Is there any way to add additional search index without algolia?

[shravanngoswamii]

I was making the website for Turing.jl, and they wanted to keep the documentation website and main website different so I made two repository for each website and there navbar and footer look exactly like each other, main website is served at turinglang.org through github pages and docs website is served at turinglang.org/docs. Now the issue is both website repos generate individual site and deploy site to there own gh-pages branch and this process generates different search.json for each site which ends up with a problem that both sites cannot search one another content!!

So what I am asking is, Is there any way to generate common search index, or maybe add one site's search index to another site in any way possible? I do not want to use algolia!

[cscheid]

So what I am asking is, Is there any way to generate common search index, or maybe add one site's search index to another site in any way possible?

I don't think so. You could try to study the contents of the .json file and write code to merge indices, but I think you're going to run into trouble with links to another top-level site.

[shravanngoswamii]

Well, I used the jq library and its solved the issue for one site but not for another! Here is what I did:
For the main site which is being served from turinglang.org/, I used this workflow:
It is adding docs sites original (rendered by quarto) search index to main site and added /docs prefix to URLs of docs search index and merged it with main sites search index and also renamed merged index as search.json and original as search_original.json and its working!

      - name: Render Quarto site
        run: quarto render

      - name: Rename original search index
        run: mv _site/search.json _site/search_original.json

      - name: Install jq
        run: sudo apt-get install jq

      - name: Fetch search_original.json from docs site
        run: curl -O https://raw.githubusercontent.com/TuringLang/docs/gh-pages/search_original.json

      - name: Fix URLs in docs site search index
        run: |
          jq 'map(if .href then .href = "docs/" + .href else . end) | 
              map(if .objectID then .objectID = "docs/" + .objectID else . end)' search_original.json > fixed_docs_search.json

      - name: Merge search indices
        run: |
          jq -s '.[0] + .[1]' _site/search_original.json fixed_docs_search.json > _site/search.json

But still this does not solve the issue for docs site as it still cannot search main sites content!
I tried to do the same process but it search URLs from turinglang.org/docs instead of directly from turinglang.org!
I also tried giving absolute URLs to main sites URLs in merged docs search index like this:
this is merged - https://github.com/shravanngoswamii/docs/blob/gh-pages/search.json
{
"objectID": "https://turinglang.org/news/posts/2022-02-17-gsoc/index.html",
"href": "https://turinglang.org/news/posts/2022-02-17-gsoc/index.html",
"title": "Google Summer of Code 2022",
"section": "",
"text": "content........"
},

but even after doing this, it searches like, https://turinglang.org/docs/https://turinglang.org/news/posts/2022-02-17-gsoc/index.html

Is there any other solution or any way to treat these absolute URLs as absolute one!

[shravanngoswamii]

update: This issue is also solved by adding ../ prefix using jq! Thanks for your time!

[shravanngoswamii]

It's solved, If any one have this common issue then you can refer these workflows:
Turing's main site: https://github.com/TuringLang/docs/blob/master/.github/workflows/publish.yml
Turing's docs site: https://github.com/TuringLang/turinglang.github.io/blob/main/.github/workflows/publish.yml