From html files?

[albertosantini]

What about generating the index also from the html files?

[jediq-ricky]

Hi Alberto

Thanks for the suggestion.

I haven't yet generated the index from HTML as the parsing of the markdown file is very rudimentary, parsing and extracting header tags from html would be a lot more involved.

[albertosantini]

Agreed.

I review my use case tomorrow and I report you a feedback, closing the issue or suggesting further steps.

[albertosantini]

Thanks for html support.

I tested Added support for html file sources and it works nice.

My use case is generating docs from markdown embedded in js with docco and creating a full text search index with lunr.

        lunr_index_generator: {
            web_docs: {
                src: ["web_docs/*.html"],
                dest: "web_docs/search_data.json"
            }
        }

(Don't forget to push a new release to npm) :)

[jediq-ricky]

Thanks Alberto, version 0.1.3 is now available in npm.

[julkue]

I don't think the HTML index generation was implemented correctly. I have a 39 KB HTML file and it will generate a 255 KB lunr.json compressed. Uncompressed it's 3906 KB 🐙
An HTML <ul> menu with a depth of e.g. 10 sub-menus will cause an nearby endless tree generated...