fix: correct comment parsing for block comments and update gogen to v1.20.2

_xtool/internal/parser/parser.go

@@ -193,11 +193,28 @@ func (ct *Converter) ParseCommentGroup(cursor clang.Cursor) (comentGroup *ast.Co
 }
 
 func (ct *Converter) ParseComment(rawComment string) *ast.CommentGroup {
-	lines := strings.Split(rawComment, "\n")
 	commentGroup := &ast.CommentGroup{}
-	for _, line := range lines {
-		commentGroup.List = append(commentGroup.List, &ast.Comment{Text: line + "\n"})
+
+	// Block comments (/* ... */ or /** ... */) should be kept as a single Comment node.
+	// Go's ast.Comment documentation states: "A Comment node represents a single //-style or /*-style comment."
+	// This means the entire block comment must be ONE ast.Comment node, not split across multiple nodes.
+	if strings.HasPrefix(rawComment, "/*") {
+		// Ensure the comment ends with newline for proper formatting
+		text := rawComment
+		if !strings.HasSuffix(text, "\n") {
+			text += "\n"
+		}
+		commentGroup.List = append(commentGroup.List, &ast.Comment{Text: text})
+	} else {
+		// Line comments (// or ///) - each line is a separate Comment node
+		lines := strings.Split(rawComment, "\n")
+		for _, line := range lines {
+			if line != "" {
+				commentGroup.List = append(commentGroup.List, &ast.Comment{Text: line + "\n"})
+			}
+		}
 	}
+
 	return commentGroup
 }
 

_xtool/internal/parser/testdata/comment/expect.json

@@ -353,19 +353,7 @@
       "Doc": {
         "List": [
           {
-            "Text": "/**\n",
-            "_Type": "Comment"
-          },
-          {
-            "Text": " * doc 1\n",
-            "_Type": "Comment"
-          },
-          {
-            "Text": " * doc 2\n",
-            "_Type": "Comment"
-          },
-          {
-            "Text": " */\n",
+            "Text": "/**\n * doc 1\n * doc 2\n */\n",
             "_Type": "Comment"
           }
         ],
@@ -524,15 +512,7 @@
               "Doc": {
                 "List": [
                   {
-                    "Text": "/**\n",
-                    "_Type": "Comment"
-                  },
-                  {
-                    "Text": "     * static field doc\n",
-                    "_Type": "Comment"
-                  },
-                  {
-                    "Text": "     */\n",
+                    "Text": "/**\n     * static field doc\n     */\n",
                     "_Type": "Comment"
                   }
                 ],
@@ -584,15 +564,7 @@
               "Doc": {
                 "List": [
                   {
-                    "Text": "/**\n",
-                    "_Type": "Comment"
-                  },
-                  {
-                    "Text": "     * field doc\n",
-                    "_Type": "Comment"
-                  },
-                  {
-                    "Text": "     */\n",
+                    "Text": "/**\n     * field doc\n     */\n",
                     "_Type": "Comment"
                   }
                 ],
@@ -672,15 +644,7 @@
             "Doc": {
               "List": [
                 {
-                  "Text": "/**\n",
-                  "_Type": "Comment"
-                },
-                {
-                  "Text": "     * method doc\n",
-                  "_Type": "Comment"
-                },
-                {
-                  "Text": "     */\n",
+                  "Text": "/**\n     * method doc\n     */\n",
                   "_Type": "Comment"
                 }
               ],

go.mod

@@ -3,7 +3,7 @@ module github.com/goplus/llcppg
 go 1.23.0
 
 require (
-	github.com/goplus/gogen v1.19.7
+	github.com/goplus/gogen v1.20.2
 	github.com/goplus/lib v0.3.1
 	github.com/goplus/llgo v0.12.0
 	github.com/goplus/mod v0.19.0

go.sum

@@ -1,5 +1,5 @@
-github.com/goplus/gogen v1.19.7 h1:0i30on0GwtYIJ+D9/I5QujswBU+mnnmNewoRk/uRVkw=
-github.com/goplus/gogen v1.19.7/go.mod h1:owX2e1EyU5WD+Nm6oH2m/GXjLdlBYcwkLO4wN8HHXZI=
+github.com/goplus/gogen v1.20.2 h1:c9wm7NzjWSrncbtH+lz4jM2j31p+6JTji8cjF1K79qg=
+github.com/goplus/gogen v1.20.2/go.mod h1:87ZJD1mdljXx2pkvBrMGynGUz6m08brj9p6xhb5aq2Y=
 github.com/goplus/lib v0.3.1 h1:Xws4DBVvgOMu58awqB972wtvTacDbk3nqcbHjdx9KSg=
 github.com/goplus/lib v0.3.1/go.mod h1:SgJv3oPqLLHCu0gcL46ejOP3x7/2ry2Jtxu7ta32kp0=
 github.com/goplus/llgo v0.12.0 h1:UhbmwR+9fSy1y944rp6fPkSP39n4YhH4TpAN2QJ15ns=

reproduce_comment_bug.go

@@ -0,0 +1,119 @@
+// Minimal reproduction for issue #618: Multi-line block comment bug with gogen
+//
+// This demonstrates how incorrectly structured ast.Comment nodes cause
+// invalid Go code generation with gogen v1.20.2.
+//
+// Run: go run reproduce_comment_bug.go
+
+package main
+
+import (
+	"bytes"
+	"fmt"
+	"go/ast"
+	"go/format"
+	"go/token"
+	"go/types"
+	"strings"
+
+	"github.com/goplus/gogen"
+)
+
+// This creates BUGGY comments - splits block comment across multiple nodes
+func createBuggyComments() *ast.CommentGroup {
+	// This is what llcppg's buggy ParseComment() does:
+	// It splits multi-line block comments by newline
+	return &ast.CommentGroup{
+		List: []*ast.Comment{
+			{Text: "/* Create an iterator for traversing a domain\n"},
+			{Text: "   The domain NULL denotes the default domain */\n"},
+		},
+	}
+}
+
+// This creates CORRECT comments - keeps block comment as single node
+func createCorrectComments() *ast.CommentGroup {
+	// This is what should be done:
+	// Keep the entire block comment as a single ast.Comment node
+	return &ast.CommentGroup{
+		List: []*ast.Comment{
+			{Text: "/* Create an iterator for traversing a domain\n   The domain NULL denotes the default domain */\n"},
+		},
+	}
+}
+
+func generateCode(comments *ast.CommentGroup, label string) {
+	fmt.Println(label)
+	fmt.Println(strings.Repeat("=", len(label)))
+
+	pkg := gogen.NewPackage("", "test", nil)
+
+	// Create a type declaration with the comment
+	tyDecl := pkg.NewTypeDefs()
+	spec := tyDecl.NewType("ExampleType")
+	spec.InitType(pkg, types.Typ[types.Int])
+
+	// Attach the comment as documentation
+	if comments != nil {
+		spec.SetComments(pkg, comments)
+	}
+
+	// Get the generated AST and try to format it
+	file := pkg.ASTFile()
+
+	var buf bytes.Buffer
+	fset := token.NewFileSet()
+	err := format.Node(&buf, fset, file)
+	if err != nil {
+		fmt.Printf("❌ ERROR: %v\n\n", err)
+		fmt.Println("This demonstrates the bug with gogen v1.20.2!")
+	} else {
+		fmt.Println("✅ Generated Go code:")
+		fmt.Println(buf.String())
+	}
+	fmt.Println()
+}
+
+func main() {
+	fmt.Println("=== Gogen Comment Bug Reproduction (Issue #618) ===\n")
+
+	fmt.Println("This reproduces the bug WITHOUT involving llcppg - pure gogen usage.\n")
+
+	// Test 1: Buggy comment structure (what llcppg currently generates)
+	fmt.Println("Test 1: BUGGY comment structure")
+	fmt.Println("--------------------------------")
+	fmt.Println("Comment split into 2 nodes:")
+	buggy := createBuggyComments()
+	for i, c := range buggy.List {
+		fmt.Printf("  [%d]: %q\n", i, c.Text)
+	}
+	fmt.Println()
+
+	generateCode(buggy, "Generated Go code (will have malformed comments):")
+
+	// Test 2: Correct comment structure
+	fmt.Println("\nTest 2: CORRECT comment structure")
+	fmt.Println("----------------------------------")
+	fmt.Println("Comment as single node:")
+	correct := createCorrectComments()
+	for i, c := range correct.List {
+		fmt.Printf("  [%d]: %q\n", i, c.Text)
+	}
+	fmt.Println()
+
+	generateCode(correct, "Generated Go code (properly formatted comments):")
+
+	fmt.Println("=== Explanation ===")
+	fmt.Println("Go's ast.Comment documentation states each Comment represents")
+	fmt.Println("'a single //-style or /*-style comment' - emphasis on SINGLE.")
+	fmt.Println()
+	fmt.Println("Block comments /* ... */ must be ONE ast.Comment node, not split")
+	fmt.Println("across multiple nodes. When gogen v1.20.2 prints malformed comment")
+	fmt.Println("structures, it generates invalid Go syntax.")
+	fmt.Println()
+	fmt.Println("This is the root cause of test failures in:")
+	fmt.Println("  - TestFromTestdata/gettext")
+	fmt.Println("  - TestFromTestdata/gpgerror")
+	fmt.Println("  - TestFromTestdata/issue507")
+	fmt.Println("  - TestFromTestdata/keepcomment")
+}