A common paradigm in most programming languages is the use of files, modules, folders, and packages to break down codebases into separate and organized segments. Well, Rust isn’t an exception to this. You can structure your code in files and modules to achieve the desired result.

While files and modules work, they come with it’s limitations, as you can only use the separated code in a single codebase. An approach Rust uses to make your code reusable across multiple projects is the use of **workspaces**.

In this guide, you will learn what Rust workspaces are, their benefits, and a practical guide on how to build a REST API using Rust workspaces.

## What are Rust Workspaces?

Rust workspace is a feature that lets you manage multiple related packages in one place. They let you structure your code for code reuse, simplify development workflow, and improve the overall developer experience.

**Benefits of Rust Workspaces:**

1. **Code reusability**: Workspaces allow you to share your code among multiple projects, which makes it easier to maintain and update.

2. **Simplified dependency management**: With workspaces, you can manage dependencies centrally for all member crates, avoid duplication, and ensure consistency.

3. **Improved organization**: Keep related projects together so that you can easily navigate and understand the codebase.

4. **Streamlined builds**: It lets you build all crates in the workspace simultaneously, which reduces build times and simplifies CI/CD processes.

In our case, you will use Rust Workspaces to build a REST API with Xata, a serverless data platform built on top of PostgreSQL for creating modern and robust applications.

![Workspace showcase](https://paper-attachments.dropboxusercontent.com/s_875B09528CC05D69F077F2DBD6CA914D522197226C156E5582878AC3B7DF4FF8_1722932707997_workspace+2.png)

The first step in creating workspaces is to architect your application by categorizing components into those that can be reused in multiple projects and those that are specific to a particular project. In this case, the data model and database service can be structured using workspaces, while the API handlers can leverage these workspaces.

## Prerequisites

To follow along with this tutorial, you’ll need to have:

-   Basic understanding of Rust
-   Xata account. Signup is [free](https://app.xata.io/signin?mode=signup?utm_source=fullstackwriter&utm_medium=fullstackwriter-blog)
-   [Postman](https://www.postman.com/downloads/) or any API testing application of your choice

## Scaffold the project

To get started, navigate to the desired directory and run the command below:

```bash
cargo new rust-worskpace && cd rust-worskpace
```

This command creates a Rust project called `rust-workspace` and navigates into the project directory.

Next, install the required dependency by modifying the `[dependencies]` section of the `Cargo.toml` file as shown below:

```toml
[dependencies]
actix-web = "4.4.1"
```

`actix-web = "4.4.1"` is a Rust-based framework for building web applications.

**Setup the database**

Log into your Xata workspace and create a `project` database with the following columns:

| **Column type** | **Column name** |
| --------------- | --------------- |
| String          | name            |
| Text            | description     |
| String          | status          |

![database](https://paper-attachments.dropboxusercontent.com/s_875B09528CC05D69F077F2DBD6CA914D522197226C156E5582878AC3B7DF4FF8_1722187264529_Screenshot+2024-07-28+at+18.18.00.png)

Next, click the **Get code snippet** button to get the database URL and navigate to your [workspace settings](https://app.xata.io/settings) to generate an API key.

![Get URL](https://paper-attachments.dropboxusercontent.com/s_875B09528CC05D69F077F2DBD6CA914D522197226C156E5582878AC3B7DF4FF8_1722187278674_image.png)

![Generate API Key](https://paper-attachments.dropboxusercontent.com/s_875B09528CC05D69F077F2DBD6CA914D522197226C156E5582878AC3B7DF4FF8_1722187482613_Screenshot+2024-07-28+at+18.23.47.png)

Finally, create `.env` file in the root directory and add the copied URL and API key.

```bash
XATA_DATABASE_URL= <REPLACE WITH THE COPIED DATABASE URL>
XATA_API_KEY=<REPLACE WITH THE COPIED API KEY>
```

With this done, you can now create the required workspaces that the application will use.

**Data model workspace**

To get started, run the command below:

```bash
cargo new --lib shared_models
```

The command above will create a **library** called `shared_models`.

Next, install the required dependency by modifying the `[dependencies]` section of the `shared_models/Cargo.toml` file as shown below:

```toml
#shared_models/Cargo.toml

[dependencies]
serde = { version = "1.0.195", features = ["derive"] }
```

Finally, update the `shared_models/src/lib.rs` file as shown below:

```rust
use serde::{Deserialize, Serialize};

#[derive(Deserialize, Serialize, Debug, Clone)]
pub struct Project {
    pub id: String,
    pub name: String,
    pub description: String,
    pub status: String,
}

#[derive(Deserialize, Serialize, Debug, Clone)]
pub struct ResponseData {
    pub records: Vec<Project>,
}

#[derive(Deserialize, Serialize, Debug, Clone)]
pub struct ProjectRequest {
    pub name: String,
    pub description: String,
    pub status: String,
}

#[derive(Deserialize, Serialize, Debug, Clone)]
pub struct ProjectResponse {
    pub id: String,
}

#[derive(Serialize, Debug, Clone)]
pub struct APIResponse<T> {
    pub status: u16,
    pub message: String,
    pub data: Option<T>,
}

#[derive(Serialize, Debug, Clone)]
pub struct APIErrorResponse {
    pub status: u16,
    pub message: String,
    pub data: Option<String>,
}
```

The snippet above does the following:

-   Imports the required dependency.
-   Creates a `Project`, `ResponseData`, `ProjectRequest`, and `ProjectResponse` structs with required properties to describe the request and response body accordingly.
-   Creates an `APIResponse`, and `APIErrorResponse` structs with the required properties needed for the API response

**Database service workspace**

To get started, run the command below:

```bash
cargo new --lib xata_client
```

The command above will create a **library** called `xata_client`.

Next, install the required dependency by modifying the `[dependencies]` section of the `xata_client/Cargo.toml` file as shown below:

```toml
#xata_client/Cargo.toml

[dependencies]
dotenv = "0.15.0"
reqwest = { version = "0.11.23", features = ["json"] }
serde_json = "1.0.111"
shared_models = { path = "../shared_models" }
```

`dotenv = "0.15.0"` is a library for loading environment variables.

`reqwest = { version = "0.11", features = ["json"] }` for making HTTP requests.

`serde_json = "1.0"` for manipulating JSON data.

`shared_models = { path = "../shared_models" }` adds the model workspace you created earlier to `xata_client` library. By doing this, you’ll be able to use the `shared_models` workspace in this newly created workspace.

Finally, update the `xata_client/src/lib.rs` file as shown below:

```rust
use std::env;

use dotenv::dotenv;
use reqwest::{header, Client, Error};
use shared_models::{Project, ProjectRequest, ProjectResponse, ResponseData};

pub struct XataClient;

impl XataClient {
    fn env_loader(key: &str) -> String {
        dotenv().ok();
        match env::var(key) {
            Ok(v) => v.to_string(),
            Err(_) => format!("Error loading environment variable"),
        }
    }

    fn init() -> Client {
        Client::new()
    }

    fn create_header() -> header::HeaderMap {
        let mut headers = header::HeaderMap::new();
        headers.insert("Content-Type", "application/json".parse().unwrap());
        headers.insert(
            header::AUTHORIZATION,
            format!("Bearer {}", XataClient::env_loader("XATA_API_KEY"))
                .parse()
                .unwrap(),
        );
        headers
    }

    pub async fn create_project(new_project: ProjectRequest) -> Result<ProjectResponse, Error> {
        let database_url = XataClient::env_loader("XATA_DATABASE_URL");
        let url = format!("{}:main/tables/Project/data", database_url);
        let client = XataClient::init()
            .post(url)
            .headers(XataClient::create_header())
            .json(&new_project)
            .send()
            .await;
        match client {
            Ok(response) => {
                let json = response.text().await?;
                let created_project: ProjectResponse = serde_json::from_str(json.as_str()).unwrap();
                Ok(created_project)
            }
            Err(error) => Err(error),
        }
    }

    pub async fn get_projects() -> Result<Vec<Project>, Error> {
        let database_url = XataClient::env_loader("XATA_DATABASE_URL");
        let url = format!("{}:main/tables/Project/query", database_url);
        let client = XataClient::init()
            .post(url)
            .headers(XataClient::create_header())
            .send()
            .await;
        match client {
            Ok(response) => {
                let json = response.text().await?;
                let response_data: ResponseData = serde_json::from_str(json.as_str()).unwrap();
                Ok(response_data.records)
            }
            Err(error) => Err(error),
        }
    }
}
```

The snippet above does the following:

-   Imports the required dependencies
-   Creates an `XataClient` struct
-   Creates an implementation block that adds `env_loader`, `init`, `create_header`, `create_project`, and `get_projects` methods. These methods are designed to load environment variables, create a connection pool for making asynchronous requests, generate request headers, create a project, and get the list of products.

## Building the APIs with the workspaces

With the workspaces fully set up, you might think you can now start using the libraries in the main project. Well, not so fast—you need to specify them as dependencies in the main project first.

While Cargo automatically recognizes the libraries as part of the workspace, you still have to explicitly add them as dependencies. If you check the `Cargo.toml` file in the root directory, you’ll see the workspaces added as shows below:

```toml
workspace = { members = ["shared_models", "xata_client"] }
```

To add the libraries as depencies, update the `[dependencies]` section by specifying the library and the associated path.

```toml
#rust-workspace/Cargo.toml

[dependencies]
actix-web = "4.4.1"
#add workspaces below
shared_models = { path = "./shared_models" }
xata_client = { path = "./xata_client" }
```

With that done, you can now use the library to build the APIs.

To do this, create a `handler.rs` file inside the `src` folder of the project and add the snippet below:

```rust
use actix_web::{http::StatusCode, post, web::Json, HttpResponse};
use shared_models::{APIErrorResponse, APIResponse, Project, ProjectRequest, ProjectResponse};
use xata_client::XataClient;

#[post("/project")]
pub async fn create_project_handler(new_project: Json<ProjectRequest>) -> HttpResponse {
    let create_project = XataClient::create_project(new_project.to_owned()).await;

    match create_project {
        Ok(data) => HttpResponse::Created().json(APIResponse::<ProjectResponse> {
            status: StatusCode::CREATED.as_u16(),
            message: "success".to_string(),
            data: Some(data),
        }),
        Err(error) => HttpResponse::InternalServerError().json(APIErrorResponse {
            status: StatusCode::INTERNAL_SERVER_ERROR.as_u16(),
            message: "failure".to_string(),
            data: Some(error.to_string()),
        }),
    }
}

#[post("/projects")]
pub async fn get_projects_handler() -> HttpResponse {
    let get_projects = XataClient::get_projects().await;

    match get_projects {
        Ok(data) => HttpResponse::Ok().json(APIResponse::<Vec<Project>> {
            status: StatusCode::OK.as_u16(),
            message: "success".to_string(),
            data: Some(data),
        }),
        Err(error) => HttpResponse::InternalServerError().json(APIErrorResponse {
            status: StatusCode::INTERNAL_SERVER_ERROR.as_u16(),
            message: "failure".to_string(),
            data: Some(error.to_string()),
        }),
    }
}
```

The snippet above does the following:

-   Imports the required dependencies (including the workspaces)
-   Creates a `create_project_handler` and `get_projects_handler` handler with corresponding API routes that use the `libraries` to perform the corresponding actions and return the appropriate response using the `APIResponse` and `APIErrorResponse`

**Putting it all together**

With that done, update the `main.rs` file to include the application entry point and use the handlers.

```rust
use actix_web::{App, HttpServer};
use handlers::{create_project_handler, get_projects_handler};
mod handlers;

#[actix_web::main]
async fn main() -> std::io::Result<()> {
    HttpServer::new(move || {
        App::new()
            .service(create_project_handler)
            .service(get_projects_handler)
    })
    .bind(("localhost", 8080))?
    .run()
    .await
}
```

The snippet above does the following:

-   Imports the required dependencies
-   Creates a new server that adds the handlers and runs on `localhost:8080`

With that done, we can start a development server using the command below:

```bash
cargo run
```

![API test](https://paper-attachments.dropboxusercontent.com/s_875B09528CC05D69F077F2DBD6CA914D522197226C156E5582878AC3B7DF4FF8_1722207963810_Screenshot+2024-07-29+at+00.05.54.png)

This post discussed what Rust workspaces are, their benefits, and a walkthrough guide on how to use them to build a REST API with Xata. Beyond what was explored above, you can extend the workspaces and the API to support deleting, editing, and getting a single project.

These resources may be helpful:

-   [Xata + Rust: A getting started guide](https://fullstackwriter.dev/post/xata-rust-a-getting-started-guide?category=rust)
-   [Xata documentation](https://xata.io/docs?utm_source=fullstackwriter&utm_medium=fullstackwriter-blog)
-   [Actix web framework](https://actix.rs/?utm_source=fullstackwriter&utm_medium=fullstackwriter-blog)


In this guide, you will learn what Rust workspaces are, their benefits, and a practical guide on how to build a REST API using Rust workspaces.

Rust Workspaces: A guide to managing your code better

As someone who actively builds applications within the JavaScript / Node.js ecosystem leveraging frameworks like Next.js, Vue, Express.js, etc., hot reloading is usually built into the framework or needs to install a package to get it up and running.

While exploring the Go ecosystem, hot reloading seems like an afterthought in most of their popular frameworks. In fact, it became a source of concern for me and led me down the rabbit hole of finding possible solutions on how to include hot reloading in my Go applications.

This guide will explore what hot reloading is and why it is needed. It will also explore two libraries to help build hot reloading into your applications.

# What is hot reloading?

Hot reloading is a technique used during application development, allowing you to observe updated changes to the source code in real-time without reloading the entire application. It preserves the application state, enabling you to continue working without losing the current state of the application. It allows you to view changes instantly, accelerating the development process and enhancing productivity.

# What are the benefits of hot reloading?

The following are the benefits of enabling hot reloading in your applications:

-   Enhanced user experience
-   Fast development
-   Minimal context switching
-   Improved feedback loop

1. **Enhanced user experience**: The ability to make changes to the codebase and see them in real-time improves the development process. With this capability, developers can rapidly prototype and design better applications.
2. **Fast development**: With hot reloading, developers don't have to manually reload the application to see the latest changes. Development time is fast because changes are reflected immediately.
3. **Minimal context switching**: A major source of distraction for developers is having to go back to the terminal to rebuild and run the application every time changes are made. However, with hot reloading, they can focus on the task without any interruptions or worries about their changes being reflected as updates occur in real-time.
4. **Improved feedback loop**: Because changes are visible in real-time, developers receive immediate feedback and can easily fix bugs, identify where the application is broken, and consequently improve the development workflow.

# How to implement hot reloading in a Go application

This section will explore two libraries that make hot reloading possible in a Go application:

-   Modd
-   Air

Before we explore how to use these libraries, let’s set up a sample Go project using the Gin-gonic framework.

Create a directory.

```bash
mkdir sample_go && cd sample_go
```

Next, initialize a Go module to manage project dependencies.

```bash
go mod init sample_go
```

Install the required dependency.

```bash
go get github.com/gin-gonic/gin
```

Create a sample application.

```go
package main

import "github.com/gin-gonic/gin"

func main() {
    r := gin.Default()
    r.GET("/", func(c *gin.Context) {
        c.JSON(200, gin.H{
            "message": "Hello world!",
        })
    })

    r.Run()
}
```

## Modd for hot reloading

[Modd](https://github.com/cortesi/modd) is a library that makes hot reloading possible in Go applications. To use it, install it on your machine using the command below:

    go install github.com/cortesi/modd/cmd/modd@latest

In your project directory, create a `modd.conf` file and add the snippet below:

```bash
**/*.go {
    prep: go test @dirmods
}
```

The command above tells **Modd** to watch for changes to any `.go` file and run the changes on `test` directories.

To enable hot reloading, run the command below in your terminal:

```bash
modd
```

> In case you get an error about `go: errors parsing go.mod: /project/path/  invalid go version '1.XX.X': must match format 1.23`. Remove the Go semantic version stated inside the `go.mod` file. For example change `go 1.22.1` to `go 1.22`

## Air for hot reloading

[Air](https://github.com/cosmtrek/air) is another library that enables hot reloading in Go applications. To use it, install it on your machine using the command below:

```
go install github.com/cosmtrek/air@latest
```

Next, initialize **Air** by running the command below:

```
air init
```

The command above will create a `.air.toml` file with default configurations to get your application up and running.

Finally, to enable hot reloading, run the command below:

```
air
```

# Conclusion

Both libraries enable hot reloading and offer extensive configuration and flexibility regarding which files and folders should be excluded or ignored, pattern matching, environment setup, and much more. To learn more, you can check out their official documentation.

Useful resources:

-   [Modd documentation](https://github.com/cortesi/modd)
-   [Air documentation](https://github.com/cosmtrek/air)


This guide will explore what hot reloading is and why it is needed. It will also explore two libraries to help build hot reloading into your applications.

How to implement hot reloading in a Go application

Go is a statically typed, compiled high-level programming language for building systems, command-line interfaces (CLI), and more. It is typically designed for use on the backend; however, there are times when you want to use the same language to build a full-stack application with a functional backend and a visual frontend.

In most cases, Go developers opt for frontend frameworks/libraries like React, Vue, Angular, etc., to build the frontend part of the application. This means they must learn JavaScript/TypeScript, framework-specific paradigms, and other frontend-related overheads.

In this guide, you’ll learn how to build a fullstack application with Go using Templ, HTMX, and Xata.

# Technology Overview

**Templ**: is a templating engine that lets you build HTML with Go. It also lets you use Go syntax like `if`, `switch`, and `for` statements to build a robust frontend. You will use Templ to build reusable components and pages for the frontend.

**HTMX**: is a frontend library that lets you access modern browser features directly using HTML rather than JavaScript. You will use HTMX to process the form submission and perform other dynamic operations.

**Xata**: is a serverless database with analytics and free-text search support that makes a wide range of applications easy to build.

# Prerequisites

To follow along with this tutorial, the following are needed:

-   [Go version 1.20 or higher installed](https://go.dev/dl/)
-   Basic understanding of Go
-   Xata account. [Signup is free](https://app.xata.io/signin?mode=signup?utm_source=fullstackwriter&utm_medium=fullstackwriter-blog)

# Getting started

To get started, you need to install Templ binary. The binary generates Go code from a Templ file.

```bash
go install github.com/a-h/templ/cmd/templ@latest
```

Create a directory.

```bash
mkdir go_fullstack && cd go_fullstack
```

Next, initialize a Go module to manage project dependencies.

```bash
go mod init go-fullstack
```

Finally, we proceed to install the required dependencies with:

```bash
go get github.com/gin-gonic/gin github.com/a-h/templ github.com/joho/godotenv
```

`github.com/gin-gonic/gin` is a framework for building web applications.

`github.com/a-h/templ` is the Templ library used in the project.

`github.com/joho/godotenv` is a library for loading environment variables.

## Structuring the application

To do this, create a `cmd`, `internals`, and `views` folder in our project directory.

`cmd` is for structuring the application entry point.

`internals` is for structuring API-related files.

`views` is for structuring frontend-related files.

# Setup the database on Xata

Log into the [Xata workspace](https://app.xata.io/workspaces) and create a `todo` database. Inside the `todo` database, create a `Todo` table and add a `description` column of type `String`.

![create project](https://paper-attachments.dropboxusercontent.com/s_724B5DB0ED02FBEDE105C37DC2102A0D29EC66EAFDCC27146EA9AEF451272040_1703592508608_Screenshot+2023-12-26+at+13.07.33.png)

## Get the Database URL and set up the API Key

To get the database URL, click the **Get code snippet** button and copy the URL. Then click the **API Key** link, add a new key, save and copy the API key.

![](https://paper-attachments.dropboxusercontent.com/s_724B5DB0ED02FBEDE105C37DC2102A0D29EC66EAFDCC27146EA9AEF451272040_1703596013324_Screenshot+2023-12-26+at+14.04.29.png)
![](https://paper-attachments.dropboxusercontent.com/s_724B5DB0ED02FBEDE105C37DC2102A0D29EC66EAFDCC27146EA9AEF451272040_1703596030165_Screenshot+2023-12-21+at+23.42.03.png)

**Setup environment variable**

Create a `.env` file in the root directory and add the copied URL and API key.

```bash
XATA_DATABASE_URL= <REPLACE WITH THE COPIED DATABASE URL>
XATA_API_KEY=<REPLACE WITH THE COPIED API KEY>
```

# Build the application Frontend

To build the frontend, you’ll use Templ and HTMX to structure the application and add dynamism to a Todo application.

## Create the application components

Inside the `views` folder, create a `components/header.templ` file and add the snippet below:

```go
package components

templ Header() {
    <head>
        <script
            src="https://unpkg.com/htmx.org@1.9.10"
            integrity="sha384-D1Kt99CQMDuVetoL1lrYwg5t+9QdHe7NLX/SoJYkXDFfX37iInKRy5xLSi8nO7UC"
            crossorigin="anonymous"
        ></script>
        <script src="https://cdn.tailwindcss.com"></script>
        <meta charset="UTF-8" />
        <meta name="viewport" content="width=device-width, initial-scale=1.0" />
        <title>GO Fullstack</title>
    </head>
}
```

The snippet creates a `Header` component and adds HTMX and TailwindCSS CDNs. TailwindCSS is a low-level framework for styling.

Next, create a `components/footer.templ` file to create the application footer and style using TailwindCSS classes.

```go
package components

templ Footer() {
    <footer class="fixed p-1 bottom-0 bg-gray-100 w-full border-t">
    <div class="rounded-lg p-4 text-xs italic text-gray-700 text-center">
        &copy; Go Fullstack
    </div>
    </footer>
}
```

Finally, create an `index.templ` file inside the same `views` folder and add the snippet below:

```go
package views

import (
    "fmt"
    "go_fullstack/views/components"
)

type Todo struct {
    Id          string
    Description string
}

templ Index(todos []*Todo) {
    <!DOCTYPE html>
    <html lang="en">
        @components.Header()
        <body>
            <main class="min-h-screen w-full">
                <nav class="flex w-full border border-b-zinc-200 px-4 py-4">
                    <h3 class="text-base lg:text-lg font-medium text-center">
                        GO Fullstack app
                    </h3>
                </nav>
                <div class="mt-6 w-full flex justify-center items-center flex-col">
                    // FORM PROCESSING
                    <form
                        hx-post="/"
                        hx-trigger="submit"
                        hx-swap="none"
                        onsubmit="reloadPage()"
                        class="w-96"
                    >
                        <textarea
                            name="description"
                            cols="30"
                            rows="2"
                            class="w-full border rounded-lg mb-2 p-4"
                            placeholder="Input todo details"
                            required
                        ></textarea>
                        <button
                            class="py-1 px-4 w-full h-10 rounded-lg text-white bg-zinc-800"
                        >
                            Create
                        </button>
                    </form>
                    <section class="border-t border-t-zinc-200 mt-6 px-2 py-4 w-96">
                        // LOOP THROUGH THE TODOS
                        <ul id="todo-list">
                            for _, todo := range todos {
                                <li class="ml-4 ml-4 border p-2 rounded-lg mb-2" id={ fmt.Sprintf("%s", todo.Id) }>
                                    <p class="font-medium text-sm">Todo item { todo.Id }</p>
                                    <p class="text-sm text-zinc-500 mb-2">
                                        { todo.Description }
                                    </p>
                                    <div class="flex gap-4 items-center mt-2">
                                        <a
                                            href="#"
                                            class="flex items-center border py-1 px-2 rounded-lg"
                                        >
                                            <p class="text-sm">Edit</p>
                                        </a>
                                        <button
                                            hx-delete={ fmt.Sprintf("/%s", todo.Id) }
                                            hx-swap="delete"
                                            hx-target={ fmt.Sprintf("#%s", todo.Id) }
                                            class="flex items-center border py-1 px-2 rounded-lg hover:bg-red-300"
                                        >
                                            <p class="text-sm">Delete</p>
                                        </button>
                                    </div>
                                </li>
                            }
                        </ul>
                    </section>
                </div>
            </main>
        </body>
        @components.Footer()
    </html>
    <script>
        function reloadPage() {
                setTimeout(function() {
            window.location.reload();
        }, 2000);
        }
    </script>
}
```

The snippet above does the following:

-   Imports the required dependencies
-   Creates a `Todo` struct to represent the response data coming from the backend
-   Creates an `Index` component that uses the `Header` and `Footer` components to structure the page. Then, it uses the HTMX attributes to process form submissions and deletion of todos by calling the respective endpoints `/` and `/{todo.Id}`

## Generating Go files from the Templ files

Next, use the Templ binary you installed earlier to Generate Go codes from the views created above by running the command below in your terminal:

```bash
templ generate
```

After you run this command, you’ll see new Go files generated for each view. You use generated file to render your frontend in next section.

![Generated files](https://paper-attachments.dropboxusercontent.com/s_F11F11938F7C8DFA1B769A23ABB6104DEEFCB6B159F515F54CB1BDB5F18F2959_1713349771513_Screenshot+2024-04-17+at+11.28.49.png)

> The generated files are not to be edited.

# Putting it together and building the backend

With that done, you can use it to build the backend and render the required page.

## Create the API models and helper function

To represent the application data, create a `model.go` file inside the `internals` folder and add the snippet below:

```go
package internals

type Todo struct {
    Id          string `json:"id,omitempty"`
    Description string `json:"description,omitempty"`
}

type TodoRequest struct {
    Description string `json:"description,omitempty"`
}

type TodoResponse struct {
    Id string `json:"id,omitempty"`
}
```

The snippet above creates a `Todo`, `TodoRequest`, and `TodoResponse` struct with the required properties to describe requests and response types.

Finally, create a `helpers.go` file with a reusable function to load environment variables.

```go
package internals

import (
    "log"
    "os"
    "github.com/joho/godotenv"
)

func GetEnvVariable(key string) string {
    err := godotenv.Load()
    if err != nil {
        log.Fatal("Error loading .env file")
    }
    return os.Getenv(key)
}
```

## Create the application and API routes

Create a `route.go` file for configuring the API routes and add the snippet below:

```go
package api

import "github.com/gin-gonic/gin"

type Config struct {
    Router *gin.Engine
}

func (app *Config) Routes() {
    //routes will come here
}
```

The snippet above does the following:

-   Imports the required dependency
-   Creates a `Config` struct with a `Router` property to configure the application methods
-   Creates a `Routes` function that takes in the `Config` struct as a pointer

## Create the API services

With that done, create a `xata_service.go` file for the application and update it by doing the following:

First, import the required dependencies and create a helper function:

```go
package internals

import (
    "bytes"
    "encoding/json"
    "fmt"
    "net/http"
)

var xataAPIKey = GetEnvVariable("XATA_API_KEY")
var baseURL = GetEnvVariable("XATA_DATABASE_URL")

func createRequest(method, url string, bodyData *bytes.Buffer) (*http.Request, error) {
    var req *http.Request
    var err error'

    if method == "GET" || method == "DELETE" || bodyData == nil {
        req, err = http.NewRequest(method, url, nil)
    } else {
        req, err = http.NewRequest(method, url, bodyData)
    }

    if err != nil {
        return nil, err
    }

    req.Header.Add("Content-Type", "application/json")
    req.Header.Add("Authorization", fmt.Sprintf("Bearer %s", xataAPIKey))

    return req, nil
}
```

The snippet above does the following:

-   Imports the required dependencies
-   Creates required environment variables
-   Creates a `createRequest` function that creates HTTP requests with the required headers

Lastly, add a `createTodoService`, `deleteTodoService`, and `getAllTodosService` methods to create, delete, and get the list of todos.

```go
//imports goes here

func createRequest(method, url string, bodyData *bytes.Buffer) (*http.Request, error) {
    //createRequest code goes here
}

func (app *Config) createTodoService(newTodo *TodoRequest) (*TodoResponse, error) {
    createTodo := TodoResponse{}
    jsonData := Todo{
        Description: newTodo.Description,
    }

    bodyData := new(bytes.Buffer)
    json.NewEncoder(bodyData).Encode(jsonData)

    fullURL := fmt.Sprintf("%s:main/tables/Todo/data", baseURL)
    req, err := createRequest("POST", fullURL, bodyData)
    if err != nil {
        return nil, err
    }

    client := &http.Client{}
    resp, err := client.Do(req)
    if err != nil {
        return nil, err
    }

    defer resp.Body.Close()
    if err := json.NewDecoder(resp.Body).Decode(&createTodo); err != nil {
        return nil, err
    }

    return &createTodo, nil
}
func (app *Config) deleteTodoService(id string) (string, error) {
    fullURL := fmt.Sprintf("%s:main/tables/Todo/data/%s", baseURL, id)
    client := &http.Client{}
    req, err := createRequest("DELETE", fullURL, nil)
    if err != nil {
        return "", err
    }

    resp, err := client.Do(req)
    if err != nil {
        return "", err
    }
    defer resp.Body.Close()

    return id, nil
}

func (app *Config) getAllTodosService() ([]*Todo, error) {
    var todos []*Todo

    fullURL := fmt.Sprintf("%s:main/tables/Todo/query", baseURL)
    client := &http.Client{}
    req, err := createRequest("POST", fullURL, nil)
    if err != nil {
        return nil, err
    }

    resp, err := client.Do(req)
    if err != nil {
        return nil, err
    }

    defer resp.Body.Close()
    var response struct {
        Records []*Todo `json:"records"`
    }

    decoder := json.NewDecoder(resp.Body)
    if err := decoder.Decode(&response); err != nil {
        return nil, err
    }

    todos = response.Records
    return todos, nil
}
```

## Create the API handlers

With that done, you can use the services to create the API handlers. Create a `handler.go` file inside `internals` folder and add the snippet below:

```go
package internals

import (
    "context"
    "fmt"
    "go_fullstack/views"
    "net/http"
    "time"
    "github.com/a-h/templ"
    "github.com/gin-gonic/gin"
)

const appTimeout = time.Second * 10

func render(ctx *gin.Context, status int, template templ.Component) error {
    ctx.Status(status)
    return template.Render(ctx.Request.Context(), ctx.Writer)
}

func (app *Config) indexPageHandler() gin.HandlerFunc {
    return func(ctx *gin.Context) {
        _, cancel := context.WithTimeout(context.Background(), appTimeout)
        defer cancel()

        todos, err := app.getAllTodosService()
        if err != nil {
            ctx.JSON(http.StatusBadRequest, err.Error())
            return
        }

        var viewsTodos []*views.Todo
        for _, todo := range todos {
            viewsTodo := &views.Todo{
                Id:          todo.Id,
                Description: todo.Description,
            }
            viewsTodos = append(viewsTodos, viewsTodo)
        }

        render(ctx, http.StatusOK, views.Index(viewsTodos))
    }
}

func (app *Config) createTodoHandler() gin.HandlerFunc {
    return func(ctx *gin.Context) {
        _, cancel := context.WithTimeout(context.Background(), appTimeout)
        description := ctx.PostForm("description")
        defer cancel()

        newTodo := TodoRequest{
            Description: description,
        }

        data, err := app.createTodoService(&newTodo)
        if err != nil {
            ctx.JSON(http.StatusBadRequest, err.Error())
            return
        }

        ctx.JSON(http.StatusCreated, data)
    }
}

func (app *Config) deleteTodoHandler() gin.HandlerFunc {
    return func(ctx *gin.Context) {
        _, cancel := context.WithTimeout(context.Background(), appTimeout)
        id := ctx.Param("id")
        defer cancel()

        data, err := app.deleteTodoService(id)
        if err != nil {
            ctx.JSON(http.StatusBadRequest, err.Error())
            return
        }

        ctx.JSON(http.StatusAccepted, fmt.Sprintf("Todo with ID: %s deleted successfully!!", data))
    }
}
```

The snippet above does the following

-   Imports the required dependencies
-   Creates a `render` function that uses the `Templ` package to render matching template
-   Creates an `indexPageHandler` function that returns a Gin-gonic handler and takes in the `Config` struct as a pointer. Inside the returned handler, use the `getAllTodosService` service to get the list of todos and then render the appropriate page using the generated code from the views package (frontend)
-   Creates a `createdTodoHandler` and `deleteProjectHandler` functions that return a Gin-gonic handler and take in the `Config` struct as a pointer. Use the service created earlier to perform the corresponding action inside the returned handler

## Update the API routes to use handlers

Update the `routes.go` file with the handlers as shown below:

```go
package internals

import (
    "github.com/gin-gonic/gin"
)

type Config struct {
    Router *gin.Engine
}

func (app *Config) Routes() {
    //views
    app.Router.GET("/", app.indexPageHandler())

    //apis
    app.Router.POST("/", app.createTodoHandler())
    app.Router.DELETE("/:id", app.deleteTodoHandler())
}
```

## Putting it all together

Create the application entry point to use to serve the routes. To do this, create a `main.go` file inside the `cmd` folder and add the snippet below:

```go
package main

import (
    "go_fullstack/internals"
    "github.com/gin-gonic/gin"
)

func main() {
    router := gin.Default()

    //initialize config
    app := internals.Config{Router: router}

    //routes
    app.Routes()

    router.Run(":8080")
}
```

The snippet above does the following:

-   Imports the required dependencies
-   Creates a Gin router using the `Default` configuration
-   Initialize the `Config` struct by passing in the `Router`
-   Adds the route and run the application on port `:8080`

With that done, you can start a development server using the command below:

    go run cmd/main.go

![](https://media.giphy.com/media/v1.Y2lkPTc5MGI3NjExdmh4OGFocG9qMndmbXpsbW1hY2Q3dXRhYjA5OG9wdXppc3d3MjRrdCZlcD12MV9pbnRlcm5hbF9naWZfYnlfaWQmY3Q9Zw/gPVi98WZGeVZuCgBG9/giphy.gif)

The complete source code can be found on [GitHub](https://github.com/Mr-Malomz/go_fullstack).

# Conclusion

This post discusses how to build a fullstack application with Go, Templ, HTMX, and Xata. You can extend the application further to support viewing and editing todos.

These resources may also be helpful:

-   [Xata documentation](https://xata.io/docs?utm_source=fullstackwriter&utm_medium=fullstackwriter-blog)
-   [Templ documentation](https://templ.guide/?utm_source=fullstackwriter&utm_medium=fullstackwriter-blog)
-   [HTMX documentation](https://htmx.org/docs/?utm_source=fullstackwriter&utm_medium=fullstackwriter-blog)
-   [Go + Xata](https://fullstackwriter.dev/post/xata-go-a-getting-started-guide?category=Golang)


In this guide, you’ll learn how to build a fullstack application with Go using Templ, HTMX, and Xata.

How to build a fullstack application with Go, Templ, and HTMX

This post discusses what Xata is and provides a detailed step-by-step guide to using it to build a basic project management application with Flutter

Read about

Fullstack Development

blog

Contact

featured post

Rust Workspaces: A guide to managing your code better

How to implement hot reloading in a Go application

How to build a fullstack application with Go, Templ, and HTMX

Xata + Flutter: A getting started guide.