// Copyright 2018 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 source provides core features for use by Go editors and tools. package source import ( "bytes" "context" "fmt" "go/ast" "go/format" "go/token" "strings" "golang.org/x/tools/go/ast/astutil" "golang.org/x/tools/imports" "golang.org/x/tools/internal/lsp/diff" ) // Format formats a file with a given range. func Format(ctx context.Context, f File, rng Range) ([]TextEdit, error) { fAST := f.GetAST() path, exact := astutil.PathEnclosingInterval(fAST, rng.Start, rng.End) if !exact || len(path) == 0 { return nil, fmt.Errorf("no exact AST node matching the specified range") } node := path[0] // format.Node can fail when the AST contains a bad expression or // statement. For now, we preemptively check for one. // TODO(rstambler): This should really return an error from format.Node. var isBad bool ast.Inspect(node, func(n ast.Node) bool { switch n.(type) { case *ast.BadDecl, *ast.BadExpr, *ast.BadStmt: isBad = true return false default: return true } }) if isBad { return nil, fmt.Errorf("unable to format file due to a badly formatted AST") } // format.Node changes slightly from one release to another, so the version // of Go used to build the LSP server will determine how it formats code. // This should be acceptable for all users, who likely be prompted to rebuild // the LSP server on each Go release. fset := f.GetFileSet() buf := &bytes.Buffer{} if err := format.Node(buf, fset, node); err != nil { return nil, err } return computeTextEdits(f, buf.String()), nil } // Imports formats a file using the goimports tool. func Imports(ctx context.Context, f File, rng Range) ([]TextEdit, error) { formatted, err := imports.Process(f.GetToken().Name(), f.GetContent(), nil) if err != nil { return nil, err } return computeTextEdits(f, string(formatted)), nil } func computeTextEdits(file File, formatted string) (edits []TextEdit) { u := strings.SplitAfter(string(file.GetContent()), "\n") tok := file.GetToken() f := strings.SplitAfter(formatted, "\n") for _, op := range diff.Operations(u, f) { start := lineStart(tok, op.I1+1) if start == token.NoPos && op.I1 == len(u) { start = tok.Pos(tok.Size()) } end := lineStart(tok, op.I2+1) if end == token.NoPos && op.I2 == len(u) { end = tok.Pos(tok.Size()) } switch op.Kind { case diff.Delete: // Delete: unformatted[i1:i2] is deleted. edits = append(edits, TextEdit{ Range: Range{ Start: start, End: end, }, }) case diff.Insert: // Insert: formatted[j1:j2] is inserted at unformatted[i1:i1]. edits = append(edits, TextEdit{ Range: Range{ Start: start, End: start, }, NewText: op.Content, }) } } return edits }