Improve In-Browser Markdown Processing Speed with Caching about shikiji HOT 4 CLOSED

Sure, PR welcome :)

from shikiji.

Sure, PR welcome :)

Adding a cache to rehypeShikiji should be pretty doable like in the snippet I showed. But i'm not very familiar with textmate's internal implementation as to how to cache the result of _loadgrammer, any guidance on where to look in the source code and how to optimize it, would adding an cache option on the top level in gethighlighter a good idea?

from shikiji.

Considering there are multiple layers to cache (code-hast cache, code-html cache, transformers cache, etc.), I think it's better to not implement the cache in core, but in the integrations instead (e.g., the remark plugin etc.). Also in mind that we might need to allow users to provide custom cache implementation like LRU, otherwise it might lead to memleak.

from shikiji.

I've just submitted a PR here: #80

For the moment, I haven't found a way to optimize or cache the grammar loading without touching the core, especially as it's invoked from the top level on each real-time render.

from shikiji.