Fix equation splitting for unicode variables

[MilesCranmer]

It seems like the code that splits an equation string into multiple lines for the progress bar was using linear indexing, whereas strings with unicode do not have indices 1, 2, 3...

e.g., the string "⋅⋅" has valid indices 1 and 4.

[github-actions]

Benchmark Results

	master	9f78093...	t[master]/t[9f78093...]
search/multithreading	0.0231 ± 0.0027 h	0.0232 ± 0.0013 h	0.997
search/serial	0.0229 ± 0.00038 h	0.0232 ± 0.0011 h	0.985
time_to_load	3.96 ± 0.1 s	4.07 ± 0.053 s	0.974
utils/best_of_sample	2.8 ± 1 μs	2.9 ± 1.1 μs	0.966
utils/check_constraints_x10	21.2 ± 5.3 μs	19.4 ± 5.2 μs	1.09
utils/compute_complexity_x10/Float64	4.1 ± 0.4 μs	3.8 ± 0.3 μs	1.08
utils/compute_complexity_x10/Int64	3.7 ± 0.3 μs	3.7 ± 0.3 μs	1
utils/compute_complexity_x10/nothing	2.6 ± 0.3 μs	2.7 ± 0.3 μs	0.963
utils/optimize_constants_x10	0.0719 ± 0.015 s	0.0731 ± 0.015 s	0.984

Benchmark Plots

A plot of the benchmark results have been uploaded as an artifact to the workflow run for this PR.
Go to "Actions"->"Benchmark a pull request"->[the most recent run]->"Artifacts" (at the bottom).