From a57d7ac44ce00c7dba0c476dd096d1c4d8af0d99 Mon Sep 17 00:00:00 2001
From: Adrian Bonpin <adrianbonpin@gmail.com>
Date: Tue, 18 Aug 2026 15:26:13 +0800
Subject: [PATCH] feat: parse study load table

---
 parsers.js           | 50 ++++++++++++++++++++++++++++++++++++++++++++
 test/parsers.test.js | 28 +++++++++++++++++++++++++
 2 files changed, 78 insertions(+)
 create mode 100644 parsers.js
 create mode 100644 test/parsers.test.js

diff --git a/parsers.js b/parsers.js
new file mode 100644
index 0000000..756b14d
--- /dev/null
+++ b/parsers.js
@@ -0,0 +1,50 @@
+(function (root, factory) {
+  if (typeof module === 'object' && module.exports) {
+    module.exports = factory();
+  } else {
+    root.ISMIStools = Object.assign(root.ISMIStools || {}, factory());
+  }
+})(typeof self !== 'undefined' ? self : this, function () {
+  'use strict';
+
+  function cleanText(s) {
+    return String(s == null ? '' : s)
+      .replace(/<[^>]+>/g, ' ')
+      .replace(/&nbsp;/gi, ' ')
+      .replace(/\s+/g, ' ')
+      .trim();
+  }
+
+  function extractRows(html) {
+    const rows = [];
+    const trRe = /<tr[^>]*>([\s\S]*?)<\/tr>/gi;
+    let trMatch;
+    while ((trMatch = trRe.exec(html)) !== null) {
+      const rowHtml = trMatch[1];
+      if (/<th/i.test(rowHtml)) continue;
+      const tds = [];
+      const tdRe = /<td[^>]*>([\s\S]*?)<\/td>/gi;
+      let tdMatch;
+      while ((tdMatch = tdRe.exec(rowHtml)) !== null) {
+        tds.push(cleanText(tdMatch[1]));
+      }
+      rows.push(tds);
+    }
+    return rows;
+  }
+
+  function parseStudyLoadTable(html) {
+    return extractRows(html)
+      .filter((tds) => tds.length >= 6)
+      .map((tds) => ({
+        group: tds[0],
+        courseCode: tds[1],
+        schedule: tds[2],
+        faculty: tds[3],
+        units: tds[4],
+        status: tds[5]
+      }));
+  }
+
+  return { parseStudyLoadTable };
+});
diff --git a/test/parsers.test.js b/test/parsers.test.js
new file mode 100644
index 0000000..fbb9f21
--- /dev/null
+++ b/test/parsers.test.js
@@ -0,0 +1,28 @@
+const { test } = require('node:test');
+const assert = require('node:assert');
+const fs = require('node:fs');
+const path = require('node:path');
+const { parseStudyLoadTable } = require('../parsers.js');
+
+const fixture = fs.readFileSync(path.join(__dirname, 'fixtures', 'study-load-table.html'), 'utf8');
+
+test('parseStudyLoadTable extracts all course rows', () => {
+  const rows = parseStudyLoadTable(fixture);
+  assert.strictEqual(rows.length, 5);
+});
+
+test('parseStudyLoadTable maps columns correctly', () => {
+  const rows = parseStudyLoadTable(fixture);
+  const first = rows[0];
+  assert.strictEqual(first.group, '1');
+  assert.strictEqual(first.courseCode, 'CS 5105');
+  assert.strictEqual(first.units, '3.00');
+  assert.strictEqual(first.status, 'DISSOLVED');
+  assert.ok(first.schedule.includes('F 01:30 PM'));
+});
+
+test('parseStudyLoadTable handles empty faculty', () => {
+  const rows = parseStudyLoadTable(fixture);
+  const it4102 = rows.find(r => r.courseCode === 'IT 4102N');
+  assert.strictEqual(it4102.faculty, '');
+});