From 1f7de306f547ecb394a34445fd6ac1d6bc8ab439 Mon Sep 17 00:00:00 2001 From: Mikhail Rakhmanov Date: Wed, 3 Jun 2020 18:57:50 +0200 Subject: Add documentation --- docs/dev/lsp-extensions.md | 24 ++++++++++++++++++++++++ 1 file changed, 24 insertions(+) diff --git a/docs/dev/lsp-extensions.md b/docs/dev/lsp-extensions.md index 647cf6107..7f7940d0b 100644 --- a/docs/dev/lsp-extensions.md +++ b/docs/dev/lsp-extensions.md @@ -97,6 +97,30 @@ Invoking code action at this position will yield two code actions for importing * Is a fixed two-level structure enough? * Should we devise a general way to encode custom interaction protocols for GUI refactorings? +## Lazy assists with `ResolveCodeAction` + +**Issue:** https://github.com/microsoft/language-server-protocol/issues/787 + +**Client Capability** `{ "resolveCodeAction": boolean }` + +If this capability is set, the assists will be computed lazily. Thus `CodeAction` returned from the server will only contain `id` but not `edit` or `command` fields. The only exclusion from the rule is the diagnostic edits. + +After the client got the id, it should then call `experimental/resolveCodeAction` command on the server and provide the following payload: + +```typescript +interface ResolveCodeActionParams { + id: string; + codeActionParams: lc.CodeActionParams; +} +``` + +As a result of the command call the client will get the respective workspace edit (`lc.WorkspaceEdit`). + +### Unresolved Questions + +* Apply smarter filtering for ids? +* Upon `resolveCodeAction` command only call the assits which should be resolved and not all of them? + ## Parent Module **Issue:** https://github.com/microsoft/language-server-protocol/issues/1002 -- cgit v1.2.3