perf(analyzer): reduce use of owned strings

[ematipico]

Summary

Address #5990

I'm not sure this PR will fix ALL the performance bottlenecks of the batch mutations. Unfortunately, there is one point inside batch.rs where we're kinda forced to allocate a new strings because we need to print an entire node with its trivia, and that's also recursive. I will have to find a smart way to emit &str when printing a node. I will leave a comment to show the line.

I also reduced the use of to_strimmed_string around lint rules. What I find saddening is that we often try to print nodes instead of extracting the token we want to check. Maybe we should have some guide that explains how to do so.

Test Plan

Current tests should still pass

[ematipico]

This is where we're diffing a node, and unfortunately we need to extract the entire string: tokens, trivia and child nodes/tokens

[github-actions]

Parser conformance results on

js/262

Test result	main count	This PR count	Difference
Total	50524	50524	0
Passed	49209	49209	0
Failed	1315	1315	0
Panics	0	0	0
Coverage	97.40%	97.40%	0.00%

jsx/babel

Test result	main count	This PR count	Difference
Total	40	40	0
Passed	37	37	0
Failed	3	3	0
Panics	0	0	0
Coverage	92.50%	92.50%	0.00%

symbols/microsoft

Test result	main count	This PR count	Difference
Total	6648	6648	0
Passed	2229	2229	0
Failed	4419	4419	0
Panics	0	0	0
Coverage	33.53%	33.53%	0.00%

ts/babel

Test result	main count	This PR count	Difference
Total	807	807	0
Passed	718	718	0
Failed	89	89	0
Panics	0	0	0
Coverage	88.97%	88.97%	0.00%

ts/microsoft

Test result	main count	This PR count	Difference
Total	18702	18702	0
Passed	14349	14349	0
Failed	4353	4353	0
Panics	0	0	0
Coverage	76.72%	76.72%	0.00%

[codspeed-hq]

CodSpeed Performance Report

Merging #5997 will not alter performance

_{Comparing perf/reduce-use-of-strings (d84b0ff) with main (73be9e1)}

Summary

✅ 95 untouched benchmarks