Optimise content parsing hook

[Dave Stewart]

Background

Right now, the module hooks into Content's content:file:afterParse hook to process image paths.

The callback hook receives the the markdown AST which – depending on the content – can be quite large.

For example, transformed code can generate many, many spans, so in order to exclude code and potentially other invalid elements, but include custom components we use a tags lookup to determine whether to walk a visited element.

There is possibly the potential to optimise this:

• using a remark plugin to replace images before code conversion has begin
• hooking into content:file:beforeParse and using much simpler parsing / regular expressions to match and replace potential paths

Proposal

Spike the regexp route and see if it's simple / reliable, then look into seeing how the remark plugin approach might work.