Speed up walkLength for narrow strings

[n8sh]

walkLength can be significantly sped up for narrow strings in the common case where they begin with a run of characters that do not require decoding.

Benchmark on my laptop for 12 character all-ASCII string:

compile command	string size	old MB/s	new MB/s	ratio
dmd	12	86.30	409.17	4.7x
dmd -O	12	92.85	840.56	9.1x
dmd -O -inline	12	298.88	842.44	2.8x
ldc2	12	75.58	474.29	6.3x
ldc2 -O3	12	727.53	2248.80	3.1x

Benchmark on my laptop for 2000 character all-ASCII string:

compile command	string size	old MB/s	new MB/s	ratio
dmd	2000	94.06	425.60	4.5x
dmd -O	2000	102.77	1047.31	10.2x
dmd -O -inline	2000	239.56	1055.37	4.4x
ldc2	2000	86.29	465.99	5.4x
ldc2 -O3	2000	626.03	2604.25	4.2x

[dlang-bot]

Thanks for your pull request, @n8sh!

Bugzilla references

Auto-close	Bugzilla	Severity	Description
✓	21559	enhancement	Speed up walkLength for narrow strings

Testing this PR locally

If you don't have a local development environment setup, you can use Digger to test this PR:

dub run digger -- build "master + phobos#7755"