exvuma / gatsby-docs

Developer turned PM

Working on updates with Dependabot, sprucing up my product skills, and dusting off my math skills

Living in ATX, but you'll find me traveling around the states in my Winnebago

gatsby-docs's Issues

Make sure all browser javascript doesn't throw errors

Errors thrown currently as an example:

Uncaught SyntaxError: Unexpected token '&'
(index):13 Uncaught SyntaxError: Unexpected token '&'
(index):1 Access to link prefetch resource at 'https://developers.cloudflare.com/workers' (redirected from 'https://staging.bigfluffycloudflare.com/workers/page-data/tutorials/build-an-application/page-data.json') from origin 'https://staging.bigfluffycloudflare.com' has been blocked by CORS policy: No 'Access-Control-Allow-Origin' header is present on the requested resource.
developers.cloudflare.com/workers:1 Failed to load resource: net::ERR_FAILED
(index):1 Access to link prefetch resource at 'https://developers.cloudflare.com/workers' (redirected from 'https://staging.bigfluffycloudflare.com/workers/page-data/tutorials/build-a-serverless-function/page-data.json') from origin 'https://staging.bigfluffycloudflare.com' has been blocked by CORS policy: No 'Access-Control-Allow-Origin' header is present on the requested resource.
developers.cloudflare.com/workers:1 Failed to load resource: net::ERR_FAILED

Search nice to have for Gallery

consider re-adding:

• type filter searches to gallery
• URL query string searches (e.g. /workers/templates?q=blah searches the gallery for blah)

//   // Handle ?q query param and set default search with it,
//   // if it exists
//   const url = new URL(window.location)
//   const initialSearch = url.searchParams.get('q')
//   if (initialSearch) {
//     handleNewSearchValue(initialSearch)
//   }

Make markdown in snippets/boilerplates render

e.g.
https://cl.ly/71410541d94e

Double Check elements that seemed useless

• Assumed we were just weighting the sidebar by weight. See in hugo version sidebar.html:25-39 {{ if not .Site.Params.noHomeIcon }}, showvisitedlinks

Test SEO and meta data for site as a whole

Set up webmaster and check the site map and all that is working fine

Port all SEO meta tags

Port everything that was in baseof.html including:

• meta tags
• scripts

Check that there are no 523 errors in network tab

Current there are 523 errors on staging

Some examples:

GET https://staging.bigfluffycloudflare.com/component---src-templates-markdown-template-tsx-e80553dd79cd2ce40efb.js net::ERR_ABORTED 523
5(index):145 GET https://staging.bigfluffycloudflare.com/webpack-runtime-38809eaa3280422b456e.js net::ERR_ABORTED 523
8staging.bigfluffycloudflare.com/:5 GET https://staging.bigfluffycloudflare.com/styles-c4c8dcf3e0a9436343be.js net::ERR_ABORTED 523
(index):1114 GET https://staging.bigfluffycloudflare.com/styles-c4c8dcf3e0a9436343be.js net::ERR_ABORTED 523

This makes sense staging is just on the path /workers/*, we should make gatsby look instead for those files under /workers

Search must work

Fix left hand margin on Template gallery

Delete all old hugo files

And make sure no configuration/value is not missed in gatsby

Make sure all config from hugo accounted for

Look into how to replace the following fields in gatsby:

• disableKinds
• canonifyURLs
• PygmentsStyle
• PygmentsCodeFences

Old hugo config:

baseURL = "https://developers.cloudflare.com/workers/"
languageCode = "en-us"
title = "Cloudflare Workers Docs"
theme = "@cloudflare/hugo-cloudflare-docs"
themesDir = "node_modules"
description = "Cloudflare Workers lets you write JavaScript on the Edge"
PygmentsCodeFences = true
PygmentsStyle = "monokai"
canonifyURLs = true
disableKinds = "taxonomyTerm"
ignoreFiles = [ "style/*", "CONTRIBUTING.md" ]

Add copy functionality to gallery?

I decided to not prioritize getting Copy to Clipboard to work for snippets in the galllery

See commit: 97def14ed34cc050bcdaa899f8f51d77707277e1

Get worker routing setup

Now both gatsby and the worker is considering the /workers prefix, we should have one or the other handler all this

Check deleted template pages with content

search through old hugo files in content/templates/pages for files with content other than the template-page partial. Then restore that content

SEO meta data for template pages - title and descriptiono

Make sure the title, description meta data is accurate for template pages.

We used to use the HTML rewriter the manually rewrite those elements, but this is clearly not best practice and actually break things in gatsby. What used to be in the worker main.js:

    // TODO remove this and just set the meta title/descriptions in Gatsby
    if (/^templates\/pages(?!.+\/)/.test(pathname)){
      //Grab the template's title from the registry
      const templateIdMatch = pathname.split(/^templates\/pages(?!.+\/)/)//pathname.replace(/.*pages\//, '')
      console.log("templateIdMatch",  templateIdMatch)
      if(!templateIdMatch[1]){
        throw 'Not in a format where a template ID can be found' + pathname
      }
      const templateId = templateIdMatch[1]
      const templateResp = await fetch(templateRegURL + templateId)
      console.log(templateRegURL + templateId)
      const templateJSON = await templateResp.json()
      const metaInfo = {
        title: templateJSON.title + ' - Cloudflare Workers Docs',
        description: templateJSON.description,
      }
      // Rewrite all meta titles/descriptions the the correct title
      return await new HTMLRewriter()
        .on('meta', new MetaHandler(metaInfo))
        .on('head>title', new TitleHandler(metaInfo))
        .transform(body)
    }

Use MDX or Remark to make sure anchor links work

e.g. on quickstart http://localhost:8000/workers/quickstart#publish-your-project link does not navigate to that spot since id="publish-your-project" does not exist