Realistically speaking it's mostly useful for scraping, since parser works only in immutable state - it creates dom elements referencing original string, so changes to it are not cost effective. This approach has the benefit of increadibly small memory usage.

Here's python api if you want to test it https://github.com/TUVIMEN/reliq-python , pypi packages are compiled for windows, x86 aarch64 armv7 linux, and macos.