Web Browser Extension Workshop - Part 4

This is the fourth part of the Web Browser Extension Workshop series. In this part, we will populate a database with the OpenGraph information parsed from the Webpage on active tabs. For this, we will implement a service that we can use later to query the OpenGraph information as well.

Overview

In the previous blog post, I concluded by saying that we would show the OpenGraph information to the user in the popup. However, I realized that we need to refactor the code a bit before we can do that, and I do not want to rush things up.

So in this part, we will implement a service that will populate a database with the OpenGraph information parsed from the Webpage on the active tab. This service will also provide a method to query the OpenGraph information and list all the OpenGraph information in the database.

First, let me clarify one thing: Why do we need such a service? Why do we need a database at all?

The information parsed from the Webpage is not persistent. It is in the background process and there are not many ways to access it from the popup. We could have used a messaging mechanism over content scripts, but it would be an overkill for the job.

Using a database-backed service, we can store the information in the database and query it whenever we need it and from wherever we need it, most importantly from the popup.

Refactoring

We will use the @/utils directory to store the utility definitions, in particular the database and service implementations. Both the database and the service need to know our data types.

Let's create a new file @/utils/types.ts and move our data types there:

diff --git a/entrypoints/background.ts b/entrypoints/background.ts
index 0fbfeb9..5c17785 100644
--- a/entrypoints/background.ts
+++ b/entrypoints/background.ts
@@ -1,9 +1,5 @@
-import {
-  ParseResult,
-  ParseResultError,
-  ParseResultSuccess,
-} from "@/utils/types";
 import ogs from "open-graph-scraper-lite";
+import type { SuccessResult } from "open-graph-scraper-lite";

 export default defineBackground(() => {
   browser.tabs.onActivated.addListener(({ tabId }) => {
@@ -80,6 +76,27 @@ export async function parse(tabId: number): Promise<ParseResult> {
   return result;
 }

+export type OgObject = SuccessResult["result"];
+
+export type ParseResult =
+  | ParseResultSuccess
+  | ParseResultError
+  | ParseResultNotApplicable;
+
+export type ParseResultSuccess = {
+  status: "success";
+  ogdata: OgObject;
+};
+
+export type ParseResultError = {
+  status: "error";
+  ogdata?: OgObject;
+};
+
+export type ParseResultNotApplicable = {
+  status: "not-applicable";
+};
+
 export const ICONS = {
   default: {
     16: "icon/16.png",
diff --git a/utils/types.ts b/utils/types.ts
deleted file mode 100644
index 2f6ffb7..0000000
--- a/utils/types.ts
+++ /dev/null
@@ -1,23 +0,0 @@
-import type { SuccessResult } from "open-graph-scraper-lite";
-
-export type OgObject = SuccessResult["result"];
-
-export type ParseResult =
-  | ParseResultSuccess
-  | ParseResultError
-  | ParseResultNotApplicable;
-
-export type ParseResultSuccess = {
-  status: "success";
-  ogdata: OgObject;
-};
-
-export type ParseResultError = {
-  status: "error";
-  ogdata?: OgObject;
-};
-
-export type ParseResultNotApplicable = {
-  status: "not-applicable";
-};
-

Database and Service

The next step will be the implementation of the database under @/utils/database.ts:

import { DBSchema, IDBPDatabase, openDB } from "idb";

interface ExtensionDatabaseSchema extends DBSchema {
  ogs: {
    key: string;
    value: ParseResultSuccess;
  };
}

export type ExtensionDatabase = IDBPDatabase<ExtensionDatabaseSchema>;

export function openExtensionDatabase(): Promise<ExtensionDatabase> {
  return openDB<ExtensionDatabaseSchema>("ogpatrol-service", 1, {
    upgrade(database) {
      database.createObjectStore("ogs", { keyPath: "source" });
    },
  });
}

As you may have noticed, we are using the idb library to interact with the IndexedDB database. The database schema is simple: we have a single object store ogs that stores the successfully parsed OpenGraph information (ParseResultSuccess) with the URL (source) as the key.

Our record, however, does not have a source field yet. We will add it to the record type:

diff --git a/utils/types.ts b/utils/types.ts
index 2f6ffb7..a0eda50 100644
--- a/utils/types.ts
+++ b/utils/types.ts
@@ -9,6 +9,7 @@ export type ParseResult =

 export type ParseResultSuccess = {
   status: "success";
+  source: string;
   ogdata: OgObject;
 };

Now we can implement the service that will populate the database with the OpenGraph information:

import { defineProxyService } from "@webext-core/proxy-service";
import type { ExtensionDatabase } from "./database";

export interface Service {
  getAll(): Promise<ParseResultSuccess[]>;
  find(key: string): Promise<ParseResultSuccess | undefined>;
  upsert(record: ParseResultSuccess): Promise<void>;
}

function createService(_db: Promise<ExtensionDatabase>): Service {
  return {
    async getAll() {
      const db = await _db;
      return await db.getAll("ogs");
    },

    async find(key) {
      const db = await _db;
      return await db.get("ogs", key);
    },

    async upsert(record) {
      const db = await _db;
      await db.put("ogs", record);
    },
  };
}

export const [registerService, getService] = defineProxyService(
  "ogpatrol-service",
  createService,
);

The service is simple, too: it provides three methods to get all the records, to find a record by key, and to upsert a record. Note that we are wrapping our service with @webext-core/proxy-service that allows us to call service methods anywhere but execute them in the background process.

Using the Service

Finally, we need to register and use the service in the background process:

diff --git a/entrypoints/background.ts b/entrypoints/background.ts
index 0fbfeb9..5e4ace4 100644
--- a/entrypoints/background.ts
+++ b/entrypoints/background.ts
@@ -1,3 +1,4 @@
+import { registerService } from "@/utils/service";
 import {
   ParseResult,
   ParseResultError,
@@ -6,18 +7,24 @@ import {
 import ogs from "open-graph-scraper-lite";

 export default defineBackground(() => {
+  // Open extension database:
+  const db = openExtensionDatabase();
+
+  // Register our service:
+  const service = registerService(db);
+
   browser.tabs.onActivated.addListener(({ tabId }) => {
-    process(tabId);
+    process(service, tabId);
   });

   browser.tabs.onUpdated.addListener((tabId, changeInfo) => {
     if (changeInfo?.status === "complete") {
-      process(tabId);
+      process(service, tabId);
     }
   });
 });

-export async function process(tabId: number): Promise<void> {
+export async function process(service: Service, tabId: number): Promise<void> {
   // Reset the icon:
   setIcon();

@@ -25,8 +32,12 @@ export async function process(tabId: number): Promise<void> {
   const result = await parse(tabId);

   // Handle the result:
-  console.log(result);
   setIcon(result);
+
+  // Store the result in the database:
+  if (result.status === "success") {
+    service.upsert(result);
+  }
 }

 export async function parse(tabId: number): Promise<ParseResult> {
@@ -69,7 +80,11 @@ export async function parse(tabId: number): Promise<ParseResult> {
       }

       // Return with success:
-      return { status: "success", ogdata: result } as ParseResultSuccess;
+      return {
+        status: "success",
+        source: url,
+        ogdata: result,
+      } as ParseResultSuccess;
     })
     .catch((error) => {
       console.error("Error while trying to extract OpenGraph data.", error);

Changes are straightforward: we open the database and register the service in the background process. We also pass the service to the process function and store the parse result in the database if it is successful. Note how we inject the URL in to the record for successful parse results.

Wrap Up

In this part, we implemented a service that populates a database with the successfully parsed OpenGraph information for each URL we visit. We also provided a method to query the OpenGraph information and list all the OpenGraph information in the database.

With these changes, we have completed the refactoring of the codebase. This time, we are really really really ready to show the OpenGraph information to the user in the popup. In the next part, we will see how.

0
Subscribe to my newsletter

Read articles from Vehbi Sinan Tunalioglu directly inside your inbox. Subscribe to the newsletter, and don't miss out.

Written by

Vehbi Sinan Tunalioglu
Vehbi Sinan Tunalioglu

My name is Sinan. I am a computer programmer and a life-style entrepreneur. You can check my LinkedIn and GitHub profile pages for more information, and send an email to vst@vsthost.com to contact me. I am re-publishing my technical blog posts on hashnode. My website is available on thenegation.com, and its source code is available on GitHub.