace(/Ｆ/g, 'F');
    txt.value = txt.value.replace(/Ｇ/g, 'G');
    txt.value = txt.value.replace(/Ｈ/g, 'H');
    txt.value = txt.value.replace(/Ｉ/g, 'I');
    txt.value = txt.value.replace(/Ｊ/g, 'J');
    txt.value = txt.value.replace(/Ｋ/g, 'K');
    txt.value = txt.value.replace(/Ｌ/g, 'L');
    txt.value = txt.value.replace(/Ｍ/g, 'M');
    txt.value = txt.value.replace(/Ｎ/g, 'N');
    txt.value = txt.value.replace(/Ｏ/g, 'O');
    txt.value = txt.value.replace(/Ｐ/g, 'P');
    txt.value = txt.value.replace(/Ｑ/g, 'Q');
    txt.value = txt.value.replace(/Ｒ/g, 'R');
    txt.value = txt.value.replace(/Ｓ/g, 'S');
    txt.value = txt.value.replace(/Ｔ/g, 'T');
    txt.value = txt.value.replace(/Ｕ/g, 'U');
    txt.value = txt.value.replace(/Ｖ/g, 'V');
    txt.value = txt.value.replace(/Ｗ/g, 'W');
    txt.value = txt.value.replace(/Ｘ/g, 'X');
    txt.value = txt.value.replace(/Ｙ/g, 'Y');
    txt.value = txt.value.replace(/Ｚ/g, 'Z');
 
    //Latin letters, lowercase
    txt.value = txt.value.replace(/ａ/g, 'a');
    txt.value = txt.value.replace(/ｂ/g, 'b');
    txt.value = txt.value.replace(/ｃ/g, 'c');
    txt.value = txt.value.replace(/ｄ/g, 'd');
    txt.value = txt.value.replace(/ｅ/g, 'e');
    txt.value = txt.value.replace(/ｆ/g, 'f');
    txt.value = txt.value.replace(/ｇ/g, 'g');
    txt.value = txt.value.replace(/ｈ/g, 'h');
    txt.value = txt.value.replace(/ｉ/g, 'i');
    txt.value = txt.value.replace(/ｊ/g, 'j');
    txt.value = txt.value.replace(/ｋ/g, 'k');
    txt.value = txt.value.replace(/ｌ/g, 'l');
    txt.value = txt.value.replace(/ｍ/g, 'm');
    txt.value = txt.value.replace(/ｎ/g, 'n');
    txt.value = txt.value.replace(/ｏ/g, 'o');
    txt.value = txt.value.replace(/ｐ/g, 'p');
    txt.value = txt.value.replace(/ｑ/g, 'q');
    txt.value = txt.value.replace(/ｒ/g, 'r');
    txt.value = txt.value.replace(/ｓ/g, 's');
    txt.value = txt.value.replace(/ｔ/g, 't');
    txt.value = txt.value.replace(/ｕ/g, 'u');
    txt.value = txt.value.replace(/ｖ/g, 'v');
    txt.value = txt.value.replace(/ｗ/g, 'w');
    txt.value = txt.value.replace(/ｘ/g, 'x');
    txt.value = txt.value.replace(/ｙ/g, 'y');
    txt.value = txt.value.replace(/ｚ/g, 'z');
 
    //Other punctuation
    txt.value = txt.value.replace(/＂/g, '"');
    txt.value = txt.value.replace(/＃/g, '#');
    txt.value = txt.value.replace(/＄/g, '$');
    txt.value = txt.value.replace(/％/g, '%');
    txt.value = txt.value.replace(/＆/g, '&');
    txt.value = txt.value.replace(/＇/g, '\'');
    txt.value = txt.value.replace(/＊/g, '*');
    txt.value = txt.value.replace(/＋/g, '+');
    txt.value = txt.value.replace(/，/g, ', ');
    //txt.value = txt.value.replace(/、/g, ', '); //replace ideographic comma with a regular comma (from TheFarix's AWB cleanup module)
    txt.value = txt.value.replace(/－/g, '-');
    txt.value = txt.value.replace(/．/g, '. ');
    txt.value = txt.value.replace(/／/g, '\/');
    txt.value = txt.value.replace(/：/g, ': ');
    txt.value = txt.value.replace(/；/g, '; ');
    txt.value = txt.value.replace(/＜/g, '<');
    txt.value = txt.value.replace(/＝/g, '='); //this replacement could break template usage, use {{=}} as an alternative inside templates
    txt.value = txt.value.replace(/＞/g, '>');
    txt.value = txt.value.replace(/＠/g, '@');
    txt.value = txt.value.replace(/［/g, '['); //this replacement could break wikimarkup usage
    txt.value = txt.value.replace(/＼/g, '\\');
    txt.value = txt.value.replace(/］/g, ']'); //this replacement could break wikimarkup usage
    txt.value = txt.value.replace(/＾/g, '^');
    txt.value = txt.value.replace(/＿/g, '_');
    txt.value = txt.value.replace(/｀/g, '`');
    txt.value = txt.value.replace(/｛/g, '{'); //this replacement could break wikimarkup usage, use {{(}} as an alternative
    txt.value = txt.value.replace(/｜/g, '|'); //this replacement could break template usage, use {{!}} as an alternative inside templates
    txt.value = txt.value.replace(/｝/g, '}'); //this replacement could break wikimarkup usage, use {{)}} as an alternative
    txt.value = txt.value.replace(/～/g, '〜'); //ASCII tilde is, apparently, treated as a diacritic in some fonts, so this should be a better replacement
    txt.value = txt.value.replace(/￠/g, '¢');
    txt.value = txt.value.replace(/￡/g, '£');
    txt.value = txt.value.replace(/￢/g, '¬');
    txt.value = txt.value.replace(/￤/g, '¦');
    txt.value = txt.value.replace(/￥/g, '¥');
    txt.value = txt.value.replace(/￦/g, '₩');
    txt.value = txt.value.replace(/￨/g, '│');
    txt.value = txt.value.replace(/￩/g, '←');
    txt.value = txt.value.replace(/￪/g, '↑');
    txt.value = txt.value.replace(/￫/g, '→');
    txt.value = txt.value.replace(/￬/g, '↓');
    txt.value = txt.value.replace(/￭/g, '■');
    txt.value = txt.value.replace(/￮/g, '○');
    txt.value = txt.value.replace(/『/g, '「'); //some Japanese quote replacement (from TheFarix's AWB cleanup module)
    txt.value = txt.value.replace(/』/g, '」'); //some Japanese quote replacement (from TheFarix's AWB cleanup module)
}
 
function fullwidthReplacerStartInEdit() { //Initiates fullwidthReplacer if you are already in edit mode
    fullwidthReplacer();
 
    // Add a tag to the summary box
    var txt = document.editform.wpSummary;
    var summary = "replaced fullwidth chars with ASCII equivalents ([[User:Dinoguy1000/scripts/fullwidth2ascii.js|FullwidthReplacer]])";
	if (txt.value.indexOf(summary) == -1) {
		if (txt.value.match(/[^\*\/\s][^\/\s]?\s*$/)) {
			txt.value += "; ";
		}
		txt.value += summary;
	}
 
    document.editform.wpDiff.click()
}
 
if(queryString('fullwidthReplacerStartFromView')) $(function() { fullwidthReplacerStartFromView(); }) //Part of next function
function fullwidthReplacerStartFromView() { //Initiates FullwidthReplacer if you are viewing the article without currently editing it
  var txt = document.getElementById('wpTextbox1');
  if(!txt) return;
  fullwidthReplacer();
  document.getElementById('wpSummary').value += 'Replaced fullwidth chars with ASCII equivalents using [[User:Dinoguy1000/scripts/fullwidth2ascii.js|FullwidthReplacer]]';
    document.editform.wpDiff.click()
}
 
function queryString(p) { //Allows URI to be properly decoded so that fullwidthReplacerStartFromView() works properly
  var re = RegExp('[&?]' + p + '=([^&]*)');
  var matches;
  if (matches = re.exec(document.location)) {
    try {
      return decodeURI(matches[1]);
    } catch (e) {
    }
  }
  return null;
}
 
$(function () { //Adds "replace fullwidth" tabs to page.
    if(mw.config.get("wgNamespaceNumber") == 0 || mw.config.get("wgNamespaceNumber") == 2) {
        if(document.forms.editform) {
            mw.util.addPortletLink('p-cactions', 'javascript:fullwidthReplacerStartInEdit()', 'replace fullwidth', 'ca-fullwidthreplaceredit', 'Replaces fullwidth ASCII characters with their ASCII equivalents', '', document.getElementById('ca-talk'));
        }
        if(mw.config.get("wgAction") == "view"){
             var url = mw.config.get("wgServer") + mw.config.get("wgScript") + '?title=' + encodeURIComponent(mw.config.get("wgPageName")) + '&action=edit';
             mw.util.addPortletLink('p-cactions', url + '&fullwidthReplacerStartFromView=true', 'replace fullwidth', 'ca-fullwidthreplacerview', 'Replaces fullwidth ASCII characters with their ASCII equivalents'), '', document.getElementById('ca-talk');
        }
    }
});