| // Copyright 2019 The Go Authors. All rights reserved. | |
| // Use of this source code is governed by a BSD-style | |
| // license that can be found in the LICENSE file. | |
| package jsonrpc2 | |
| import ( | |
| "context" | |
| "fmt" | |
| "sync" | |
| "golang.org/x/tools/internal/event" | |
| ) | |
| // Handler is invoked to handle incoming requests. | |
| // The Replier sends a reply to the request and must be called exactly once. | |
| type Handler func(ctx context.Context, reply Replier, req Request) error | |
| // Replier is passed to handlers to allow them to reply to the request. | |
| // If err is set then result will be ignored. | |
| type Replier func(ctx context.Context, result any, err error) error | |
| // MethodNotFound is a Handler that replies to all call requests with the | |
| // standard method not found response. | |
| // This should normally be the final handler in a chain. | |
| func MethodNotFound(ctx context.Context, reply Replier, req Request) error { | |
| return reply(ctx, nil, fmt.Errorf("%w: %q", ErrMethodNotFound, req.Method())) | |
| } | |
| // MustReplyHandler is a middleware that creates a Handler that panics if the | |
| // wrapped handler does not call Reply for every request that it is passed. | |
| func MustReplyHandler(handler Handler) Handler { | |
| return func(ctx context.Context, reply Replier, req Request) error { | |
| called := false | |
| err := handler(ctx, func(ctx context.Context, result any, err error) error { | |
| if called { | |
| panic(fmt.Errorf("request %q replied to more than once", req.Method())) | |
| } | |
| called = true | |
| return reply(ctx, result, err) | |
| }, req) | |
| if !called { | |
| panic(fmt.Errorf("request %q was never replied to", req.Method())) | |
| } | |
| return err | |
| } | |
| } | |
| // CancelHandler returns a handler that supports cancellation, and a function | |
| // that can be used to trigger canceling in progress requests. | |
| func CancelHandler(handler Handler) (Handler, func(id ID)) { | |
| var mu sync.Mutex | |
| handling := make(map[ID]context.CancelFunc) | |
| wrapped := func(ctx context.Context, reply Replier, req Request) error { | |
| if call, ok := req.(*Call); ok { | |
| cancelCtx, cancel := context.WithCancel(ctx) | |
| ctx = cancelCtx | |
| mu.Lock() | |
| handling[call.ID()] = cancel | |
| mu.Unlock() | |
| innerReply := reply | |
| reply = func(ctx context.Context, result any, err error) error { | |
| mu.Lock() | |
| delete(handling, call.ID()) | |
| mu.Unlock() | |
| return innerReply(ctx, result, err) | |
| } | |
| } | |
| return handler(ctx, reply, req) | |
| } | |
| return wrapped, func(id ID) { | |
| mu.Lock() | |
| cancel, found := handling[id] | |
| mu.Unlock() | |
| if found { | |
| cancel() | |
| } | |
| } | |
| } | |
| // AsyncHandler is a middleware that returns a handler that processes each | |
| // request goes in its own goroutine. | |
| // The handler returns immediately, without the request being processed. | |
| // Each request then waits for the previous request to finish before it starts. | |
| // This allows the stream to unblock at the cost of unbounded goroutines | |
| // all stalled on the previous one. | |
| func AsyncHandler(handler Handler) Handler { | |
| nextRequest := make(chan struct{}) | |
| close(nextRequest) | |
| return func(ctx context.Context, reply Replier, req Request) error { | |
| waitForPrevious := nextRequest | |
| nextRequest = make(chan struct{}) | |
| releaser := &releaser{ch: nextRequest} | |
| innerReply := reply | |
| reply = func(ctx context.Context, result any, err error) error { | |
| releaser.release(true) | |
| return innerReply(ctx, result, err) | |
| } | |
| _, queueDone := event.Start(ctx, "queued") | |
| ctx = context.WithValue(ctx, asyncKey, releaser) | |
| go func() { | |
| <-waitForPrevious | |
| queueDone() | |
| if err := handler(ctx, reply, req); err != nil { | |
| event.Error(ctx, "jsonrpc2 async message delivery failed", err) | |
| } | |
| }() | |
| return nil | |
| } | |
| } | |
| // Async, when used with the [AsyncHandler] middleware, indicates that the | |
| // current jsonrpc2 request may be handled asynchronously to subsequent | |
| // requests. | |
| // | |
| // When not used with an AsyncHandler, Async is a no-op. | |
| // | |
| // Async must be called at most once on each request's context (and its | |
| // descendants). | |
| func Async(ctx context.Context) { | |
| if r, ok := ctx.Value(asyncKey).(*releaser); ok { | |
| r.release(false) | |
| } | |
| } | |
| type asyncKeyType struct{} | |
| var asyncKey = asyncKeyType{} | |
| // A releaser implements concurrency safe 'releasing' of async requests. (A | |
| // request is released when it is allowed to run concurrent with other | |
| // requests, via a call to [Async].) | |
| type releaser struct { | |
| mu sync.Mutex | |
| ch chan struct{} | |
| released bool | |
| } | |
| // release closes the associated channel. If soft is set, multiple calls to | |
| // release are allowed. | |
| func (r *releaser) release(soft bool) { | |
| r.mu.Lock() | |
| defer r.mu.Unlock() | |
| if r.released { | |
| if !soft { | |
| panic("jsonrpc2.Async called multiple times") | |
| } | |
| } else { | |
| close(r.ch) | |
| r.released = true | |
| } | |
| } |