Providing free IGCSE resources for everyone!
This website uses Material Design 2 for components and design guidelines because Material 3 does not support the web yet and looks worse.
The SSR system was built using my own TypeScript library called webity which can integrate with express neatly.
A FORK and STAR shows that you support this project!
There is a SvelteKit version of this web app available that has structured questions.
The license is a modified GNU license. You must:
- include a copy of the
LICENSEfile in your own version - release your product under the same open-source license
Essentially, you can do anything you want to the code as long as your version is open-source and you credit the author. The whole point is to keep things free and transparent.
The /api route is a JSON API can be used to fetch information about website content like notes or questions.
The /api/search route is meant for searching pages and extracting useful information. There is no need to use Algolia or Google Search to index pages.
// file searching might need some tweaking if different file types are included other than .html in views
app.get('/api/search', (req, res) => {
const regex = /<\/*[A-z0-9 ="#%{}.\$/:-]+>/g
const { query } = req
console.log(query)
if (!('q' in query) || !query.q) return res.status(404).json({ error: 'empty query' })
if (query.q === '*') query.q = ''
const result = {}
const icons = {
'mathematics': 'calculate',
'computer_science': 'terminal',
'biology': 'microbiology',
'chemistry': 'science',
'physics': 'speed',
}
if ('scopes' in query) {
for (const scope of query.scopes.split(',')) {
if (!Object.keys(icons).includes(scope)) return res.status(404).json({ error: 'invalid scope' })
const icon = icons[scope]
const topics = fs.readdirSync(`views/courses/${scope}`).filter(file => !file.endsWith('.html'))
for (const topic of topics) {
const notes = fs.readFileSync(`views/courses/${scope}/${topic}/content.html`, 'utf-8').replaceAll(regex, '')
if (contains(notes, query.q)) {
result[topic] = {
text: notes,
path: `/courses/${scope}/${topic}`,
icon,
topic: true,
}
}
}
}
return res.json(result)
}
const courses = fs.readdirSync('views/courses')
for (const course of courses) {
const content = fs.readFileSync(`views/courses/${course}/content.html`, 'utf-8').replaceAll(regex, '')
const icon = icons[course]
if (contains(content, query.q)) {
result[course] = {
text: content,
path: `/courses/${course}`,
icon,
topic: false,
}
}
const topics = fs.readdirSync(`views/courses/${course}`).filter(file => !file.endsWith('.html'))
for (const topic of topics) {
const notes = fs.readFileSync(`views/courses/${course}/${topic}/content.html`, 'utf-8').replaceAll(regex, '')
if (contains(notes, query.q)) {
result[topic] = {
text: notes,
path: `/courses/${course}/${topic}`,
icon,
topic: true,
}
}
}
}
res.status(200).json(result)
})It might be a good idea to view the webity documentation before you continue development.
Templates are placed inside
%{}%special brackets that return strings as plain text.
Make sure to install all dependencies before you run your project.
npm installThere are a lot of folders in this repository.
publicstatic scripts and stylesheetsviewsHTML templates and widgetscoursessubjects and course contentbiologychemistrycomputer_sciencemathematicsphysics
The webity templates are extremely flexible. The $ selector can be used to identify a file by its path starting from the views directory.
<!DOCTYPE html>
<html lang="en">
<head>
<!-- renders views/meta.html -->
%{ include('$meta.html') }%
<!-- renders the title string -->
<title>%{ title }%</title>
%{ include('$lib.html') }%
</head>
<body>
%{ include('$appbar.html') }%
%{ include('$drawer.html') }%
<main class="mdc-top-app-bar--fixed-adjust">
<!-- renders ./content.html relative to the current webity path -->
%{ include('content.html') }%
</main>
%{ include('$root.html') }%
</body>
</html>Other parameters can be passed through webity to dynamically set src and href properties.
<!-- renders ./animations.js for HOME and ../../animations.js for COURSES -->
<script src="%{ path + 'animations.js' }%"></script>
<script src="%{ path + 'script.js' }%"></script>Revision notes are under construction!
Selective questions and theory questions can be accessed through the /api/questions route.
subject: the snake case subject nametopic: the snake case topic name
The response is a JSON file with a selective and a theory field. An example with the expected data types is shown below. These fields may slightly change over time.
interface Response {
selective: [
{
intro: string,
statements: string[],
image: string,
question: string, // length > 0
difficulty: number, // [0, 1, 2, 3]
options: string[], // length = 4
scheme: number, // [0, 1, 2, 3]
explanation: string // length > 0
}
],
theory: [
{
intro: string,
statements: string[],
image: string,
question: string, // length > 0
difficulty: number, // [0, 1, 2, 3]
scheme: string[], // length > 0
explanation: string // length > 0
}
]
}This the rough idea for text-davinci marking.
Mark this student's answer. Your response should only include a JSON object with a "marks" field. The bold words in the "scheme" field mean that the student should not be awarded the mark if their answer does not contain the exact phrase in bold!
{
"answer": "aerobic respiration releases more energy than anaerobic respiration and produces no lactic acid",
"scheme": [
"aerobic respiration releases more energy **per glucose molecule**",
"no lactic acid is produced in aerobic respiration"
]
}The expected result is a JSON object with a marks field of 1 after applying the IGCSE marking rules.
There are free hosting platforms like render and cyclic! Make sure to set environment variables before you deploy your app.
Contributors are welcome. The goal of this project is to produce a website like Save My Exams but without ads and completely free. A few objectives are listed below.
- Quality Revision Notes
- Custom Diagrams
- Multiple Choice Questions
- Theory Questions & AI Marking