User:Opencooper/showKanji-dev.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:Opencooper/showKanji-dev.

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.

// This script shows, if found, the kanji and kana for an article
// For configuration, please see the documentation

/* Sample pages:
    https://en.wikipedia.org/wiki/Tamio_Kawachi - kana on wikidata
    https://en.wikipedia.org/wiki/A_Fantastic_Tale_of_Naruto - kanji from wikidata only
    https://en.wikipedia.org/wiki/What_a_Wonderful_World! - kana from wikidata only
    https://en.wikipedia.org/wiki/Asako_I_%26_II - from redirect
    https://en.wikipedia.org/wiki/Bokura_ga_Ita_(film) - interwiki to subsection
    https://ja.wikipedia.org/wiki/%E7%B4%AF - kana part of bolded title
    https://en.wikipedia.org/wiki/Bokutachi_no_Koukan_Nikki - kana not in first sentence
    https://ja.wikipedia.org/wiki/SCP%E8%B2%A1%E5%9B%A3 - bolded term w/ kana past first sentence
    https://en.wikipedia.org/wiki/Nuclear_fusion - different term w/ kana in lead
    https://en.wikipedia.org/wiki/20th_Century_Boys - overcapturing because title is subset
    https://en.wikipedia.org/wiki/Kanji_Furutachi - kanji only
    https://en.wikipedia.org/wiki/Anata_e - hiragana only
    https://en.wikipedia.org/wiki/Anatahan_(film) - katakana only
    https://en.wikipedia.org/wiki/A.LI.CE - latin only
    https://en.wikipedia.org/wiki/0.5_mm - numeric
    https://en.wikipedia.org/wiki/Comic_Magazine - exclamation point
    https://en.wikipedia.org/wiki/Flare_(film) - wave dash
    https://en.wikipedia.org/wiki/Dog%C3%97Police - multiplication sign
    https://en.wikipedia.org/wiki/Foreboding_(film) - spaces
    https://en.wikipedia.org/wiki/Age_12 - period in title
    https://en.wikipedia.org/wiki/After_the_Rain_(manga) - kanji + hiragana
    https://en.wikipedia.org/wiki/Afro_Tanaka - kanji + katakana
    https://en.wikipedia.org/wiki/Battle_Girl:_The_Living_Dead_in_Tokyo_Bay - katakana + latin
    https://en.wikipedia.org/wiki/Calling_You_(short_story_collection) - kanji + hiragana + latin
    https://en.wikipedia.org/wiki/Ashita_no_Joe - hiragana + katakana
    https://en.wikipedia.org/wiki/Arcadia_of_My_Youth - kanji + hiragana + katakana
*/

function setup() {
    // If we're not reading an article, do nothing
    if (!(mw.config.get( 'wgAction' ) === 'view'
          && mw.config.get( 'wgIsArticle' )
          && !___location.search.split('oldid=')[1]
          && mw.config.get("wgPageName") !== "Main_Page")) {
        return;
    }

    // Assuming that if there's no wikidata, there're no 1:1 interlanguage links,
    // and we don't want cases where a page links to a subsection of a jawiki
    // article
    if (wikidataId === null) {
        return;
    }

    // Placeholder so other elements don't push it down later
    $('#firstHeading').append("<div id='kanjiInfo' lang='ja'></div>");

    // Get the Japanese label from wikidata
    // API docs: https://www.wikidata.org/w/api.php?action=help&modules=wbgetentities
    $.ajax({
        url: "https://www.wikidata.org/w/api.php",
        data: {
            action: "wbgetentities",
            ids: wikidataId,
            props: "labels",
            languages: "ja",
            format: "json",
            origin: "*"
        },
        success: parseJaLabel
    });
}

function parseJaLabel(response) {
    var wikidataInfo = response.entities[wikidataId];
    var jaLabel;
    if (!jQuery.isEmptyObject(wikidataInfo.labels.ja)) {
        jaLabel = wikidataInfo.labels.ja.value;
    }

    if (jaLabel) {
        displayKanji(jaLabel);
    } else {
        return;
    }

    // If the japanese title is not just only kana, get the reading
    if (!kanaOnlyRe.test(jaLabel)) {
        requestKana();
    }
}

function displayKanji(kanji) {
    $('#kanjiInfo').append("<ruby>" + kanji + "</ruby>");

    // Add some classes so users can choose to not display for example
    // katakana-only kanji in their CSS
    if (latinOnlyRe.test(kanji)) {
        $("#kanjiInfo").addClass("kanjiInfo-latin-only");
        $("#kanjiInfo").css("display", "none");
    } else if (hiraganaOnlyRe.test(kanji)) {
        $("#kanjiInfo").addClass("kanjiInfo-hiragana-only");
    } else if (katakanaOnlyRe.test(kanji)) {
        $("#kanjiInfo").addClass("kanjiInfo-katakana-only");
    }
}

function requestKana() {
    // API docs: https://www.wikidata.org/w/api.php?action=help&modules=wbgetclaims
    // We have to wholesale get all the claims instead of just one because the
    // kana might be present as a qualifier to another claim
    $.ajax({
        url: "https://www.wikidata.org/w/api.php",
        data: {
            action: "wbgetclaims",
            entity: wikidataId,
            format: "json",
            origin: "*"
        },
        success: parseKanaClaim
    });
}

function parseKanaClaim(response) {
    var kana;
    if (response.claims.P1814) {
        // name in kana
        kana = response.claims.P1814[0].mainsnak.datavalue.value;
    } else if (response.claims.P1476 && response.claims.P1476[0].qualifiers
               && response.claims.P1476[0].qualifiers.P1814) {
        // title w/ name in kana
        kana = response.claims.P1476[0].qualifiers.P1814[0].datavalue.value;
    } else if (response.claims.P1705 && response.claims.P1705[0].qualifiers
               && response.claims.P1705[0].qualifiers.P1814) {
        // native label w/ name in kana
        kana = response.claims.P1705[0].qualifiers.P1814[0].datavalue.value;
    } else {
        scrapeKana();
        return;
    }

    displayKana(kana);
    $("#kanjiInfo rt").addClass("kanjiInfo-wikidata");
}

function scrapeKana() {
    var jaLabel;
    if ($(".interwiki-ja").length) {
        jaLabel = getInterlanguageJa();
    } else {
        return;
    }

    // Get jawiki article's lead wikitext
    // API docs: https://www.mediawiki.org/wiki/API:Revisions
    $.ajax({
        url: "https://ja.wikipedia.org/w/api.php",
        data: {
            action: "query",
            prop: "revisions",
            format: "json",
            titles: jaLabel,
            redirects: "true",
            rvprop: "content",
            rvsection: "0",
            rvslots: "main",
            origin: "*"
        },
        success: parseJaLead
    });
}

function getInterlanguageJa() {
    var jaLabel = $(".interwiki-ja .interlanguage-link-target").attr("href");
    jaLabel = jaLabel.split("\/wiki\/")[1];
    jaLabel = decodeURIComponent(jaLabel);
    jaLabel = jaLabel.replace(/(.*)#.*/, "$1"); // rm anchors
    return jaLabel;
}

function parseJaLead(response) {
    var responsePart = response.query.pages;
    // Have to split parsing into two parts since jawiki pageid is unknown
    var pageId = Object.keys(responsePart)[0];
    var leadText = responsePart[pageId].revisions[0].slots.main["*"];

    // Only first sentence
    var introSearch = leadText.match(/'''.*?。/);
    var wikitext;
    if (introSearch) {
        wikitext = introSearch[0];
    } else {
        return;
    }
    wikitext = wikitext.replace(/\{\{.*?\}\}/g, "{{}}"); // Remove templates

    var kana;
    var kanji = $("#kanjiInfo").text();
    var kanjiEscaped = mw.RegExp.escape(kanji);
    kanjiEscaped = kanjiEscaped.replace(/\B/g, " ?"); // Account for spaces
    // Add kanji to regex to make sure we're not getting the reading of some
    // other term
    var leadKanaRe = new RegExp(kanjiEscaped + ".*?" + leadRe);

    console.log("showKanji-dev.js: lead: " + wikitext);
    console.log("showKanji-dev.js: kanji: " + kanji);
    console.log("showKanji-dev.js: regex: " + leadKanaRe);

    var kanaSearch = wikitext.match(leadKanaRe);
    if (kanaSearch && kanaSearch.length == 2) {
        kana = kanaSearch[1];
    } else {
        return;
    }

    // Rm trailing characters
    kana = kana.replace(/[・、]$/, "");

    displayKana(kana);
    $("#kanjiInfo rt").addClass("kanjiInfo-jawiki");
}

function displayKana(kana) {
    $("#kanjiInfo ruby").append("<rt>" + kana + "</rt>");
}

var wikidataId = mw.config.get( 'wgWikibaseItemId' );
// Regexes
var latinOnlyRe = /^[A-Za-z0-9\-.?!,:;@#$%&+=*'" ]+$/;
// Stolen from https://www.wikidata.org/wiki/Property:P1814#P1793
var extrasRe = "0-9「」・、 　\\-.?!！~〜×";
var kanaOnlyRe = new RegExp("^[ぁ-ゔァ-ヴー" + extrasRe + "]+$");
var hiraganaOnlyRe = new RegExp("^[ぁ-ゔーA-Za-z" + extrasRe + "]+$");
var katakanaOnlyRe = new RegExp("^[ァ-ヴーA-Za-z" + extrasRe + "]+$");
var leadRe = "'''.*?（([ぁ-ゔァ-ヴー" + extrasRe + "]+)"; // brittle
$(setup);