Speedup ntriples and nquads parsing.

[rchateauneu]

This is the same code change as in #1317 but on a brand new branch and #1300

The performance test is make with parsing:

g.parse("orkg.nt", format="nt")

The general idea is to use a single regular expression to parse the line into its three or four individual components.
nquads parsing reuse much more ntriples codes than before.

Windows:
14.71s => 12.20 : 17% faster
Linux: 13.12 => 12.45: 5% faster.

Please note that about 15% of time is spent in codecs.readline(), implying UTF-8 conversions due to opening the file in "rb" mode. handling only str instead of bytes would easily speed things up.