initial commit

.dockerignore

@@ -0,0 +1 @@
+volume/

.gitignore

@@ -0,0 +1,4 @@
+node_modules
+volume/browser
+db
+downloads

Dockerfile

@@ -0,0 +1,20 @@
+# Use official Node.js image as base
+FROM node:18
+
+# Create app directory
+WORKDIR /usr/src/app
+
+# Copy package.json and package-lock.json (if available)
+COPY package*.json ./
+
+# Install dependencies
+RUN npm install
+
+# Copy app source code
+COPY . .
+
+# Expose the port your app runs on (optional, depends on your app)
+EXPOSE 3000
+
+# Start the app
+CMD ["npm", "start"]

cron-runner.js

@@ -0,0 +1,22 @@
+const cron = require('node-cron');
+const { exec } = require('child_process');
+
+function runScraper() {
+  console.log('Running scraper...');
+  exec('node scrape/scraper.js', (error, stdout, stderr) => {
+    if (error) {
+      console.error(`Scraper error: ${error.message}`);
+      return;
+    }
+    if (stderr) console.error(`Scraper stderr: ${stderr}`);
+    if (stdout) console.log(`Scraper output:\n${stdout}`);
+  });
+}
+
+// Run immediately at start
+runScraper();
+
+// Schedule to run every 10 minutes
+cron.schedule('*/10 * * * *', runScraper);
+
+console.log('Cron scheduler started. Scraper will run every 10 minutes.');

docker-compose.yml

@@ -0,0 +1,12 @@
+version: "3.9"
+
+services:
+  app:
+    build: .
+    ports:
+      - "3000:3000"
+    environment:
+      NODE_ENV: development
+    volumes:
+      - ./volume:./usr/src/app/volume
+    command: npm start

package.json

@@ -0,0 +1,22 @@
+{
+  "name": "tablescraper",
+  "version": "1.0.0",
+  "description": "",
+  "license": "ISC",
+  "author": "Jakub Žitník",
+  "type": "commonjs",
+  "main": "server.js",
+  "scripts": {
+    "test": "echo \"Error: no test specified\" && exit 1",
+    "start": "concurrently \"node server.js\" \"node cron-runner.js\""
+  },
+  "dependencies": {
+    "axios": "^1.9.0",
+    "concurrently": "^9.2.0",
+    "express": "^5.1.0",
+    "node-cron": "^4.2.1",
+    "openai": "^5.0.2",
+    "puppeteer": "^24.10.0",
+    "xlsx": "^0.18.5"
+  }
+}

scrape/parse.js

@@ -0,0 +1,119 @@
+const XLSX = require('xlsx');
+const fs = require("fs");
+
+function parseThisShit(downloadedFilePath) {
+  const workbook = XLSX.readFile(downloadedFilePath);
+  const sheetNames = workbook.SheetNames;
+
+  const dateRegex = /^(pondělí|úterý|středa|čtvrtek|pátek)\s+(\d{1,2})\.\s*(\d{1,2})\.\s*(20\d{2})/i;
+
+  // Get today's date for comparison
+  function getCurrentDateObject() {
+    const now = new Date();
+    return new Date(now.getFullYear(), now.getMonth(), now.getDate());
+  }
+
+  const today = getCurrentDateObject();
+
+  const upcomingSheets = sheetNames.filter(name => {
+    const match = name.match(dateRegex);
+    if (!match) return false;
+
+    const day = parseInt(match[2], 10);
+    const month = parseInt(match[3], 10) - 1; // JavaScript months are 0-indexed
+    const year = parseInt(match[4], 10);
+
+    const sheetDate = new Date(year, month, day);
+
+    return sheetDate >= today;
+  });
+
+  const final = [];
+
+  let finalIndex = 0;
+  for (const key of upcomingSheets) {
+    const currentSheet = workbook.Sheets[key];
+    final.push({});
+
+    const regex = /[AEC][0-4][a-c]?\/.*/s;
+    const prefixRegex = /[AEC][0-4][a-c]?/;
+    const classes = [];
+    const matchingKeys = Object.keys(currentSheet).filter(key => {
+      const value = currentSheet[key];
+
+      const testResult = regex.test(value.v);
+      if (testResult) {
+        const prefixMatch = value.v.match(prefixRegex);
+        if (prefixMatch) {
+          const prefix = prefixMatch[0];
+          classes.push(prefix);
+        }
+      }
+
+      return testResult;
+    });
+
+    function letterToNumber(letter) {
+      return letter.toLowerCase().charCodeAt(0) - 'a'.charCodeAt(0);
+    }
+
+    // For each class
+    let classI = 0;
+    for (const matchingKey of matchingKeys) {
+      const allKeys = Object.keys(currentSheet).filter(key => key !== matchingKey && key.endsWith(matchingKey.replace(/[a-z]/gi, '')));
+
+      const final2 = [];
+
+      for (const key of allKeys) {
+        const parsedKey = letterToNumber(key.replace(/[0-9]/gi, ''));
+        let d = true;
+
+        try {
+          if (currentSheet[key]["w"].startsWith("úklid")) {
+            d = false;
+          }
+        } catch {}
+
+        if (d) {
+          final2[parsedKey] = currentSheet[key]["w"];
+        } else {
+          final2[parsedKey] = null;
+        }
+      }
+
+      final[finalIndex][classes[classI]] = final2.slice(1, 11);;
+
+      classI++;
+    }
+
+    finalIndex++;
+  }
+
+  const data = {
+    schedule: final,
+    props: upcomingSheets.map(str => {
+      const dateMatch = str.match(/(\d{1,2})\.\s*(\d{1,2})\.\s*(\d{4})/);
+
+      let date = null;
+
+      if (dateMatch) {
+        const day = parseInt(dateMatch[1], 10);
+        const month = parseInt(dateMatch[2], 10);
+        const year = parseInt(dateMatch[3], 10);
+
+        date = new Date(year, month - 1, day);
+      }
+
+      const isPriprava = /priprava/i.test(str);
+
+      return {
+        date: date ? date.toISOString().slice(0,10) : null, // ISO yyyy-mm-dd string for easy use
+        priprava: isPriprava,
+      };
+    }),
+  }
+
+  fs.writeFileSync('db/current.json', JSON.stringify(data));
+}
+
+module.exports = parseThisShit;

scrape/scraper.js

@@ -0,0 +1,133 @@
+const puppeteer = require('puppeteer');
+const path = require('path');
+const fs = require('fs');
+const parseThisShit = require('./parse');
+
+const EMAIL = process.env.EMAIL;
+const PASSWORD = process.env.PASSWORD;
+//const SHAREPOINT_URL = 'https://onedrive.live.com/personal/7d8c4d9baeeebde3/_layouts/15/doc2.aspx?resid=2bddf9b7-8613-4ae3-a684-0be6d73d90bf&cid=7d8c4d9baeeebde3&ct=1748937302474&wdOrigin=OFFICECOM-WEB.START.UPLOAD&wdPreviousSessionSrc=HarmonyWeb&wdPreviousSession=ce7df0ab-aade-4df2-9e2e-492e99049666';
+const SHAREPOINT_URL = 'https://spsejecnacz.sharepoint.com/:x:/s/nastenka/EbA_RcWKRdRNlB8YU1iuWM4BnMetCQlVm8toHuuyW-TPyA?e=uu3iPR&CID=2686cea0-2d06-3304-4519-087fb9e06fd0';
+const VOLUME_PATH = path.resolve('./volume/browser');
+
+async function clearDownloadsFolder() {
+  try {
+    await fs.promises.rm('./downloads', { recursive: true, force: true });
+    await fs.promises.mkdir('./downloads');
+  } catch (err) {
+    console.error('Error:', err);
+  }
+}
+
+(async () => {
+  const browser = await puppeteer.launch({
+    headless: 'new',
+    //headless: false,
+    userDataDir: VOLUME_PATH
+  });
+  const [page] = await browser.pages();
+
+  const downloadPath = path.resolve('./downloads');
+  if (!fs.existsSync(downloadPath)) fs.mkdirSync(downloadPath);
+
+  const client = await page.target().createCDPSession();
+  await client.send('Page.setDownloadBehavior', {
+    behavior: 'allow',
+    downloadPath: downloadPath,
+  });
+
+  await page.goto(SHAREPOINT_URL, { waitUntil: 'networkidle2' });
+
+  await new Promise(r => setTimeout(r, 3000));
+
+  if (page.url().includes('login.')) {
+    console.log('Logging in...');
+
+    await page.waitForSelector('input[type="email"]', { timeout: 10000 });
+    await page.type('input[type="email"]', EMAIL, { delay: 50 });
+    await page.keyboard.press('Enter');
+    await new Promise(r => setTimeout(r, 2000));
+
+    await page.waitForSelector('input[type="password"]', { timeout: 10000 });
+    await page.type('input[type="password"]', PASSWORD, { delay: 50 });
+    await page.keyboard.press('Enter');
+
+    try {
+      await page.waitForSelector('input[type="submit"]', { timeout: 10000 });
+      await page.click('input[type="submit"]');
+    } catch {
+      console.log('No stay signed in prompt.');
+    }
+
+    // wait for navigation after login
+    await page.waitForNavigation({ waitUntil: 'networkidle2' });
+  }
+
+  // Wait for iframe containing file options
+  const frameHandle = await page.waitForSelector('iframe');
+  const frame = await frameHandle.contentFrame();
+
+  await frame.waitForSelector('button[title="File"]', { timeout: 60000 });
+  await frame.click('button[title="File"]');
+
+  await new Promise(r => setTimeout(r, 500));
+
+  // Click "Create a Copy"
+  try {
+    await frame.waitForSelector('div[role="menuitem"][name="Create a Copy"]', { visible: true });
+  } catch {
+    await frame.click('button[title="File"]');
+    await new Promise(r => setTimeout(r, 500));
+  }
+  await frame.click('div[role="menuitem"][name="Create a Copy"]');
+
+  await new Promise(r => setTimeout(r, 500));
+
+  // Click "Download a Copy"
+  await frame.waitForSelector('div[role="menuitem"][name="Download a Copy"]', { visible: true });
+  await frame.click('div[role="menuitem"][name="Download a Copy"]');
+
+  // Wait some seconds for download to start
+  await new Promise(r => setTimeout(r, 10000));
+
+  // Helper: wait for file to appear in download folder
+  function waitForFile(filename, timeout = 30000) {
+    return new Promise((resolve, reject) => {
+      const start = Date.now();
+      const interval = setInterval(() => {
+        if (fs.existsSync(filename)) {
+          clearInterval(interval);
+          resolve();
+        } else if (Date.now() - start > timeout) {
+          clearInterval(interval);
+          reject(new Error('Timeout waiting for file'));
+        }
+      }, 500);
+    });
+  }
+
+  // Helper: get newest .xlsx file in downloads folder
+  function getNewestFile(dir, ext = '.xlsx') {
+    const files = fs.readdirSync(dir)
+      .filter(f => f.endsWith(ext))
+      .map(f => ({
+        name: f,
+        time: fs.statSync(path.join(dir, f)).mtime.getTime()
+      }))
+      .sort((a, b) => b.time - a.time);
+    return files.length ? path.join(dir, files[0].name) : null;
+  }
+
+  // Wait for the downloaded file
+  const downloadedFilePath = getNewestFile(downloadPath, '.xlsx');
+  if (!downloadedFilePath) {
+    throw new Error('No XLSX file found in download folder');
+  }
+  console.log('Waiting for file:', downloadedFilePath);
+  await waitForFile(downloadedFilePath);
+
+  parseThisShit(downloadedFilePath);
+
+  await clearDownloadsFolder();
+
+  await browser.close();
+})();

server.js

@@ -0,0 +1,27 @@
+const express = require("express");
+const path = require("path");
+const app = express();
+const fs = require("fs/promises");
+
+const PORT = 3000;
+
+app.get('/', (_, res) => {
+  res.sendFile(path.join(__dirname, "db", "current.json"));
+});
+
+app.get("/status", async (_, res) => {
+  const dataStr = await fs.readFile(path.resolve("./volume/customState.json"), {encoding: "utf8"});
+  const data = JSON.parse(dataStr);
+
+  if (data.working) {
+    res.json({working: true})
+  } else {
+    res.json({working: data.working, message: data.message})
+  }
+})
+
+// TODO: Reporting errors
+
+app.listen(PORT, () => {
+  console.log(`Server is running at http://localhost:${PORT}`);
+});

volume/customState.json

@@ -0,0 +1,4 @@
+{
+  "working": true,
+  "message": ""
+}