MapComplete/scripts/downloadCommons.ts

/**
 * Script to download images from Wikimedia Commons, and save them together with license information.
 */

import { existsSync, mkdirSync, readFileSync, writeFileSync } from "fs"
import { unescape } from "querystring"
import SmallLicense from "../src/Models/smallLicense"
import Wikimedia, { ImageQueryAPIResponse } from "../src/Logic/Web/Wikimedia"

interface CategoryMember {
    pageid: number
    ns: number
    title: string
}

interface CategoryQueryAPIResponse {
    batchcomplete: string
    query: {
        categorymembers: CategoryMember[]
    }
}

interface ImagesQueryAPIResponse {
    continue: {
        imcontinue: string
        continue: string
    }
    query: {
        normalized?: {
            from: string
            to: string
        }[]
        pages: {
            [key: string]: {
                pageid: number
                ns: number
                title: string
                images?: {
                    ns: number
                    title: string
                }[]
            }
        }
    }
}

export async function main(args: string[]) {
    if (args.length < 2) {
        console.log("Usage: downloadCommons.ts <output folder> <url> <?url> <?url> .. ")
        console.log(
            "Example: npx vite-node scripts/downloadCommons.ts -- assets/svg https://commons.wikimedia.org/wiki/File:Example.jpg"
        )
        process.exit(1)
    }
    const [outputFolder, ...urls] = args

    for (const url of urls) {
        // Download details from the API
        const commonsFileNamePath = url.split("/").pop()
        if (commonsFileNamePath === undefined) {
            console.log(
                "\x1b[31m%s\x1b[0m",
                `URL ${url} doesn't seem to be a valid URL! Skipping...`
            )
            continue
        }

        const commonsFileName = commonsFileNamePath.split("?").shift()

        if (commonsFileName === undefined) {
            console.log(
                "\x1b[31m%s\x1b[0m",
                `URL ${url} doesn't seem to contain a filename or category! Skipping...`
            )
            continue
        }

        console.log(`Processing ${commonsFileName}...`)

        const baseUrl = url.split("/").slice(0, 3).join("/")

        // Check if it is a file or a category
        if (url.includes("Category:")) {
            // Download all files in the category
            const apiUrl = `${baseUrl}/w/api.php?action=query&format=json&list=categorymembers&cmtitle=${commonsFileName}&cmlimit=250&cmtype=file`
            const response = await fetch(apiUrl)
            const apiDetails: CategoryQueryAPIResponse = await response.json()
            for (const member of apiDetails.query.categorymembers) {
                await downloadImage(member.title, outputFolder, baseUrl)
            }
        } else if (url.includes("File:")) {
            await downloadImage(commonsFileName, outputFolder, baseUrl)
        } else {
            // Probably a page url, try to get all images from the page
            const apiUrl = `${baseUrl}/w/api.php?action=query&format=json&prop=images&titles=${commonsFileName}&imlimit=250`
            const response = await fetch(apiUrl)
            const apiDetails: ImagesQueryAPIResponse = await response.json()
            const page = apiDetails.query.pages[Object.keys(apiDetails.query.pages)[0]]
            if (page.images) {
                for (const image of page.images) {
                    await downloadImage(image.title, outputFolder, baseUrl)
                }
            } else {
                console.log(
                    "\x1b[31m%s\x1b[0m",
                    `URL ${url} doesn't seem to contain any images! Skipping...`
                )
            }
        }
    }
}

async function downloadImage(filename: string, outputFolder: string, baseUrl: string) {
    const apiUrl = `${baseUrl}/w/api.php?action=query&format=json&prop=imageinfo&iiprop=url|extmetadata|user&iimetadataversion=latest&titles=${filename}`
    const response = await fetch(apiUrl)
    const apiDetails: ImageQueryAPIResponse = await response.json()
    const missingPage = apiDetails.query.pages["-1"]

    // Check if the local file already exists, if it does, skip it
    if (existsSync(`${outputFolder}/${filename}`)) {
        console.log(`\x1b[33m%s\x1b[0m`, `${filename} already exists, skipping...`)
        return
    }

    // Check if the file exists, locally or externally
    if (missingPage !== undefined) {
        // Image does not exist locally, check if it exists externally
        if (
            apiDetails.query.pages["-1"].imagerepository !== "local" &&
            apiDetails.query.pages["-1"].imagerepository !== ""
        ) {
            // Check if we actually have image info
            if (missingPage.imageinfo?.length !== undefined && missingPage.imageinfo.length > 0) {
                const externalUrl = missingPage.imageinfo[0].descriptionurl
                const externalBase = externalUrl.split("/").slice(0, 3).join("/")

                const externalFilenamePath = externalUrl.split("/").pop()
                if (externalFilenamePath !== undefined) {
                    const externalFilename = externalFilenamePath.split("?").shift()
                    console.log(
                        `\x1b[33m%s\x1b[0m`,
                        `${filename} is external, re-running with ${externalUrl}...`
                    )
                    if (externalFilename !== undefined) {
                        await downloadImage(externalFilename, outputFolder, externalBase)
                        return
                    } else {
                        // Edge case
                        console.log(
                            `\x1b[33m%s\x1b[0m`,
                            `External URL ${externalUrl} doesn't seem to contain a filename or category! Skipping...`
                        )
                    }
                } else {
                    // Edge case
                    console.log(
                        `\x1b[33m%s\x1b[0m`,
                        `External URL ${externalUrl} doesn't seem to be a valid URL! Skipping...`
                    )
                    return
                }
            } else {
                console.log(
                    `\x1b[33m%s\x1b[0m`,
                    `${filename} does not have image info!, skipping...`
                )
            }
        }
        console.log(`\x1b[33m%s\x1b[0m`, `${filename} does not exist!, skipping...`)
    } else {
        // Harvest useful information
        const wikiPage = apiDetails.query.pages[Object.keys(apiDetails.query.pages)[0]]

        // Check if we actually have image info
        if (wikiPage.imageinfo?.length !== undefined && wikiPage.imageinfo.length > 0) {
            // Check if the output folder exists
            if (!existsSync(outputFolder)) {
                const parts = outputFolder.split("/")
                for (let i = 0; i < parts.length; i++) {
                    const part = parts.slice(0, i + 1).join("/")
                    if (!existsSync(part)) {
                        console.log(`Creating folder ${part}`)
                        mkdirSync(part)
                    }
                }
            }

            // Download the file and save it
            const cleanFileName = unescape(filename).replace("File:", "")
            const fileUrl = wikiPage.imageinfo[0].url
            console.log(
                `Downloading ${cleanFileName} from ${fileUrl} and saving it to ${outputFolder}/${cleanFileName}...`
            )
            const fileResponse = await fetch(fileUrl)
            const fileBuffer = await fileResponse.arrayBuffer()
            const file = Buffer.from(fileBuffer)
            const filePath = `${outputFolder}/${cleanFileName}`
            writeFileSync(filePath, file)

            // Save the license information
            const licenseInfo: SmallLicense = await Wikimedia.getLicenseFor(baseUrl, filename)

            const licensePath = `${outputFolder}/license_info.json`
            if (!existsSync(licensePath)) {
                // Create the file if it doesn't exist
                writeFileSync(licensePath, JSON.stringify([licenseInfo], null, 2))
            } else {
                // Append to the file if it does exist
                const licenseFile = await readFileSync(licensePath, "utf8")
                const licenseData = JSON.parse(licenseFile)
                licenseData.push(licenseInfo)
                writeFileSync(licensePath, JSON.stringify(licenseData, null, 2))
            }
        } else {
            console.log(`\x1b[33m%s\x1b[0m`, `${filename} does not have image info!, skipping...`)
        }
    }
}

// main(process.argv.slice(2))