惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

钛媒体:引领未来商业与生活新知
钛媒体:引领未来商业与生活新知
The GitHub Blog
The GitHub Blog
J
Java Code Geeks
Engineering at Meta
Engineering at Meta
N
Netflix TechBlog - Medium
A
About on SuperTechFans
博客园 - 三生石上(FineUI控件)
罗磊的独立博客
MongoDB | Blog
MongoDB | Blog
B
Blog RSS Feed
让小产品的独立变现更简单 - ezindie.com
让小产品的独立变现更简单 - ezindie.com
freeCodeCamp Programming Tutorials: Python, JavaScript, Git & More
有赞技术团队
有赞技术团队
T
Tailwind CSS Blog
酷 壳 – CoolShell
酷 壳 – CoolShell
Vercel News
Vercel News
腾讯CDC
博客园 - 聂微东
The Cloudflare Blog
F
Fortinet All Blogs
OSCHINA 社区最新新闻
OSCHINA 社区最新新闻
V
Visual Studio Blog
Last Week in AI
Last Week in AI
B
Blog

DEV Community

Authentication Security Deep Dive: From Brute Force to Salted Hashing (With Java Examples) Why AI Systems Don’t Fail — They Drift Spilling beans for how i learn for exam😁"Reinforcement Learning Cheat Sheet" I Replaced Chrome with Safari for AI Browser Automation. Here's What Broke (and What Finally Worked) How Python Borrows Other People's Work The $40 Architecture: Processing 1 Billion API Requests with 99.99% Uptime Vibe Coding: A Workflow Guide (From Zero to SaaS) Most webhook security guides protect the wrong side. The scary part is delivery. Headless CMS for TanStack Start: Build a Blog with Cosmic EU Age Verification App "Hacked in 2 Minutes" — What Actually Happened Comfy Cloud’s delete function does not actually remove files Running AI Models on GPU Cloud Servers: A Beginner Guide Event-driven media intelligence with AWS Step Functions and Bedrock I scored 500 AI prompts across 8 quality dimensions — here's what broke How to Call Google Gemini API from Next.js (Free Tier, No Backend Needed) The Portal Protocol: Reclaiming Human Connection in the Age of AI How to Fix Your Team's Scattered Knowledge Problem With a Self-Hosted Forum Intro to tc Cloud Functors: A Graph-First Mental Model for the Modern Cloud Designing Multi-Tenant Backends With Both Ownership and Team Access I Built a Neumorphic CSS Library with 77+ Components — Here's What I Learned PostgreSQL Performance Optimization: Why Connection Pooling Is Critical at Scale Cómo construí un SaaS multi-rubro para gestionar expensas en Argentina con FastAPI + Vue 3 🚀 I Built an Ethical Hacking Scanner Tool – Open Source Project I Replaced /usage and /context in Claude Code With a Single Statusline A Pythonic Way to Handle Emails (IMAP/SMTP) with Auto-Discovery and AI-Ready Design I Collected 8.9 Million Polymarket Price Points — Here's What I Found About How Markets Really Move EcoTrack AI — Carbon Footprint Tracker & Dashboard Everyone's Using AI. No One Agrees How. 5 self-hosted ebook managers worth trying in 2026 Building Your First AI Agent with LangChain: From Chatbot to Autonomous Assistant
How I Added Pre-Rendering to a Vite Multi-Page App Withou...
Bright Agbom · 2026-05-11 · via DEV Community

I run RelahConvert, an online file conversion site with 50+ tools across 25 languages. Last week I noticed a problem in my Ahrefs audit: every single one of my 1,449 pages was flagged as "H1 missing" and "Low word count."
The pages weren't empty. They had H1s, descriptions, FAQs — all rendering correctly when you visited them. But here's what crawlers were seeing:

Yep. Empty.

The Setup

The site is a Vite MPA (multi-page app), not a SPA. Each tool has its own .html file. Tool-specific JavaScript injects the page content at runtime via document.querySelector('#app').innerHTML = template.
This works great for users. The page loads, JS runs, content appears. Fast and clean.
But it breaks for crawlers that don't execute JavaScript (Ahrefs, basic bots, social media scrapers like Facebook and X). They see an empty body. Even Google, which does run JS, has a delayed second-pass rendering that costs crawl budget and makes new pages slow to index.
The Usual Solutions Didn't Fit
The standard fixes for this problem:

Server-Side Rendering (SSR) — requires a runtime backend. I'm on Cloudflare Pages (static hosting). Not happening.
Static Site Generation (SSG) — would mean restructuring around a framework like Astro or Vike. Massive refactor for a site already shipping.
Pre-rendering with Puppeteer — spinning up a headless browser for every route during build. ~3-5 minute build time hit for 1,449 pages, plus heavy dependencies.

I wanted something lighter: just inject the SEO-relevant content (H1, description, FAQs) into the static HTML at build time, without re-implementing my entire UI in Node.
The Approach: Hybrid Pre-Render via Vite Plugin
I extended the existing build plugin to read my i18n dictionaries and write the SEO content directly into each HTML file. The interactive tool UI (drag-drop, canvas operations, conversion logic) stays JS-rendered — only the content that matters for crawlers gets baked in.
Rough shape of the plugin:

function preRenderPlugin() {
  return {
    name: 'pre-render-seo',
    apply: 'build',
    closeBundle() {
      const tools = ['compress', 'resize', 'merge-pdf', ...]
      const langs = ['en', 'fr', 'es', 'de', 'ar', ...]

      for (const tool of tools) {
        for (const lang of langs) {
          const i18n = loadI18n(lang)
          const seo = i18n.seo[tool]

          const html = readHTML(`dist/${lang}/${slugFor(tool, lang)}/index.html`)
          const injected = injectSEOContent(html, {
            h1: i18n.nav_short[tool],
            description: i18n.heroDesc,
            body: seo.body,
            faqs: seo.faqs,
          })

          writeHTML(path, injected)
        }
      }
    }

The content gets injected into the

placeholder. When JS runs at runtime, it replaces the contents with the interactive UI — but the pre-rendered version is what crawlers see.

Three Gotchas

  1. Per-language metadata. My tool pages had hardcoded English

    and , even on French URLs. Pre-rendering exposed this — crawlers were seeing French H1s with English titles. I had to extend the same plugin to resolve title and meta description per language from i18n at build time.
  2. FOUC on per-language URLs. Initially I derived per-language URLs from the homepage template. JS at runtime detected the route and wiped the body to inject the tool. Result: brief flash of French content, then disappearance, then return. Fixed by switching per-language URLs to use the tool's own template as the base.

  3. Cloudflare bot protection blocking social scrapers. Unrelated discovery — after fixing the OG tags, Facebook's Sharing Debugger returned 403. Cloudflare's Browser Integrity Check was challenging the Facebook scraper. The fix was a Cloudflare dashboard config, not code.

Result

Build time went from 22.7s to ~23s. Negligible. Every tool page now ships with proper H1, description, FAQs, and internal links in the static HTML across all 25 languages. Social previews work. Ahrefs and Google can read content on first crawl.

You can see the live result on the image compressor — view-source on the page shows the pre-rendered content.

The takeaway: if you're on Vite without a framework and need crawler-friendly HTML without going full SSG, a build-time plugin that injects SEO content into your existing structure is a surprisingly clean middle ground.