perf(reactivity): refactor effects

[Simon-He95]

No description provided.

[pikax]

I don't think this is a performance improvement, you not running 2 loops but you creating a new array, push to it and then looping it, not sure how this will affect the performance, I'd recommend benchmarking to prove there's actually performance benefits here.

[Simon-He95]

I don't think this is a performance improvement, you not running 2 loops but you creating a new array, push to it and then looping it, not sure how this will affect the performance, I'd recommend benchmarking to prove there's actually performance benefits here.我不认为这是性能改进，您没有运行 2 个循环，而是创建一个新数组，推送到它然后循环它，不确定这将如何影响性能，我建议进行基准测试以证明这里确实有性能优势。

Its execution will definitely be faster than before, because the previous logic is to traverse the entire data again, and each time it has to judge effect.computed, and an array is added so that the data that has been filtered in the first loop can be directly Loop execution, the amount of data you have the second time will be less, and no conditional judgment is needed anymore.

https://perf.link/%EF%BC%89%EF%BC%9AJSPerf#eyJpZCI6ImFvejkyNjFqa3kyIiwidGl0bGUiOiJGaW5kaW5nIG51bWJlcnMgaW4gYW4gYXJyYXkgb2YgMTAwMCIsImJlZm9yZSI6ImNvbnN0IGVmZmVjdHMgPSBbXVxuZm9yKGxldCBpPTA7aTwxMDA7aSsrKXtcbiAgZWZmZWN0cy5wdXNoKHtcbiAgIGNvbXB1dGVkOiBNYXRoLnJhbmRvbSgpPD0wLjVcbiAgfSlcbn1cbmNvbnN0IHRyaWdnZXJFZmZlY3QgPSAoKT0%2Be30iLCJ0ZXN0cyI6W3sibmFtZSI6IkZpbmQgaXRlbSAxMDAiLCJjb2RlIjoiICBjb25zdCBub3RDb21wdXRlZEVmZmVjdCA9IFtdXG4gIGZvciAoY29uc3QgZWZmZWN0IG9mIGVmZmVjdHMpIHtcbiAgICBpZiAoZWZmZWN0LmNvbXB1dGVkKSB7XG4gICAgICB0cmlnZ2VyRWZmZWN0KGVmZmVjdClcbiAgICB9IGVsc2Uge1xuICAgICAgbm90Q29tcHV0ZWRFZmZlY3QucHVzaChlZmZlY3QpXG4gICAgfVxuICB9XG4gIG5vdENvbXB1dGVkRWZmZWN0LmZvckVhY2goZWZmZWN0ID0%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%3D

[pikax]

Thank you for the link, based on that you can optimise it even further see perf

[Simon-He95]

Quite good, but you need to ensure that it is consistent with the original logic, and you need to execute the effect.computed effect first.

[skirtles-code]

The title for this PR seems to be off. It doesn't seem to be related to compiler-sfc or defineOptions. It looks like you've copied the commit message from #9453.

[Simon-He95]

The title for this PR seems to be off. [The title for this PR seems to be off.] It doesn't seem to be related to compiler-sfc or defineOptions. [It doesn't seem to be related to compiler-sfc or defineOptions.] It looks like you've copied the commit message from #9453 [#9453] . [It looks like you've copied the commit message from #9453.] 此 PR 的标题似乎已关闭。 [此 PR 的标题似乎已关闭。它似乎与 compiler-sfc 或 defineOptions 无关。[它似乎与编译器-sfc或defineOptions无关。看起来您已经从 #9453 [#9453] 复制了提交消息。[看起来您已经从 #9453 复制了提交消息。

Do you have any better suggestions? Because compared to the previous logic, it does skip the second loop of the entire effects elements that need to be judged not to be effect.computed.

[sxzz]

Please update your PR title. Seems this PR is not related to compiler-sfc.

[github-actions]

Size Report

Bundles

File	Size	Gzip	Brotli
runtime-dom.global.prod.js	86.3 kB (-106 B)	32.9 kB (-31 B)	29.7 kB (-70 B)
vue.global.prod.js	132 kB (-149 B)	49.6 kB (-48 B)	44.5 kB (-5 B)

Usages

Name	Size	Gzip	Brotli
createApp	47.9 kB (-62 B)	18.9 kB (-15 B)	17.2 kB (-8 B)
createSSRApp	51.1 kB (-94 B)	20.1 kB (-32 B)	18.4 kB (-22 B)
defineCustomElement	50.3 kB (-62 B)	19.7 kB (-14 B)	18 kB (+28 B)
overall	61.3 kB (-62 B)	23.7 kB (-17 B)	21.6 kB (+26 B)