SchemaCraft (Java/Maven)

Lightweight, in‑memory relational toolkit that lets you define schemas, create typed tables, insert tuples/rows, and perform lookups, projection, and natural joins. Types are pluggable via a small IType interface; the project ships with INT and VARCHAR(n) implementations that support fixed‑width binary serialization using ByteBuffer.

At a Glance

• Project Name: SchemaCraft
• Modules: single‑module Maven project (root)
• Maven: wrapper present (mvnw, .mvn/wrapper)
• Coordinates: schemacraft:schemacraft:1.0-SNAPSHOT
• Java Toolchain: 17+ (set via maven.compiler.release or compatible)
• Testing: JUnit Jupiter 5.11.0
• Entry Point: schemacraft.MainApplication (demo); library API is under schemacraft.*
• Key Package: schemacraft

Project Structure

SchemaCraft/
├── .mvn/
│   └── wrapper/
│       ├── maven-wrapper.jar
│       ├── maven-wrapper.properties
│       └── MavenWrapperDownloader.java
├── src/
│   ├── main/
│   │   └── java/
│   │       └── schemacraft/
│   │           ├── Constants.java
│   │           ├── ITable.java
│   │           ├── IType.java
│   │           ├── MainApplication.java
│   │           ├── Schema.java
│   │           ├── Table.java
│   │           ├── Tuple.java
│   │           ├── TypeInt.java
│   │           └── TypeVarchar.java
│   └── test/
│       └── java/
│           └── schemacraft/
│               ├── TableTest.java
│               └── TupleTest.java
├── .gitignore
├── mvnw
├── mvnw.cmd
├── pom.xml
└── README.md

Build & Test

Maven (wrapper)

# from the project root (folder containing pom.xml)
./mvnw -q clean package
./mvnw -q test

This produces target/schemacraft-1.0-SNAPSHOT.jar (no external runtime deps).

Features

• Schemas & Types

  • Schema defines an ordered set of columns (optionally a key column).
  • Built‑in types:
    • TypeInt = INT
    • TypeVarchar(max) = VARCHAR(n) with fixed maximum length and padded binary layout.
  • Add columns with helpers: addIntType(name), addVarCharType(name, max), and addKey* variants.
  • Utilities: size(), getTupleSizeInBytes(), getColumnIndex(name), getType(i), getName(i).

• Tables & Tuples

  • Table holds a collection of Tuples for a Schema.
  • Core ops: insert(Tuple), delete(key), lookup(key) (by key), lookup(colName, value) (by predicate), iterator(), size().
  • Tuple gives field‑level accessors: get(i), get(String), getInt(i), getString(i), set(i, value), getKey().

• Relational Operations

  • Schema.project(String... attrs) → new projected Schema.
  • Schema.naturaljoin(Schema other) → combined Schema (no key) for natural join.
  • Tuple.project(Schema projected) and Tuple.joinTuple(Schema joinSchema, Tuple t1, Tuple t2) offer row‑wise transforms.
  • Table.lookup(col, value) returns a new Table with matching tuples.

• Binary Serialization

  • IType.readValue(ByteBuffer) / writeValue(Object, ByteBuffer) permit fixed‑width row serialization.
  • Schema.serialize(ByteBuffer) and Tuple.serialize(ByteBuffer) support compact storage.
  • Constants.BLOCK_SIZE = 4096 suggests page/block alignment for future storage engines.

• Developer Ergonomics

  • Constants.MAX_COLUMN_NAME_LENGTH = 24 and Constants.DEBUG flag.
  • Clean, minimal API — easy to extend with custom IType implementations.

Code Map

schemacraft.Schema

• Add columns: addIntType, addVarCharType, addKeyIntType, addKeyVarCharType
• Inspect: getKey, getColumnIndex, getType, getName, getMaxSQLSize, size, getTupleSizeInBytes
• Algebra: project(String[]), naturaljoin(Schema)
• IO: serialize(ByteBuffer)
• toString(): human‑readable schema listing

schemacraft.Table

• Lifecycle: Table(Schema), close()
• Mutation: insert(Tuple), delete(key)
• Query: lookup(key), lookup(colName, value), iterator(), size()
• toString(): pretty prints rows

schemacraft.Tuple

• Construct: Tuple(Schema, Object... values); validates arity and types
• Access: get(int), get(String), getInt(int), getString(int), set(int, Object), getKey()
• Transform: project(Schema), joinTuple(Schema, Tuple, Tuple)
• IO: serialize(ByteBuffer), deserialize(Schema, ByteBuffer)

schemacraft.IType (SPI)

String getColumnName();
int getMaxSizeBytes();
int getMaxSQLLength();
String getExternalName();
int getInternalType();
Object readValue(ByteBuffer buf);
void writeValue(Object value, ByteBuffer buf);

schemacraft.TypeInt / schemacraft.TypeVarchar

• Fixed‑size ByteBuffer encoding; VARCHAR pads/truncates to max bytes.

schemacraft.Constants

• BLOCK_SIZE = 4096, MAX_COLUMN_NAME_LENGTH = 24, DEBUG

schemacraft.MainApplication

• Small demo showing schema creation, inserts, and lookup by column/key.

Usage

Below is a minimal example using the public API:

import schemacraft.*;

public class Demo {
    public static void main(String[] args) {
        // 1) Define a schema with a key column and two attributes
        Schema schema = new Schema();
        schema.addKeyIntType("ID");
        schema.addVarCharType("dept_name", 20);
        schema.addVarCharType("building", 16);

        // 2) Create a table and insert tuples
        Table table = new Table(schema);
        table.insert(new Tuple(schema, 19803, "Comp. Sci.", "SIG"));
        table.insert(new Tuple(schema, 19901, "Comp. Sci.", "GHC"));
        table.insert(new Tuple(schema, 20123, "Math",      "Wean"));

        // 3) Lookup by key and by non-key column
        Table row = table.lookup(19803);
        Table cs  = table.lookup("dept_name", "Comp. Sci.");

        // 4) Project columns
        Schema proj = schema.project(new String[] {"dept_name"});

        System.out.println(row);
        System.out.println(cs);
        System.out.println(proj);
    }
}

Compile & run your own demo (outside Maven build):

# compile (point to target jar if you also want to reuse build outputs)
javac -cp target/schemacraft-1.0-SNAPSHOT.jar Demo.java
java  Demo

Testing

JUnit 5 tests (e.g., tuple construction & schema validation):

./mvnw -q test