hold-on DOM structure for improving React SEO
npm install react-hydratablereact-hydratable is nodejs module for improving SEO performance of SPA project.
In React, npm run build create build outputs.
But index.html has just blank body. Because in SPA framework(ex: react) javascript file will render html elements.
In SPA with CSR, SEO performance is very poor. Of course, SSR framework(ex: nextjs) can solve this problem, but some project cannot be moved from CSR to SSR because of various problems.
In this case, react-hydratable can be very powerful weapon.
react-hydratable just create simple http webserver based on your SPA build output, and starting crawling html after executing javascript. Finally crawled html data is saved to appropriate build path.
So, After executing react-hydratable, index.html is prerendered.
It has fullfilled body.
Now every search bot(that cannot execute javascript, just read html file) can collect your website informations!!
1. install module from npm
npm install -D react-hydratable
2. setup
Go to your project root and open src/index.js
``javascript
//react 16, 17
import React from 'react';
import { hydrate, render } from 'react-dom';
// ...
//update here
//optional render/hydrate
if (rootElement.hasChildNodes()) {
hydrate(element, rootElement);
} else {
render(element, rootElement);
}
`
javascript
//react 18
import React from "react";
import { createRoot, hydrateRoot } from "react-dom/client";
if (rootElement.hasChildNodes()){
hydrateRoot(rootElement, element);
} else {
const root = createRoot(rootElement);
root.render(element, rootElement);
}
`
If user access crawled page, ???/html has been filled body. So browser need not break all DOM and re-render all elements.
Browser can use already-made-DOM objects. hydrate()/hydrateRoot() function support this.
That's all! react-hydratable is ready!!
react-hydratable use /build files.
Add react-hydratable script to your package.json.
package.json
`json`
{
"...": "...",
"scripts": {
"...": "...",
//add new scripts
"postbuild": "react-hydratable"
},
"...": "..."
}
This script(postbuild) means that run react-hydratable after build script is finished.
What a so simple usage.
You just type npm run build to terminal.
react-hydratable will be executed automatically after building project.
Do you wonder your prerendered pages are working correctly?
react-hydratable support "preview mode".
// package.json
`json`
{
"...": "...",
"scripts": {
"...": "...",
//add new scripts
"preview": "react-hydratable --preview"
},
"...": "..."
}
Now, you can check your prerendered pages after type npm run preview to terminal.
react-hydratable has default config.
`json`
{
"webroot": "./build",
"host": "http://localhost",
"port": 3000,
"crawlingUrls": ["/"],
"delay": 1500,
"htmlPrefix": ""
}
If you want customize settings(ex: I want to crawl html about more url), create hydratable.config.json to project root(same level with package.json).
Explaination of all configurations is at page bottom.
#### Set crawling targets(urls)
`json`
//hydratable.config.json
{
"crawlingUrls": ["/", "/copyrights"]
}
As this configuration, react-hydratable start crawling '/' and '/copyrights'.
And also, create new html file(build/copyrights/index.html) and refresh default index.html file(build/index.html).
Default crawlingUrls value is ['/'] (only crawling /build/index.html and refresh it)
Keep in mind "all type of crawlingUrl must be text/html"
Other Content-Type is not accepted
In crawling phase, load html and wait for delay.
After finishing loading, maybe html is not perfecttly loadded.
For example, if target page uses ajax, html is not perfect before ajax is finished.
So react-hydratable solve this problem as waiting some time.
Default waiting time is 1500ms.
If you want increase delay from 1500ms to 2000ms, override delay property.
hydratable.config.json
`json`
{
"delay": 2000
}
Now react-hydratable wait 2000 ms after loading page.
If you have a lot of page that will be crawling(crawlingUrls), it takes many time.
In this case, you can use multi-page crawling.
Default value is 1, this means one page(one window) is used to crawl your crawlingUrls.
You can increase page(window) count as increase pageCount value.
But if pageCount gets too big, server may be slow and also crawling result be wrong.
Recommend set between 1~4.
`json`
{
"webroot": "/your_project_root/build",
"host": "http://localhost",
"port": 3000,
"crawlingUrls": ["/"],
"delay": 1500,
"userAgent": "react-hydratable",
"htmlPrefix": "",
"pageCount": 1,
"retryCount": 1,
}
webroot: react build directory path
host: webserver host (http only)
port: webserver port
crawlingUrls: crawling target url list
delay: wait time(ms) after page is loaded
userAgent: crawler user-agent (request header)
htmlPrefix: prefix string that is added crawling result (if blank, is not included)
pageCount: multi-page crawling is support. if you increase this value, crawling speed is up, but may stress to server.
retryCount`: if crawling is fail, do retry as retryCount times automatically.