Reorder arguments in quote_token and quote_token_spanned

[lqd]

This PR implements the suggestion in #209 (review) to reorder the arguments to these macros, and avoid some parsing of idents during rule matching.

The tests succeed but I'm not familiar with quote's internals, and may have made a mistake, especially on the more complicated rules.

If it's correct, this would result in a small improvement:

Benchmark & Profile	Scenario	% Change	Significance Factor ?
ctor-0.1.21 check	incr-unchanged	-3.00%	15.00x
pest_generator-2.1.3 check	incr-unchanged	-2.20%	11.02x
num-derive-0.3.3 check	incr-unchanged	-2.05%	10.25x
ctor-0.1.21 check	full	-1.88%	9.39x
mockall_derive-0.11.0 check	incr-unchanged	-1.82%	9.08x
ctor-0.1.21 check	incr-full	-1.64%	8.21x
scroll_derive-0.11.0 check	incr-unchanged	-1.56%	7.82x
pest_generator-2.1.3 check	full	-1.32%	6.59x
num-derive-0.3.3 check	full	-1.19%	5.97x
futures-macro-0.3.19 check	incr-unchanged	-1.19%	5.97x
pest_generator-2.1.3 check	incr-full	-1.14%	5.70x
num-derive-0.3.3 check	incr-full	-1.09%	5.47x
mockall_derive-0.11.0 check	full	-0.94%	4.72x
scroll_derive-0.11.0 check	full	-0.85%	4.26x
mockall_derive-0.11.0 check	incr-full	-0.78%	3.91x
scroll_derive-0.11.0 check	incr-full	-0.69%	3.44x
futures-macro-0.3.19 check	incr-full	0.05%	0.27x
futures-macro-0.3.19 check	full	-0.04%	0.21x

[dtolnay]

Make sure you're changing every $crate::quote_token_spanned! call site. There are some that are missed in this PR. (Maybe also some $crate::quote_token!? I didn't look too carefully)

[nnethercote]

It looks just how I expected, modulo the missed changes.

I was also going to try putting $tokens last for quote_token_with_context and quote_token_with_context_spanned, to see if that helped as well. (No point doing anything with the higher-up macros like quote_each_token and quote_tokens_with_context because they only have one rule, so there's no need to fail fast.)

I would also put a brief comment above quote_tokens!, something like "$tokens comes at the end so that failing rules will fail as quickly as possible. This helps reduce compile times for crates that use quote! heavily.`

[lqd]

Thanks to the both of you for the review, I'll make the changes asap.

[dtolnay]

Thanks!

[lqd]

Oh you were super quick, I was completing the latest couple changes nick mentioned :)

I'll open another PR in a few minutes.

[lqd]

Testing on the same 6 crates, and in a micro-benchmark: the quote_token_with_context and quote_token_with_context_spanned reorderings don't seem to make much of a difference. The branch is here if anyone is interested, but it's probably no use to open a PR.