React Components to PDF API with CloudRun and Terraform

Mac Chicano — Mon, 26 Dec 2022 15:24:50 +0000

Hi! Are you curious about how to create an API endpoint that generates PDF templates using React components and uploads them to Google Cloud Storage with Terraform as Infrastructure as Code? Or have you been assigned a task to do just that? If so, you've come to the right place!

Architecture

Techstack to use

Express.js
Typescript
Google Cloud Storage
react-pdf/renderer

Requirements

NPM CLI - https://docs.npmjs.com/downloading-and-installing-node-js-and-npm
Terraform CLI - https://developer.hashicorp.com/terraform/cli/commands
- If you are using M1 mac and having a trouble setting up, you can follow this GIST https://gist.github.com/mharrvic/12b46934c608b0e21d6dd3e9fdeb1669
GCLOUD CLI - https://cloud.google.com/sdk/gcloud

Express.js API Setup

You can skip this step and just use some templates available https://github.com/topics/express-typescript

If you'd like to proceed with the manual setup, let's do it!

Initialize the package.json file, open terminal

# Create a new folder
mkdir pdf-service

# Change the directory to our newly created folder
cd pdf-service

# Create package.json file
npm init --y

Install packages

# Install Dependencies
npm install express dotenv esbuild ulid rimraf

# Install DevDependencies
npm install -D @types/express @types/node typescript ts-node nodemon

Create a tsconfig.json in the root folder, and add this

{
  "compilerOptions": {
    "jsx": "react",
    "rootDir": "./src",
    "outDir": "./build",
    "target": "es6" /* Set the JavaScript language version for emitted JavaScript and include compatible library declarations. */,
    "module": "commonjs" /* Specify what module code is generated. */,
    "esModuleInterop": true /* Emit additional JavaScript to ease support for importing CommonJS modules. This enables 'allowSyntheticDefaultImports' for type compatibility. */,
    "forceConsistentCasingInFileNames": true /* Ensure that casing is correct in imports. */,
    "strict": true /* Enable all strict type-checking options. */,
    "skipLibCheck": true /* Skip type checking all .d.ts files. */
  }
}

Create a new src folder, then add an app.ts file inside it.

import * as dotenv from "dotenv";
import express from "express";

dotenv.config();

const app = express();
app.use(express.json());

app.get("/health-check", async (req, res) => {
  res.status(200).send({ message: "PDF Generator Service" });
});

app.post("/", async (req, res) => {
  try {
    res.status(200).send({
      message: "PDF Generator Service",
      url: "test",
    });
  } catch (error) {
    res.status(500).send(error);
  }
});

export { app };

Create an index.ts file inside the src folder

import { app } from "./app";

const PORT = process.env.PORT || 8080;

/**
 * Initialize app and start Express server
 */
const main = async () => {
  // Start server listening on PORT env var
  app.listen(PORT, () => console.log(`Listening on port ${PORT}`));
};

main();

The current folder structure should look like this

├── node_modules/
├── package-lock.json
├── package.json
├── src
│   ├── app.ts
│   └── index.ts
└── tsconfig.json

Let’s add dev , serve:build, and build command
Open package.json file and add these inside the scripts

"dev": "NODE_ENV=development ./node_modules/.bin/nodemon --exec './node_modules/.bin/ts-node' src/index.ts",
"serve:build": "cd build && node index.js",
"build": "rimraf build && esbuild src/index.ts --platform=node --loader:.png=file --loader:.svg=file --loader:.woff=file --bundle --minify --outfile=build/index.js"

Let’s try the dev command if it works

npm run dev

# The output should be like this
[nodemon] 2.0.20
[nodemon] to restart at any time, enter `rs`
[nodemon] watching path(s): *.*
[nodemon] watching extensions: ts,json
[nodemon] starting `./node_modules/.bin/ts-node src/index.ts`
Listening on port 8080

Open the browser with http://localhost:8080/health-check, and the output should be like this

{"message":"PDF Generator Service"}

Commit: https://github.com/mharrvic/pdf-generator-service-template/commit/bf665b3f834413a95fde548bae716dff52d5490f

React-PDF Template

Let’s now setup our API with react-pdf template

Install react-pdf/renderer package

npm install @react-pdf/renderer react@latest react-dom@latest --legacy-peer-deps

npm install -D @types/react @types/react-dom --legacy-peer-deps

Create a new template folder under src, then add a pdf-template.tsx file inside it.

// Contents from https://react-pdf.org/repl

import ReactPDF, {
  Document,
  Font,
  Image,
  Page,
  StyleSheet,
  Text,
} from "@react-pdf/renderer";

import React from "react";

Font.register({
  family: "Oswald",
  src: "https://fonts.gstatic.com/s/oswald/v13/Y_TKV6o8WovbUd3m_X9aAA.ttf",
});

const styles = StyleSheet.create({
  body: {
    paddingTop: 35,
    paddingBottom: 65,
    paddingHorizontal: 35,
  },
  title: {
    fontSize: 24,
    textAlign: "center",
    fontFamily: "Oswald",
  },
  author: {
    fontSize: 12,
    textAlign: "center",
    marginBottom: 40,
  },
  subtitle: {
    fontSize: 18,
    margin: 12,
    fontFamily: "Oswald",
  },
  text: {
    margin: 12,
    fontSize: 14,
    textAlign: "justify",
    fontFamily: "Times-Roman",
  },
  image: {
    marginVertical: 15,
    marginHorizontal: 100,
  },
  header: {
    fontSize: 12,
    marginBottom: 20,
    textAlign: "center",
    color: "grey",
  },
  pageNumber: {
    position: "absolute",
    fontSize: 12,
    bottom: 30,
    left: 0,
    right: 0,
    textAlign: "center",
    color: "grey",
  },
});

export const Quixote = () => (
  <Document>
    <Page style={styles.body}>
      <Text style={styles.header} fixed>
        ~ Created with react-pdf ~
      </Text>
      <Text style={styles.title}>Don Quijote de la Mancha</Text>
      <Text style={styles.author}>Miguel de Cervantes</Text>
      <Image
        style={styles.image}
        src="https://cdn.britannica.com/12/154812-050-D4E47005/Don-Quixote-Sancho-Panza-illustration-Miguel-de.jpg?w=300&h=169&c=crop"
      />
      <Text style={styles.subtitle}>
        Capítulo I: Que trata de la condición y ejercicio del famoso hidalgo D.
        Quijote de la Mancha
      </Text>
      <Text style={styles.text}>
        En un lugar de la Mancha, de cuyo nombre no quiero acordarme, no ha...
      </Text>
      <Text style={styles.text}>
        Es, pues, de saber, que este sobredicho hidalgo...
      </Text>
      <Text style={styles.text}>
        Con estas y semejantes razones perdía el pobre caballero el...
      </Text>
      <Text style={styles.text}>
        En resolución, él se enfrascó tanto en su lectura...
      </Text>
      <Text style={styles.subtitle} break>
        Capítulo II: Que trata de la primera salida que de su tierra hizo el
        ingenioso Don Quijote
      </Text>
      <Image
        style={styles.image}
        src="https://cdn.britannica.com/12/154812-050-D4E47005/Don-Quixote-Sancho-Panza-illustration-Miguel-de.jpg?w=300&h=169&c=crop"
      />
      <Text style={styles.text}>Hechas, pues, estas prevenciones...</Text>
      <Text style={styles.text}>
        Yendo, pues, caminando nuestro flamante aventurero...
      </Text>
      <Text style={styles.text}>Y era la verdad que por él caminaba...</Text>
      <Text style={styles.text}>Luego volvía diciendo...</Text>
      <Text style={styles.text}>
        Casi todo aquel día caminó sin acontecerle cosa que de contar fuese...
      </Text>
      <Text
        style={styles.pageNumber}
        render={({ pageNumber, totalPages }) => `${pageNumber} / ${totalPages}`}
        fixed
      />
    </Page>
  </Document>
);

export const pdfTemplate = async () => {
  return await ReactPDF.renderToStream(<Quixote />);
};

Back to our app.ts , let’s import the pdfTemplate and temporarily comment out the response with NodeJS.ReadableStream Pipe


import { pdfTemplate } from "./template/pdf-template";

....

app.post("/", async (req, res) => {
  try {
    const pdfStream = await pdfTemplate();
    res.setHeader("Content-Type", "application/pdf");
    pdfStream.pipe(res);
    pdfStream.on("end", () => console.log("Done streaming, response sent."));

    // res.status(200).send({
    //   message: "PDF Generator Service",
    //   url: "test",
    // });
  } catch (error) {
    res.status(500).send(error);
  }
});

Open your Postman/Insomnia/hoppscotch or any API Client and make a POST request to http://localhost:8080, the response should be in PDF like this

Or you can just create another GET test endpoint and directly access it to the browser

app.get("/test", async (req, res) => {
  try {
    const pdfStream = await pdfTemplate();
    res.setHeader("Content-Type", "application/pdf");
    pdfStream.pipe(res);
    pdfStream.on("end", () => console.log("Done streaming, response sent."));
  } catch (error) {
    res.status(500).send(error);
  }
});

// Open it in the browser http://localhost:8080/test

UI Template Preview

We can now access the PDF with our test endpoint. However, debugging with react-pdf might be difficult. A UI previewer could be useful.

Let’s use Vite for this use case

npm install -D vite @vitejs/plugin-react @esbuild-plugins/node-globals-polyfill @esbuild-plugins/node-modules-polyfill --legacy-peer-deps

Create a new client-preview folder under src, then add an App.tsx and index.tsx file inside it.

// App.tsx

import React from "react";

import { PDFViewer } from "@react-pdf/renderer";
import { Quixote } from "../template/pdf-template";

function App() {
  return (
    <PDFViewer height={800} width={1000}>
      <Quixote />
    </PDFViewer>
  );
}

export default App;

// index.tsx

import React from "react";

import ReactDOM from "react-dom/client";
import App from "./App";

ReactDOM.createRoot(document.getElementById("root") as HTMLElement).render(
  <App />
);

Create vite.config.ts in our root folder

import react from "@vitejs/plugin-react";
import { defineConfig } from "vite";

// You don't need to add this to deps, it's included by @esbuild-plugins/node-modules-polyfill

export default defineConfig({
  plugins: [react()],
  server: {
    port: 3001,
  },
  resolve: {
    alias: {
      // This Rollup aliases are extracted from @esbuild-plugins/node-modules-polyfill,
      // see https://github.com/remorses/esbuild-plugins/blob/master/node-modules-polyfill/src/polyfills.ts
      // process and buffer are excluded because already managed
      // by node-globals-polyfill
      util: "rollup-plugin-node-polyfills/polyfills/util",
      sys: "util",
      events: "rollup-plugin-node-polyfills/polyfills/events",
      stream: "rollup-plugin-node-polyfills/polyfills/stream",
      path: "rollup-plugin-node-polyfills/polyfills/path",
      querystring: "rollup-plugin-node-polyfills/polyfills/qs",
      punycode: "rollup-plugin-node-polyfills/polyfills/punycode",
      url: "rollup-plugin-node-polyfills/polyfills/url",
      string_decoder: "rollup-plugin-node-polyfills/polyfills/string-decoder",
      buffer: "rollup-plugin-node-polyfills/polyfills/buffer-es6",
      process: "rollup-plugin-node-polyfills/polyfills/process-es6",
      http: "rollup-plugin-node-polyfills/polyfills/http",
      https: "rollup-plugin-node-polyfills/polyfills/http",
      os: "rollup-plugin-node-polyfills/polyfills/os",
      assert: "rollup-plugin-node-polyfills/polyfills/assert",
      constants: "rollup-plugin-node-polyfills/polyfills/constants",
      _stream_duplex:
        "rollup-plugin-node-polyfills/polyfills/readable-stream/duplex",
      _stream_passthrough:
        "rollup-plugin-node-polyfills/polyfills/readable-stream/passthrough",
      _stream_readable:
        "rollup-plugin-node-polyfills/polyfills/readable-stream/readable",
      _stream_writable:
        "rollup-plugin-node-polyfills/polyfills/readable-stream/writable",
      _stream_transform:
        "rollup-plugin-node-polyfills/polyfills/readable-stream/transform",
      timers: "rollup-plugin-node-polyfills/polyfills/timers",
      console: "rollup-plugin-node-polyfills/polyfills/console",
      vm: "rollup-plugin-node-polyfills/polyfills/vm",
      zlib: "rollup-plugin-node-polyfills/polyfills/zlib",
      tty: "rollup-plugin-node-polyfills/polyfills/tty",
      domain: "rollup-plugin-node-polyfills/polyfills/domain",
    },
  },
});

Create index.html in our root folder

<!DOCTYPE html>
<html lang="en">
  <head>
    <meta charset="UTF-8" />
    <link rel="icon" type="image/svg+xml" href="/vite.svg" />
    <meta name="viewport" content="width=device-width, initial-scale=1.0" />
    <title>React PDF Service</title>
  </head>
  <body>
    <div id="root"></div>
    <script type="module" src="./src/client-preview/index.tsx"></script>
  </body>
</html>

Open the package.json and add "preview": "vite", inside scripts

...
"scripts": {
    "preview": "vite",
    "dev": "NODE_ENV=development ./node_modules/.bin/nodemon --exec './node_modules/.bin/ts-node' src/index.ts",
    "serve:build": "cd build && node index.js",
    "build": "rimraf build && esbuild src/index.ts --platform=node --loader:.png=file --loader:.svg=file --loader:.woff=file --bundle --minify --outfile=build/index.js"
  },
...

Run npm run preview, then open the URL in your browser. You should now be able to preview your PDF!

Upload Stream to GCS

Let’s create a utility that uploads the pdf stream to GCS

Install the cloud storage package (we are using --legacy-peer-deps here because we used the latest version of React and React-DOM. However, the current peer-deps of React-PDF is using the old version of React. Don't worry, we know what we're doing, so it's okay to put it there!)

npm install @google-cloud/storage --legacy-peer-deps

Create a new utils folder under src, then add uploadStreamToGcs.ts

import { Storage } from "@google-cloud/storage";
import { ulid } from "ulid";
import { format } from "util";

const storage = new Storage();

const bucketName =
  process.env.GCS_BUCKET || "any-bucket-name-may-multo-sa-likod-mo-huhu";

const storageBucket = storage.bucket(bucketName);

export const uploadStreamToGcs = async (data: NodeJS.ReadableStream) => {
  try {
    return new Promise((resolve, reject) => {
      const blob = storageBucket.file(`${ulid()}/test.pdf`);

      const blobStream = data.pipe(
        blob.createWriteStream({
          resumable: false,
        })
      );

      blobStream
        .on("finish", () => {
          const url = format(`gs://${storageBucket.name}/${blob.name}`);
          resolve(url);
        })
        .on("error", (error) => {
          reject(error);
        });
    });
  } catch (error) {
    throw new Error("Error uploading to GCS");
  }
};

Let’s modify our app.ts and use the upload to stream util to our POST endpoint

...
import { uploadStreamToGcs } from "./utils/uploadStreamToGcs";

...
app.post("/", async (req, res) => {
  try {
    // Feel free to use this body variable to pass data to your template
    const body = req.body;

    const pdfStream = await pdfTemplate();
    const url = await uploadStreamToGcs(pdfStream);

    res.status(200).send({
      message: "PDF Generator Service",
      url,
    });
  } catch (error) {
    res.status(500).send(error);
  }
});
...

To test it locally, make sure to configure your gcloud cli and authenticate. Then, try to hit the POST endpoint. It should now be able to upload the PDF file to Google Cloud Storage – Yey!

Docker Setup

We need to set up our API with Docker, as we plan to deploy it to CloudRun.

Create Dockerfile to the root directory

FROM node:16

WORKDIR /app
COPY package*.json /app/

RUN npm ci --force --only=production
COPY . /app/

RUN npm run build
CMD ["npm", "run", "serve"]

Done! So simple right? haha

Alright, I'll explain each line with the help of ChatGPT.

# Specifies the base image to use for the Docker container.
# In this case, the base image is the latest version of Node.js with a version number of 16.
FROM node:16

# Sets the working directory for the rest of the instructions in the Dockerfile.
# This means that any commands that follow will be executed within the /app directory inside the container.
WORKDIR /app

# Copies the package.json file and any other files that match the pattern package*.json from the host system
# (i.e., the system where the Dockerfile is located) to the /app directory inside the container.
COPY package*.json /app/

# Runs the npm ci command inside the container to install the dependencies specified in the package.json file.
# The --force flag is used to force npm to install the dependencies even if there are warnings,
# and the --only=production flag tells npm to only install the dependencies needed for production (i.e., not development dependencies).
RUN npm ci --force --only=production

# Copies the entire project directory (i.e., all files and subdirectories)
# from the host system to the /app directory inside the container.
COPY . /app/

# Runs the npm run build command inside the container to build the project.
RUN npm run build

# Specifies the default command to run when the container is started.
# In this case, the command is npm run serve, which starts the server for the project.
CMD ["npm", "run", "serve:build"]

Whew! Thank you, ChatGPT!

Deployment

Registry Artifact Setup with Docker

Create an artifact registry for the pdf service

Copy, modify, and paste this to your terminal

   export PROJECT_ID=your-project-id
   export REGION=us-central1
   export REPO_NAME=pdf-service

Create artifact repo

   gcloud artifacts repositories create ${REPO_NAME} --repository-format=docker \
   --location=${REGION} --description="Docker image for pdf service"

Authenticate your Docker

   gcloud auth configure-docker ${REGION}-docker.pkg.dev

Docker Build

   docker build . --tag ${REGION}-docker.pkg.dev/${PROJECT_ID}/${REPO_NAME}/${REPO_NAME}:dev

Docker Push to Artifact Registry

   docker push ${REGION}-docker.pkg.dev/${PROJECT_ID}/${REPO_NAME}/${REPO_NAME}:dev

Terraform (Infrastructure as Code)

Terraform is currently the best way to convert infrastructure to code for me, as it supports a wide range of providers (multi-cloud). It is similar to the mantra of React Native, "Learn once, write anywhere".

Let's create an infrastructure/terraform folder in the root directory. Then, create two more folders: environment and modules. We will use the reusable modules and set up different environments.

The folder tree should look something like this:

├── infrastructure
│   └── terraform
│       ├── environment
│       └── modules

CloudRun Module

Inside the modules directory, create a cloudrun folder and create two files: main.tf and variables.tf.

# variables.tf

variable "region" {
  type = string
}

variable "project" {
  type = string
}

variable "pdf_service_account_email" {
  type = string
}

variable "artifact_repo" {
  type = string
}

variable "artifact_registry_url" {
  type = string
}

variable "gcs_bucket" {
  type = string
}

# main.tf

resource "google_cloud_run_service" "pdf_service" {
  provider = google-beta
  name     = "pdf-service"
  location = var.region
  project  = var.project

  template {
    spec {
      containers {
        image = "${artifact_registry_url}/${var.project}/${artifact_repo}/${artifact_repo}:dev"
        env {
          name  = "GCS_BUCKET"
          value = var.gcs_bucket
        }
      }

      service_account_name = var.pdf_service_account_email
    }

    metadata {
      annotations = {
        "run.googleapis.com/client-name"   = "terraform"
        "autoscaling.knative.dev/maxScale" = "1000"
      }
    }
  }

  traffic {
    percent         = 100
    latest_revision = true
  }

  autogenerate_revision_name = true
}

Service Account Module

Inside the modules directory, create a service-account folder and create three files: main.tf, outputs.tf, and variables.tf.

# variables.tf

variable "project" {
  type = string
}

variable "region" {
  type = string
}

# outputs.tf

output "pdf_service_account_email" {
  value = google_service_account.pdf_service_account.email
}

# main.tf

resource "google_service_account" "pdf_service_account" {
  provider     = google-beta
  account_id   = "pdf-service-identity"
  project      = var.project
  display_name = "PDF Service Identity"
}

resource "google_project_iam_member" "pdf_service_account" {
  project = var.project
  for_each = toset([
    "roles/storage.objectCreator",
    "roles/artifactregistry.writer",
    "roles/run.developer",
    "roles/iam.serviceAccountUser"
  ])
  role   = each.key
  member = "serviceAccount:${google_service_account.pdf_service_account.email}"
}

Dev Environment

Inside the environment directory, create a dev folder and create main.tf file

Make sure to update the local values with your configuration

provider "google" {
  project = local.project
}

locals {
  project               = "your-gcloud-project-id"
  region                = "us-central1"
  zone                  = "us-central1-a"
  artifact_registry_url = "us-central1-docker.pkg.dev"
  artifact_repo         = "pdf-service"
  gcs_bucket            = "any-gcs-bucket-name"
}

module "service_account" {
  source = "../../modules/service-account"

  project = local.project
  region  = local.region
}

module "cloudrun" {
  source = "../../modules/cloudrun"

  project                   = local.project
  region                    = local.region
  pdf_service_account_email = module.service_account.pdf_service_account_email
  artifact_registry_url     = var.artifact_registry_url
  artifact_repo             = var.artifact_repo
  gcs_bucket                = var.gcs_bucket
}

The updated folder tree should look something like this:

├── infrastructure
│   └── terraform
│       ├── environment
│       │   └── dev
│       │       └── main.tf
│       └── modules
│           ├── cloudrun
│           │   ├── main.tf
│           │   └── variables.tf
│           └── service-accounts
│               ├── main.tf
│               ├── outputs.tf
│               └── variables.tf

Run terraform init to initialize the terraform configuration

Run terraform plan to generate the execution plan

Run terraform apply to create or update infrastructure according to the execution plan

Source Code

Github Repository

DEV Community: Mac Chicano

Semantic Search with OpenAI Embeddings and Pinecone

Tech Stack

Learn More

Input record

Query