我认为应该支持使用 "${lib:idx}"和${workspaceFolder:name}

Sumneko Lua translate bot

I think the use of "${lib:idx}" and ${workspaceFolder:name} should be supported

我认为应该支持使用 "${lib:idx}"

${lib:idx} 這個不好吧 🤔
因為這個受 workspace.library[] 中對應 entry 的次序影響

• 如果 user 在 array 中間插值 => 那後續的 index 就亂套了
• 並且完全不直觀,比如當看到 ignoreDir: [ "${lib:2}/xxx/yyy" ]
  => 一眼看過去看不出 lib:2 是哪1個 library path
  => 還得手動數一下 workspace.library[] 中對應的 element 是什麼? 😂

... 和 ${workspaceFolder:name}

確實是個好想法 👍

• 這個 placeholder 應該會在 files.normalize() 時
  由裡邊的 m.resolvePathPlaceholders 負責處理了的
  

  lua-language-server/script/files.lua

  Lines 984 to 989 in 32fec3c

  ---@param path string

  ---@return string

  function m.normalize(path)

  path = m.resolvePathPlaceholders(path)

  path = util.expandPath(path)

  path = path:gsub('^%.[/\\]+', '')
• 但問題是這 PR 中處理 per library ignoreDir 的方式
  是因為 m.getLibraryMatchers 本身就有為 每1個 library 創建1個 matcher
  => 我按 library path 對應的 matcher 來寫入不同的 ignore patterns
• 而對於 workspace root 自身的 path matcher 來說
  我暫不確定他的運作機制是如何 😕
  我看到 m.getNativeMatcher 最終只會返回 1個 matcher?
  不確定可以怎樣支持到 🙈

如果有必要支持 multi root workspace 中針對不同 workspaceFolder 的 ignoreDir
感覺應該再開個新 PR 較好?

Sumneko Lua translate bot

I think the use of `"${lib:idx}" should be supported

${lib:idx} This is not good 🤔
Because this is affected by the order of the corresponding entry in workspace.library[]

• If user interpolates => in the middle of array, then the subsequent index will be messed up
• And it is completely unintuitive, for example when you see ignoreDir: [ "${lib:2}/xxx/yyy" ]
  => I can't see which library path lib:2 is
  => You have to manually count what is the corresponding element in workspace.library[]? 😂

... and ${workspaceFolder:name}

It's a good idea indeed

• This placeholder should be in files.normalize()
  The m.resolvePathPlaceholders is handled by
  

  lua-language-server/script/files.lua

  Lines 984 to 989 in 32fec3c

  ---@param path string

  ---@return string

  function m.normalize(path)

  path = m.resolvePathPlaceholders(path)

  path = util.expandPath(path)

  path = path:gsub('^%.[/\\]+', '')
• But the problem is how to deal with per library ignoreDir in this PR
  Because m.getLibraryMatchers itself has a chance Create 1 matcher for every library
  => I press the matcher corresponding to library path to write different ignore patterns
• And for the workspace root's own path matcher
  I'm not sure how it works 😕
  I see that m.getNativeMatcher will only return 1 matcher in the end?
  Not sure how to support it 🙈

If necessary, support ignoreDir for different workspaceFolder in multi root workspace
I feel like it should be better to open a new PR?