feat/adding-h5p-tasks #10

Merged
DrunkenPinguin merged 6 commits from feat/adding-h5p-tasks into main 2026-09-15 13:00:22 +00:00
17 changed files with 6513 additions and 51 deletions

1
.gitignore vendored
View file

@ -11,3 +11,4 @@ build/
*.iws
*.ipr
out/
build/

118
README.md
View file

@ -13,8 +13,9 @@ Kinder können durch das Lösen von Lernaufgaben, beispielsweise Mathematikaufga
| Feature | Beschreibung |
| :--------------------------------- | :-------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| **JWT-Authentifizierung** | Sichere API-Endpunkte für authentifizierte Benutzer über JSON Web Tokens (JWT). |
| **Lernaufgaben (Tasks)** | Abrufen von interaktiven Lernaufgaben mit vordefinierten Antwortmöglichkeiten und Belohnungen in Minuten. |
| **Lernaufgaben (Tasks)** | Abrufen von interaktiven Lernaufgaben mit vordefinierten Antwortmöglichkeiten, Filterung nach Fach und Alter sowie Belohnungen in Minuten. |
| **Antwort-Verifizierung** | Überprüfung der vom Kind ausgewählten Antworten. Bei korrekter Antwort wird die Belohnung direkt auf die verbleibende Bildschirmzeit aufgerechnet. |
| **OER / H5P Ingestion Pipeline** | Automatisierter Live-Crawler und Batch-Import von freien Bildungsinhalten (ZUM.de / H5P) inklusive HTML-Bereinigung und Lizenzattributierung (CC-BY). |
| **Bildschirmzeit-Synchronisation** | Laufender Datenabgleich der verbrauchten Bildschirmzeit. Sinkt die verbleibende Zeit auf oder unter null, wird der Status automatisch auf `isBlocked = true` gesetzt. |
| **Zwei-Ebenen-Speicher** | PostgreSQL als persistente Datenbank, kombiniert mit Redis-Caching für schnelle Screentime-Abfragen. |
| **Fallback-Mechanismus** | Ist Redis nicht erreichbar, fällt das System automatisch und transparent auf PostgreSQL zurück. |
@ -98,7 +99,8 @@ Prüft, ob der Backend-Server ordnungsgemäß läuft.
```json
{
"token": "eyJhbGciOiJIUzI1NiIsIn...",
"familyName": "Familie Schmidt"
"userId": "6380776f-88d3-4e9f-a2be-0d9cacae447e",
"message": "Registrierung erfolgreich"
}
```
@ -128,7 +130,9 @@ Prüft, ob der Backend-Server ordnungsgemäß läuft.
```json
{
"token": "eyJhbGciOiJIUzI1NiIsIn...",
"familyName": "Familie Schmidt"
"userId": "6380776f-88d3-4e9f-a2be-0d9cacae447e",
"familyName": "Familie Schmidt",
"childAge": 8
}
```
@ -156,9 +160,9 @@ Authorization: Bearer <Token>
**Beschreibung:**
Liefert die aktuell verbleibende Bildschirmzeit sowie den aktuellen Sperrstatus.
Liefert die aktuell verbleibende Bildschirmzeit sowie den aktuellen Sperrstatus des authentifizierten Benutzers.
**Antwort:**
**Antwort – `200 OK`:**
```json
{
@ -181,29 +185,28 @@ Liefert die aktuell verbleibende Bildschirmzeit sowie den aktuellen Sperrstatus.
**Beschreibung:**
Zieht die seit dem letzten Abgleich verbrauchten Sekunden von der verfügbaren Bildschirmzeit ab.
Zieht die seit dem letzten Abgleich vom Endgerät verbrauchten Sekunden von der verfügbaren Bildschirmzeit ab.
**Request-Body:**
```json
{
"userId": "default_user",
"usedSecondsSinceLastSync": 200
"usedSecondsSinceLastSync": 30
}
```
**Antwort:**
**Antwort – `200 OK`:**
```json
{
"remainingSeconds": 1600,
"remainingSeconds": 1770,
"isBlocked": false
}
```
---
#### 6. Alle verfügbaren Aufgaben abrufen
#### 6. Lernaufgaben abrufen (mit Fach- & Altersfilter)
**Methode:** `GET`
@ -213,24 +216,42 @@ Zieht die seit dem letzten Abgleich verbrauchten Sekunden von der verfügbaren B
/api/v1/tasks
```
**Optionale Query-Parameter:**
| Parameter | Typ | Standardwert | Beschreibung |
| :--- | :--- | :--- | :--- |
| `subject` | String | `null` | Schulfach (z. B. `Mathematik`, `Deutsch`, `Sachunterricht`, `Englisch`, `Biologie`, `Geographie`, `Geschichte`). |
| `age` | Integer | `null` | Zielalter des Kindes (filtert nach `minAge <= age AND maxAge >= age`). |
| `limit` | Integer | `10` | Maximale Anzahl an zurückgegebenen Aufgaben. |
**Beispiel-Anfrage:**
```text
GET /api/v1/tasks?subject=Mathematik&age=8&limit=5
```
**Beschreibung:**
Liefert eine Liste aller verfügbaren Lernaufgaben.
Liefert kuratierte oder aus OER/H5P importierte Lernaufgaben. Die korrekte Lösung (`correctAnswer`) wird zum Schutz vor Manipulationen serverseitig geheim gehalten und niemals an den Client übertragen.
**Antwort:**
**Antwort – `200 OK`:**
```json
[
{
"id": "t1",
"question": "Was ist 12 + 15?",
"id": "zum-math-001",
"subject": "Mathematik",
"minAge": 8,
"maxAge": 9,
"question": "Was ist das Ergebnis von 6 × 8?",
"options": [
"25",
"27",
"30",
"22"
"42",
"48",
"54",
"56"
],
"rewardMinutes": 10
"rewardMinutes": 5,
"license": "CC BY-SA 4.0"
}
]
```
@ -251,22 +272,22 @@ Liefert eine Liste aller verfügbaren Lernaufgaben.
```json
{
"taskId": "t1",
"selectedAnswer": "27"
"taskId": "zum-math-001",
"selectedAnswer": "48"
}
```
**Antwort – korrekte Lösung:**
**Antwort – korrekte Lösung (`200 OK`):**
```json
{
"isCorrect": true,
"earnedMinutes": 10,
"message": "Super gemacht! Du hast 10 Minuten gewonnen."
"earnedMinutes": 5,
"message": "Super gemacht! Du hast 5 Minuten gewonnen."
}
```
**Antwort – falsche Lösung:**
**Antwort – falsche Lösung (`200 OK`):**
```json
{
@ -278,6 +299,38 @@ Liefert eine Liste aller verfügbaren Lernaufgaben.
---
## H5P / ZUM.de Aufgaben-Import (Standalone CLI)
Das Backend verfügt über ein integriertes CLI-Tool zum automatisierten Importieren und Crawlen interaktiver Lernaufgaben von **ZUM.de / OER (H5P)**.
### Batch-Import ausführen
```bash
# Alle Fächer importieren (z. B. 100 Aufgaben)
./gradlew runIngestion --args="--limit=100"
# Gezielt ein Schulfach importieren (z. B. Mathematik)
./gradlew runIngestion --args="--subject=Mathematik --limit=50"
# Deutsch für Grundschüler importieren
./gradlew runIngestion --args="--subject=Deutsch --limit=30"
```
### CLI-Optionen
| Argument | Standardwert | Beschreibung |
| :--- | :--- | :--- |
| `--subject=<Name>` | *(alle Fächer)* | Filtert Aufgaben nach Fach (z. B. `Mathematik`, `Deutsch`, `Englisch`, `Sachunterricht`). |
| `--limit=<Zahl>` | `50` | Maximale Anzahl an Aufgaben, die im Durchlauf importiert/aktualisiert werden. |
### Funktionsweise der Ingestion-Pipeline
1. **Live-Crawler & Discovery:** Durchsucht `apps.zum.de` nach interaktiven Inhalten (`H5P.MultiChoice`, `H5P.TrueFalse`, `H5P.QuestionSet`).
2. **Textbereinigung:** Entfernt störende HTML-Tags (`<p>`, `<strong>`, `&nbsp;`) für native mobile Darstellung.
3. **Taxonomie- & Alters-Mapping:** Weist Altersgrenzen (`minAge`, `maxAge`) basierend auf Klassenstufen automatisch zu.
4. **Idempotenter Upsert:** Speichert Fragen, Optionen, geheime Lösungen und CC-Lizenzen direkt relational in PostgreSQL (`TasksTable`). Duplikate werden vermieden.
---
## Projektkonfiguration und Umgebungsvariablen
Die Anwendung wird über folgende Konfigurationsdatei konfiguriert:
@ -517,3 +570,16 @@ MIT License
```
Eine entsprechende `LICENSE`-Datei sollte sich im Root-Verzeichnis des Repositorys befinden.
---
## Kidio Ökosystem
Für Kinder (Der spielerische Teil - "Kid"):
* Kidio Learn: Interaktive Lernspiele (Mathe, Sprachen, Logik), die sich dem Alter und Lernfortschritt des Kindes anpassen.
* Kidio Create: Tools und Plattformen, um eigene Ideen, Kunstwerke oder kleine Projekte zu entwickeln und zu teilen.
* Kidio Media: Ein sicherer Bereich mit kindgerechten Videos, Hörspielen und Büchern (ähnlich wie YouTube Kids).
![img.png](img.png)

View file

@ -11,6 +11,12 @@ version = "1.0.0-SNAPSHOT"
application {
mainClass = "io.ktor.server.netty.EngineMain"
}
ktor {
openApi {
enabled = true // Aktiviert die OpenAPI-Code-Generierung
codeInferenceEnabled = true
}
}
kotlin {
jvmToolchain(21)
@ -20,7 +26,7 @@ dependencies {
implementation(ktorLibs.server.config.yaml)
implementation(ktorLibs.server.core)
implementation(ktorLibs.server.netty)
implementation("io.ktor:ktor-server-openapi")
// JWT Authentication
implementation("io.ktor:ktor-server-auth")
implementation("io.ktor:ktor-server-auth-jwt")
@ -46,3 +52,11 @@ dependencies {
// redis
implementation("redis.clients:jedis:5.1.2")
}
tasks.register<JavaExec>("runIngestion") {
group = "application"
description = "Runs the standalone H5P/ZUM task ingestion CLI tool"
mainClass.set("com.oliver.kidio.backend.ingestion.H5PIngestionCli")
classpath = sourceSets["main"].runtimeClasspath
standardInput = System.`in`
}

View file

@ -0,0 +1,23 @@
# Swagger Codegen Ignore
# Generated by swagger-codegen https://github.com/swagger-api/swagger-codegen
# Use this file to prevent files from being overwritten by the generator.
# The patterns follow closely to .gitignore or .dockerignore.
# As an example, the C# client generator defines ApiClient.cs.
# You can make changes and tell Swagger Codgen to ignore just this file by uncommenting the following line:
#ApiClient.cs
# You can match any string of characters against a directory, file or extension with a single asterisk (*):
#foo/*/qux
# The above matches foo/bar/qux and foo/baz/qux, but not foo/bar/baz/qux
# You can recursively match patterns against a directory, file or extension with a double asterisk (**):
#foo/**/qux
# This matches foo/bar/qux, foo/baz/qux, and foo/bar/baz/qux
# You can also negate patterns with an exclamation (!).
# For example, you can ignore all files in a docs folder with the file extension .md:
#docs/*.md
# Then explicitly reverse the ignore rule for a single file:
#!docs/README.md

View file

@ -0,0 +1 @@
3.0.81

4987
docs/index.html Normal file

File diff suppressed because one or more lines are too long

BIN
img.png Normal file

Binary file not shown.

After

Width:  |  Height:  |  Size: 453 KiB

View file

@ -7,13 +7,24 @@ import org.jetbrains.exposed.sql.transactions.experimental.newSuspendedTransacti
import org.jetbrains.exposed.sql.transactions.transaction
object TasksTable : Table("tasks") {
val id = varchar("id", 50)
val id = varchar("id", 64)
val externalSource = varchar("external_source", 32).default("local")
val sourceUrl = text("source_url").nullable()
val license = varchar("license", 100).default("CC BY-SA 4.0")
val subject = varchar("subject", 64).default("Mathematik")
val minAge = integer("min_age").default(6)
val maxAge = integer("max_age").default(12)
val taskType = varchar("task_type", 32).default("MULTIPLE_CHOICE")
val question = text("question")
val options = text("options")
val correctAnswer = varchar("correct_answer", 255)
val rewardMinutes = integer("reward_minutes")
val rewardMinutes = integer("reward_minutes").default(5)
override val primaryKey = PrimaryKey(id)
init {
index("idx_tasks_filter", false, subject, minAge, maxAge)
}
}
object ScreentimeTable : Table("user_screentime") {
@ -27,9 +38,9 @@ object UserTable : Table("users") {
val id = varchar("id", 36)
val email = varchar("email", 255).uniqueIndex()
val passwordHash = varchar("password_hash", 255)
val familyName = varchar("family_name", 100)
val childAge = integer("child_age")
val parentPin = varchar("parent_pin", 4)
val familyName = varchar("family_name", 100).default("Familie")
val childAge = integer("child_age").default(8)
val parentPin = varchar("parent_pin", 4).default("1234")
override val primaryKey = PrimaryKey(id)
}
@ -55,11 +66,42 @@ object DatabaseFactory {
)
transaction(database) {
SchemaUtils.create(TasksTable, ScreentimeTable, UserTable)
// Fill existing NULL values if columns exist before SchemaUtils alters constraints
exec(
"""
DO ${'$'}${'$'}
BEGIN
IF EXISTS (SELECT 1 FROM information_schema.tables WHERE table_name='users') THEN
IF EXISTS (SELECT 1 FROM information_schema.columns WHERE table_name='users' AND column_name='family_name') THEN
UPDATE users SET family_name = 'Familie' WHERE family_name IS NULL;
END IF;
IF EXISTS (SELECT 1 FROM information_schema.columns WHERE table_name='users' AND column_name='child_age') THEN
UPDATE users SET child_age = 8 WHERE child_age IS NULL;
END IF;
IF EXISTS (SELECT 1 FROM information_schema.columns WHERE table_name='users' AND column_name='parent_pin') THEN
UPDATE users SET parent_pin = '1234' WHERE parent_pin IS NULL;
END IF;
END IF;
IF EXISTS (SELECT 1 FROM information_schema.tables WHERE table_name='tasks') THEN
IF EXISTS (SELECT 1 FROM information_schema.columns WHERE table_name='tasks' AND column_name='subject') THEN
UPDATE tasks SET subject = 'Mathematik' WHERE subject IS NULL;
END IF;
END IF;
END ${'$'}${'$'};
""".trimIndent(),
)
SchemaUtils.createMissingTablesAndColumns(TasksTable, ScreentimeTable, UserTable)
if (TasksTable.selectAll().empty()) {
TasksTable.insert {
it[id] = "t1"
it[externalSource] = "local"
it[license] = "CC BY-SA 4.0"
it[subject] = "Mathematik"
it[minAge] = 6
it[maxAge] = 10
it[taskType] = "MULTIPLE_CHOICE"
it[question] = "Was ist 12 + 15?"
it[options] = "25,27,30,22"
it[correctAnswer] = "27"
@ -67,6 +109,12 @@ object DatabaseFactory {
}
TasksTable.insert {
it[id] = "t2"
it[externalSource] = "local"
it[license] = "CC BY-SA 4.0"
it[subject] = "Mathematik"
it[minAge] = 7
it[maxAge] = 12
it[taskType] = "MULTIPLE_CHOICE"
it[question] = "Was ist 8 x 7?"
it[options] = "54,56,64,48"
it[correctAnswer] = "56"

View file

@ -6,9 +6,29 @@ import kotlinx.serialization.Serializable
@Serializable
data class Task(
val id: String,
val subject: String = "Mathematik",
val minAge: Int = 6,
val maxAge: Int = 12,
val question: String,
val options: List<String>,
val rewardMinutes: Int,
val license: String? = null,
)
// Interne Datenklasse für importierte Aufgaben (Ingestion)
data class ImportedTask(
val id: String,
val externalSource: String = "zum.de",
val sourceUrl: String? = null,
val license: String = "CC BY-SA 4.0",
val subject: String,
val minAge: Int,
val maxAge: Int,
val taskType: String = "MULTIPLE_CHOICE",
val question: String,
val options: List<String>,
val correctAnswer: String,
val rewardMinutes: Int = 5,
)
// Anfrage des Kindes, wenn es eine Aufgabe löst

View file

@ -1,5 +1,6 @@
package com.oliver.kidio.backend.domain.repository
import com.oliver.kidio.backend.ImportedTask
import com.oliver.kidio.backend.Task
import com.oliver.kidio.backend.User
import com.oliver.kidio.backend.data.database.DatabaseFactory
@ -8,10 +9,27 @@ import com.oliver.kidio.backend.data.database.ScreentimeTable
import com.oliver.kidio.backend.data.database.TasksTable
import com.oliver.kidio.backend.data.database.UserTable
import org.jetbrains.exposed.sql.*
import org.jetbrains.exposed.sql.SqlExpressionBuilder.eq
import org.jetbrains.exposed.sql.SqlExpressionBuilder.greaterEq
import org.jetbrains.exposed.sql.SqlExpressionBuilder.lessEq
import org.jetbrains.exposed.sql.lowerCase
import org.mindrot.jbcrypt.BCrypt
interface KidioRepository {
suspend fun getAllTasks(): List<Task>
suspend fun getAllTasks(): List<Task> = getTasks(limit = 100)
suspend fun getUserById(userId: String): User?
suspend fun getTasks(
subject: String? = null,
age: Int? = null,
limit: Int = 10,
): List<Task>
// Mit suspend markiert, damit es zur DatabaseFactory.dbQuery Architektur passt
suspend fun getExistingTaskIds(externalSource: String = "zum.de"): Set<String>
suspend fun upsertTask(task: ImportedTask): Boolean
suspend fun getTaskAnswerAndReward(taskId: String): Pair<String, Int>?
@ -33,18 +51,103 @@ interface KidioRepository {
}
class ExposedKidioRepository : KidioRepository {
override suspend fun getAllTasks(): List<Task> =
override suspend fun getTasks(
subject: String?,
age: Int?,
limit: Int,
): List<Task> =
DatabaseFactory.dbQuery {
TasksTable.selectAll().map {
var query = TasksTable.selectAll()
if (!subject.isNullOrBlank()) {
query = query.where { TasksTable.subject.lowerCase() eq subject.trim().lowercase() }
}
if (age != null) {
query = query.andWhere { (TasksTable.minAge lessEq age) and (TasksTable.maxAge greaterEq age) }
}
val safeLimit = if (limit <= 0) 10 else limit
query.limit(safeLimit).map {
Task(
id = it[TasksTable.id],
subject = it[TasksTable.subject],
minAge = it[TasksTable.minAge],
maxAge = it[TasksTable.maxAge],
question = it[TasksTable.question],
options = it[TasksTable.options].split(",").map { opt -> opt.trim() },
rewardMinutes = it[TasksTable.rewardMinutes],
license = it[TasksTable.license],
)
}
}
override suspend fun getUserById(userId: String): User? =
DatabaseFactory.dbQuery {
UserTable.selectAll()
.where { UserTable.id eq userId }
.map {
User(
id = it[UserTable.id],
email = it[UserTable.email],
passwordHash = it[UserTable.passwordHash],
familyName = it[UserTable.familyName],
childAge = it[UserTable.childAge],
parentPin = it[UserTable.parentPin],
)
}
.singleOrNull()
}
override suspend fun getAllTasks(): List<Task> = getTasks(limit = 100)
// NEU: Implementierung für Exposed (PostgreSQL)
override suspend fun getExistingTaskIds(externalSource: String): Set<String> =
DatabaseFactory.dbQuery {
TasksTable
.select(TasksTable.id) // Wir brauchen nur die ID-Spalte
.where { TasksTable.externalSource eq externalSource }
.map { it[TasksTable.id] }
.toSet()
}
override suspend fun upsertTask(task: ImportedTask): Boolean =
DatabaseFactory.dbQuery {
try {
val updated =
TasksTable.update({ TasksTable.id eq task.id }) {
it[externalSource] = task.externalSource
it[sourceUrl] = task.sourceUrl
it[license] = task.license
it[subject] = task.subject
it[minAge] = task.minAge
it[maxAge] = task.maxAge
it[taskType] = task.taskType
it[question] = task.question
it[options] = task.options.joinToString(",")
it[correctAnswer] = task.correctAnswer
it[rewardMinutes] = task.rewardMinutes
}
if (updated == 0) {
TasksTable.insert {
it[id] = task.id
it[externalSource] = task.externalSource
it[sourceUrl] = task.sourceUrl
it[license] = task.license
it[subject] = task.subject
it[minAge] = task.minAge
it[maxAge] = task.maxAge
it[taskType] = task.taskType
it[question] = task.question
it[options] = task.options.joinToString(",")
it[correctAnswer] = task.correctAnswer
it[rewardMinutes] = task.rewardMinutes
}
}
true
} catch (e: Exception) {
println("FEHLER beim Upsert von Task ${task.id}: ${e.message}")
false
}
}
override suspend fun registerUser(user: User): Boolean =
DatabaseFactory.dbQuery {
try {
@ -164,13 +267,34 @@ class ExposedKidioRepository : KidioRepository {
class InMemoryKidioRepository : KidioRepository {
private val tasks =
mutableListOf(
Task("t1", "Was ist 12 + 15?", listOf("25", "27", "30", "22"), 10),
Task("t2", "Was ist 8 x 7?", listOf("54", "56", "64", "48"), 15),
)
private val answers =
mutableMapOf(
"t1" to ("27" to 10),
"t2" to ("56" to 15),
ImportedTask(
DrunkenPinguin marked this conversation as resolved
Review

Eigene Aufgaben rausnehmen oder in eine JSON Datei packen

Eigene Aufgaben rausnehmen oder in eine JSON Datei packen
id = "t1",
externalSource = "manual",
subject = "Mathematik",
minAge = 6,
maxAge = 10,
taskType = "MULTIPLE_CHOICE",
question = "Was ist 12 + 15?",
options = listOf("25", "27", "30", "22"),
correctAnswer = "27",
rewardMinutes = 10,
license = "CC BY-SA 4.0",
sourceUrl = "local",
),
ImportedTask(
id = "t2",
externalSource = "manual",
subject = "Mathematik",
minAge = 7,
maxAge = 12,
taskType = "MULTIPLE_CHOICE",
question = "Was ist 8 x 7?",
options = listOf("54", "56", "64", "48"),
correctAnswer = "56",
rewardMinutes = 15,
license = "CC BY-SA 4.0",
sourceUrl = "local",
),
)
private val screentimes =
mutableMapOf(
@ -178,9 +302,54 @@ class InMemoryKidioRepository : KidioRepository {
)
private val users = mutableListOf<User>()
override suspend fun getAllTasks(): List<Task> = tasks
override suspend fun getAllTasks(): List<Task> = getTasks(limit = 100)
override suspend fun getTaskAnswerAndReward(taskId: String): Pair<String, Int>? = answers[taskId]
override suspend fun getTasks(
subject: String?,
age: Int?,
limit: Int,
): List<Task> {
val safeLimit = if (limit <= 0) 10 else limit
return tasks.filter { task ->
val matchSubject = subject.isNullOrBlank() || task.subject.equals(subject.trim(), ignoreCase = true)
val matchAge = age == null || (task.minAge <= age && task.maxAge >= age)
matchSubject && matchAge
}.take(safeLimit).map {
Task(
id = it.id,
subject = it.subject,
minAge = it.minAge,
maxAge = it.maxAge,
question = it.question,
options = it.options,
rewardMinutes = it.rewardMinutes,
license = it.license,
)
}
}
override suspend fun getUserById(userId: String): User? {
return users.find { it.id == userId }
}
// NEU: Implementierung für InMemory (Mock-Daten)
override suspend fun getExistingTaskIds(externalSource: String): Set<String> {
return tasks
.filter { it.externalSource == externalSource }
.map { it.id }
.toSet()
}
override suspend fun upsertTask(task: ImportedTask): Boolean {
tasks.removeIf { it.id == task.id }
tasks.add(task)
return true
}
override suspend fun getTaskAnswerAndReward(taskId: String): Pair<String, Int>? {
val task = tasks.find { it.id == taskId } ?: return null
return task.correctAnswer to task.rewardMinutes
}
override suspend fun getScreentime(userId: String): Int = screentimes[userId] ?: 1800

View file

@ -0,0 +1,133 @@
package com.oliver.kidio.backend.ingestion
import com.oliver.kidio.backend.ImportedTask
import com.oliver.kidio.backend.data.database.TasksTable
import com.oliver.kidio.backend.domain.repository.ExposedKidioRepository
import com.oliver.kidio.backend.domain.repository.KidioRepository
import kotlinx.coroutines.runBlocking
import org.jetbrains.exposed.sql.Database
import org.jetbrains.exposed.sql.SchemaUtils
import org.jetbrains.exposed.sql.transactions.transaction
import java.io.ByteArrayInputStream
object H5PIngestionCli {
/**
* Executes batch ingestion pipeline and upserts tasks into the target repository.
*/
fun ingestBatch(
repository: KidioRepository,
client: ZumApiClient = ZumApiClient(),
subject: String? = null,
limit: Int = 50,
): Int =
runBlocking {
// Bereits vorhandene Task-IDs aus der Datenbank auslesen
val existingIds = repository.getExistingTaskIds("zum.de")
println("Bereits in DB vorhandene ZUM-Aufgaben: ${existingIds.size}")
var importedCount = 0
// existingIds an fetchLiveOerItems übergeben, um Duplikate vor dem Download zu filtern
val items =
client.fetchLiveOerItems(
subject = subject,
limit = limit,
existingIds = existingIds,
)
for (item in items) {
var parsedList = emptyList<ParsedH5PTask>()
// 1. Try parsing JSON content directly (fast, no ZIP decompression needed)
if (item.rawJsonContent != null) {
parsedList =
H5PParser.parseAllTasks(
contentJson = item.rawJsonContent,
library = item.library ?: "H5P.MultiChoice",
taskId = item.id,
title = item.title,
license = item.license,
sourceUrl = item.downloadUrl,
)
}
// 2. Fallback to downloading .h5p ZIP archive if remote archive is specified
if (parsedList.isEmpty() && item.downloadUrl.endsWith(".h5p", ignoreCase = true)) {
val h5pBytes = client.downloadH5P(item.downloadUrl)
if (h5pBytes != null) {
val single = H5PParser.parseH5PZip(ByteArrayInputStream(h5pBytes), item.id, item.downloadUrl)
if (single != null) {
parsedList = listOf(single)
}
}
}
for (parsed in parsedList) {
val (minAge, maxAge) = ZumApiClient.mapGradeToAgeRange(item.grade)
val task =
ImportedTask(
id = parsed.id,
externalSource = "zum.de",
sourceUrl = parsed.sourceUrl ?: item.downloadUrl,
license = parsed.license,
subject = ZumApiClient.mapSubject(item.subject),
minAge = minAge,
maxAge = maxAge,
taskType = if (parsed.options.size == 2 && parsed.options.contains("Wahr")) "TRUE_FALSE" else "MULTIPLE_CHOICE",
question = parsed.question,
options = parsed.options,
correctAnswer = parsed.correctAnswer,
rewardMinutes = 5,
)
val success = repository.upsertTask(task)
if (success) {
importedCount++
val shortQ = task.question.take(30)
val ageRange = "${task.minAge}-${task.maxAge}"
println(" -> [Importiert] ${task.id} (${task.subject}, $ageRange): \"$shortQ...\"")
}
}
}
println("Ingestion completed: $importedCount tasks successfully imported/updated.")
importedCount
}
@JvmStatic
fun main(args: Array<String>) {
println("=== Kidio H5P/ZUM Ingestion Tool ===")
val host = System.getenv("DB_HOST") ?: "localhost"
val port = System.getenv("DB_PORT") ?: "5432"
val dbName = System.getenv("DB_NAME") ?: "kidio_db"
val user = System.getenv("DB_USER") ?: "kidio_admin"
val password = System.getenv("DB_PASSWORD") ?: "learning"
var subject: String? = null
var limit = 50
for (arg in args) {
when {
arg.startsWith("--subject=") -> subject = arg.substringAfter("--subject=")
arg.startsWith("--limit=") -> limit = arg.substringAfter("--limit=").toIntOrNull() ?: 50
}
}
val jdbcUrl = "jdbc:postgresql://$host:$port/$dbName"
println("Connecting to database: $jdbcUrl")
val database =
Database.connect(
url = jdbcUrl,
driver = "org.postgresql.Driver",
user = user,
password = password,
)
transaction(database) {
SchemaUtils.createMissingTablesAndColumns(TasksTable)
}
val repository = ExposedKidioRepository()
ingestBatch(repository = repository, subject = subject, limit = limit)
}
}

View file

@ -0,0 +1,270 @@
package com.oliver.kidio.backend.ingestion
import kotlinx.serialization.json.*
import java.io.InputStream
import java.util.zip.ZipInputStream
data class ParsedH5PTask(
val id: String,
val title: String,
val library: String,
val question: String,
val options: List<String>,
val correctAnswer: String,
val license: String = "CC BY-SA 4.0",
val sourceUrl: String? = null,
)
object H5PParser {
private val json =
Json {
ignoreUnknownKeys = true
isLenient = true
}
/**
* Strips HTML tags and decodes common HTML entities to provide clean text.
*/
fun stripHtml(html: String): String {
var text = html.replace(Regex("(?i)<(br|p|div|li|tr|h[1-6])[^>]*>"), " ")
text = text.replace(Regex("<[^>]*>"), "")
text =
text
.replace("&nbsp;", " ")
.replace("&amp;", "&")
.replace("&quot;", "\"")
.replace("&#39;", "'")
.replace("&apos;", "'")
.replace("&lt;", "<")
.replace("&gt;", ">")
.replace("&times;", "×")
.replace("&bdquo;", "„")
.replace("&ldquo;", "“")
.replace("&auml;", "ä")
.replace("&Auml;", "Ä")
.replace("&ouml;", "ö")
.replace("&Ouml;", "Ö")
.replace("&uuml;", "ü")
.replace("&Uuml;", "Ü")
.replace("&szlig;", "ß")
text = text.replace(Regex("\\s+([?.!,:;])"), "$1")
return text.replace(Regex("\\s+"), " ").trim()
}
/**
* Parses an .h5p ZIP archive stream.
*/
fun parseH5PZip(
inputStream: InputStream,
taskId: String,
sourceUrl: String? = null,
): ParsedH5PTask? {
val zip = ZipInputStream(inputStream)
var h5pJsonString: String? = null
var contentJsonString: String? = null
var entry = zip.nextEntry
while (entry != null) {
val name = entry.name
if (name == "h5p.json" || name.endsWith("/h5p.json")) {
h5pJsonString = zip.readBytes().toString(Charsets.UTF_8)
} else if (name == "content/content.json" || name == "content.json" || name.endsWith("/content.json")) {
contentJsonString = zip.readBytes().toString(Charsets.UTF_8)
}
zip.closeEntry()
entry = zip.nextEntry
}
if (contentJsonString == null) {
return null
}
var library = "H5P.MultiChoice"
var title = "H5P Task"
var license = "CC BY-SA 4.0"
if (h5pJsonString != null) {
try {
val h5pTree = json.parseToJsonElement(h5pJsonString).jsonObject
library = h5pTree["mainLibrary"]?.jsonPrimitive?.contentOrNull ?: library
title = h5pTree["title"]?.jsonPrimitive?.contentOrNull ?: title
license = h5pTree["license"]?.jsonPrimitive?.contentOrNull ?: license
} catch (_: Exception) {
}
}
return parseContentJson(
contentJson = contentJsonString,
library = library,
taskId = taskId,
title = title,
license = license,
sourceUrl = sourceUrl,
)
}
/**
* Parses content.json representation for supported H5P libraries (MultiChoice, TrueFalse).
*/
fun parseContentJson(
contentJson: String,
library: String = "H5P.MultiChoice",
taskId: String = "task-1",
title: String = "H5P Task",
license: String = "CC BY-SA 4.0",
sourceUrl: String? = null,
): ParsedH5PTask? {
val element =
try {
json.parseToJsonElement(contentJson).jsonObject
} catch (e: Exception) {
return null
}
if (library.contains("TrueFalse", ignoreCase = true) || (element.containsKey("correct") && !element.containsKey("answers"))) {
return parseTrueFalse(element, taskId, title, library, license, sourceUrl)
}
return parseMultiChoice(element, taskId, title, library, license, sourceUrl)
}
/**
* Parses one or multiple tasks from content.json (supports MultiChoice, TrueFalse, and QuestionSet).
*/
fun parseAllTasks(
contentJson: String,
library: String = "H5P.MultiChoice",
taskId: String = "task-1",
title: String = "H5P Task",
license: String = "CC BY-SA 4.0",
sourceUrl: String? = null,
): List<ParsedH5PTask> {
val element =
try {
json.parseToJsonElement(contentJson).jsonObject
} catch (e: Exception) {
return emptyList()
}
if (library.contains("QuestionSet", ignoreCase = true) && element.containsKey("questions")) {
val questionsArray = element["questions"]?.jsonArray ?: return emptyList()
val results = mutableListOf<ParsedH5PTask>()
for ((idx, qElement) in questionsArray.withIndex()) {
val qObj = qElement.jsonObject
val subLib = qObj["library"]?.jsonPrimitive?.contentOrNull ?: "H5P.MultiChoice"
val subParams = qObj["params"]?.jsonObject ?: continue
val subId = "$taskId-q${idx + 1}"
val isTrueFalse =
subLib.contains("TrueFalse", ignoreCase = true) ||
(subParams.containsKey("correct") && !subParams.containsKey("answers"))
val isMultiChoice =
subLib.contains("MultiChoice", ignoreCase = true) || subParams.containsKey("answers")
val parsedSub =
if (isTrueFalse) {
parseTrueFalse(subParams, subId, "$title #${idx + 1}", subLib, license, sourceUrl)
} else if (isMultiChoice) {
parseMultiChoice(subParams, subId, "$title #${idx + 1}", subLib, license, sourceUrl)
} else {
null
}
if (parsedSub != null) {
results.add(parsedSub)
}
}
return results
}
val single = parseContentJson(contentJson, library, taskId, title, license, sourceUrl)
return if (single != null) listOf(single) else emptyList()
}
private fun parseMultiChoice(
element: JsonObject,
taskId: String,
title: String,
library: String,
license: String,
sourceUrl: String?,
): ParsedH5PTask? {
val questionRaw =
element["question"]?.jsonPrimitive?.contentOrNull
?: element["text"]?.jsonPrimitive?.contentOrNull
?: return null
val questionClean = stripHtml(questionRaw)
if (questionClean.isBlank()) return null
val answersArray = element["answers"]?.jsonArray ?: return null
val options = mutableListOf<String>()
var correctAnswer: String? = null
for (item in answersArray) {
val itemObj = item.jsonObject
val textRaw = itemObj["text"]?.jsonPrimitive?.contentOrNull ?: continue
val textClean = stripHtml(textRaw)
if (textClean.isNotBlank()) {
options.add(textClean)
val isCorrect = itemObj["correct"]?.jsonPrimitive?.booleanOrNull ?: false
if (isCorrect && correctAnswer == null) {
correctAnswer = textClean
}
}
}
if (options.isEmpty() || correctAnswer == null) {
return null
}
return ParsedH5PTask(
id = taskId,
title = title,
library = library,
question = questionClean,
options = options,
correctAnswer = correctAnswer,
license = license,
sourceUrl = sourceUrl,
)
}
private fun parseTrueFalse(
element: JsonObject,
taskId: String,
title: String,
library: String,
license: String,
sourceUrl: String?,
): ParsedH5PTask? {
val questionRaw =
element["question"]?.jsonPrimitive?.contentOrNull
?: element["text"]?.jsonPrimitive?.contentOrNull
?: return null
val questionClean = stripHtml(questionRaw)
if (questionClean.isBlank()) return null
val correctPrimitive = element["correct"]?.jsonPrimitive
val isTrueCorrect =
when {
correctPrimitive?.booleanOrNull != null -> correctPrimitive.boolean
correctPrimitive?.contentOrNull != null -> correctPrimitive.content.equals("true", ignoreCase = true)
else -> true
}
val options = listOf("Wahr", "Falsch")
val correctAnswer = if (isTrueCorrect) "Wahr" else "Falsch"
return ParsedH5PTask(
id = taskId,
title = title,
library = library,
question = questionClean,
options = options,
correctAnswer = correctAnswer,
license = license,
sourceUrl = sourceUrl,
)
}
}

View file

@ -0,0 +1,514 @@
package com.oliver.kidio.backend.ingestion
import kotlinx.serialization.json.*
import java.net.URI
import java.net.URLEncoder
import java.net.http.HttpClient
import java.net.http.HttpRequest
import java.net.http.HttpResponse
import java.time.Duration
data class ZumOerItem(
val id: String,
val title: String,
val subject: String,
val grade: Int? = null,
val downloadUrl: String,
val license: String = "CC BY-SA 4.0",
val library: String? = null,
val rawJsonContent: String? = null,
)
class ZumApiClient(
private val httpClient: HttpClient =
HttpClient.newBuilder()
.connectTimeout(Duration.ofSeconds(10))
.followRedirects(HttpClient.Redirect.NORMAL)
.build(),
) {
private val json =
Json {
ignoreUnknownKeys = true
isLenient = true
}
companion object {
fun mapSubject(raw: String): String {
val lower = raw.trim().lowercase()
return when {
lower.contains("mathe") -> "Mathematik"
lower.contains("deutsch") || lower.contains("german") -> "Deutsch"
lower.contains("sach") || lower.contains("hsu") || lower.contains("natur") || lower.contains("heimat") -> "Sachunterricht"
lower.contains("englisch") || lower.contains("english") -> "Englisch"
lower.contains("bio") -> "Biologie"
lower.contains("geo") || lower.contains("erdkunde") -> "Geographie"
lower.contains("gesch") -> "Geschichte"
lower.contains("chemie") -> "Chemie"
lower.contains("physik") -> "Physik"
lower.contains("musik") -> "Musik"
lower.contains("kunst") -> "Kunst"
lower.contains("informatik") -> "Informatik"
else -> raw.trim().replaceFirstChar { if (it.isLowerCase()) it.titlecase() else it.toString() }
}
}
fun mapGradeToAgeRange(grade: Int?): Pair<Int, Int> {
return when (grade) {
1 -> 6 to 7
2 -> 7 to 8
3 -> 8 to 9
4 -> 9 to 10
5 -> 10 to 11
6 -> 11 to 12
7 -> 12 to 13
8 -> 13 to 14
9 -> 14 to 15
10 -> 15 to 16
11, 12, 13 -> 16 to 18
else -> 6 to 12
}
}
}
/**
* Downloads an .h5p package file from a remote URL.
*/
fun downloadH5P(downloadUrl: String): ByteArray? {
return try {
val request =
HttpRequest.newBuilder()
.uri(URI.create(downloadUrl))
.timeout(Duration.ofSeconds(15))
.header("User-Agent", "Mozilla/5.0 (compatible; KidioBot/1.0; +https://kidio.oliver.com)")
.GET()
.build()
val response = httpClient.send(request, HttpResponse.BodyHandlers.ofByteArray())
if (response.statusCode() == 200) {
response.body()
} else {
null
}
} catch (e: Exception) {
println("WARNUNG: Download fehlgeschlagen für $downloadUrl: ${e.message}")
null
}
}
/**
* Fetches live OER items from apps.zum.de via search and Drupal settings extraction.
* Skips items whose IDs are already present in existingIds.
* Falls back to curated items if remote catalog is unreachable or returns 0 matching interactive items.
*/
fun fetchLiveOerItems(
subject: String? = null,
limit: Int = 50,
maxPages: Int = 15,
existingIds: Set<String> = emptySet(),
): List<ZumOerItem> {
val results = mutableListOf<ZumOerItem>()
val seenAppIds = mutableSetOf<String>()
println("[ZUM-Apps Live] Starte Live-Abfrage auf apps.zum.de (Fach: ${subject ?: "Alle Fächer"}, Limit: $limit)...")
for (page in 0 until maxPages) {
if (results.size >= limit) break
val searchUrl =
if (subject.isNullOrBlank()) {
"https://apps.zum.de/apps?page=$page"
} else {
val encoded = URLEncoder.encode(subject, "UTF-8")
"https://apps.zum.de/apps?keys=$encoded&page=$page"
}
val html = fetchString(searchUrl)
if (html == null) {
println("[ZUM-Apps Live] Warnung: Seite $page konnte nicht geladen werden ($searchUrl)")
continue
}
// Filtert bereits verarbeitete oder in DB vorhandene Apps heraus
val appIds =
Regex("""href=["']/apps/(\d+)["']""")
.findAll(html)
.map { it.groupValues[1] }
.distinct()
.filter { appId ->
val fullId = "zum-$appId"
!seenAppIds.contains(appId) && !existingIds.contains(fullId) && !existingIds.contains(appId)
}
.toList()
if (appIds.isEmpty()) {
println("[ZUM-Apps Live] Seite $page enthält keine neuen Aufgaben. Blättere weiter...")
continue
}
for (appId in appIds) {
if (results.size >= limit) break
seenAppIds.add(appId)
val item = fetchAppDetails(appId, subject)
if (item != null) {
results.add(item)
val lib = item.library ?: "H5P"
val gr = item.grade ?: "k.A."
println(
"[ZUM-Apps Live] Neue App #$appId: \"${item.title}\" ($lib) [${item.subject}, Kl. $gr]",
)
}
}
}
if (results.isEmpty()) {
println("[ZUM-Apps Live] Keine neuen Live-Inhalte gefunden oder Seite nicht erreichbar. Nutze kuratierte OER-Fallback-Daten.")
return getCuratedOerItems(subject = subject, limit = limit)
}
println("[ZUM-Apps Live] ${results.size} neue interaktive Aufgaben erfolgreich von apps.zum.de geladen.")
return results
}
private fun fetchAppDetails(
appId: String,
fallbackSubject: String?,
): ZumOerItem? {
val appUrl = "https://apps.zum.de/apps/$appId"
val html = fetchString(appUrl) ?: return null
val settingsMatch = Regex("""data-drupal-selector="drupal-settings-json">([^<]+)<""").find(html) ?: return null
val jsonStr = settingsMatch.groupValues[1]
val drupalSettings =
try {
json.parseToJsonElement(jsonStr).jsonObject
} catch (_: Exception) {
return null
}
val h5pContents =
drupalSettings["h5p"]?.jsonObject
?.get("H5PIntegration")?.jsonObject
?.get("contents")?.jsonObject
?: return null
val contentObj =
h5pContents["cid-$appId"]?.jsonObject
?: h5pContents.values.firstOrNull()?.jsonObject
?: return null
val library = contentObj["library"]?.jsonPrimitive?.contentOrNull ?: ""
if (!library.contains("MultiChoice", ignoreCase = true) &&
!library.contains("TrueFalse", ignoreCase = true) &&
!library.contains("QuestionSet", ignoreCase = true)
) {
return null
}
val jsonContent = contentObj["jsonContent"]?.jsonPrimitive?.contentOrNull ?: return null
val metadata = contentObj["metadata"]?.jsonObject
val title =
metadata?.get("title")?.jsonPrimitive?.contentOrNull
?: Regex("""<title>(.*?)(?:\|.*)?</title>""").find(html)?.groupValues?.get(1)?.trim()
?: "ZUM App #$appId"
val rawLicense = metadata?.get("license")?.jsonPrimitive?.contentOrNull
val license =
when {
rawLicense == "U" || rawLicense == null || rawLicense.isBlank() -> "CC BY-SA 4.0"
rawLicense.contains("CC", ignoreCase = true) -> rawLicense
else -> "CC BY-SA 4.0 ($rawLicense)"
}
val exportUrl = contentObj["exportUrl"]?.jsonPrimitive?.contentOrNull ?: appUrl
// Breitgefächerter Abruf aller HTML-Linktexte (Kategorien, Schulstufen, Tags)
val tags =
Regex("""<a[^>]*>([^<]+)</a>""")
.findAll(html)
.map { it.groupValues[1].trim() }
.filter { it.isNotBlank() && it.length < 60 }
.toList()
val subject = inferSubject(title, tags, fallbackSubject)
val grade = inferGrade(title, html, tags)
return ZumOerItem(
id = "zum-$appId",
title = title,
subject = subject,
grade = grade,
downloadUrl = exportUrl,
license = license,
library = library,
rawJsonContent = jsonContent,
)
}
private fun inferSubject(
title: String,
tags: List<String>,
fallbackSubject: String?,
): String {
for (tag in tags) {
val mapped = mapSubject(tag)
if (mapped != tag ||
tag.equals("Mathematik", ignoreCase = true) ||
tag.equals("Deutsch", ignoreCase = true) ||
tag.equals("Biologie", ignoreCase = true) ||
tag.equals("Englisch", ignoreCase = true)
) {
return mapped
}
}
if (!fallbackSubject.isNullOrBlank()) {
return mapSubject(fallbackSubject)
}
return mapSubject(title)
}
private fun inferGrade(
title: String,
html: String,
tags: List<String>,
): Int? {
val primaryText = (tags + title).joinToString(" ")
// 1. Muster: "3. Klasse", "3. Schuljahr", "3. Kl."
Regex("""\b(\d+)\.\s*(?:Klasse|Schuljahr|Kl)\b""", RegexOption.IGNORE_CASE)
.find(primaryText)?.groupValues?.get(1)?.toIntOrNull()?.let { return it }
// 2. Muster: "Klasse 3", "Kl. 3", "Schuljahr 3"
Regex("""\b(?:Klasse|Schuljahr|Kl\.)\s*(\d+)\b""", RegexOption.IGNORE_CASE)
.find(primaryText)?.groupValues?.get(1)?.toIntOrNull()?.let { return it }
// 3. Muster: "Klasse 3-4", "Klasse 3/4", "3.-4. Klasse"
Regex("""\b(?:Klasse|Kl\.)?\s*(\d+)\s*[-/]\s*(\d+)\b""", RegexOption.IGNORE_CASE)
.find(primaryText)?.groupValues?.get(1)?.toIntOrNull()?.let { return it }
// 4. Fallback: Gesamten HTML-Text der Seite nach Klassenangaben absuchen
Regex("""\b(\d+)\.\s*(?:Klasse|Schuljahr|Kl)\b""", RegexOption.IGNORE_CASE)
.find(html)?.groupValues?.get(1)?.toIntOrNull()?.let { return it }
Regex("""\b(?:Klasse|Schuljahr|Kl\.)\s*(\d+)\b""", RegexOption.IGNORE_CASE)
.find(html)?.groupValues?.get(1)?.toIntOrNull()?.let { return it }
// 5. Fallback: Schulstufen-Schlüsselwörter
val combinedLower = (primaryText + " " + html).lowercase()
if (combinedLower.contains("primarstufe") || combinedLower.contains("grundschule")) return 3
val isSek1 =
combinedLower.contains("sekundarstufe i") ||
combinedLower.contains("sek i") ||
combinedLower.contains("realschule") ||
combinedLower.contains("hauptschule")
if (isSek1) return 6
val isSek2 =
combinedLower.contains("sekundarstufe ii") ||
combinedLower.contains("sek ii") ||
combinedLower.contains("gymnasium") ||
combinedLower.contains("abitur")
if (isSek2) return 10
return null
}
private fun fetchString(url: String): String? {
return try {
val request =
HttpRequest.newBuilder()
.uri(URI.create(url))
.timeout(Duration.ofSeconds(12))
.header("User-Agent", "Mozilla/5.0 (compatible; KidioBot/1.0; +https://kidio.oliver.com)")
.header("Accept", "text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8")
.header("Accept-Language", "de-DE,de;q=0.9,en;q=0.8")
.GET()
.build()
val response = httpClient.send(request, HttpResponse.BodyHandlers.ofString(Charsets.UTF_8))
if (response.statusCode() == 200) {
response.body()
} else {
null
}
} catch (e: Exception) {
null
}
}
/**
* Returns curated ZUM/OER sample metadata items, filterable by subject.
*/
fun getCuratedOerItems(
subject: String? = null,
limit: Int = 20,
): List<ZumOerItem> {
val allItems =
listOf(
ZumOerItem(
id = "zum-math-001",
title = "Grundrechenarten: Multiplikation 3. Klasse",
subject = "Mathematik",
grade = 3,
downloadUrl = "https://apps.zum.de/api/h5p/math-001.h5p",
license = "CC BY-SA 4.0",
rawJsonContent =
"""
{
"question": "<p>Was ist das Ergebnis von <strong>6 &times; 8</strong>?</p>",
"answers": [
{"text": "42", "correct": false},
{"text": "48", "correct": true},
{"text": "54", "correct": false},
{"text": "56", "correct": false}
]
}
""".trimIndent(),
),
ZumOerItem(
id = "zum-math-002",
title = "Zahlenraum bis 100: Addition",
subject = "Mathematik",
grade = 2,
downloadUrl = "https://apps.zum.de/api/h5p/math-002.h5p",
license = "CC BY-SA 4.0",
rawJsonContent =
"""
{
"question": "<p>Berechne: 34 + 28</p>",
"answers": [
{"text": "52", "correct": false},
{"text": "62", "correct": true},
{"text": "64", "correct": false}
]
}
""".trimIndent(),
),
ZumOerItem(
id = "zum-math-003",
title = "Geometrie: Formen erkennen",
subject = "Mathematik",
grade = 4,
downloadUrl = "https://apps.zum.de/api/h5p/math-003.h5p",
license = "CC BY-SA 4.0",
rawJsonContent =
"""
{
"question": "<p>Ein Quadrat hat vier gleich lange Seiten und vier rechte Winkel.</p>",
"correct": true
}
""".trimIndent(),
),
ZumOerItem(
id = "zum-de-001",
title = "Wortarten: Nomen erkennen",
subject = "Deutsch",
grade = 2,
downloadUrl = "https://apps.zum.de/api/h5p/de-001.h5p",
license = "CC BY-SA 4.0",
rawJsonContent =
"""
{
"question": "<p>Welches der folgenden W&ouml;rter ist ein <strong>Nomen (Namenwort)</strong>?</p>",
"answers": [
{"text": "schnell", "correct": false},
{"text": "laufen", "correct": false},
{"text": "Schule", "correct": true},
{"text": "sch&ouml;n", "correct": false}
]
}
""".trimIndent(),
),
ZumOerItem(
id = "zum-de-002",
title = "Rechtschreibung: Doppel-s oder ss",
subject = "Deutsch",
grade = 4,
downloadUrl = "https://apps.zum.de/api/h5p/de-002.h5p",
license = "CC BY-SA 4.0",
rawJsonContent =
"""
{
"question": "<p>Wie wird das Wort richtig geschrieben? &bdquo;Der Flu... ist tief.&ldquo;</p>",
"answers": [
{"text": "Flus", "correct": false},
{"text": "Fluss", "correct": true},
{"text": "Fluß", "correct": false}
]
}
""".trimIndent(),
),
ZumOerItem(
id = "zum-sach-001",
title = "Natur & Tiere: Winterschlaf",
subject = "Sachunterricht",
grade = 3,
downloadUrl = "https://apps.zum.de/api/h5p/sach-001.h5p",
license = "CC BY-SA 4.0",
rawJsonContent =
"""
{
"question": "<p>Welches dieser Tiere h&auml;lt einen echten <strong>Winterschlaf</strong>?</p>",
"answers": [
{"text": "Igel", "correct": true},
{"text": "Eichh&ouml;rnchen", "correct": false},
{"text": "Amsel", "correct": false},
{"text": "Fuchs", "correct": false}
]
}
""".trimIndent(),
),
ZumOerItem(
id = "zum-sach-002",
title = "Planetensystem: Sonnennähe",
subject = "Sachunterricht",
grade = 4,
downloadUrl = "https://apps.zum.de/api/h5p/sach-002.h5p",
license = "CC BY-SA 4.0",
rawJsonContent =
"""
{
"question": "<p>Welcher Planet ist der Sonne am n&auml;chsten?</p>",
"answers": [
{"text": "Venus", "correct": false},
{"text": "Mars", "correct": false},
{"text": "Merkur", "correct": true},
{"text": "Jupiter", "correct": false}
]
}
""".trimIndent(),
),
ZumOerItem(
id = "zum-eng-001",
title = "English Vocabulary: Colors & Animals",
subject = "Englisch",
grade = 4,
downloadUrl = "https://apps.zum.de/api/h5p/eng-001.h5p",
license = "CC BY-SA 4.0",
rawJsonContent =
"""
{
"question": "<p>What is the English word for &bdquo;Hund&ldquo;?</p>",
"answers": [
{"text": "Cat", "correct": false},
{"text": "Dog", "correct": true},
{"text": "Bird", "correct": false},
{"text": "Rabbit", "correct": false}
]
}
""".trimIndent(),
),
)
val normalizedSubject = subject?.let { mapSubject(it) }
val filtered =
if (normalizedSubject.isNullOrBlank()) {
allItems
} else {
allItems.filter { it.subject.equals(normalizedSubject, ignoreCase = true) }
}
return filtered.take(limit)
}
}

View file

@ -7,6 +7,7 @@ import io.ktor.http.HttpStatusCode
import io.ktor.server.application.*
import io.ktor.server.auth.authenticate
import io.ktor.server.auth.principal
import io.ktor.server.plugins.openapi.*
import io.ktor.server.request.*
import io.ktor.server.response.*
import io.ktor.server.routing.*
@ -38,6 +39,8 @@ fun Application.configureRouting(
jwtService: JwtService,
) {
routing {
openAPI(path = "openapi")
post("/api/v1/auth/register") {
val registerRequest = call.receive<RegisterRequest>()
if (registerRequest.email.isBlank() || registerRequest.passwordPlain.isBlank()) {
@ -119,9 +122,21 @@ fun Application.configureRouting(
}
authenticate("auth-jwt") {
// 1. Alle verfügbaren Aufgaben abrufen
// 1. Alle verfügbaren Aufgaben abrufen (mit optionaler Filterung nach Fach und Alter)
get("/api/v1/tasks") {
val tasks = repository.getAllTasks()
val principal = call.principal<UserPrincipal>()
val userId = principal?.userId
val subject = call.request.queryParameters["subject"]
// Expliziter Query-Parameter 'age' ODER Fallback auf das Alter des Users
val queryAge = call.request.queryParameters["age"]?.toIntOrNull()
val userAge = userId?.let { repository.getUserById(it)?.childAge }
val effectiveAge = queryAge ?: userAge
val limit = call.request.queryParameters["limit"]?.toIntOrNull() ?: 10
val tasks = repository.getTasks(subject = subject, age = effectiveAge, limit = limit)
call.respond(tasks)
}
@ -174,6 +189,7 @@ fun Application.configureRouting(
)
}
}
get("/api/v1/screentime") {
val principal = call.principal<UserPrincipal>()
val userid =

View file

@ -1,13 +1,13 @@
ktor:
deployment:
port: 8080
port: 8081
application:
modules:
- MainKt.module
storage:
driverClassName: "org.postgresql.Driver"
host: "82.165.11.162"
host: "localhost"
port: "5432"
database: "kidio_db"
username: "kidio_admin"

View file

@ -0,0 +1,178 @@
package com.oliver.kidio.backend
import com.oliver.kidio.backend.domain.repository.InMemoryKidioRepository
import com.oliver.kidio.backend.ingestion.H5PIngestionCli
import com.oliver.kidio.backend.ingestion.H5PParser
import com.oliver.kidio.backend.ingestion.ZumApiClient
import kotlinx.coroutines.runBlocking
import java.io.ByteArrayInputStream
import java.io.ByteArrayOutputStream
import java.util.zip.ZipEntry
import java.util.zip.ZipOutputStream
import kotlin.test.Test
import kotlin.test.assertEquals
import kotlin.test.assertNotNull
import kotlin.test.assertTrue
class H5PParserTest {
@Test
fun `test HTML stripping and entity decoding`() {
val html = "<p>Was ist <strong>12 &times; 4</strong>? &nbsp; &bdquo;Test&ldquo; &auml;&ouml;&uuml;</p>"
val clean = H5PParser.stripHtml(html)
assertTrue(clean.contains("Was ist 12"))
assertTrue(clean.contains("äöü"))
}
@Test
fun `test parse MultiChoice content json`() {
val json =
"""
{
"question": "<p>Was ist die Hauptstadt von <strong>Frankreich</strong>?</p>",
"answers": [
{"text": "<div>Berlin</div>", "correct": false},
{"text": "<div>Paris</div>", "correct": true},
{"text": "<div>Rom</div>", "correct": false}
]
}
""".trimIndent()
val parsed = H5PParser.parseContentJson(json, library = "H5P.MultiChoice", taskId = "test-1")
assertNotNull(parsed)
assertEquals("test-1", parsed.id)
assertEquals("Was ist die Hauptstadt von Frankreich?", parsed.question)
assertEquals(3, parsed.options.size)
assertEquals("Paris", parsed.correctAnswer)
}
@Test
fun `test parse TrueFalse content json`() {
val json =
"""
{
"question": "<p>Die Erde ist eine Scheibe.</p>",
"correct": false
}
""".trimIndent()
val parsed = H5PParser.parseContentJson(json, library = "H5P.TrueFalse", taskId = "tf-1")
assertNotNull(parsed)
assertEquals("tf-1", parsed.id)
assertEquals("Die Erde ist eine Scheibe.", parsed.question)
assertEquals(listOf("Wahr", "Falsch"), parsed.options)
assertEquals("Falsch", parsed.correctAnswer)
}
@Test
fun `test parse ZIP stream with h5p metadata and content`() {
val baos = ByteArrayOutputStream()
ZipOutputStream(baos).use { zip ->
// Add h5p.json
val h5pEntry = ZipEntry("h5p.json")
zip.putNextEntry(h5pEntry)
zip.write(
"""
{
"title": "Mathematik Quiz",
"mainLibrary": "H5P.MultiChoice",
"license": "CC BY-SA 4.0"
}
""".trimIndent().toByteArray(Charsets.UTF_8),
)
zip.closeEntry()
// Add content/content.json
val contentEntry = ZipEntry("content/content.json")
zip.putNextEntry(contentEntry)
zip.write(
"""
{
"question": "<p>Wie viele Seiten hat ein Dreieck?</p>",
"answers": [
{"text": "3", "correct": true},
{"text": "4", "correct": false},
{"text": "5", "correct": false}
]
}
""".trimIndent().toByteArray(Charsets.UTF_8),
)
zip.closeEntry()
}
val parsed = H5PParser.parseH5PZip(ByteArrayInputStream(baos.toByteArray()), "zip-task-1")
assertNotNull(parsed)
assertEquals("zip-task-1", parsed.id)
assertEquals("Mathematik Quiz", parsed.title)
assertEquals("Wie viele Seiten hat ein Dreieck?", parsed.question)
assertEquals("3", parsed.correctAnswer)
assertEquals(listOf("3", "4", "5"), parsed.options)
assertEquals("CC BY-SA 4.0", parsed.license)
}
@Test
fun `test parse QuestionSet with multiple questions`() {
val json =
"""
{
"questions": [
{
"library": "H5P.MultiChoice 1.16",
"params": {
"question": "<p>Wie viel ist 5 + 5?</p>",
"answers": [
{"text": "10", "correct": true},
{"text": "11", "correct": false}
]
}
},
{
"library": "H5P.TrueFalse 1.8",
"params": {
"question": "<p>Wasser kocht bei 100 Grad Celsius.</p>",
"correct": true
}
}
]
}
""".trimIndent()
val parsedList = H5PParser.parseAllTasks(json, library = "H5P.QuestionSet", taskId = "qs-1", title = "Mini Quiz")
assertEquals(2, parsedList.size)
val q1 = parsedList[0]
assertEquals("qs-1-q1", q1.id)
assertEquals("Wie viel ist 5 + 5?", q1.question)
assertEquals("10", q1.correctAnswer)
val q2 = parsedList[1]
assertEquals("qs-1-q2", q2.id)
assertEquals("Wasser kocht bei 100 Grad Celsius.", q2.question)
assertEquals("Wahr", q2.correctAnswer)
}
@Test
fun `test taxonomy and grade mapping`() {
assertEquals("Mathematik", ZumApiClient.mapSubject("mathe"))
assertEquals("Deutsch", ZumApiClient.mapSubject("german language"))
assertEquals("Sachunterricht", ZumApiClient.mapSubject("Sachkunde"))
assertEquals("Englisch", ZumApiClient.mapSubject("english"))
assertEquals(6 to 7, ZumApiClient.mapGradeToAgeRange(1))
assertEquals(8 to 9, ZumApiClient.mapGradeToAgeRange(3))
assertEquals(6 to 12, ZumApiClient.mapGradeToAgeRange(null))
}
@Test
fun `test ingestion CLI batch process with in-memory repository`() =
runBlocking {
val repository = InMemoryKidioRepository()
val importedCount = H5PIngestionCli.ingestBatch(repository = repository, limit = 10)
assertTrue(importedCount > 0)
val mathTasks = repository.getTasks(subject = "Mathematik", age = 8)
assertTrue(mathTasks.isNotEmpty())
val deTasks = repository.getTasks(subject = "Deutsch")
assertTrue(deTasks.isNotEmpty())
}
}

View file

@ -210,7 +210,29 @@ class ServerTest {
assertEquals("t1", tasksJson[0].id)
assertEquals("Was ist 12 + 15?", tasksJson[0].question)
// 2. Verify Task correct answer
// 2. Filter nach Fach & Alter
val filteredResponse =
client.get("/api/v1/tasks?subject=Mathematik&age=8") {
header(HttpHeaders.Authorization, "Bearer $token")
}
assertEquals(HttpStatusCode.OK, filteredResponse.status)
val filteredJson = Json.decodeFromString<List<Task>>(filteredResponse.bodyAsText())
assertEquals(2, filteredJson.size)
// 3. Filter mit Alter außerhalb des Bereichs
val age15Response =
client.get("/api/v1/tasks?subject=Mathematik&age=15") {
header(HttpHeaders.Authorization, "Bearer $token")
}
assertEquals(HttpStatusCode.OK, age15Response.status)
val age15Json = Json.decodeFromString<List<Task>>(age15Response.bodyAsText())
assertEquals(0, age15Json.size)
// 4. Antwort enthält NIEMALS correctAnswer
assertFalse(tasksResponse.bodyAsText().contains("correctAnswer"))
assertFalse(tasksResponse.bodyAsText().contains("correct_answer"))
// 5. Verify Task correct answer
val verifyResponse =
client.post("/api/v1/tasks/verify") {
header(HttpHeaders.Authorization, "Bearer $token")
@ -222,7 +244,7 @@ class ServerTest {
assertTrue(verifyJson.isCorrect)
assertEquals(10, verifyJson.earnedMinutes)
// 3. Verify Task incorrect answer
// 6. Verify Task incorrect answer
val verifyWrongResponse =
client.post("/api/v1/tasks/verify") {
header(HttpHeaders.Authorization, "Bearer $token")