Add a benchmark for helpers.ShellEscape and optimise it
What does this MR do?
Improve the performance of helpers.ShellEscape - 1MiB of data goes from taking 430 seconds to taking 0.02 seconds. 4 orders of magnitude.
Why was this MR needed?
Encoding large blobs of data - such as an extensive tls-ca-file - is ridiculously expensive at the moment. This implementation produces the same output, just faster.
What are the relevant issue numbers?
Closes #1801 (closed)