Parsing **_* **_* **_* … takes quadratic time (or worse?)

[andersk]

$ time node -e 'require("markdown-it")().render("**_* ".repeat(10000))'
1.01user 0.02system 0:01.00elapsed 104%CPU (0avgtext+0avgdata 57200maxresident)k
0inputs+0outputs (0major+8600minor)pagefaults 0swaps
$ time node -e 'require("markdown-it")().render("**_* ".repeat(20000))'
3.56user 0.02system 0:03.53elapsed 101%CPU (0avgtext+0avgdata 74648maxresident)k
0inputs+0outputs (0major+13727minor)pagefaults 0swaps
$ time node -e 'require("markdown-it")().render("**_* ".repeat(40000))'
19.54user 0.05system 0:19.54elapsed 100%CPU (0avgtext+0avgdata 118080maxresident)k
0inputs+0outputs (0major+25527minor)pagefaults 0swaps
$ time node -e 'require("markdown-it")().render("**_* ".repeat(80000))'
127.43user 0.07system 2:07.49elapsed 100%CPU (0avgtext+0avgdata 177084maxresident)k
0inputs+0outputs (0major+46482minor)pagefaults 0swaps
$ time node -e 'require("markdown-it")().render("**_* ".repeat(160000))'
679.73user 0.15system 11:20.66elapsed 99%CPU (0avgtext+0avgdata 299880maxresident)k
0inputs+0outputs (0major+89443minor)pagefaults 0swaps

[rlidwka]

Not worse, just quadratic. Fixed now, thanks for reporting.