dothething.js

import { By, Builder } from "selenium-webdriver";
import chrome from 'selenium-webdriver/chrome.js';

/**
 * Extract target links from the given cards.
 * @param {WebElement[]} cards - elements of CSS selector `.card`
 * @returns {Promise<string[]>} - target links
 */
async function extractWeeklySalesLinks(cards) {
    const weeklySalesLinks = [];
    for (const card of cards) {
        const text = await card.getText();
        if (text.includes('【一週99書單】')) {
            const link = await card.findElement(By.css('a.card__link')).getAttribute('href');
            weeklySalesLinks.push(link);
        }
    }
    if (weeklySalesLinks.length === 0) {
        throw new Error('No weekly sales links found. Possible reasons: 1. The page structure has changed; 2. The target text "【一週99書單】" has changed; 3. The latest weekly post is not on page 1 anymore.');
    }
    return weeklySalesLinks;
}


/**
 * Format a date string in the format "M/D" into a string in the format "YYYY-MM-DD"
 * @param {string} dateStr - date string in the format "M/D"
 * @returns {string} - formatted date string in the format "YYYY-MM-DD"
 */
function formatDate(dateStr) {
    const currentYear = new Date().getFullYear();
    const [month, day] = dateStr.split('/').map(num => num.padStart(2, '0'));
    return `${currentYear}-${month}-${day}`;
}

/**
 * Extract daily deals from the given content blocks.
 * @param {WebElement[]} blocks - elements of CSS selector `.book__info`
 * @returns {Promise<Object[]>} - array of daily deals, each as an object with properties:
 *   - date: {string} - date string in the format "YYYY-MM-DD"
 *   - title: {string} - title of the book
 *   - author: {string} - author of the book
 *   - salesCopy: {string} - sales copy of the book
 *   - link: {string} - link to the book page
 *   - bookCover: {string} - link to the book cover image
 */
async function extractDailyDeals(blocks) {
    const dailyDeals = [];
    const booksOf8 = [];
    [0, 2, 4, 6, 8, 10, 12, 14].forEach(i => {
        booksOf8.push([blocks[i], blocks[i + 1]]);
    });
    for (const book of booksOf8) {
        const deal = {};
        const titleLine = await book[0].findElement(By.css('h3')).getText();
        deal.date = formatDate(titleLine.split('週').shift());
        const title = await book[0].findElement(By.css('a')).getText();
        deal.title = title;
        const author = await book[1].findElement(By.css('.author')).getText();
        deal.author = author.slice(2, -2);
        const salesCopy = await book[0].findElement(By.css('p')).getText();
        deal.salesCopy = salesCopy;
        const link = await book[0].findElement(By.css('a')).getAttribute('href');
        const parsedUrl = new URL(link);
        deal.link = parsedUrl.origin + parsedUrl.pathname;
        const bookCover = await book[1].findElement(By.css('img')).getAttribute('src');
        deal.bookCover = bookCover;
        dailyDeals.push(deal);
    }
    return dailyDeals;
}

/**
 * Extract the ISBN (Book ID) from the metadata.
 * @param {WebElement[]} metadata - elements containing book metadata
 * @returns {Promise<string>} - extracted ISBN as a string, or an empty string if not found
 * 
 * The metadata is expected to contain lines of text, one of which includes
 * the '書籍ID' identifier followed by the ISBN. This function searches through
 * the metadata to find and return the ISBN.
 */

async function extractISBN(metadata) {    
    for (const line of metadata) {
        const text = await line.getText();
        if (text.includes('書籍ID')) {
            return text.split('：').pop().trim();
        }
    }
    // If ISBN not found after checking all lines, return empty string.
    // Errors during getText() or other operations will propagate up.
    return "";
}

(async () => {
    console.log("Starting the job");
    let driver;
    try {
        // Determine whether to load images based on environment variable
        const loadImages = process.env.LOAD_IMAGES !== 'false'; // Default to true if not set or not 'false'
        const chromeOptions = new chrome.Options();
        if (!loadImages) {
            chromeOptions.addArguments('--blink-settings=imagesEnabled=false');
            chromeOptions.addArguments('--disable-images');
            console.log("Image loading disabled.");
        } else {
            console.log("Image loading enabled.");
        }
        driver = await new Builder()
            .forBrowser('chrome')
            .setChromeOptions(chromeOptions)
            .build();
        console.log("Navigating to the Kobo blog page");
        await driver.get('https://www.kobo.com/zh/blog');
        const cards = await driver.findElements(By.css('.card'));
        if (!cards || cards.length === 0) {
            throw new Error('No cards found on the blog page.');
        }
        console.log("Extracting blog posts on the page");
        const weeklySalesLinks = await extractWeeklySalesLinks(cards);
        console.log("Navigating to the latest Weekly Sales link");
        await driver.get(weeklySalesLinks.shift());
        console.log("Latest Weekly Sales:", await driver.getTitle());

        const blocks = await driver.findElements(By.css('.content-block, .book-block'));
        if (!blocks || blocks.length < 16) {
            // Specific check for expected content structure
            throw new Error(`Error collecting post content. Expected at least 16 blocks. Found ${blocks.length}. Please check the blog page structure.`);
        }
        blocks.shift(); // Remove the first element because it's a header
        const dailyDeals = await extractDailyDeals(blocks);

        //Amending dailyDeals with the Kobo Book ID
        for (const deal of dailyDeals) {
            console.log("Navigating to the individual book page to retrieve Book ID: " + deal.title);
            await driver.get(deal.link);
            const metadata = await driver.findElements(By.css('.bookitem-secondary-metadata > ul > li'));
            const isbn = await extractISBN(metadata);
            if (!isbn) {
                // Throw an error if ISBN is not found, as it's considered fatal.
                throw new Error(`Fatal: ISBN not found for book: ${deal.title} (${deal.link})`);
            }
            deal.isbn = isbn;
        }
        console.log("Daily Deals:");
        console.log(JSON.stringify(dailyDeals, null, 2));

    } catch (err) {
        // Centralized error handling
        console.error('An error occurred during script execution:', err);
        // Attempt to quit driver before exiting
        if (driver) {
            try {
                console.log("Attempting to quit the web driver after error...");
                await driver.quit();
            } catch (quitErr) {
                console.error('Error quitting driver after script error:', quitErr);
            }
        }
        process.exit(1); // Exit with error code
    } finally {
        // Final check to ensure driver is quit if it still exists (e.g., error occurred before main try block)
        if (driver) {
            try {
                console.log("Ensuring web driver is quit in finally block...");
                await driver.quit();
            } catch (quitErr) {
                console.error('Error quitting driver in final finally block:', quitErr);
                // Exit even if quit fails in finally, as an error already occurred or script finished.
                if (process.exitCode !== 1) { // Avoid double exit if already exiting due to error
                   process.exit(1);
                }
            }
        }
    }
    console.log("Job finished successfully."); // Log success if no errors caused an exit
})();
// Removed the outer try-catch as the main one now handles initialization errors too.
// The IIFE remains for structure.