User:Est. 2021/Script/AutoCleanup.js
< User:Est. 2021 | Script
| 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:Est. 2021/Script/AutoCleanup. |
/*
Fork of [[User:Qwerfjkl/scripts/autoCleanup.js]], based on [[User:Tol/AutoCleanup.js]]
*/
// <syntaxhighlight lang="js">
$(function() {
function clean_wikitext(text) {
const replacements = [
// SPACING
[/(\u00A0|\u00AD|\u200A|\u200B|\u0009|\u0000-\u001F|\u0080-\u009F)/g, ' '], // Replace non-standard spacing with normal spacing #OWN // [[Help:CS1 errors#invisible char]]
[/( )+/gm, '$1'], // Remove excess spacing #1 #OWN
[/([^=\|\u3131-\uD79D]) (,|;|:[^:]|\.[^\.]|\![^\!]|\?)([^\w])/g, '$1$2$3'], // Remove excess spacing before punctuation #OWN // [[MOS:PUNCTSPACE]] (except: Korean)
[/ː /g, ': '], // Replace misused IPA colons #OWN // [[MOS:COLON]]
[/\[\[(.*)\|(\1)([\w']*)\]\]/gi, '[[$2]]$3'], // Fix piped links #OWN // [[MOS:PIPESTYLE]] & [[WP:NOPIPE]]
[/(\n[^\|\}].+) +\n/g, '$1\n'], // Remove lines' trailing spaces (that don't start with "|" or "}"
[/ \n/g, '\n'], // Remove excess spacing before line break #1 #OWN
// SPACING +
// [/([^\n\{ ])\| ([^\]])/g, '$1|$2'], // Remove excess spacing after "|" #OWN
[/(\w) \.(\s|\n)/g, '$1.$2'], // Remove spacing before dot #OWN
[/(\n *\| *\w*) *\= */g, '$1 = '], // Fix spacing around equal signs, pt.1/2 #OWN
[/(align|(col|row)span|scope|style) *\= */g, '$1='], // Fix spacing around equal signs, pt.2/2 #OWN
// TAGS (SPACING)
[/ *\<[\/]*br *[\/]*\> */gi, '<br />'], // Remove excess spacing around line breaks #OWN // [[H:BR]]
[/([^=\|]) (\<\!)/gi, '$1$2'], // Remove excess spacing before hidden tags #OWN
[/([^=\|]) (\<ref|\{\{(sfn|r)\|)/gi, '$1$2'], // Remove excess spacing before refs #OWN
[/(\!\!|\/ref\>)(\w)/g, '$1 $2'], // Add spacing after refs and "!!" #OWN
// TAGS & QUOTES
[/(\n)*<(small)>([^\2\/]*)<\/?\2>/gis, '$1{{$2|$3}}'], // [[Template:Small]] #OWN
[/(\n)*\:?<(block(quote))>([^\2\/]*)(\n)*<\/?\2>/gis, '$1{{$3|$4}}'], // [[Template:Quote]] #OWN
[/\{\{block(quote)\|/gi, '{{$1|'], // #OWN
[/\{\{i(ll)/gi, '{{i$1'], // #OWN
[/\{\{l(ang)/gi, '{{l$1'], // #OWN
[/\{\{n(otelist)/gi, '{{n$1'], // #OWN
[/\{\{r(eflist)/gi, '{{r$1'], // #OWN
[/\<references *\/\>/gi, '{{reflist}}'], // #OWN
// [[MOS:']] & [[MOS:"]] // [[MOS:CQ]]
[/(‘|’|`|´|˹|˺)/g, '\''], // Replace curly single quotes with straight ones #CUSTOM
[/(“|”)/g, '"'], // Replace curly double quotes with straight ones #CUSTOM
// [[MOS:NBSP]] & [[MOS:DASH]]
[/(–)/g, '–'], // Replace "–" with "–" #OWN
[/( |\{\{nbs\}\})(–|—)/gi, '{{snd}}'], // Replace "{{nbs}}–" with "{{dash}}" #OWN
[/(–|—)( |\{\{nbs\}\})/gi, '{{snd}}'], // Replace "–{{nbs}}" with "{{dash}}" #OWN
[/ *(\{\{(snd|dash)\}\}) */gi, '$1'], // Remove spacing around "{{dash}}" #OWN
[/([^\=\!\"\'])\}\}(\w)/g, '$1}} $2'], // Add spacing after templates #OWN
// [[MOS:NUM]]
// Birthdates
[/([^\w-–])b\. *(Jan|Feb|Mar|Apr|May|Jun|Jul|Aug|Sep|Oct|Nov|Dec)/g, '$1born $2'], // replace "b. Jan" with "born Jan" #CUSTOM
[/([^\w-–])b\. *([0-9]{1,2}) *(Jan|Feb|Mar|Apr|May|Jun|Jul|Aug|Sep|Oct|Nov|Dec)/g, '$1born $2 $3'], // replace "b. 31 Jan" with "born 31 Jan" #CUSTOM
[/([^\w-–])b\. *([0-9]{1,4})/g, '$1born $2'], // replace "b. 1987" with "born 1987" #OWN
[/(born|b\.) \{\{ *birth ?date *\|/ig, 'born {{SUBST:User:Zyxw/bd|'], // fix duplicated birth date templates
[/(\| *height *= *\{\{convert\| *[1-2]),([0-9]{1,2} *\| *m)/ig, '$1.$2'], // fix height with comma in {{convert}}
// Date formats
[/(\| *)([0-9]{2})\/([0-9]{2})\/([0-9]{4})/g, '$1$4-$3-$2'], // replace "| 31/12/2015" with "| 2015-12-31"
[/(\| *)([0-9]{4})[–—−\/]([0-9]{2})[–—−\/]([0-9]{2})/g, '$1$2-$3-$4'], // replace "| 2015–12–31" with "| 2015-12-31" (endash/emdash/minus/slash to hyphen)
// [[MOS:HEAD]]
[/^(=+) *\<(big\>)(.*?)\<\/\2 *\1$/gm, '$1 $3 $1'], // Remove "<big>" tags from headers #OWN
[/^(=+) *(.*?) *\1$/gm, '$1 $2 $1'], // Standardise spacing inside headers
[/\n+^((=+) .*? \2)$/gm, '\n\n$1'], // Standardise spacing before headers
[/^((=+) .*? \2)$\n+/gm, '$1\n'], // Standardise spacing after headers #CUSTOM
[/\n{3,}/g, '\n\n'], // Replace three or more newlines with two newlines #1
// [[MOS:FOOTERS]]
[/\n(== *)External (Links?)( *==)/i, '\n$1External links$3'], // #CUSTOM
[/\n(== *)Further Reading( *==)/i, '\n$1Further reading$2'],
[/\n(== *)Published Papers( *==)/i, '\n$1Published papers$2'],
[/\n(== *)See Also( *==)/i, '\n$1See also$2'],
[/\n(== *)Selected Works( *==)/i, '\n$1Selected works$2'],
// Add Ref section
// [/\.\n\n== ?External links ?==/g, '.\n\n== References =='], // #OFF
// References
[/((?:<ref[^>]*?>[^>]*?<\/ref>){1,})(\)?(,|;|:|\.)?)(\s|\n)/gm, '$2$1$4'], // Fix ref punctuation // Taken from [[User:Novem Linguae/Scripts/DraftCleaner.js]] #CUSTOM
[/>\[(http)([^ ]*) "*([^\]"]*)"*\]</g, '>{{Citation |title=$3 |url=$1$2}}<'], // Format external links #OWN
[/\| *(last|author)(\d*) *\= *(\w*), /g, '|last$2=$3 |first$2='], // Replace "last=Smith, John" with "last=Smith |first=John" #OWN
// Links
[/(?<!\[)\[https?:\/\/en\.wikipedia\.org\/wiki\/([^ \]]*)(?: ([^\]]*))?\]/gs, function(match, p1, p2) {
p1 = decodeURIComponent(p1);
p1 = p1.replace(/_/g, ' ');
if (p2 == undefined) {
return `[[${p1}]]`;
} else {
return `[[${p1}|${p2}]]`; // format wikilinks
} }],
[/\[\[([^\]]*)\]\]/gs, function(match, p1) {
p1 = decodeURIComponent(p1);
p1 = p1.replace(/_/g, ' ');
return `[[${p1}]]`;}], // remove "_" from wikilinks #OWN
// image cleanup
[/((Image:|File:)([^\.\n])*(\.\w*))/g, function(match, p1) {
p1 = decodeURIComponent(p1);
p1 = p1.replace(/_/g, ' ');
return `${p1}`;}], // remove "_" from file names #OWN
[/(\| *(image\d*|image_\w*|cover|flag|insignia|logo|photo|seal|shield) *= *)((Image|File):)?(([^\.\n])*(\.\w*))/g, function(match, p1, p2, p3, p4, p5) {
p5 = decodeURIComponent(p5);
p5 = p5.replace(/_/g, ' ');
return `${p1}${p5}`;}], // if "| image = File:name" is used, remove unneeded "File:" #OWN
[/(\| *(image\d*|image_\w*|cover|flag|insignia|logo|photo|seal|shield) *= *)\[\[((Image|File):)?(([^\.\n])*(\.\w*))\]\]/g, function(match, p1, p2, p3, p4, p5) {
p5 = decodeURIComponent(p5);
p5 = p5.replace(/_/g, ' ');
return `${p1}${p5}`;}], // if "| image = [[File:name]]" is used, remove unneeded "[[File:]]" #OWN
// Taken from pywikibot safe-syntax replaces, from the fixes.py script
[/\[\[(?<url>https?:\/\/[^\]]+?)\]\]/g, '[$1]'], // external link in double brackets
[/\[\[(?<url>https?:\/\/.+?)\]/g, '[$1]'], // external link starting with double bracket
[/\[(?<url>https?:\/\/[^\|\] \r\n]+?) +\| *(?<label>[^\|\]]+?)\]/, '[$1 $2]'], // external link and description separated by a dash, with whitespace in front of the dash, so that it is clear that the dash is not a legitimate part of the URL.
[/\[(?<url>https?:\/\/[^\|\] ]+?(\.pdf|\.html|\.htm|\.php|\.asp|\.aspx|\.jsp)) *\| *(?<label>[^\|\]]+?)\]/, '[$1 $3]'], // dash in external link, where the correct end of the URL can be detected from the file extension. It is very unlikely that this will cause mistakes.
[/\[\[([^\|\]]+)\|\1\]\]/, '[[$1]]'], // link equal to linktext
// Taken from [[User:Sam Sailor/Scripts/autoFormatterSettings.js]]
// [/ =\n/g, ' = \n'], // add trailing space after = for empty infobox parameters #OFF
[/\{\{ /g, '{{'], // remove space after {{
[/\|tau\]\]/g, '|τ]]'], // fix unwanted change in category sorting
// spaces after bullets
[/(\n[*#]+) *([^\s*#:;R])/g, '$1 $2'], // space after one or more * or # at start of line
// End cleaning
[/( )+/gm, '$1'], // Remove excess spacing #2 #OWN
[/ \n/g, '\n'], // Remove excess spacing before line break #2 #OWN
[/\n{3,}/g, '\n\n'], // Replace three or more newlines with two newlines #2
// [/\[\.\.\.\]/g, '[…]'], // Replace "[...]" with "[…]" #OWN #OFF // [[MOS:...]]
];
for (const replacement of replacements) {
text = text.replace(...replacement);
}
return text;
}
function process_summary(summary) {
const ac_summary = 'cleanup via [[User:Est. 2021/Script/AutoCleanup.js|script]]';
if (summary) {
if (/^\/\*.*?\*\/ $/.test(summary)) { // auto summary
return summary + ac_summary;
} else {
return summary + '; ' + ac_summary;
}
} else {
return ac_summary;
}
}
function main() {
mw.notify(
'Processing...',
{
tag: 'tol-autocleanup',
title: 'AutoCleanup',
type: 'info',
autoHide: true,
autoHideSeconds: 2
}
);
let text_box = $('#wpTextbox1');
let old_text = text_box.textSelection('getContents');
let new_text = clean_wikitext(old_text);
if (old_text != new_text) {
text_box.textSelection('setContents', new_text);
if (mw.util.getParamValue('section') != 'new') { // not section title instead of summary
let summary_box = $('#wpSummary');
summary_box.val(
process_summary(summary_box.val())
)}
mw.notify(
'Done',
{
tag: 'tol-autocleanup',
title: 'AutoCleanup',
type: 'success'
}
);
} else {
mw.notify(
'No changes made',
{
tag: 'tol-autocleanup',
title: 'AutoCleanup',
type: 'info'
}
);
}
}
function add_button() {
try {
if ($('editform').length) {
mw.loader.using(['oojs-ui', 'oojs-ui-widgets'], function() {
let button = new OO.ui.ButtonWidget({label: 'AutoCleanup'});
button.on('click', main);
$('.editButtons').append(button.$element);
});
}
} catch(error) {
alert(error);
}
}
//$(add_button);
if (mw.config.get('wgPageContentModel') === 'wikitext' && mw.config.get('wgAction') === 'edit' && mw.util.getParamValue( "ACdisable" ) != "1" ) {
mw.util.addPortletLink( "p-cactions", window.location.href + '&ACdisable=1', "Disable AutoCleanup", "pt-disableAC" );
$(main);
}
});
importScript('User:Est. 2021/Script/AutoCleanup/Tables.js'); // Backlink: [[User:Est. 2021/Script/AutoCleanup/Tables.js]]
importScript('User:Est. 2021/Script/AutoCleanup/Templates.js'); // Backlink: [[User:Est. 2021/Script/AutoCleanup/Templates.js]]
importScript('User:Est. 2021/Script/AutoCleanup/sandbox.js'); // Backlink: [[User:Est. 2021/Script/AutoCleanup/sandbox.js]]
// </syntaxhighlight>