Tokenizer: support beginning of a string ^ in actual regexp

[DmitrySoshnikov]

Currently tokenizer automatically add ^ at the beginning of all regexes. Some regexes, e.g. lookbehind assertions, can use it explicitly:

'/(?<=^[[:space:]])A(?=[[:space:]])/'

In this case we don't need to append the ^.

Note: this doesn't work:

'/^(?<=[[:space:]])A(?=[[:space:]])/'