<!-- What / Why -->05T03:41:30.3592718Z [36;1mEOF[0m 2024-06-05T03:41:3

[Annie602942]

This pull request optimizes the Range class in the following ways:

1. Produce fewer intermediate objects when reducing a range's space characters to single spaces. This seems to improve bench-subset scores by up to 5%, and bench-satisfies scores to a lesser degree.
2. Optimize Range formatting with explicit for loops instead, avoiding an intermediate array creation. This seems to improve bench-satisfies and bench-subset scores by up to 20%.
3. Calculate Range's .range string (used by .format() and .toString()) lazily. This seems to improve bench-satisfies and bench-subset scores by up to 9%.

The external interface for the class stays the same, except for the new internal .formatted property used to cache its lazily calculated string. Range#range property is now also read-only.

There is a new test lazy formatting to ensure full test coverage.

The benchmarks bench-satisfies and bench-subset benefit from these changes, sometimes by up to 40%. Other benchmark results seem to stay the same. Here are the affected benchmarks before:

$ node benchmarks/bench-satisfies.js
satisfies(1.0.6, 1.0.3||^2.0.0) x 695,094 ops/sec ±0.68% (97 runs sampled)
satisfies(1.0.6, 2.2.2||~3.0.0) x 764,115 ops/sec ±0.40% (99 runs sampled)
satisfies(1.0.6, 2.3.0||<4.0.0) x 805,593 ops/sec ±0.62% (97 runs sampled)
satisfies(1.0.6, 1.0.3||^2.0.0, {"includePrelease":true}) x 695,045 ops/sec ±0.73% (95 runs sampled)
satisfies(1.0.6, 2.2.2||~3.0.0, {"includePrelease":true}) x 750,433 ops/sec ±0.66% (99 runs sampled)
satisfies(1.0.6, 2.3.0||<4.0.0, {"includePrelease":true}) x 787,903 ops/sec ±0.39% (99 runs sampled)
satisfies(1.0.6, 1.0.3||^2.0.0, {"includePrelease":true,"loose":true}) x 652,166 ops/sec ±0.34% (99 runs sampled)
satisfies(1.0.6, 2.2.2||~3.0.0, {"includePrelease":true,"loose":true}) x 696,377 ops/sec ±0.36% (96 runs sampled)
satisfies(1.0.6, 2.3.0||<4.0.0, {"includePrelease":true,"loose":true}) x 721,729 ops/sec ±0.35% (98 runs sampled)
satisfies(1.0.6, 1.0.3||^2.0.0, {"includePrelease":true,"loose":true,"rtl":true}) x 585,692 ops/sec ±0.75% (95 runs sampled)
satisfies(1.0.6, 2.2.2||~3.0.0, {"includePrelease":true,"loose":true,"rtl":true}) x 631,653 ops/sec ±0.33% (96 runs sampled)
satisfies(1.0.6, 2.3.0||<4.0.0, {"includePrelease":true,"loose":true,"rtl":true}) x 650,110 ops/sec ±0.64% (95 runs sampled)

$ node benchmarks/bench-subset.js
subset(1.2.3, *) x 633,342 ops/sec ±0.61% (95 runs sampled)
subset(^1.2.3, *) x 743,036 ops/sec ±0.47% (97 runs sampled)
subset(^1.2.3-pre.0, *) x 680,087 ops/sec ±0.76% (98 runs sampled)
subset(^1.2.3-pre.0, *) x 680,948 ops/sec ±0.46% (96 runs sampled)
subset(1 || 2 || 3, *) x 330,669 ops/sec ±0.53% (98 runs sampled)

And after:

$ node benchmarks/bench-satisfies.js
satisfies(1.0.6, 1.0.3||^2.0.0) x 896,936 ops/sec ±0.53% (94 runs sampled)
satisfies(1.0.6, 2.2.2||~3.0.0) x 998,214 ops/sec ±0.40% (95 runs sampled)
satisfies(1.0.6, 2.3.0||<4.0.0) x 1,000,593 ops/sec ±0.43% (97 runs sampled)
satisfies(1.0.6, 1.0.3||^2.0.0, {"includePrelease":true}) x 890,369 ops/sec ±0.41% (100 runs sampled)
satisfies(1.0.6, 2.2.2||~3.0.0, {"includePrelease":true}) x 977,239 ops/sec ±0.48% (97 runs sampled)
satisfies(1.0.6, 2.3.0||<4.0.0, {"includePrelease":true}) x 983,682 ops/sec ±0.95% (96 runs sampled)
satisfies(1.0.6, 1.0.3||^2.0.0, {"includePrelease":true,"loose":true}) x 805,330 ops/sec ±0.84% (98 runs sampled)
satisfies(1.0.6, 2.2.2||~3.0.0, {"includePrelease":true,"loose":true}) x 894,117 ops/sec ±0.43% (99 runs sampled)
satisfies(1.0.6, 2.3.0||<4.0.0, {"includePrelease":true,"loose":true}) x 911,742 ops/sec ±0.42% (96 runs sampled)
satisfies(1.0.6, 1.0.3||^2.0.0, {"includePrelease":true,"loose":true,"rtl":true}) x 741,254 ops/sec ±0.35% (97 runs sampled)
satisfies(1.0.6, 2.2.2||~3.0.0, {"includePrelease":true,"loose":true,"rtl":true}) x 807,380 ops/sec ±0.42% (99 runs sampled)
satisfies(1.0.6, 2.3.0||<4.0.0, {"includePrelease":true,"loose":true,"rtl":true}) x 820,390 ops/sec ±0.37% (99 runs sampled)

$ node benchmarks/bench-subset.js
subset(1.2.3, *) x 905,030 ops/sec ±0.63% (96 runs sampled)
subset(^1.2.3, *) x 1,026,457 ops/sec ±0.63% (95 runs sampled)
subset(^1.2.3-pre.0, *) x 923,789 ops/sec ±0.41% (97 runs sampled)
subset(^1.2.3-pre.0, *) x 923,136 ops/sec ±0.44% (96 runs sampled)
subset(1 || 2 || 3, *) x 432,037 ops/sec ±0.67% (96 runs sampled)

Originally posted by @jviide in npm/node-semver#726 05T03:41:30.3592718Z �[36;1mEOF�[0m
2024-06-05T03:41:3