无法提示ipair中的的类型 #435

sniper00 · 2021-03-02T07:55:57Z

Describe the bug
A clear and concise description of what the bug is.

---@class TypeA
---@field public a integer
---@field public b integer
---@field public c integer

---@class TypeB
---@field public data TypeA[]

---@type TypeB
local arr = {}

for _, v in ipairs(arr.data) do
    if v.b == 10 then ---v 提示类型是any
        print("hello")
    end
end

To Reproduce
Steps to reproduce the behavior:

Go to '...'
Click on '....'
Scroll down to '....'
See error

Expected behavior
A clear and concise description of what you expected to happen.

Screenshots
If applicable, add screenshots to help explain your problem.

Environment (please complete the following information):

OS: Windows
Client: VSCode

Additional context
Add any other context about the problem here.

sniper00 · 2021-03-03T02:11:13Z

怎么手动更新最新版？

sumneko · 2021-03-03T02:12:27Z

自己手动替换文件/应用更改

sniper00 · 2021-03-03T02:24:01Z

pairs 也有同样问题

sumneko · 2021-03-03T02:26:45Z

pairs暂时只支持 table<K, V> 的注解

sniper00 · 2021-03-03T02:32:59Z

替换之后报这个错误

    [1] = "ECode",
    finish = 21249,
    loc = {
        [1] = "ECode",
        effect = 252,
        finish = 223,
        parent = "<Deep>",
        range = 251,
        ref = "<Deep>",
        start = 219,
        type = "local",
        value = "<Deep>",
    },
    next = {
        dot = "<Deep>",
        field = "<Deep>",
        finish = 21258,
        node = "<Loop>",
        parent = "<Deep>",
        start = 21245,
        type = "getfield",
    },
    node = {
        [1] = "ECode",
        effect = 252,
        finish = 223,
        parent = "<Deep>",
        range = 251,
        ref = "<Deep>",
        start = 219,
        type = "local",
        value = "<Deep>",
    },
    parent = {
        dot = "<Deep>",
        field = "<Deep>",
        finish = 21258,
        node = "<Loop>",
        parent = "<Deep>",
        start = 21245,
        type = "getfield",
    },
    start = 21245,
    type = "getlocal",
}
[10:30:51.675][warn] [#0:script.parser.guide:29]: status.depth overflow

sumneko · 2021-03-03T02:36:43Z

这个一般是你的代码比较复杂，搜索深度达到上限了（也有可能是插件bug导致搜索时发生了死循环）。如果不影响结果的话无视即可。
如果你觉得代码并没有那么复杂的话，请尝试制作一个演示脚本给我。

sniper00 · 2021-03-03T02:38:15Z

项目代码，不好分享。替换前是没这个问题的，替换后解析文件会超级慢。

sumneko · 2021-03-03T02:50:20Z

你可以尝试做个演示脚本

sniper00 · 2021-03-03T02:50:45Z

嗯，我试试

sniper00 · 2021-03-03T06:34:26Z

可不可以通过添加日志定位下问题？我发现只要代码多一些基本就一直卡在：正在诊断状态，lua-language-server进程使用的内存能达到 4G, 还继续增长，并且跑满了一个核CPU

sumneko · 2021-03-03T06:36:04Z

听起来很有可能是死循环了，我加一些限制试试

sumneko · 2021-03-03T06:52:22Z

cc6ec7e
试试这个

sniper00 · 2021-03-03T06:53:41Z

我重启了下，这次没出现内存爆涨，大概等了2分钟，诊断结束，但结果好像不对。

sniper00 · 2021-03-03T07:01:13Z

替换了你刚才发的最新的，我们项目300多个lua文件，每个1k-30K, 改动一行代码, 2分钟左右才诊断结束。

sumneko · 2021-03-03T07:02:58Z

在你第一次手动更新这个issue提到的问题前，诊断速度是合理的对吗？

sniper00 · 2021-03-03T07:04:55Z

vscode 只调整过文件大小限制其它配置是默认的, 确认了下速度值是 100

sniper00 · 2021-03-03T07:21:03Z

替换了你刚才发的最新的，还是会出现一直在诊断某个文件， cpu 20%, 内存缓慢增长，15分钟涨到了1GB

sumneko · 2021-03-03T07:24:23Z

我还在尝试复现这个问题，先不要应用我的新提交

sumneko · 2021-03-03T09:47:22Z

试试这个提交

sniper00 · 2021-03-03T12:50:03Z

还是很慢

[20:48:16.624][warn] [#0:script.parser.guide:29]: Queue too large!
[20:48:16.624][warn] [#0:script.vm.eachRef:22]: requestReference	100000	1.472	file:///d%3a/
  [1] = "allianceData",
    finish = 13734,
    loc = {
        [1] = "allianceData",
        effect = 13476,
        finish = 13430,
        parent = "<Deep>",
        range = 13475,
        ref = "<Deep>",
        start = 13419,
        type = "local",
        value = "<Deep>",
    },
    next = {
        dot = "<Deep>",
        field = "<Deep>",
        finish = 13742,
        next = "<Deep>",
        node = "<Loop>",
        parent = "<Deep>",
        start = 13723,
        type = "getfield",
    },
    node = {
        [1] = "allianceData",
        effect = 13476,
        finish = 13430,
        parent = "<Deep>",
        range = 13475,
        ref = "<Deep>",
        start = 13419,
        type = "local",
        value = "<Deep>",
    },
    parent = {
        dot = "<Deep>",
        field = "<Deep>",
        finish = 13742,
        next = "<Deep>",
        node = "<Loop>",
        parent = "<Deep>",
        start = 13723,
        type = "getfield",
    },
    start = 13723,
    type = "getlocal",
}

sniper00 · 2021-03-03T12:55:55Z

建议用 hash 来处理超长队列, table.remove(t, 1)的方式算法复杂度是n平方

local queue = {h = 1,t = 0}

---push
local t = queue.t + 1
queue.t = t
queue[t] = {}---value
---push

---pop
if queue.h > queue.t then
    -- queue is empty
    queue.h = 1
    queue.t = 0
    ---reutrn nil
else
    -- pop queue
    local h = queue.h
    local value = queue[h]
    queue[h] = nil
    queue.h = h + 1
    ---return value
end
---pop

sumneko · 2021-03-03T13:01:53Z

建议用 hash 来处理超长队列, table.remove(t, 1)的方式算法复杂度是n平方

local queue = {h = 1,t = 0}

---push
local t = queue.t + 1
queue.t = t
queue[t] = {}---value
---push

---pop
if queue.h > queue.t then
    -- queue is empty
    queue.h = 1
    queue.t = 0
    ---reutrn nil
else
    -- pop queue
    local h = queue.h
    local value = queue[h]
    queue[h] = nil
    queue.h = h + 1
    ---return value
end
---pop

不确定你说的是什么意思，插件代码中并没有这样的用法

sniper00 · 2021-03-03T13:03:47Z

ast.lua:1432 行， local func = tableRemove(exp, 1), 不确定你这个exp 是否是个超长数组，table.remove(t, 1)的方式算法复杂度是n平方，会极大影响性能

sniper00 · 2021-03-03T13:05:43Z

我贴的那段代码是用 hash模拟队列，算法复杂度不会随着元素变多n平方增长

sumneko · 2021-03-03T13:06:53Z

你说的这个地方长度有多大，取决于你 for _ in a1, a2, a3, a4, a5, a6 ... do 写的有多长。
这些代码和问题没有太大关系，主要还是我无法复现，最好是你能想办法提供一个能复现的例子。

sniper00 · 2021-03-03T13:08:29Z

感觉和 #437 这个问题差不多

sumneko · 2021-03-03T13:10:45Z

#437 (comment) 提供了一个演示项目，我在1.17.3中修复了，不过依然有人表示他用1.17.3打开这个项目有问题。不确定是真的有问题还是他自己侧错了，你也来测试一下看看？

sniper00 · 2021-03-03T13:11:36Z

我试试

sniper00 · 2021-03-03T13:13:29Z

我这里依然很慢，一直显示正在诊断中

sniper00 · 2021-03-03T13:14:50Z

[21:13:34.316][warn] [#0:script.parser.guide:29]: Queue too large!
[21:13:34.316][warn] [#0:script.vm.eachRef:22]: requestReference	100000	2.963	file:///d%3a/	{
    dot = {
        finish = 14705,
        start = 14705,
        type = ".",
    },
    field = {
        [1] = "info",
        finish = 14709,
        parent = "<Loop>",
        start = 14706,
        type = "field",
    },
    finish = 14709,
    next = {
        dot = "<Deep>",
        field = "<Deep>",
        finish = 14715,
        node = "<Loop>",
        parent = "<Deep>",
        start = 14693,
        type = "getfield",
    },
    node = {
        [1] = "allianceData",
        finish = 14704,
        loc = "<Deep>",
        next = "<Loop>",
        node = "<Deep>",
        parent = "<Loop>",
        start = 14693,
        type = "getlocal",
    },
    parent = {
        dot = "<Deep>",
        field = "<Deep>",
        finish = 14715,
        node = "<Loop>",
        parent = "<Deep>",
        start = 14693,
        type = "getfield",
    },
    start = 14693,
    type = "getfield",
}
[21:13:35.066][warn] [#0:script.await:166]: Await step takes [0.223] sec.

sumneko · 2021-03-03T15:09:25Z

你的问题应该和 #437 不同，把日志发我看看吧（打开 #437 中提供的演示项目）

sumneko · 2021-03-05T03:20:28Z

此 issue 标题的部分已经修复，关于性能问题请单独开个新的 issue ，并提供上述测试的日志，谢谢！

sniper00 · 2021-03-05T03:30:47Z

好的，有空我重新开个，这几天github上不去发自我的iPhone 在 2021年3月5日，上午11:20，最萌小汐 <notifications@github.com> 写道：此 issue 标题的部分已经修复，关于性能问题请单独开个新的 issue ，并提供上述测试的日志，谢谢！ — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub<#435 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/ADDQZK26RF64PYQ6QNE7S73TCBEYTANCNFSM4YOLXNRA>.

sumneko added the bug Something isn't working label Mar 2, 2021

sumneko closed this as completed in 679689b Mar 2, 2021

sumneko reopened this Mar 3, 2021

sumneko added a commit that referenced this issue Mar 3, 2021

#435 treat V[] as table<integer, V> in pairs

b9e9b9e

sumneko added a commit that referenced this issue Mar 3, 2021

#435 #437 why not cache generic results?

b06ba33

sumneko closed this as completed Mar 5, 2021

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

无法提示ipair中的的类型 #435

无法提示ipair中的的类型 #435

sniper00 commented Mar 2, 2021 •

edited

Loading

sniper00 commented Mar 3, 2021

sumneko commented Mar 3, 2021

sniper00 commented Mar 3, 2021

sumneko commented Mar 3, 2021

sniper00 commented Mar 3, 2021

sumneko commented Mar 3, 2021

sniper00 commented Mar 3, 2021

sumneko commented Mar 3, 2021

sniper00 commented Mar 3, 2021

sniper00 commented Mar 3, 2021 •

edited

Loading

sumneko commented Mar 3, 2021

sumneko commented Mar 3, 2021

sniper00 commented Mar 3, 2021

sniper00 commented Mar 3, 2021 •

edited

Loading

sumneko commented Mar 3, 2021

sniper00 commented Mar 3, 2021 •

edited

Loading

sniper00 commented Mar 3, 2021

sumneko commented Mar 3, 2021

sumneko commented Mar 3, 2021

sniper00 commented Mar 3, 2021

sniper00 commented Mar 3, 2021

sumneko commented Mar 3, 2021

sniper00 commented Mar 3, 2021

sniper00 commented Mar 3, 2021

sumneko commented Mar 3, 2021

sniper00 commented Mar 3, 2021

sumneko commented Mar 3, 2021

sniper00 commented Mar 3, 2021

sniper00 commented Mar 3, 2021

sniper00 commented Mar 3, 2021

sumneko commented Mar 3, 2021

sumneko commented Mar 5, 2021

sniper00 commented Mar 5, 2021 via email

无法提示ipair中的的类型 #435

无法提示ipair中的的类型 #435

Comments

sniper00 commented Mar 2, 2021 • edited Loading

sniper00 commented Mar 3, 2021

sumneko commented Mar 3, 2021

sniper00 commented Mar 3, 2021

sumneko commented Mar 3, 2021

sniper00 commented Mar 3, 2021

sumneko commented Mar 3, 2021

sniper00 commented Mar 3, 2021

sumneko commented Mar 3, 2021

sniper00 commented Mar 3, 2021

sniper00 commented Mar 3, 2021 • edited Loading

sumneko commented Mar 3, 2021

sumneko commented Mar 3, 2021

sniper00 commented Mar 3, 2021

sniper00 commented Mar 3, 2021 • edited Loading

sumneko commented Mar 3, 2021

sniper00 commented Mar 3, 2021 • edited Loading

sniper00 commented Mar 3, 2021

sumneko commented Mar 3, 2021

sumneko commented Mar 3, 2021

sniper00 commented Mar 3, 2021

sniper00 commented Mar 3, 2021

sumneko commented Mar 3, 2021

sniper00 commented Mar 3, 2021

sniper00 commented Mar 3, 2021

sumneko commented Mar 3, 2021

sniper00 commented Mar 3, 2021

sumneko commented Mar 3, 2021

sniper00 commented Mar 3, 2021

sniper00 commented Mar 3, 2021

sniper00 commented Mar 3, 2021

sumneko commented Mar 3, 2021

sumneko commented Mar 5, 2021

sniper00 commented Mar 5, 2021 via email

sniper00 commented Mar 2, 2021 •

edited

Loading

sniper00 commented Mar 3, 2021 •

edited

Loading

sniper00 commented Mar 3, 2021 •

edited

Loading

sniper00 commented Mar 3, 2021 •

edited

Loading