User:Novem Linguae/Scripts/DraftCleaner.js
//
// === Compiled with Novem Linguae's publish.php script ======================
// === modules/DraftCleaner.js ======================================================
class DraftCleaner {
cleanDraft( wikicode, namespaceNumber, titleWithNamespaceAndSpaces ) {
// run before other stuff
wikicode = this.deleteSomeHTMLTags( wikicode );
wikicode = this.deleteNonAFCDraftTags( wikicode );
wikicode = this.deleteAFCDraftTagsIfMainspace( wikicode, mw.config.get( 'wgNamespaceNumber' ) );
wikicode = this.fixWikilinksContainingURL( wikicode );
wikicode = this.fixExternalLinksToWikipediaArticles( wikicode );
wikicode = this.deleteWeirdUnicodeCharacters( wikicode );
wikicode = this.trimEveryLine( wikicode );
wikicode = this.convertH1ToH2( wikicode );
wikicode = this.convertVeryLongHeadingToParagraph( wikicode );
wikicode = this.deleteHeadingsWithTitle( wikicode, titleWithNamespaceAndSpaces );
wikicode = this.unlinkWikilinksToThisTitle( wikicode, titleWithNamespaceAndSpaces );
wikicode = this.capitalizeCOVID19( wikicode );
wikicode = this.removeBoldFromHeadings( wikicode );
wikicode = this.convertReferenceToReferences( wikicode );
wikicode = this.deleteMultipleReferenceTags( wikicode );
wikicode = this.addReferencesSectionIfMissing( wikicode );
wikicode = this.fixEmptyReferencesSection( wikicode );
wikicode = this.deleteWhitespaceAtEndOfLines( wikicode );
wikicode = this.convertSmartQuotesToRegularQuotes( wikicode );
// wikicode = this.fixWordEmphasizedWithSingleQuotes(wikicode); // most of these appear in citations as names of newspaper articles, arguably should keep these single quotes
wikicode = this.convertDoubleSpacesToSingleSpaces( wikicode );
wikicode = this.deleteBlankHeadings( wikicode );
wikicode = this.changeYearRangeDashToNDash( wikicode );
wikicode = this.disableCategoriesInDraftspace( wikicode, namespaceNumber );
// wikicode = this.deleteBRTagsOutsideInfoboxes(wikicode, namespaceNumber); // edge case in image captions, and probably other places
// wikicode = this.rightAlignImages(wikicode); // commenting out, too many false positives in featured articles
wikicode = this.correctCapitalizationOfEndMatterHeaders( wikicode );
wikicode = this.ifNoLeadSectionDeleteFirstHeading( wikicode );
wikicode = this.deleteCopyPastedEditAndEditSource( wikicode );
wikicode = this.replaceUnicodeBulletsWithAsterisks( wikicode );
wikicode = this.deleteEmptySections( wikicode );
wikicode = this.fixHeadingsInAllCaps( wikicode );
wikicode = this.deleteDuplicateReferencesSection( wikicode );
wikicode = this.deleteBlankLinesBetweenBullets( wikicode );
wikicode = this.removeUnderscoresFromWikilinks( wikicode );
wikicode = this.fixPipedWikilinksWithIdenticalParameters( wikicode );
wikicode = this.removeBorderFromImagesInInfoboxes( wikicode );
wikicode = this.removeExtraAFCSubmissionTemplates( wikicode );
wikicode = this.moveAFCSubmissionTemplatesToTop( wikicode );
// all ==sections== should start with a capital letter
// after swap, if citation has no spaces on either side, and is not touching two other citations, add a space on the right
// strip [[File: from infobox's image field
// example 1: | image = File:SAMIR 1626.png
// example 2: | image = File:SAMIR 1626.png
// trim whitespace inside refs, e.g. abc
// replace unreliable sources with {{cn}}.
// if adjacent to other sources, just delete
// if ref is used multiple times, account for that
// duplicate citation fixer
// move refs that are below {{Reflist}}, to above {{Reflist}}
// move refs out of headings
// delete AFC submission templates located mid-article, they end up self-hiding then appear as inexplicable whitespace. example: {{AfC submission|t||ts=20211212134609|u=Doezdemir|ns=118|demo=}}
// fix redundant wikilinks, e.g. Spotify
// change youtu.be to youtube.com, to avoid the blacklist. test: https://youtu.be/bnWHeRNIPiA
// delete ©®™
// convert all
// remove px from images, should use default
// convert refs toward the end. we want deleteSomeHTMLTags() to run first, to get rid of tags around URLs
wikicode = this.bareURLToRef( wikicode );
wikicode = this.refShortLinkToLongLink( wikicode );
wikicode = this.inlineExternalLinksToRefs( wikicode );
wikicode = this.moveRefsOutsideOfItalics( wikicode );
wikicode = this.deleteSpacesInFrontOfRefs( wikicode );
wikicode = this.deleteNewLinesBetweenRefs( wikicode );
wikicode = this.swapRefPeriodWithPeriodRef( wikicode );
wikicode = this.swapRefCommaWithCommaRef( wikicode );
// stuff we want to run at the end
wikicode = this.fixDoublePeriod( wikicode ); // need test cases. I've seen this one not work.
wikicode = this.boldArticleTitle( wikicode, titleWithNamespaceAndSpaces );
wikicode = this.trimEmptyLines( wikicode );
wikicode = this.deleteMoreThanTwoEntersInARow( wikicode );
return wikicode;
}
// Useful for seeing all URL's in the reflist section, and for CiteHighlighter ref quality highlighting
bareURLToRef( wikicode ) {
return wikicode.replace( /^(http[^\n ]*) {0,}$/gm, '$1' );
}
// in refs, turn [short links] into long links, so you can see the domain
// also fixes link issues with extra spaces in the ref
refShortLinkToLongLink( wikicode ) {
wikicode = wikicode.replace( /(]*>) {0,}\[ {0,}([^'\]]*)(''[^\]]*)\] {0,}(<\/ref>)/gm, '$1$2 $3$4' );
wikicode = wikicode.replace( /(]*>) {0,}\[ {0,}([^\]]*) {0,}\] {0,}(<\/ref>)/gm, '$1$2$3' );
return wikicode;
}
// convert inline external links to references
inlineExternalLinksToRefs( wikicode ) {
const sectionsToSkip = [ 'External link', 'Further reading', 'Links' ];
let regExString = '== ?(?:';
for ( const sectionToSkip of sectionsToSkip ) {
regExString += sectionToSkip + '|';
}
regExString = regExString.slice( 0, -1 ) + ')';
const hasSectionToSkip = wikicode.match( new RegExp( regExString, 'i' ) );
const sf = new StringFilter();
if ( hasSectionToSkip ) {
const regExToSplitArticle = new RegExp( '((' + regExString + ').*$)', 'is' );
const topHalf = wikicode.replace( regExToSplitArticle, '' );
const bottomHalf = wikicode.match( regExToSplitArticle )[ 1 ];
const buffer = sf.surgicalReplaceOutsideTags(
/(?|> )\[(http[^ \]]+) ?(.*?)\](?!<\/ref>| <\/ref>)/gm,
'$2$1',
topHalf,
[ '
[ '
);
wikicode = buffer + bottomHalf;
} else {
wikicode = sf.surgicalReplaceOutsideTags(
/(?|> )\[(http[^ \]]+) ?(.*?)\](?!<\/ref>| <\/ref>)/gm,
'$2$1',
wikicode,
[ '
[ '
);
}
return wikicode;
}
/**
*/
moveRefsOutsideOfItalics( wikicode ) {
wikicode = wikicode.replace( /([^']+)([^<]+<\/ref>)/gm, '\'\'$1\'\'$2' );
return wikicode;
}
// get rid of spaces in front of
deleteSpacesInFrontOfRefs( wikicode ) {
return wikicode.replace( /(?
}
// get rid of any level 2 heading that contains the article's title
// this takes care of 2 common cases: heading at the bottom next to the {{AFC Submission}} template, and heading at the top above the lead
deleteHeadingsWithTitle( wikicode, titleWithNamespaceAndSpaces ) {
let headingNameToLookFor = titleWithNamespaceAndSpaces;
headingNameToLookFor = headingNameToLookFor.replace( /^Draft:/, '' );
headingNameToLookFor = this._escapeRegEx( headingNameToLookFor );
let regEx = new RegExp( '^== ?' + headingNameToLookFor + ' ?==\n', 'gmi' );
wikicode = wikicode.replace( regEx, '' );
// now look for titles that contain Draft: at the beginning, too
headingNameToLookFor = titleWithNamespaceAndSpaces;
headingNameToLookFor = this._escapeRegEx( headingNameToLookFor );
regEx = new RegExp( '^== ?' + headingNameToLookFor + ' ?==\n', 'gmi' );
wikicode = wikicode.replace( regEx, '' );
return wikicode;
}
// remove wikilinks to article name
// Example: if title is Draft:Menna Shahin, change Menna Shahin to Menna Shahin
unlinkWikilinksToThisTitle( wikicode, titleWithNamespaceAndSpaces ) {
let wikilinkToLookFor = titleWithNamespaceAndSpaces;
wikilinkToLookFor = wikilinkToLookFor.replace( /^Draft:/, '' );
wikilinkToLookFor = this._escapeRegEx( wikilinkToLookFor );
const regEx = new RegExp( '\\[\\[(' + wikilinkToLookFor + ')\\]\\]', 'gm' );
wikicode = wikicode.replace( regEx, '$1' );
return wikicode;
}
// if located in the first paragraph, bold the article title
boldArticleTitle( wikicode, titleWithNamespaceAndSpaces ) {
let titleToLookFor = titleWithNamespaceAndSpaces;
titleToLookFor = titleToLookFor.replace( /^Draft:/, '' );
titleToLookFor = titleToLookFor.replace( / \(.*?\)$/, '' );
titleToLookFor = this._escapeRegEx( titleToLookFor );
// Don't bold the title if it's already bolded. Fixes a "bold twice" bug.
const hasBoldedTitle = wikicode.match( new RegExp( `${ titleToLookFor }`, 'i' ) );
if ( hasBoldedTitle ) {
return wikicode;
}
// Be pretty strict, to avoid adding ''' to image paths and infoboxes, which messes up the image. Also, only replace first match.
const regEx = new RegExp( '^(The )?(' + titleToLookFor + ')([ <,])', 'mi' );
wikicode = wikicode.replace( regEx, "$1$2$3" );
return wikicode;
}
// /covid-19/i -> COVID-19
// Careful of this string in URLs.
capitalizeCOVID19( wikicode ) {
const sf = new StringFilter();
wikicode = sf.surgicalReplaceOutsideTags( / covid-19/gmi, ' COVID-19', wikicode, [ '{{', '' ], [ '}}', '' ] );
wikicode = sf.surgicalReplaceOutsideTags( /\ncovid-19/gmi, '\nCOVID-19', wikicode, [ '{{', '' ], [ '}}', '' ] );
return wikicode;
}
// remove bold from headings
removeBoldFromHeadings( wikicode ) {
return wikicode.replace( /^(=.*)(.*)(.*=)$/gm, '$1$2$3' );
}
// remove enter characters between s
deleteNewLinesBetweenRefs( wikicode ) {
return wikicode.replace( /<\/ref>\n{1,}/gm, '' );
}
// convert ==Reference== to ==References==
convertReferenceToReferences( wikicode ) {
return wikicode.replace( /^== ?Reference ?==$/gmi, '== References ==' );
}
// TOOL - swap ref period with period ref
swapRefPeriodWithPeriodRef( wikicode ) {
wikicode = wikicode.replace( /((?:]*?>[^>]*?<\/ref>){1,})\. /gm, '.$1 ' );
wikicode = wikicode.replace( /((?:]*?>[^>]*?<\/ref>){1,})\.\n/gm, '.$1\n' );
return wikicode;
}
swapRefCommaWithCommaRef( wikicode ) {
wikicode = wikicode.replace( /((?:]*?>[^>]*?<\/ref>){1,}), /gm, ',$1 ' );
wikicode = wikicode.replace( /((?:]*?>[^>]*?<\/ref>){1,}),\n/gm, ',$1\n' );
return wikicode;
}
// fix errant spaces at beginning of lines, which makes a blockquote looking thing (AFCH does it)
trimEveryLine( wikicode ) {
let output = '';
const lines = wikicode.split( '\n' );
const lineCount = lines.length;
let i = 0;
for ( const line of lines ) {
i++;
const trimmed = line.trim();
if ( trimmed.startsWith( '|' ) || trimmed.startsWith( '}' ) ) { // don't trim lines that start with | or }. It is common in FAs to indent these a bit.
output += line;
} else {
output += line.trim();
}
if ( i !== lineCount ) {
output += '\n';
}
}
return output;
}
// add references section if missing
addReferencesSectionIfMissing( wikicode ) {
const hasRefSection = wikicode.match( /^== ?References ?==$/mi );
const hasReflist = wikicode.match( /(?:{{Reflist| if ( !hasRefSection && !hasReflist ) { const hasBottomAFCTemplate = wikicode.match( /(\n{{AfC submission[^}]*}}\s*)$/ ); if ( hasBottomAFCTemplate ) { wikicode = wikicode.replace( /(\n{{AfC submission[^}]*}}\s*)$/, '\n\n== References ==\n{{Reflist}}$1' ); } else { wikicode = wikicode.replace( /$/, '\n\n== References ==\n{{Reflist}}' ); } } return wikicode; } // fix empty references section fixEmptyReferencesSection( wikicode ) { const hasRefSection = wikicode.match( /^== ?References ?==$/mi ); const hasReflist = wikicode.match( /(?:{{Reflist| if ( !hasReflist && hasRefSection ) { wikicode = wikicode.replace( /(?<=== ?References ?==)/gmi, '\n{{Reflist}}' ); } return wikicode; } // delete whitespace at the end of lines // (?!\|)(?!\}\}) is to stop this from deleting spaces after = in infoboxes deleteWhitespaceAtEndOfLines( wikicode ) { return wikicode.replace( /[ \t]+\n(?!\|)(?!\}\})/g, '\n' ); } // convert smart quotes to regular quotes convertSmartQuotesToRegularQuotes( wikicode ) { const sf = new StringFilter(); wikicode = sf.surgicalReplaceOutsideTags( /”/g, '"', wikicode, [ 'File:' ], [ '' ] ); wikicode = sf.surgicalReplaceOutsideTags( /“/g, '"', wikicode, [ 'File:' ], [ '' ] ); wikicode = sf.surgicalReplaceOutsideTags( /‘/g, "'", wikicode, [ 'File:' ], [ '' ] ); wikicode = sf.surgicalReplaceOutsideTags( /’/g, "'", wikicode, [ 'File:' ], [ '' ] ); wikicode = sf.surgicalReplaceOutsideTags( /…/g, '...', wikicode, [ 'File:' ], [ '' ] ); return wikicode; } // convert double spaces to single spaces convertDoubleSpacesToSingleSpaces( wikicode ) { return wikicode.replace( /\. {2,}/g, '. ' ); } // remove blank heading deleteBlankHeadings( wikicode ) { return wikicode.replace( /\n={2,} {0,}={2,}\n/g, '\n' ); } // Change year range dash to ndash. Skip text inside of [[File: changeYearRangeDashToNDash( wikicode ) { const sf = new StringFilter(); // (1111-1111) wikicode = sf.surgicalReplaceOutsideTags( /(\(\d{4}) ?- ?(\d{4}\))/gm, '$1–$2', wikicode, [ 'File:' ], [ '' ] ); // 1839 - 1926) wikicode = sf.surgicalReplaceOutsideTags( /( \d{4}) ?- ?(\d{4}\))/gm, '$1–$2', wikicode, [ 'File:' ], [ '' ] ); return wikicode; } // if in draftspace, and draft has categories, disable the categories disableCategoriesInDraftspace( wikicode, namespace ) { const draft = ( namespace == 118 ); if ( draft ) { wikicode = wikicode.replace( /:?(\[\[)(Category:[^\]]*\]\])/gm, '$1:$2' ); } wikicode = wikicode.replace( /\[\[:Category:Created via preloaddraft\]\]/gi, 'Category:Created via preloaddraft' ); return wikicode; } // delete deleteBRTagsOutsideInfoboxes( wikicode ) { let output = ''; const lines = wikicode.split( '\n' ); const lineCount = lines.length; let i = 0; for ( const line of lines ) { i++; // Skip lines that start with { or |. This is the easiest way to detect infoboxes if ( line.startsWith( '{' ) || line.startsWith( '|' ) ) { output += line; } else { output += line.replace( / } if ( i !== lineCount ) { output += '\n'; } } return output; } // right align images rightAlignImages( wikicode ) { return wikicode.replace( /(\[\[File:[^\]]*\|)left(\|[^\]]*\]\])/gm, '$1right$2' ); } // correct capitalization of see also, references, further reading, external links correctCapitalizationOfEndMatterHeaders( wikicode ) { wikicode = wikicode.replace( /^(== ?)References( ?==)$/gmi, '$1References$2' ); wikicode = wikicode.replace( /^(== ?)External links( ?==)$/gmi, '$1External links$2' ); wikicode = wikicode.replace( /^(== ?)Further reading( ?==)$/gmi, '$1Further reading$2' ); wikicode = wikicode.replace( /^(== ?)See also( ?==)$/gmi, '$1See also$2' ); return wikicode; } // if article has headings but no lead, remove first heading ifNoLeadSectionDeleteFirstHeading( wikicode ) { let output = ''; const lines = wikicode.split( '\n' ); const lineCount = lines.length; let i = 0; let textCount = 0; for ( const line of lines ) { i++; // scan for first heading. // empty lines, lines with templates, or lines with images do not count. if ( line.startsWith( '{' ) || line.length === 0 || line.startsWith( '[[File:' ) ) { output += line; } else if ( line.startsWith( '==' ) && !textCount ) { continue; // delete this line by not putting it in the output string } else { textCount++; output += line; } if ( i !== lineCount ) { output += '\n'; } } return output; } // delete [edit], [edit source], and [editar] from headings deleteCopyPastedEditAndEditSource( wikicode ) { wikicode = wikicode.replace( /\[edit\]( ?={2,})$/gm, '$1' ); wikicode = wikicode.replace( /\[edit source\]( ?={2,})$/gm, '$1' ); wikicode = wikicode.replace( /\[editar\]( ?={2,})$/gm, '$1' ); return wikicode; } // at beginning of lines, replace unicode bullets with asterisks replaceUnicodeBulletsWithAsterisks( wikicode ) { return wikicode.replace( /^\s{0,}[·•●]\s{0,}/gm, '* ' ); } // remove whitespace if that is the only character on a line trimEmptyLines( wikicode ) { return wikicode.replace( /^\s*$/gm, '' ); } // no more than 2 newlines (1 blank line) in a row. // Note: AFCH does this too deleteMoreThanTwoEntersInARow( wikicode ) { wikicode = wikicode.replace( /\n{3,}/gm, '\n\n' ); return wikicode; } // convert =TitleHeading= to ==H2Heading== convertH1ToH2( wikicode ) { return wikicode.replace( /^= ?([^=]*?) ?=$/gm, '== $1 ==' ); } convertVeryLongHeadingToParagraph( wikicode ) { let output = ''; const lines = wikicode.split( '\n' ); const lineCount = lines.length; let i = 0; for ( const line of lines ) { i++; if ( line.length > 150 && line.match( /^==.*==$/gm ) && !line.match( /
output += line.replace( /^={1,}\s*(.*?)\s*={1,}$/m, '$1' ); } else { output += line; } if ( i !== lineCount ) { output += '\n'; } } return output; } fixWordEmphasizedWithSingleQuotes( wikicode ) { return wikicode.replace( / '(\w+)' /g, ' "$1" ' ); } fixDoublePeriod( wikicode ) { return wikicode.replace( /(?<=[A-Za-z\]])\.\.(?=
} fixWikilinksContainingURL( wikicode ) { // non-piped wikilink wikicode = wikicode.replace( /\[\[https?:\/\/en\.(?:m\.)?wikipedia\.org\/wiki\/([^|]*)\]\]/g, '$1' ); // piped wikilink wikicode = wikicode.replace( /\[\[https?:\/\/en\.(?:m\.)?wikipedia\.org\/wiki\/([^|]*)\|([^\]]*)\]\]/g, '$2' ); // non-piped external link wikicode = wikicode.replace( /\[\[(http[^|]*)\]\]/g, '[$1]' ); // piped external link wikicode = wikicode.replace( /\[\[(http[^|]*)\|([^\]]*)\]\]/g, '[$1 $2]' ); return wikicode; } fixExternalLinksToWikipediaArticles( wikicode ) { // [https://en.wikipedia.org/wiki/Article] and [https://en.wikipedia.org/wiki/Article Article name] return wikicode.replace( /(? { p1 = decodeURIComponent( p1 ); p1 = p1.replace( /_/g, ' ' ); return `${ p1 }`; } ); } deleteBlankLinesBetweenBullets( wikicode ) { const lines = wikicode.split( '\n' ); const buffer = []; const length = lines.length; for ( let i = 0; i < length; i++ ) { const previous = lines[ i - 1 ]; const current = lines[ i ]; const next = lines[ i + 1 ]; if ( typeof previous !== 'undefined' && typeof next !== 'undefined' && previous.startsWith( '*' ) && current === '' && next.startsWith( '*' ) ) { continue; } buffer.push( current ); } return buffer.join( '\n' ); } deleteWeirdUnicodeCharacters( wikicode ) { return wikicode.replace( /[]/g, '' ); } deleteSomeHTMLTags( wikicode ) { wikicode = wikicode.replace( /<\/?p( [^>]*)?\/?>/g, '' ); wikicode = wikicode.replace( /<\/?strong( [^>]*)?\/?>/g, '' ); wikicode = wikicode.replace( /<\/?em( [^>]*)?\/?>/g, '' ); wikicode = wikicode.replace( /<\/?nowiki( [^>]*)?\/?>/g, '' ); wikicode = wikicode.replace( /<\/?u( [^>]*)?\/?>/g, '' ); wikicode = wikicode.replace( /(?:|<\/big>)/g, '' ); return wikicode; } deleteNonAFCDraftTags( wikicode ) { wikicode = wikicode.replace( /{{Preloaddraft submit}}\n{0,2}/gi, '' ); wikicode = wikicode.replace( /\n{0,2}/gi, '' ); wikicode = wikicode.replace( /{{Draft}}\n{0,2}/gi, '' ); return wikicode; } deleteAFCDraftTagsIfMainspace( wikicode, namespaceNumber ) { const isMainspace = namespaceNumber == 0; if ( isMainspace ) { // {{AfC submission}}, {{AfC topic}}, {{AfC comment}}, etc. wikicode = wikicode.replace( /{{AfC [^}]*}}\n?/g, '' ); wikicode = wikicode.replace( /{{Draft topics[^}]*}}\n?/g, '' ); } return wikicode; } fixHeadingsInAllCaps( wikicode ) { // create a concatenated string with the text from every heading const matches = wikicode.matchAll( /== {0,}(.+) {0,}==/g ); let headingString = ''; for ( const match of matches ) { headingString += match[ 1 ]; } // if string only contains caps if ( this._isUpperCase( headingString ) ) { // convert all headings to sentence case const matches = wikicode.matchAll( /== {0,}(.+) {0,}==/g ); for ( const match of matches ) { let matchRegex = this._escapeRegEx( match[ 1 ].trim() ); matchRegex = new RegExp( '== {0,}' + matchRegex + ' {0,}==', 'g' ); const sentenceCase = this._toSentenceCase( match[ 1 ].trim() ); wikicode = wikicode.replace( matchRegex, '== ' + sentenceCase + ' ==' ); } } return wikicode; } deleteEmptySections( wikicode ) { return wikicode.replace( /\n*== ?(?:See also|External links) ?==\n*$/, '' ); } deleteDuplicateReferencesSection( wikicode ) { const matches = wikicode.match( /==\s*References\s*==/gi ); if ( matches !== null && matches.length > 1 ) { // run regexes that are likely to delete the extra section const attempt = wikicode.replace( `== References == {{reflist}}` , '' ); const matches2 = attempt.match( /==\s*References\s*==/gi ); if ( matches2.length === 1 ) { wikicode = attempt.trim(); wikicode = wikicode.replace( /==\s*References\s*==/gi, '== References ==' ); } } return wikicode; } removeUnderscoresFromWikilinks( wikicode ) { const sf = new StringFilter(); wikicode = sf.surgicalReplaceInsideTags( /_/g, ' ', wikicode, [ '' ], [ '' ] ); return wikicode; } fixPipedWikilinksWithIdenticalParameters( wikicode ) { const matches = wikicode.matchAll( /\[\[([^|\]]+)\|([^\]]+)\]\]/g ); for ( const match of matches ) { if ( match[ 1 ] === match[ 2 ] ) { wikicode = this._replaceAll( wikicode, `${ match[ 1 ] }`, `${ match[ 1 ] }` ); } } return wikicode; } removeBorderFromImagesInInfoboxes( wikicode ) { wikicode = wikicode.replace( /(\|\s*logo\s*=\s*)\[\[File:([^\]|]*)[^\]\]]*\]\]/g, '$1$2' ); wikicode = wikicode.replace( /(\|\s*cover\s*=\s*)\[\[File:([^\]|]*)[^\]\]]*\]\]/g, '$1$2' ); return wikicode; } /** These often hide towards the bottom of a draft. When the draft is submitted, unsubmitted templates (t) detect this and show up as blank, creating a weird extra line break. So this basically fixes the line break. */ removeExtraAFCSubmissionTemplates( wikicode ) { const hasSubmittedTemplate = wikicode.match( /{{AfC submission\|\|/ ); const hasUnsubmittedTemplate = wikicode.match( /{{AfC submission\|t\|/ ); if ( hasSubmittedTemplate && hasUnsubmittedTemplate ) { wikicode = wikicode.replace( /{{AfC submission\|t\|[^}}]*\}\}\n?/gm, '' ); } return wikicode; } moveAFCSubmissionTemplatesToTop( wikicode ) { const hasTemplateAtBottom = wikicode.match( /\n[^\n]+\n*({{AfC submission[^}]*}})\s*$/i ); if ( hasTemplateAtBottom ) { // delete all submission templates wikicode = wikicode.replace( /{{AfC submission[^}}]*\}\}\n?/gm, '' ); // insert template at top wikicode = hasTemplateAtBottom[ 1 ] + '\n----\n\n' + wikicode; } return wikicode; } deleteMultipleReferenceTags( wikicode ) { const hasReflist = wikicode.match( /{{Reflist}}/i ); const hasReferencesTag = wikicode.match( / if ( hasReflist && hasReferencesTag ) { // delete all references tags wikicode = wikicode.replace( / } return wikicode; } _isUpperCase( str ) { return str === str.toUpperCase(); } _toSentenceCase( string ) { return string.charAt( 0 ).toUpperCase() + string.slice( 1 ).toLowerCase(); } _replaceAll( haystack, needle, replacement ) { const regex = new RegExp( this._escapeRegEx( needle ), 'g' ); haystack = haystack.replace( regex, replacement ); return haystack; } _escapeRegEx( string ) { return string.replace( /[.*+?^${}()|[\]\\]/g, '\\$&' ); // $& means the whole matched string } } // === modules/StringFilter.js ====================================================== /** * Lets you use regex to specify what parts of a very long string you want to specify as "off limits", then you can do additional regex's and search/replace to the remaining parts of the string. */ class StringFilter { /** * Does a replace, but specifies areas of the file that should NOT be replaced. Those areas are specified by providing an openingTag and a closingTag, and those areas are marked as off limits. */ surgicalReplaceOutsideTags( regex, replacement, haystack, openingTags, closingTags ) { const allTags = [ ...openingTags, ...closingTags ]; const parts = this._splitStringUsingMultiplePatterns( haystack, allTags ); const resultArray = []; for ( let part of parts ) { let openingTagMatch = false; for ( const tag of openingTags ) { if ( part.startsWith( tag ) ) { openingTagMatch = true; break; } } if ( !openingTagMatch ) { part = part.replace( regex, replacement ); } resultArray.push( part ); } return resultArray.join( '' ); } /** * Does a replace, but specifies areas of the file that SHOULD be replaced, then skips the rest of the file. The area that should be replaced is specified by providing an openingTag and a closingTag. */ surgicalReplaceInsideTags( regex, replacement, haystack, openingTags, closingTags ) { const allTags = [ ...openingTags, ...closingTags ]; const parts = this._splitStringUsingMultiplePatterns( haystack, allTags ); const resultArray = []; for ( let part of parts ) { for ( const tag of openingTags ) { if ( part.startsWith( tag ) ) { part = part.replace( regex, replacement ); } } resultArray.push( part ); } return resultArray.join( '' ); } /** * Also keeps the pattern in the result, unlike string.prototype.split. Algorithm isn't perfect, will fail with this pattern: Test/>Test. But should be good enough for DraftCleaner stuff. * * @param {string} string * @param {string[]} patterns * @return {string[]} */ _splitStringUsingMultiplePatterns( string, patterns ) { const length = string.length; const result = []; let positionOfLastMatch = 0; for ( let i = 0; i < length; i++ ) { const lookAhead = string.slice( i ); // the rest of the string after current position let patternMatch = false; for ( const pattern of patterns ) { if ( lookAhead.startsWith( pattern ) ) { patternMatch = true; break; } } if ( patternMatch ) { const chunk = string.slice( positionOfLastMatch, i ); // if blank (happens if i=0 matches), continue instead of putting an empty "" into the array if ( !chunk ) { continue; } result.push( chunk ); positionOfLastMatch = i; } } // Don't forget the last chunk. result.push( string.slice( positionOfLastMatch ) ); return result; } } $(async function() { // === main.js ====================================================== /* THIS SCRIPT IS BUGGY ABOUT 10% OF THE TIME. Be sure to check the diff that pops up before submitting. - Adds "Run DraftCleaner" link to the left sidebar - Top uses: - remove extra line breaks (for example, 3 enters in a row) - in the first sentence, bold the title - convert curly quotes to regular quotes - put s after periods - clean external links out of the main article area (turn them into references) - add ==References== section - remove bold from headings - Other uses: - converts [inline external links] to s - removes spaces in front of s - get rid of any level 2 heading that contains the article's title - converts =TitleHeading= to ==H2Heading== - replaces Covid-19 with COVID-19 - removes enter characters between s - trims whitespace at beginning and end - remove self wikilinks to the article title - convert ==Reference== to ==References== - turn bare URLs into references - fix errant spaces at beginning of lines, which makes a blockquote looking thing - delete whitespace at the end of lines - convert double spaces to single spaces - remove blank heading - in refs, turn short links into long links, so you can see the domain - change year range dash to ndash - if in draftspace, and draft in categories, disable the categories - delete - fix empty references section - right align images - remove whitespace if that is the only character on a line - correct capitalization of see also, references, further reading, external links - if article has headings but no lead, remove first heading - replace unicode bullets with asterisks Add one of the following to your User:yourName/common.js (at the top) to change the position where DraftCleaner puts its link: window.draftCleanerPutInToolsMenu = true; window.draftCleanerPutInMoreMenu = true; This page was assembled from 3 files using my publish.php script. I have an offline test suite with around 100 unit tests for the DraftCleaner and StringFilter classes. ( function () { async function getWikicode( title ) { const pageIsDeleted = !mw.config.get( 'wgCurRevisionId' ); if ( pageIsDeleted ) { return ''; } let wikicode = ''; title = encodeURIComponent( title ); await $.ajax( { url: 'https://en.wikipedia.org/w/api.php?action=parse&page=' + title + '&prop=wikitext&formatversion=2&format=json', success: function ( result ) { wikicode = result.parse.wikitext; }, dataType: 'json' } ); return wikicode; } function goToShowChangesScreen( titleWithNamespaceAndUnderscores, wikicode, editSummary ) { const titleEncoded = encodeURIComponent( titleWithNamespaceAndUnderscores ); const wgServer = mw.config.get( 'wgServer' ); const wgScriptPath = mw.config.get( 'wgScriptPath' ); const baseURL = wgServer + wgScriptPath + '/'; // https://stackoverflow.com/a/12464290/3480193 $( ` .append( $( '' ).val( wikicode ) ) .append( $( '' ).val( editSummary ) ) .append( $( '' ).val( 'preview' ) ) .append( $( '' ).val( 'Show changes' ) ) .append( $( '' ).val( '1' ) ) .appendTo( $( document.body ) ) // it has to be added somewhere into the .trigger( 'submit' ); } /** returns the pagename, including the namespace name, but with spaces replaced by underscores */ function getArticleName() { return mw.config.get( 'wgPageName' ); } // don't run when not viewing articles const action = mw.config.get( 'wgAction' ); const isNotViewing = action != 'view'; if ( isNotViewing ) { return; } // don't run when viewing diffs const isDiff = mw.config.get( 'wgDiffNewId' ); if ( isDiff ) { return; } // Don't run in virtual namespaces const isVirtualNamespace = mw.config.get( 'wgNamespaceNumber' ) < 0; if ( isVirtualNamespace ) { return; } let menuID = 'p-navigation'; // @ts-ignore if ( window.draftCleanerPutInToolsMenu ) { menuID = 'p-tb'; // @ts-ignore } else if ( window.draftCleanerPutInMoreMenu ) { menuID = 'p-cactions'; } const titleWithNamespaceAndUnderscores = getArticleName(); const namespaceNumber = mw.config.get( 'wgNamespaceNumber' ); let running = false; // Add DraftCleaner to the toolbar mw.loader.using( [ 'mediawiki.util' ], () => { mw.util.addPortletLink( menuID, '#', 'Run DraftCleaner', 'DraftCleanerLink' ); $( '#DraftCleanerLink' ).on( 'click', async () => { // prevent running the script while script is already in progress if ( running ) { return; } running = true; mw.notify( 'Parsing page content...' ); // get page wikicode const titleWithNamespaceAndSpaces = titleWithNamespaceAndUnderscores.replace( /_/g, ' ' ); const originalWikicode = await getWikicode( titleWithNamespaceAndUnderscores ); let wikicode = originalWikicode; const dc = new DraftCleaner(); wikicode = dc.cleanDraft( wikicode, namespaceNumber, titleWithNamespaceAndSpaces ); const needsChanges = wikicode != originalWikicode; if ( needsChanges ) { const summary = 'clean up (DraftCleaner)'; await goToShowChangesScreen( titleWithNamespaceAndUnderscores, wikicode, summary ); } else { mw.notify( 'No changes needed!' ); } } ); } ); }() ); }); //
in drafts, these are usually poorly placed
/gm, '' );
. in drafts, these are usually poorly placed