chore: Started implementing

scrape/parse/v3/call.js

@@ -0,0 +1,73 @@
+/*
+ * Copyright (C) 2025 Jakub Žitník
+ *
+ * This program is free software: you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation, either version 3 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+import { GoogleGenAI } from "@google/genai";
+import fs from "fs/promises";
+
+const TIMETABLE_PATH = "db/persistent/timetables.json";
+
+export async function setup() {
+  const timetable = JSON.parse(
+    await fs.readFile(TIMETABLE_PATH, { encoding: "utf8" })
+  );
+
+  const ai = new GoogleGenAI({
+    apiKey: process.env.GEMINI_API_KEY
+  });
+
+  const systemPrompt = await fs.readFile("./prompt.txt", "utf-8");
+
+  /**
+   * @param {Object} changesByClass
+   *   {
+   *     "1A": [ ...changes... ],
+   *     "2B": [ ...changes... ]
+   *   }
+   * @param {number} dayIndex
+   */
+  return async (changesByClass, dayIndex) => {
+    const input = {};
+
+    for (const cls of Object.keys(changesByClass)) {
+      input[cls] = {
+        stableSchedule: timetable[cls][dayIndex],
+        changes: changesByClass[cls]
+      };
+    }
+
+    const response = await ai.models.generateContent({
+      model: "gemini-3-flash-preview",
+      config: {
+        systemInstruction: {
+          parts: [{ text: systemPrompt }]
+        },
+        temperature: 0
+      },
+      contents: [
+        {
+          role: "user",
+          parts: [{ text: JSON.stringify(input) }]
+        }
+      ]
+    });
+
+    const aiOutput = response.text ?? "";
+
+    try {
+      return JSON.parse(aiOutput);
+    } catch {
+      return { invalid: true, reason: "AI output could not be parsed" };
+    }
+  };
+}

scrape/parse/v3/v3.js

@@ -0,0 +1,130 @@
+/*
+ * Copyright (C) 2025 Jakub Žitník
+ *
+ * This program is free software: you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation, either version 3 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+import fs from "fs/promises";
+import { setup } from "./call.js";
+
+const PREVIOUS = "db/v3/_previous.json";
+const FINAL = "db/v3/v3.json";
+const EXCLUDE_CLASSES = new Set(["ABSENCE"]);
+
+async function checkFileExists(filePath) {
+  try {
+    await fs.access(filePath);
+    return true;
+  } catch (err) {
+    return false;
+  }
+}
+
+function arraysAreEqual(arr1, arr2) {
+  if (arr1.length !== arr2.length) return false;
+  for (let i = 0; i < arr1.length; i++) {
+    if (arr1[i] !== arr2[i]) return false;
+  }
+  return true;
+}
+
+function getTime() {
+  const currentDate = new Date();
+  return currentDate.getHours().toString().padStart(2, "0") + ":" + currentDate.getMinutes().toString().padStart(2, "0");
+}
+
+function setupFinal() {
+  return {
+    schedule: [],
+    status: {
+      lastUpdated: getTime(),
+    }
+  }
+}
+
+export default async function parseV3(fileV2Path) {
+  const call = await setup();
+
+  let clearRun = false;
+  let previousStr = "{}";
+
+  if (await checkFileExists(PREVIOUS)) {
+    previousStr = await fs.readFile(PREVIOUS, "utf8");
+  } else {
+    clearRun = true;
+  }
+
+  const now = JSON.parse(await fs.readFile(fileV2Path, "utf8"));
+  const previous = JSON.parse(previousStr);
+  const previousDays = previous.props?.map(p => p.date) || [];
+
+  let final;
+  if (await checkFileExists(FINAL)) {
+    final = JSON.parse(await fs.readFile(FINAL, "utf8"));
+  } else {
+    final = setupFinal();
+    clearRun = true;
+  }
+
+  let i = 0;
+  for (const prop of now.props) {
+    const date = new Date(prop.date);
+    const dayIndex = (date.getDay() + 6) % 7;
+
+    if (!final.schedule[i]) {
+      final.schedule[i] = {};
+    }
+
+    const day = now.schedule[i];
+    const batch = {};
+
+    for (const cls of Object.keys(day)) {
+      if (EXCLUDE_CLASSES.has(cls)) continue;
+
+      const newClass = day[cls];
+
+      if (clearRun || !previousDays.includes(prop.date)) {
+        batch[cls] = newClass;
+        continue;
+      }
+
+      const oldPropIndex = previous.props.findIndex(
+        p => p.date === prop.date
+      );
+      const oldClass = previous.schedule[oldPropIndex]?.[cls] || [];
+
+      if (!arraysAreEqual(oldClass, newClass)) {
+        batch[cls] = newClass;
+      }
+    }
+
+    if (Object.keys(batch).length > 0) {
+      const results = await call(batch, dayIndex);
+
+      for (const cls of Object.keys(results)) {
+        final.schedule[i][cls] = results[cls];
+      }
+    }
+
+    i++;
+  }
+
+  if (!clearRun) {
+    final.status.lastUpdated = getTime();
+  }
+
+  final.props = now.props;
+
+  await fs.writeFile(FINAL, JSON.stringify(final), "utf8");
+  await fs.copyFile(fileV2Path, PREVIOUS);
+}
+
+parseV3("db/v2.json");