// @name Wikipedia translation attribution checker
// @namespace https://en.wikipedia.org/
// @version 1.4
// @description Checks if a page is a potential unattributed translation, has suspicious access dates, or was created using the ContentTranslation tool.
// @author [[User:Vanderwaalforces]]
// @match https://en.wikipedia.org/wiki/*
// @match https://en.wikipedia.org/w/index.php?title=*
(function() {
'use strict';
// Ensure the script only runs in mainspace (0) or draftspace (118)
const namespace = mw.config.get('wgNamespaceNumber');
if (namespace !== 0 && namespace !== 118) return;
const apiUrl = "https://en.wikipedia.org/w/api.php";
const pageTitle = mw.config.get('wgPageName');
const talkPageTitle = "Talk:" + pageTitle;
// List of valid two-letter and three-letter language codes for interwiki links
const validLanguageCodes = [
'ab', 'ar', 'az', 'en', 'bn', 'de', 'fa', 'fi', 'fr', 'es', 'et', 'ru', 'ha', 'he', 'ig', 'is', 'id', 'it', 'ja', 'ka', 'lv', 'pt', 'ml', 'nl', 'pl', 'sk', 'sv', 'sq', 'tr', 'no', 'da', 'cs', 'ko', 'uk', 'zh',
'ace', 'arc', 'arz', 'ast', 'bat', 'bcl', 'bjn', 'bpy', 'bug', 'cbk', 'ceb', 'crh', 'csb', 'diq', 'dsb', 'eml',
'fiu', 'gag', 'glk', 'hif', 'hsb', 'ilo', 'jv', 'kab', 'kbd', 'ksh', 'lez', 'lmo', 'ltg', 'mai', 'map', 'mdf', 'mhr',
'min', 'mrj', 'mwl', 'nds', 'nov', 'nrm', 'pag', 'pam', 'pcm', 'pdc', 'pfl', 'pnb', 'roa', 'rue', 'sah', 'scn', 'skr',
'srn', 'szl', 'tpi', 'vec', 'vep', 'vls', 'war', 'wuu', 'xmf', 'yo', 'yue', 'zea'
];
// Function to fetch edit summaries and the first revision date (first 100 revisions)
function fetchEditSummaries() {
return new Promise((resolve, reject) => {
$.ajax({
url: apiUrl,
data: {
action: "query",
format: "json",
prop: "revisions",
titles: pageTitle,
rvprop: "comment|timestamp",
rvlimit: 100,
origin: "*"
},
success: function(data) {
const pages = data.query.pages;
const revisions = pages[Object.keys(pages)[0]].revisions;
const firstRevisionDate = revisions[revisions.length - 1].timestamp;
const comments = revisions.map(rev => rev.comment);
// Log the fetched comments and first revision date for debugging
console.log("Fetched edit summaries:", comments);
console.log("First revision date (from 100 revisions):", firstRevisionDate);
resolve({ comments, firstRevisionDate });
},
error: function(err) {
reject(err);
}
});
});
}
// Function to fetch the very first revision of the article (creation edit)
function fetchFirstRevision() {
return new Promise((resolve, reject) => {
$.ajax({
url: apiUrl,
data: {
action: "query",
format: "json",
prop: "revisions",
titles: pageTitle,
rvprop: "comment|tags|timestamp", // Include timestamp to get the creation date
rvdir: "newer", // Fetch from the oldest revision
rvlimit: 1, // Only fetch the first revision
origin: "*"
},
success: function(data) {
const pages = data.query.pages;
const revisions = pages[Object.keys(pages)[0]].revisions;
const firstRevision = revisions[0]; // Get the first revision
console.log("First revision fetched:", firstRevision);
resolve(firstRevision); // Now includes the creation timestamp
},
error: function(err) {
reject(err);
}
});
});
}
// Function to fetch article wikitext for citation checks
function fetchWikitext() {
return new Promise((resolve, reject) => {
$.ajax({
url: apiUrl,
data: {
action: "query",
format: "json",
prop: "revisions",
titles: pageTitle,
rvprop: "content",
origin: "*"
},
success: function(data) {
const pages = data.query.pages;
const pageData = pages[Object.keys(pages)[0]];
if (pageData.revisions && pageData.revisions[0]) {
const wikitext = pageData.revisions[0]['*'];
// Log the wikitext for debugging
console.log("Fetched wikitext:", wikitext);
resolve(wikitext);
} else {
console.log("No wikitext found");
resolve(null); // Return null if wikitext is missing
}
},
error: function(err) {
reject(err);
}
});
});
}
// Function to check if talk page contains the word "translat"
function fetchTalkPageContent() {
return new Promise((resolve, reject) => {
$.ajax({
url: apiUrl,
data: {
action: "query",
format: "json",
prop: "revisions",
titles: talkPageTitle,
rvprop: "content",
origin: "*"
},
success: function(data) {
const pages = data.query.pages;
const revisions = pages[Object.keys(pages)[0]].revisions;
if (revisions && revisions[0] && revisions[0]['*']) {
const talkPageContent = revisions[0]['*'].toLowerCase();
// Log the talk page content for debugging
console.log("Fetched talk page content:", talkPageContent);
resolve(talkPageContent.includes("translat"));
} else {
resolve(false);
}
},
error: function(err) {
reject(err);
}
});
});
}
// Helper function to check if a comment contains a valid interwiki link
function containsInterwikiLink(comment) {
const interwikiRegex = new RegExp(`\\b(${validLanguageCodes.join('|')}):`, 'i');
return interwikiRegex.test(comment);
}
// Helper function to check if an edit summary contains both "translat" or "import" and "from"
function containsTranslationKeywords(comment) {
const lowerComment = comment.toLowerCase();
const hasTranslatAndFrom = lowerComment.includes("translat") && lowerComment.includes("from");
const hasImportedAndFrom = lowerComment.includes("imported") && lowerComment.includes("from");
const hasImportingAndFrom = lowerComment.includes("importing") && lowerComment.includes("from");
// Log the status of each summary's keyword check
console.log(`Summary: ${comment}, Has translat + from: ${hasTranslatAndFrom}, Has imported + from: ${hasImportedAndFrom}, Has importing + from: ${hasImportingAndFrom}`);
return hasTranslatAndFrom || hasImportedAndFrom || hasImportingAndFrom;
}
// Function to classify edit summaries based on the refined algorithm
function classifyEditSummaries(editSummaries) {
let hasTranslatNoInterwiki = false;
let hasTranslatWithInterwiki = false;
editSummaries.forEach(summary => {
if (containsTranslationKeywords(summary)) {
const hasInterwiki = containsInterwikiLink(summary);
if (hasInterwiki) {
hasTranslatWithInterwiki = true;
} else {
hasTranslatNoInterwiki = true;
}
}
});
return { hasTranslatNoInterwiki, hasTranslatWithInterwiki };
}
// Improved date regex to handle various date formats
function parseDate(dateString) {
const regexes = [
/\b(\d{1,2})\s(Jan|Feb|Mar|Apr|May|Jun|Jul|Aug|Sep|Oct|Nov|Dec)[a-z]*\s(\d{4})\b/, // e.g. "21 January 2024"
/\b(\d{4})-(\d{2})-(\d{2})\b/, // e.g. "2024-01-21"
/\b(\d{1,2})-(\d{1,2})-(\d{4})\b/, // e.g. "21-01-2024"
/\b(Jan|Feb|Mar|Apr|May|Jun|Jul|Aug|Sep|Oct|Nov|Dec)\s(\d{1,2}),?\s(\d{4})\b/ // e.g. "January 21, 2024"
];
for (const regex of regexes) {
const match = regex.exec(dateString);
if (match) {
return new Date(match[0]);
}
}
return null; // Return null if no valid date format is found
}
// Function to check for suspicious access dates in the article's wikitext
function checkSuspiciousAccessDates(wikitext, firstRevisionDate) {
if (!wikitext) {
console.log("No wikitext to check for access dates");
return false;
}
const accessDateRegex = /\|\s*access[- ]date\s*=\s*([A-Za-z0-9, -]+)/g;
const matches = [...wikitext.matchAll(accessDateRegex)];
const suspiciousDates = [];
const firstRevisionParsed = new Date(firstRevisionDate);
const firstRevisionDateOnly = new Date(firstRevisionParsed.toISOString().split('T')[0]); // Strip time part
matches.forEach(match => {
const accessDate = match[1];
const accessDateParsed = parseDate(accessDate);
if (accessDateParsed) {
const accessDateOnly = new Date(accessDateParsed.toISOString().split('T')[0]); // Strip time part
// Check if the access date is strictly before the article's creation date
if (!isNaN(accessDateParsed.getTime()) && accessDateOnly < firstRevisionDateOnly) {
suspiciousDates.push(accessDate);
}
}
});
// Log any suspicious dates for debugging
console.log("Suspicious access dates:", suspiciousDates);
return suspiciousDates.length > 0;
}
// Function to display a message before the #contentSub element, with a dismiss button in the top-right corner
function displayMessage(text, color) {
const messageDiv = document.createElement("div");
messageDiv.style.backgroundColor = color;
messageDiv.style.color = "white";
messageDiv.style.padding = "15px";
messageDiv.style.textAlign = "center";
messageDiv.style.fontSize = "16px";
messageDiv.style.fontWeight = "bold";
messageDiv.style.position = "relative"; // Required for dismiss button positioning
messageDiv.innerHTML = text; // Use innerHTML for parsing link elements
// Add dismiss button at the very top-right corner
const dismissButton = document.createElement("button");
dismissButton.textContent = "x";
dismissButton.style.position = "absolute";
dismissButton.style.top = "5px";
dismissButton.style.right = "10px";
dismissButton.style.backgroundColor = "#ff5f5f";
dismissButton.style.border = "none";
dismissButton.style.color = "white";
dismissButton.style.padding = "5px 10px";
dismissButton.style.cursor = "pointer";
dismissButton.style.fontWeight = "bold";
dismissButton.style.borderRadius = "5px";
dismissButton.onclick = () => {
messageDiv.style.display = "none";
};
messageDiv.appendChild(dismissButton);
// Insert the message before the #contentSub element to make it compatible with all skins
$('#contentSub').before(messageDiv);
}
// Main logic
async function checkTranslationAttribution() {
try {
// Fetch the first revision to check for ContentTranslation tool and creation date
const firstRevision = await fetchFirstRevision();
const firstComment = firstRevision.comment;
const firstTags = firstRevision.tags;
const firstRevisionDate = firstRevision.timestamp; // Fetch the timestamp of the first revision
// Define the constants for links to WP:TFOLWP and the Translated from template
const wpShortcutLink = mw.html.element('a', {
href: '/wiki/WP:TFOLWP',
target: '_blank',
style: 'color:white; text-decoration:underline;'
}, '[[WP:TFOLWP]]');
const templateLink = mw.html.element('a', {
href: '/wiki/Template:Translated_from',
target: '_blank',
style: 'color:white; text-decoration:underline;'
}, '{{Translated from}}');
// If the first edit used the ContentTranslation tool, display the green banner and skip other checks
if (firstComment.includes("Created by translating the page") && firstTags.includes("contenttranslation")) {
const contentTranslationLink = mw.html.element('a', {
href: 'https://www.mediawiki.org/wiki/Special:MyLanguage/Content_translation',
target: '_blank',
style: 'color:white; text-decoration:underline;'
}, 'ContentTranslation');
displayMessage(`This article was created using the ${contentTranslationLink} module and is correctly attributed to the source Wikipedia.`, "green");
return; // Skip all other checks if ContentTranslation tool was used
}
// Continue with other checks for suspicious access dates and translation attribution
const { comments } = await fetchEditSummaries(); // Use only for edit summaries now
const { hasTranslatNoInterwiki, hasTranslatWithInterwiki } = classifyEditSummaries(comments);
const wikitext = await fetchWikitext();
const hasSuspiciousAccessDates = checkSuspiciousAccessDates(wikitext, firstRevisionDate); // Use actual creation date
// Display singular orange banner if there are suspicious access dates (no translation detected)
if (hasSuspiciousAccessDates && !hasTranslatNoInterwiki && !hasTranslatWithInterwiki) {
displayMessage("Warning: There are citations in this article that have access dates from before the article was created. This suggests the article may have been copy-pasted from somewhere.", "orange");
return; // Only show the orange banner and stop further checks
}
// If suspicious dates are found and the article is correctly attributed
if (hasSuspiciousAccessDates && hasTranslatWithInterwiki) {
displayMessage("Notice: Despite some citations having access dates before the article's creation, indicating possible copy-pasting or interwiki translation, proper attribution has been given.", "green");
}
// If suspicious dates are found and no proper attribution
else if (hasSuspiciousAccessDates && hasTranslatNoInterwiki) {
displayMessage(`Warning: This article is likely an unattributed translation. Please see ${wpShortcutLink} for proper attribution, and consider adding ${templateLink} to the talk page.`, "red");
displayMessage("Warning: There are citations in this article that have access dates from before the article was created. This suggests the article may have been copy-pasted from somewhere.", "orange");
}
// If there are no suspicious dates and the article is correctly attributed
else if (!hasSuspiciousAccessDates && hasTranslatWithInterwiki) {
const hasTranslatInTalkPage = await fetchTalkPageContent();
if (!hasTranslatInTalkPage) {
displayMessage(`Notice: This translated article has been correctly attributed. Consider optionally adding ${templateLink} to the talk page.`, "green");
} else {
displayMessage("Notice: This translated article has been correctly attributed.", "green");
}
}
// If there are no suspicious dates and no proper attribution
else if (!hasSuspiciousAccessDates && hasTranslatNoInterwiki) {
displayMessage(`Warning: This article is likely an unattributed translation. Please see ${wpShortcutLink} for proper attribution, and consider adding ${templateLink} to the talk page.`, "red");
}
} catch (error) {
console.error("Error checking translation attribution:", error);
}
}
// Run the check
checkTranslationAttribution();
})();
