The cache build process generally follows these steps:
- render the whole document as usual (for example a page) but insert special markers before and after the rendered segments
- parse the rendered document and extract segments by the previously added markers
This results in two artifacts:
- an array of content segments which are later stored as cache entries (if they may be cached)
- a string called "output" which is the originally rendered output but without the markers
We use non-visible ASCII characters as markers / tokens in order to minimize potential conflicts with the actual content.
Note: If you choose a different cache backend for this content cache, make sure that it is one implementing
TaggableBackendInterface.