Fix Mitaku

Author: TheNetsky

.gitignore

@@ -17,3 +17,5 @@ docs
 # mac stuff
 .DS_Store
 package-lock.json
+notes
+/ignore_this

src/Mitaku/Mitaku.ts

@@ -10,7 +10,6 @@ import {
     BadgeColor,
     Request,
     Response,
-    TagSection,
     SourceIntents,
     HomeSectionType,
     ChapterProviding,
@@ -198,29 +197,6 @@ export class Mitaku implements SearchResultsProviding, MangaProviding, ChapterPr
         })
     }
 
-    async getSearchTags(): Promise<TagSection[]> {
-        const request = App.createRequest({
-            url: `${MT_DOMAIN}/wp-json/wp/v2/tags?per_page=100`,
-            method: 'GET'
-        })
-
-        const response = await this.requestManager.schedule(request, 1)
-        const tagJSON = JSON.parse(response.data as string)
-
-        const arrayTags: Tag[] = []
-
-        for (const tag of tagJSON) {
-            const label = tag.name
-            const id = tag.slug
-
-            if (!id || !label) continue
-            arrayTags.push({ id: id, label: label })
-        }
-
-        const tagSections: TagSection[] = [App.createTagSection({ id: '0', label: 'genres', tags: arrayTags.map(x => App.createTag(x)) })]
-        return tagSections
-    }
-
     async getSearchResults(query: SearchRequest, metadata: any): Promise<PagedResults> {
         const page: number = metadata?.page ?? 1
 

src/Mitaku/MitakuParser.ts

@@ -16,20 +16,24 @@ import {
 
 import { type Element } from 'domhandler'
 
+const MT_DOMAIN = 'https://mitaku.net'
+
 export const parseMangaDetails = ($: CheerioAPI, mangaId: string): SourceManga => {
     const images = parseImages($)
 
-    const title = decodeHTMLEntity($('h1.cm-entry-title').first().text().trim())
+    const title = decodeHTMLEntity($('.cm-entry-title').first().text().trim())
     const artist = decodeHTMLEntity($('p:contains(Cosplayer:)').text().trim().replace('Cosplayer:', '').trim())
     const description = `Cosplayer: ${artist}\n\nGallery: ${title}\n\nImages: ${images.length}`
 
 
     const arrayTags: Tag[] = []
-    for (const tag of $('a', 'span.tag-links').toArray()) {
-        const id = $(tag).attr('href')?.split('/tag/')[1]?.replace(/\//g, '')
-        const label = $(tag).text().trim()
-        if (!id || !label) continue
-        arrayTags.push({ id: id, label: label })
+    for (const obj of $('div.genres-content a').toArray()) {
+        const id = idCleaner($(obj).attr('href') ?? '')
+        const title = $(obj).text().trim()
+
+        if (!title || !id) continue
+
+        arrayTags.push({ label: title, id: id })
     }
     const tagSections: TagSection[] = [App.createTagSection({ id: '0', label: 'genres', tags: arrayTags.map(x => App.createTag(x)) })]
 
@@ -70,14 +74,17 @@ export const parseHomeSections = ($: CheerioAPI): PartialSourceManga[] => {
     const collectedIds: string[] = []
     const itemArray: PartialSourceManga[] = []
 
-    for (const item of $('article', 'div#cm-primary').toArray()) {
+    for (const item of $('article').toArray()) {
         const postId = $(item).attr('id')
         const id = postId?.split('post-').pop()
+
         const image: string = getImageSrc($('img', item).first()) ?? ''
-        const title: string = $('h2.cm-entry-title', item).text().trim() ?? ''
-        const subtitle: string = $('span.cm-tag-links > a', item).toArray().map(x => $(x).text().trim()).join(', ')
+        const title: string = $('a', item).first().attr('title')?.trim() ?? ''
+
+        const subtitle = $('a[rel="tag"]', item).map((i, el) => $(el).text().trim()).get().join(', ')
+
+        if (!id || isNaN(Number(id)) || !title || !subtitle) continue
 
-        if (!id || isNaN(Number(id)) || !title || collectedIds.includes(id) || !subtitle) continue
         itemArray.push(App.createPartialSourceManga({
             image: encodeURI(image),
             title: decodeHTMLEntity(title),
@@ -105,27 +112,57 @@ const parseImages = ($: CheerioAPI): string[] => {
     return images
 }
 
-const getImageSrc = (imageObj: Cheerio<Element>): string => {
+// Utils
+const getImageSrc = (imageObj: Cheerio<Element> | undefined): string => {
     let image: string | undefined
-    if ((typeof imageObj?.attr('data-src')) != 'undefined') {
-        image = imageObj?.attr('data-src')
-    }
-    else if ((typeof imageObj?.attr('data-lazy-src')) != 'undefined') {
-        image = imageObj?.attr('data-lazy-src')
-    }
-    else if ((typeof imageObj?.attr('srcset')) != 'undefined') {
-        image = imageObj?.attr('srcset')?.split(' ')[0] ?? ''
+    const sources = [
+        'data-src',
+        'data-lazy-src',
+        'srcset',
+        'src',
+        'data-cfsrc'
+    ]
+
+    for (const attr of sources) {
+        const val = imageObj?.attr(attr)
+
+        if (val == null || val.trim() === '') continue
+
+        // If it's srcset, extract the first URL
+        if (attr === 'srcset') {
+            image = val.split(',')[0]?.trim().split(' ')[0] ?? ''
+        } else {
+            image = val
+        }
+
+        break
     }
-    else if ((typeof imageObj?.attr('src')) != 'undefined') {
-        image = imageObj?.attr('src')
-    }
-    else if ((typeof imageObj?.attr('data-cfsrc')) != 'undefined') {
-        image = imageObj?.attr('data-cfsrc')
-    } else {
-        image = ''
+
+    image = image?.replace(/-\d+x\d+/g, '')
+
+    if (image?.startsWith('/')) {
+        image = MT_DOMAIN + image
     }
 
-    return encodeURI(decodeURI(decodeHTMLEntity(image?.trim() ?? '')))
+    image = image
+        ?.trim()
+        .replace(/(\s{2,})/gi, '')
+
+    image = image?.replace(/http:\/\/\//g, 'http://') // only changes urls with http protocol
+    image = image?.replace(/http:\/\//g, 'https://')
+    // Malforumed url fix (Turns https:///example.com into https://example.com (or the http:// equivalent))
+    image = image?.replace(/https:\/\/\//g, 'https://') // only changes urls with https protocol
+
+    return decodeURI(decodeHTMLEntity(image ?? ''))
+}
+
+const idCleaner = (str: string): string => {
+    let cleanId: string | null = str
+    cleanId = cleanId.replace(/\/$/, '')
+    cleanId = cleanId.split('/').pop() ?? null
+
+    if (!cleanId) throw new Error(`Unable to parse id for ${str}`)
+    return cleanId
 }
 
 export const isLastPage = ($: CheerioAPI): boolean => {