refactor: change miniz_oxide to flate2 with zlib-rs backend

[Boshen]

Reason for change:

• unify usage across our tools
  • consistent behaviour
  • reduce compilation size
• reason for switching from miniz_oxide to zlib-rs:
  • more performant Change default backend to zlib-rs instead rust-lang/flate2-rs#469
  • miniz_oxide uses simd-adler32 https://github.com/rust-lang/flate2-rs/releases/tag/1.1.3
  • simd-adler32 breaks on wasm:

error[E0433]: failed to resolve: use of unresolved module or unlinked crate `std`
   --> /home/runner/.cargo/registry/src/index.crates.io-1949cf8c6b5b557f/simd-adler32-0.3.7/src/imp/wasm.rs:146:34
    |
146 |     let arr: [u32; 4] = unsafe { std::mem::transmute(v) };
    |                                  ^^^ use of unresolved module or unlinked crate `std`
    |
    = help: if you wanted to use a crate named `std`, use `cargo add std` to add it to your `Cargo.toml`
help: consider importing this module
    |
 27 +   use core::mem;
    |
help: if you import `mem`, refer to it directly
    |
146 -     let arr: [u32; 4] = unsafe { std::mem::transmute(v) };
146 +     let arr: [u32; 4] = unsafe { mem::transmute(v) };
    |

[github-actions]

Benchmark for c667375

Click to view benchmark

Test	Base	PR	%
vpath_edge_cases	2.1±0.06µs	2.2±0.07µs	+4.76%
vpath_mixed_workload	4.4±0.20µs	4.6±0.10µs	+4.55%
vpath_native	1353.9±64.18ns	1380.0±66.56ns	+1.93%
vpath_virtual	3.3±0.13µs	3.4±0.16µs	+3.03%
vpath_virtual_zip	2.7±0.08µs	2.9±0.09µs	+7.41%
vpath_zip	1828.8±66.98ns	1875.8±62.50ns	+2.57%

[github-actions]

Benchmark for ad7cec1

Click to view benchmark

Test	Base	PR	%
vpath_edge_cases	2.0±0.07µs	2.1±0.04µs	+5.00%
vpath_mixed_workload	4.2±0.19µs	4.2±0.11µs	0.00%
vpath_native	1310.5±45.20ns	1329.3±49.56ns	+1.43%
vpath_virtual	3.2±0.06µs	3.3±0.12µs	+3.12%
vpath_virtual_zip	2.7±0.09µs	2.8±0.07µs	+3.70%
vpath_zip	1794.1±50.22ns	1824.3±48.77ns	+1.68%