feat: implement {cd,pwd} -{L,P}

[reubeno]

This integrates the compat tests, clap annotation updates, and some of the core builtin updates from #219.

There are 2 notable differences with this PR's approach:

• In place of a custom implementation of path normalization, this PR instead opts for adopting the normalize-path crate with the acknowledged trade-off that this latter crate likely does not necessarily handle Windows paths.

• This version does not persistently maintain a logical and physical representation of the current working dir. Rather, it maintains the existing single representation, and only computes physical paths on-demand when required. We still may very well need to go with maintaining both, but for now this keeps the change more minimal and seems to be more in line with handling the edge case of how, for example, pwd -P should behave on a directory that has been renamed/moved/altered since being cd'd into.

We will leave the older PR open for now, and re-evaluate how best to preserve/use its Windows-specific logic in subsequent changes.

(Note: @39555 -- the commit tags you as a contributor because you had done the heavy lifting of identifying the work needed and also because this PR directly incorporates some of your changes from the above-linked PR.)

Resolves: #202.

[github-actions]

Test Results

    3 files     14 suites   3m 44s ⏱️
  723 tests   723 ✅ 0 💤 0 ❌
2 155 runs  2 155 ✅ 0 💤 0 ❌

Results for commit d6ffb8e.

[github-actions]

Performance Benchmark Report

Benchmark name	Baseline (μs)	Test/PR (μs)	Delta (μs)	Delta %
clone_shell_object	18.10 μs	18.81 μs	0.71 μs	🟠 +3.94%
eval_arithmetic	0.17 μs	0.17 μs	-0.00 μs	⚪ Unchanged
expand_one_string	1.87 μs	1.83 μs	-0.05 μs	⚪ Unchanged
for_loop	22.97 μs	22.04 μs	-0.93 μs	🟢 -4.04%
function_call	2.47 μs	2.40 μs	-0.08 μs	⚪ Unchanged
instantiate_shell	57.10 μs	56.84 μs	-0.25 μs	⚪ Unchanged
instantiate_shell_with_init_scripts	24232.32 μs	23975.26 μs	-257.06 μs	⚪ Unchanged
parse_bash_completion	1665.57 μs	1667.42 μs	1.85 μs	⚪ Unchanged
parse_sample_script	1.76 μs	1.73 μs	-0.02 μs	⚪ Unchanged
run_echo_builtin_command	15.32 μs	15.19 μs	-0.12 μs	⚪ Unchanged
run_one_external_command	2338.50 μs	2397.61 μs	59.11 μs	⚪ Unchanged
tokenize_sample_script	2.89 μs	2.84 μs	-0.05 μs	🟢 -1.76%

Code Coverage Report: Only Changed Files listed

Package	Base Coverage	New Coverage	Difference
brush-core/src/builtins/cd.rs	🟢 81.82%	🟢 82.35%	🟢 0.53%
brush-core/src/builtins/pwd.rs	🟢 80.95%	🟢 88.89%	🟢 7.94%
brush-core/src/shell.rs	🟢 82.99%	🟢 83.01%	🟢 0.02%
Overall Coverage	🟢 76.39%	🟢 76.41%	🟢 0.02%

Minimum allowed coverage is 70%, this run produced 76.41%

Test Summary: bash-completion test suite

Outcome	Count	Percentage
✅ Pass	1289	61.12
❗️ Error	226	10.72
❌ Fail	200	9.48
⏩ Skip	379	17.97
❎ Expected Fail	14	0.66
✔️ Unexpected Pass	1	0.05
📊 Total	2109	100.00