In-Memory Database (Tree-Based Key-Value Store)

📌 Overview

This is a simple tree-based in-memory database that organizes data hierarchically, similar to a filesystem.

• Nodes (directories) represent hierarchical paths (/a, /a/b).
• Leaves (key-value pairs) store actual data (/a/b/c = "MOHIT").
• Dynamic node creation ensures missing parent directories are automatically created.
• Supports TCP communication for client-server interaction.

---

🏗️ How It Works

The database is structured as a tree, where:

• Root (/) is the starting point.
• Each directory (/a, /a/b) is a node.
• Each key-value pair (/a/b/c = "MOHIT") is a leaf.

🔹 Example Structure

(root)
  ├── /a
       ├── /a/b
            ├── c -> "MOHIT"

---

⚙️ Core Components

Component	Description
tree.c	Implements the in-memory database using a tree structure.
redis.c	Handles TCP communication and processes client commands.
Client	Sends commands like ADD and FIND to interact with the database.

---

📝 How Data is Stored

1️⃣ Creating Nodes (Directories)

Nodes are created dynamically when a user adds a new path.

Node *create_node(Node *parent, int8 *path);

Example

Node *users = create_node((Node *)&root, "/users");

Creates /users under root.

---

2️⃣ Storing Key-Value Pairs

Key-value pairs are stored as leaves under a node.

Leaf *create_leaf(Node *parent, int8 *key, int8 *value, int16 count);

Example

create_leaf(users, "john", "25", 2);

Stores john -> 25 under /users.

---

3️⃣ Retrieving Data

Users can retrieve values using linear lookup.

int8 *lookup_linear(int8 *path, int8 *key);

Example

lookup_linear("/users", "john");

🔹 Returns: "25"

---

🛠️ TCP Communication

Users interact with the database over TCP using commands.

Supported Commands

Command	Description
ADD /a/b key value	Adds a key-value pair under /a/b.
FIND /a/b key	Retrieves the value of key under /a/b.

Example TCP Requests

Client Sends

add /users/john age 25

Server Response

Added: john -> 25 under /users

Client Sends

find /users john

Server Response

Found: john -> 25

---

🚀 Handling Missing Parent Nodes

The system automatically creates missing parent directories when adding data.

Before Fix

❌ Fails if /a is missing before /a/b is created.

After Fix

✅ Automatically creates /a when adding /a/b.

Node *create_node(Node *parent, int8 *path)
{
    if (!parent) {
        char parent_path[256];
        strncpy(parent_path, path, 255);
        char *last_slash = strrchr(parent_path, '/');
        if (last_slash) {
            *last_slash = '\0';
            parent = find_node_linear(parent_path);
            if (!parent) {
                parent = create_node((Node *)&root, parent_path);
            }
        }
    }
    ...
}

🎯 How to Run

1️⃣ Compile the Code

make

2️⃣ Run the Server

./redis 

3️⃣ Connect Using a Client (e.g., Netcat)

telnet localhost 6379

4️⃣ Send Commands

add /users/john age 25
find /users john

---

📝 Conclusion

• Tree-based hierarchical database
• Stores key-value pairs inside nodes
• Handles TCP requests for ADD and FIND
• Automatically creates missing parent directories
• Optimizable for larger datasets

Would you like hash-based lookups or disk storage integration next? 🚀