Merge branch 'main' into thrift

Author: Skyenought

docs/parser-zh.md

@@ -31,11 +31,9 @@
 
 ### LaunguageSpec
 
-```
 用于在 LSP 符号收集过程中转换为 UniAST 所需信息，并且这些信息非 LSP 通用定义
 
 ```go
-
 // Detailed implementation used for collect LSP symbols and transform them to UniAST
 type LanguageSpec interface {
     // initialize a root workspace, and return all modules [modulename=>abs-path] inside
@@ -50,7 +48,7 @@ type LanguageSpec interface {
     ShouldSkip(path string) bool
 
     // FileImports parse file codes to get its imports
-	FileImports(content []byte) ([]uniast.Import, error)
+    FileImports(content []byte) ([]uniast.Import, error)
 
     // return the first declaration token of a symbol, as Type-Name
     DeclareTokenOfSymbol(sym DocumentSymbol) int

docs/uniast-en.md

@@ -1,4 +1,4 @@
-# Universal Abstract-Syntax-Tree Specification (v0.1.2)
+# Universal Abstract-Syntax-Tree Specification (v0.1.3)
 
 Universal Abstract-Syntax-Tree is a LLM-friendly, language-agnostic code context data structure established by ABCoder. It represents a unified abstract syntax tree of a repository's code, collecting definitions of language entities (functions, types, constants/variables) and their interdependencies for subsequent AI understanding and coding-workflow development.
 
@@ -105,6 +105,9 @@ A repository consists of entity Modules and relationship Graph
 
 - Graph: Dependency topology graph of AST Nodes, see [Graph] below
 
+- Path: The file directory of the repository, usually should be an absolute path
+
+- ASTVersion: The UniAST version used to parse
 
 ### Module
 

docs/uniast-zh.md

@@ -1,4 +1,4 @@
-# Universal Abstract-Syntax-Tree Specification (v0.1.2)
+# Universal Abstract-Syntax-Tree Specification (v0.1.3)
 
 Universal Abstract-Syntax-Tree 是 ABCoder 建立的一种 LLM 亲和、语言无关的代码上下文数据结构，表示某个仓库代码的统一抽象语法树。收集了语言实体（函数、类型、常（变）量）的定义及其相互依赖关系，用于后续的 AI 理解、coding-workflow 开发。
 
@@ -105,6 +105,10 @@ Universal Abstract-Syntax-Tree 是 ABCoder 建立的一种 LLM 亲和、语言
 
 - Graph: AST Node 的依赖拓扑图，见下文【Graph】
 
+- Path: 仓库的文件目录，通常应该为绝对路径
+
+- ASTVersion: 解析时使用的 UniAST 版本
+
 
 ### Module
 

lang/collect/collect.go

@@ -168,7 +168,11 @@ func (c *Collector) Collect(ctx context.Context) error {
 
 		file := c.files[path]
 		if file == nil {
-			file = uniast.NewFile(path)
+			rel, err := filepath.Rel(c.repo, path)
+			if err != nil {
+				return err
+			}
+			file = uniast.NewFile(rel)
 			c.files[path] = file
 		}
 
@@ -180,8 +184,9 @@ func (c *Collector) Collect(ctx context.Context) error {
 		uses, err := c.spec.FileImports(content)
 		if err != nil {
 			log.Error("parse file %s use statements failed: %v", path, err)
+		} else {
+			file.Imports = uses
 		}
-		file.Imports = uses
 
 		// collect symbols
 		uri := NewURI(path)
@@ -210,7 +215,7 @@ func (c *Collector) Collect(ctx context.Context) error {
 		return nil
 	}
 	if err := filepath.Walk(c.repo, scanner); err != nil {
-		return err
+		log.Error("scan files failed: %v", err)
 	}
 
 	// collect some extra metadata

lang/collect/export.go

@@ -79,46 +79,34 @@ func (c *Collector) Export(ctx context.Context) (*uniast.Repository, error) {
 		_, _ = c.exportSymbol(&repo, symbol, "", visited)
 	}
 
-	// connect file with package on demands
-	// for p, m := range repo.Modules {
-	// 	if p == "" || strings.Contains(p, "@") {
-	// 		continue
-	// 	}
-	// 	for _, f := range m.Files {
-	// 		if f.Package != "" {
-	// 			continue
-	// 		}
-	// 		_, pkgpath, err := c.spec.NameSpace(filepath.Join(c.repo, f.Path))
-	// 		if err != nil {
-	// 			continue
-	// 		}
-	// 		f.Package = pkgpath
-	// 	}
-	// }
 	for fp, f := range c.files {
-		if f.Package != "" {
-			continue
-		}
 		rel, err := filepath.Rel(c.repo, fp)
 		if err != nil {
 			continue
 		}
+
 		modpath, pkgpath, err := c.spec.NameSpace(fp)
 		if err != nil {
 			continue
 		}
+
+		// connect file to package
 		if modpath == "" || strings.Contains(modpath, "@") {
 			continue
 		}
 		m, ok := repo.Modules[modpath]
 		if !ok {
 			continue
 		}
+
+		m.Files[rel] = f
+		if pkgpath == "" || f.Package != "" {
+			continue
+		}
 		if _, ok := m.Packages[pkgpath]; !ok {
 			continue
 		}
 		f.Package = pkgpath
-		m.Files[rel] = f
 	}
 
 	return &repo, nil
@@ -199,25 +187,15 @@ func (c *Collector) exportSymbol(repo *uniast.Repository, symbol *DocumentSymbol
 		repo.Modules[mod] = newModule(mod, "", c.Language)
 	}
 	module := repo.Modules[mod]
-	if repo.Modules[mod].Packages[path] == nil {
-		repo.Modules[mod].Packages[path] = uniast.NewPackage(path)
+	if module.Packages[path] == nil {
+		module.Packages[path] = uniast.NewPackage(path)
 	}
 	pkg := repo.Modules[mod].Packages[path]
 	if c.spec.IsMainFunction(*symbol) {
 		pkg.IsMain = true
 	}
 
-	var relfile string
-	if c.internal(symbol.Location) {
-		relfile, _ = filepath.Rel(c.repo, file)
-	} else {
-		relfile = filepath.Base(file)
-	}
 	fileLine := c.fileLine(symbol.Location)
-	// collect files
-	if module.Files[relfile] == nil {
-		module.Files[relfile] = uniast.NewFile(relfile)
-	}
 
 	content := symbol.Text
 	public := c.spec.IsPublicSymbol(*symbol)

lang/golang/parser/ctx.go

@@ -446,7 +446,8 @@ func (p *GoParser) collectTypes(ctx *fileContext, typ ast.Expr, st *Type, inline
 
 // get type id and tells if it is std or builtin
 func (ctx *fileContext) getTypeinfo(typ types.Type) (ti typeInfo) {
-	tobjs, isPointer, isNamed := getNamedTypes(typ)
+	visited := make(map[types.Type]bool)
+	tobjs, isPointer, isNamed := getNamedTypes(typ, visited)
 	ti.IsPointer = isPointer
 	ti.Ty = typ
 	ti.IsNamed = isNamed

lang/golang/parser/utils.go

@@ -183,63 +183,69 @@ func getTypeKind(n ast.Expr) TypeKind {
 	}
 }
 
-func getNamedTypes(typ types.Type) (tys []types.Object, isPointer bool, isNamed bool) {
+func getNamedTypes(typ types.Type, visited map[types.Type]bool) (tys []types.Object, isPointer bool, isNamed bool) {
+	if visited[typ] {
+		return nil, false, false
+	}
+
+	visited[typ] = true
+
 	switch t := typ.(type) {
 	case *types.Pointer:
 		isPointer = true
-		typs, _, isNamed2 := getNamedTypes(t.Elem())
+		var typs []types.Object
+		typs, _, isNamed = getNamedTypes(t.Elem(), visited)
 		tys = append(tys, typs...)
-		isNamed = isNamed2
 	case *types.Slice:
-		typs, _, _ := getNamedTypes(t.Elem())
+		typs, _, _ := getNamedTypes(t.Elem(), visited)
 		tys = append(tys, typs...)
 	case *types.Array:
-		typs, _, _ := getNamedTypes(t.Elem())
+		typs, _, _ := getNamedTypes(t.Elem(), visited)
 		tys = append(tys, typs...)
 	case *types.Chan:
-		typs, _, _ := getNamedTypes(t.Elem())
+		typs, _, _ := getNamedTypes(t.Elem(), visited)
 		tys = append(tys, typs...)
 	case *types.Tuple:
 		for i := 0; i < t.Len(); i++ {
-			typs, _, _ := getNamedTypes(t.At(i).Type())
+			typs, _, _ := getNamedTypes(t.At(i).Type(), visited)
 			tys = append(tys, typs...)
 		}
 	case *types.Map:
-		typs2, _, _ := getNamedTypes(t.Elem())
-		typs1, _, _ := getNamedTypes(t.Key())
+		typs2, _, _ := getNamedTypes(t.Elem(), visited)
+		typs1, _, _ := getNamedTypes(t.Key(), visited)
 		tys = append(tys, typs1...)
 		tys = append(tys, typs2...)
 	case *types.Named:
 		tys = append(tys, t.Obj())
 		isNamed = true
 	case *types.Struct:
 		for i := 0; i < t.NumFields(); i++ {
-			typs, _, _ := getNamedTypes(t.Field(i).Type())
+			typs, _, _ := getNamedTypes(t.Field(i).Type(), visited)
 			tys = append(tys, typs...)
 		}
 	case *types.Interface:
 		for i := 0; i < t.NumEmbeddeds(); i++ {
-			typs, _, _ := getNamedTypes(t.EmbeddedType(i))
+			typs, _, _ := getNamedTypes(t.EmbeddedType(i), visited)
 			tys = append(tys, typs...)
 		}
 		for i := 0; i < t.NumExplicitMethods(); i++ {
-			typs, _, _ := getNamedTypes(t.ExplicitMethod(i).Type())
+			typs, _, _ := getNamedTypes(t.ExplicitMethod(i).Type(), visited)
 			tys = append(tys, typs...)
 		}
 	case *types.TypeParam:
-		typs, _, _ := getNamedTypes(t.Constraint())
+		typs, _, _ := getNamedTypes(t.Constraint(), visited)
 		tys = append(tys, typs...)
 	case *types.Alias:
 		var typs []types.Object
-		typs, isPointer, isNamed = getNamedTypes(t.Rhs())
+		typs, isPointer, isNamed = getNamedTypes(t.Rhs(), visited)
 		tys = append(tys, typs...)
 	case *types.Signature:
 		for i := 0; i < t.Params().Len(); i++ {
-			typs, _, _ := getNamedTypes(t.Params().At(i).Type())
+			typs, _, _ := getNamedTypes(t.Params().At(i).Type(), visited)
 			tys = append(tys, typs...)
 		}
 		for i := 0; i < t.Results().Len(); i++ {
-			typs, _, _ := getNamedTypes(t.Results().At(i).Type())
+			typs, _, _ := getNamedTypes(t.Results().At(i).Type(), visited)
 			tys = append(tys, typs...)
 		}
 	}