feat(参照标准): 规范资源处理，完善版本编辑

ofdgo_color_profile.go

@@ -68,7 +68,7 @@ func (r *Reader) colorProfile(space *ColorSpace) (*colorProfile, error) {
 	}
 	name := r.ResPath(space.Profile)
 	for _, other := range r.colorSpaceCache {
-		if other != nil && other.profile != nil && strings.EqualFold(other.Type, space.Type) && strings.EqualFold(r.ResPath(other.Profile), name) {
+		if other != nil && other.profile != nil && strings.EqualFold(other.Type, space.Type) && r.ResPath(other.Profile) == name {
 			space.profile = other.profile
 			return space.profile, space.profile.err
 		}

ofdgo_compression.go

@@ -136,7 +136,7 @@ func (e *Editor) outputSnapshot(ctx context.Context, options WriteOptions) (*Edi
 	}
 	for _, resource := range e.resources {
 		if resource.image != nil {
-			snapshot.output.images[strings.ToLower(cleanPackagePath(resource.name))] = true
+			snapshot.output.images[cleanPackagePath(resource.name)] = true
 		}
 	}
 	if e.source != nil {
@@ -154,7 +154,7 @@ func (e *Editor) outputSnapshot(ctx context.Context, options WriteOptions) (*Edi
 			return nil, err
 		}
 		for _, img := range images {
-			snapshot.output.images[strings.ToLower(cleanPackagePath(img.Location))] = true
+			snapshot.output.images[cleanPackagePath(img.Location)] = true
 		}
 	}
 	if options.Compression.Mode == CompressionLossy {
@@ -188,9 +188,8 @@ func (r *Reader) compressionImageSafety(ctx context.Context, images []ImageInfo)
 			}
 			safe = false
 		}
-		for key := range dependencies {
+		for name := range dependencies {
-			name, exists := r.fileNamesFold[key]
+			if !r.fileNames[name] {
-			if !exists {
 				safe = false
 				break
 			}
@@ -269,7 +268,7 @@ func (r *Reader) compressionImageSafety(ctx context.Context, images []ImageInfo)
 	paths := make(map[string]bool)
 	for _, img := range images {
 		if !safe || masks[editorResourceID(img.ID)] {
-			paths[strings.ToLower(cleanPackagePath(img.Location))] = true
+			paths[cleanPackagePath(img.Location)] = true
 		}
 	}
 	return paths, safe && geometry, ctx.Err()

ofdgo_compression_images.go

@@ -19,7 +19,6 @@ import (
 	"fmt"
 	"image"
 	"math"
-	"strings"
 )
 
 // compressionImages 记录图片各处使用的最大像素需求
@@ -71,7 +70,7 @@ func (r *Reader) compressionImagePlan(ctx context.Context, options CompressionOp
 			return nil, nil, err
 		}
 		for _, img := range catalog {
-			name := strings.ToLower(cleanPackagePath(img.Location))
+			name := cleanPackagePath(img.Location)
 			if old, ok := images[name]; !ok || old {
 				images[name] = !masks[name]
 			}
@@ -150,7 +149,7 @@ func (r *Reader) compressionImageSizes(ctx context.Context, images []ImageInfo,
 	}
 	result := make(map[string]image.Point)
 	for _, img := range images {
-		name := strings.ToLower(cleanPackagePath(img.Location))
+		name := cleanPackagePath(img.Location)
 		size := visitor.sizes[editorResourceID(img.ID)]
 		old := result[name]
 		result[name] = image.Pt(max(old.X, size.X), max(old.Y, size.Y))

ofdgo_compression_resources.go

@@ -76,7 +76,7 @@ func (e *Editor) compressResourceReferences(parts map[string][]byte, reader *Rea
 		if err := e.output.ctx.Err(); err != nil {
 			return err
 		}
-		key := strings.ToLower(cleanPackagePath(name))
+		key := cleanPackagePath(name)
 		if removed[key] || strings.HasSuffix(name, "/") {
 			continue
 		}
@@ -126,6 +126,7 @@ func (e *Editor) compressResourceReferences(parts map[string][]byte, reader *Rea
 		if err != nil {
 			return err
 		}
+		preserved := refs.versions[cleanPackagePath(name)]
 		for _, group := range root.children {
 			for _, node := range group.children {
 				field := ""
@@ -147,11 +148,11 @@ func (e *Editor) compressResourceReferences(parts map[string][]byte, reader *Rea
 					return err
 				}
 				key := editorResourceLocation(name, root.attr("BaseLoc"), value)
-				if field == "MediaFile" && node.attr("Type") != "Image" {
+				if preserved || field == "MediaFile" && node.attr("Type") != "Image" {
 					refs.files[key] = true
 					continue
 				}
-				if key != strings.ToLower(cleanPackagePath(resolveResourcePath(name, root.attr("BaseLoc"), value))) {
+				if key != cleanPackagePath(resolveResourcePath(name, root.attr("BaseLoc"), value)) {
 					return nil
 				}
 				if actual[key] != "" && !refs.files[key] {
@@ -259,10 +260,10 @@ func (e *Editor) compressResourceReferences(parts map[string][]byte, reader *Rea
 						}
 						stem := strings.TrimSuffix(actual[key], ext)
 						name := stem + ".jpg"
-						for n := 1; reserved[strings.ToLower(name)]; n++ {
+						for n := 1; reserved[name]; n++ {
 							name = fmt.Sprintf("%s-%d.jpg", stem, n)
 						}
-						reserved[strings.ToLower(name)] = true
+						reserved[name] = true
 						locations[key] = name
 						parts[name] = data
 						delete(parts, actual[key])
@@ -286,6 +287,9 @@ func (e *Editor) compressResourceReferences(parts map[string][]byte, reader *Rea
 		}
 	}
 	for _, name := range resources {
+		if refs.versions[cleanPackagePath(name)] {
+			continue
+		}
 		data, err := read(name)
 		if err != nil {
 			return err

ofdgo_crypto_package.go

@@ -124,7 +124,7 @@ func (r *Reader) open(options []ReaderOption) error {
 	if err := r.indexPackage(); err != nil {
 		return err
 	}
-	if _, ok := r.fileNamesFold["encryptions.xml"]; ok {
+	if r.fileNames["Encryptions.xml"] {
 		if err := r.decryptPackage(settings); err != nil {
 			return err
 		}
@@ -135,8 +135,8 @@ func (r *Reader) open(options []ReaderOption) error {
 // packageData 获取包内所有条目的实际字节，保留未识别文件
 // 返回: map[string][]byte 包内条目, error 错误信息
 func (r *Reader) packageData() (map[string][]byte, error) {
-	parts := make(map[string][]byte, len(r.fileNamesFold))
+	parts := make(map[string][]byte, len(r.fileNames))
-	for _, name := range r.fileNamesFold {
+	for name := range r.fileNames {
 		data, err := r.readFile(name)
 		if err != nil {
 			return nil, err
@@ -198,8 +198,7 @@ func (r *Reader) decryptPackage(settings readerOptions) error {
 		extraSeeds := make(map[string]bool)
 		for _, entry := range entries.Entries {
 			name := cleanPackagePath(entry.Path)
-			fold := strings.ToLower(name)
+			if !validPackagePath(name) || used[name] {
-			if !validPackagePath(name) || used[fold] {
 				clear(key)
 				return fmt.Errorf("ambiguous encrypted package path: %q", entry.Path)
 			}
@@ -231,7 +230,7 @@ func (r *Reader) decryptPackage(settings readerOptions) error {
 				return ErrInvalidCredentials
 			}
 			delete(parts, actual)
-			used[fold] = true
+			used[name] = true
 		}
 		clear(key)
 		delete(parts, seedName)
@@ -239,16 +238,7 @@ func (r *Reader) decryptPackage(settings readerOptions) error {
 		for name := range extraSeeds {
 			delete(parts, name)
 		}
-		remaining := make(map[string]string, len(parts))
+		maps.Copy(parts, decrypted)
-		for name := range parts {
-			remaining[strings.ToLower(name)] = name
-		}
-		for name, plain := range decrypted {
-			if existing, ok := remaining[strings.ToLower(name)]; ok {
-				delete(parts, existing)
-			}
-			parts[name] = plain
-		}
 		state.info.Method = seed.Method
 		for _, user := range seed.Users {
 			state.info.Users = append(state.info.Users, user.Name)
@@ -257,11 +247,7 @@ func (r *Reader) decryptPackage(settings readerOptions) error {
 			state.options = inheritedEncryptionOptions(seed, credential, settings.provider)
 		}
 	}
-	for name := range parts {
+	delete(parts, "Encryptions.xml")
-		if strings.ToLower(name) == "encryptions.xml" {
-			delete(parts, name)
-		}
-	}
 	r.Zip, r.files, r.encryption = nil, parts, state
 	return nil
 }
@@ -289,7 +275,7 @@ func decodeEncryptionEntries(data, key, iv []byte, provider CryptoProvider) (enc
 	return entries, nil
 }
 
-// encryptionPart 按统一大小写规则读取加密元数据引用
+// encryptionPart 按大小写敏感路径读取加密元数据引用
 // 入参: parts 包内条目, name 引用路径
 // 返回: []byte 条目内容, string 实际路径, error 错误信息
 func encryptionPart(parts map[string][]byte, name string) ([]byte, string, error) {
@@ -300,11 +286,6 @@ func encryptionPart(parts map[string][]byte, name string) ([]byte, string, error
 	if data, ok := parts[name]; ok {
 		return data, name, nil
 	}
-	for actual, data := range parts {
-		if strings.ToLower(actual) == strings.ToLower(name) {
-			return data, actual, nil
-		}
-	}
 	return nil, "", fmt.Errorf("missing encrypted entry: %s", name)
 }
 

ofdgo_editor_action.go

@@ -49,7 +49,7 @@ func (e *Editor) SetDocumentActions(actions []Action) error {
 			return err
 		}
 	}
-	name := cleanPackagePath(base.reader.OFD.DocBody[base.reader.documentIndex].DocRoot)
+	name := base.reader.documentRoot
 	data, err := base.reader.readFile(name)
 	if err != nil {
 		return err

ofdgo_editor_annot.go

@@ -269,7 +269,7 @@ func (e *Editor) editAnnotations(index int, ids []string, edit func([]byte, *edi
 	if len(patches) != 0 {
 		parts[name] = editorPatchXML(data, patches)
 		if remaining == 0 {
-			docName, err := editorPageLocation(reader, nil, "", "/"+cleanPackagePath(reader.OFD.DocBody[reader.documentIndex].DocRoot))
+			docName, err := editorPageLocation(reader, nil, "", "/"+reader.documentRoot)
 			if err != nil {
 				return err
 			}

ofdgo_editor_annot_content.go

@@ -16,6 +16,7 @@ package ofdgo
 
 import (
 	"bytes"
+	"context"
 	"encoding/xml"
 	"fmt"
 	"maps"
@@ -379,7 +380,7 @@ func (e *Editor) appendAnnotations(index int, annotations []byte, generated bool
 	}
 	parts := map[string][]byte{name: data, file: content}
 	if base.document.Annotations == "" {
-		docName := cleanPackagePath(reader.OFD.DocBody[reader.documentIndex].DocRoot)
+		docName := reader.documentRoot
 		docData, err := reader.readFile(docName)
 		if err != nil {
 			return err
@@ -448,6 +449,10 @@ func (e *Editor) annotationFiles() []string {
 // 入参: base 原包, parts 修改的文件
 // 返回: error 错误信息
 func (e *Editor) commitAnnotationParts(base *editorSource, parts map[string][]byte) error {
+	parts, err := base.reader.versionChanges(context.Background(), parts)
+	if err != nil {
+		return err
+	}
 	files := maps.Clone(base.reader.files)
 	if files == nil {
 		files = make(map[string][]byte)
@@ -458,7 +463,7 @@ func (e *Editor) commitAnnotationParts(base *editorSource, parts map[string][]by
 		}
 		files[name] = data
 	}
-	reader := &Reader{Zip: base.reader.Zip, files: files, encryption: base.reader.encryption, documentIndex: base.reader.documentIndex}
+	reader := &Reader{Zip: base.reader.Zip, files: files, encryption: base.reader.encryption, documentIndex: base.reader.documentIndex, selectedVersion: base.reader.selectedVersion}
 	if err := reader.initRoot(); err != nil {
 		return err
 	}

ofdgo_editor_annot_ops.go

@@ -141,7 +141,7 @@ func (e *Editor) PasteAnnotations(page int, selection *AnnotationSelection, dx,
 		edit.source.annotationStates = states
 		if len(selection.resources) != 0 {
 			reader := edit.source.reader
-			name := cleanPackagePath(reader.OFD.DocBody[reader.documentIndex].DocRoot)
+			name := reader.documentRoot
 			data, err := reader.readFile(name)
 			if err != nil {
 				return err

ofdgo_editor_attachment.go

@@ -73,7 +73,7 @@ func (e *Editor) AddAttachment(name string, data []byte) (string, error) {
 		return "", err
 	}
 	entry = bytes.TrimPrefix(entry, []byte(xml.Header))
-	docName := cleanPackagePath(base.reader.OFD.DocBody[base.reader.documentIndex].DocRoot)
+	docName := base.reader.documentRoot
 	docData, err := base.reader.readFile(docName)
 	if err != nil {
 		return "", err
@@ -220,7 +220,7 @@ func (e *Editor) editAttachment(id string, change func([]byte, *editorXML, map[s
 	if base == nil {
 		return fmt.Errorf("attachment not found: %s", id)
 	}
-	name := cleanPackagePath(base.reader.OFD.DocBody[base.reader.documentIndex].DocRoot)
+	name := base.reader.documentRoot
 	external := strings.TrimSpace(base.document.Attachments.Path) != ""
 	if external {
 		name = base.reader.ResPath(base.document.Attachments.Path)
@@ -277,7 +277,7 @@ func (e *Editor) editAttachment(id string, change func([]byte, *editorXML, map[s
 			if retired == nil {
 				retired = make(map[string]bool)
 			}
-			retired[strings.ToLower(cleanPackagePath(base.reader.ResPath(location)))] = true
+			retired[editorResourceLocation(name, "", location)] = true
 			e.source.retiredAttachments = retired
 		}
 	}

ofdgo_editor_documents.go

@@ -93,7 +93,7 @@ func (e *Editor) SetDocumentInfo(index int, info DocInfo) error {
 		if err != nil {
 			return 0, err
 		}
-		reader.files[reader.fileNamesFold["ofd.xml"]] = updated
+		reader.files["OFD.xml"] = updated
 		return reader.DocumentIndex(), nil
 	})
 }
@@ -116,8 +116,8 @@ func (e *Editor) AddDocument(title string, width, height float64, at int) (int,
 		for number := 0; ; number++ {
 			candidate := fmt.Sprintf("Doc_%d", number)
 			occupied := false
-			for name := range reader.fileNamesFold {
+			for name := range reader.fileNames {
-				if name == strings.ToLower(candidate) || strings.HasPrefix(name, strings.ToLower(candidate)+"/") {
+				if name == candidate || strings.HasPrefix(name, candidate+"/") {
 					occupied = true
 					break
 				}
@@ -156,7 +156,7 @@ func (e *Editor) AddDocument(title string, width, height float64, at int) (int,
 		if existing := root.childAt("DocBody", at); existing != nil {
 			position = existing.start
 		}
-		reader.files[reader.fileNamesFold["ofd.xml"]] = editorPatchXML(data, []editorXMLPatch{{position, position, bytes.TrimPrefix(body, []byte(xml.Header))}})
+		reader.files["OFD.xml"] = editorPatchXML(data, []editorXMLPatch{{position, position, bytes.TrimPrefix(body, []byte(xml.Header))}})
 		return at, nil
 	})
 	if err != nil {
@@ -199,7 +199,7 @@ func (e *Editor) MoveDocument(from, to int) error {
 			body := root.childAt("DocBody", index)
 			patches = append(patches, editorXMLPatch{body.start, body.end, content})
 		}
-		reader.files[reader.fileNamesFold["ofd.xml"]] = editorPatchXML(data, patches)
+		reader.files["OFD.xml"] = editorPatchXML(data, patches)
 		selected := reader.DocumentIndex()
 		if selected == from {
 			selected = to
@@ -232,10 +232,10 @@ func (e *Editor) DeleteDocument(index int) error {
 		directory := path.Dir(cleanPackagePath(reader.OFD.DocBody[index].DocRoot))
 		if directory != "." && !slices.ContainsFunc(reader.OFD.DocBody, func(body DocBody) bool {
 			name := cleanPackagePath(body.DocRoot)
-			return name != cleanPackagePath(reader.OFD.DocBody[index].DocRoot) && strings.HasPrefix(strings.ToLower(name), strings.ToLower(directory)+"/")
+			return name != cleanPackagePath(reader.OFD.DocBody[index].DocRoot) && strings.HasPrefix(name, directory+"/")
 		}) {
-			for name := range reader.fileNamesFold {
+			for name := range reader.fileNames {
-				if strings.HasPrefix(name, strings.ToLower(directory)+"/") {
+				if strings.HasPrefix(name, directory+"/") {
 					removed[name] = true
 				}
 			}
@@ -252,19 +252,19 @@ func (e *Editor) DeleteDocument(index int) error {
 				delete(removed, name)
 			}
 		}
-		delete(removed, "ofd.xml")
+		delete(removed, "OFD.xml")
 		for name := range reader.files {
-			if removed[strings.ToLower(name)] {
+			if removed[name] {
 				delete(reader.files, name)
 			}
 		}
 		if reader.Zip != nil {
 			reader.Zip = &zip.Reader{Comment: reader.Zip.Comment, File: slices.DeleteFunc(slices.Clone(reader.Zip.File), func(file *zip.File) bool {
-				return removed[strings.ToLower(cleanPackagePath(file.Name))]
+				return removed[cleanPackagePath(file.Name)]
 			})}
 		}
 		body := root.childAt("DocBody", index)
-		reader.files[reader.fileNamesFold["ofd.xml"]] = editorPatchXML(data, []editorXMLPatch{{start: body.start, end: body.end}})
+		reader.files["OFD.xml"] = editorPatchXML(data, []editorXMLPatch{{start: body.start, end: body.end}})
 		selected := reader.DocumentIndex()
 		if selected > index {
 			selected--
@@ -292,6 +292,7 @@ func (e *Editor) editDocuments(edit func(*Reader, []byte, *editorXML) (int, erro
 		return err
 	}
 	reader.files = maps.Clone(reader.files)
+	previous := reader.OFD.DocBody[reader.documentIndex]
 	index, err := edit(reader, data, root)
 	if err != nil {
 		return err
@@ -301,10 +302,14 @@ func (e *Editor) editDocuments(edit func(*Reader, []byte, *editorXML) (int, erro
 	if err := reader.initRoot(); err != nil {
 		return err
 	}
+	current := reader.OFD.DocBody[index]
+	if previous.DocRoot != current.DocRoot || previous.DocInfo.DocID != current.DocInfo.DocID {
+		reader.selectedVersion = nil
+	}
 	if _, err := reader.Doc(); err != nil {
 		return err
 	}
-	if e.source != nil && e.documentRoot() == reader.OFD.DocBody[index].DocRoot && e.Info.DocID == reader.OFD.DocBody[index].DocInfo.DocID {
+	if e.source != nil && e.documentRoot() == reader.documentRoot && e.Info.DocID == reader.OFD.DocBody[index].DocInfo.DocID {
 		source := *e.source
 		source.reader = reader
 		e.source = &source
@@ -331,7 +336,7 @@ func (e *Editor) documentRoot() string {
 	if e.source == nil {
 		return ""
 	}
-	return e.source.reader.OFD.DocBody[e.DocumentIndex()].DocRoot
+	return e.source.reader.documentRoot
 }
 
 // sameDocument 判断快照是否属于同一文档，区别删除后复用目录的新文档

ofdgo_editor_font_binding.go

@@ -407,6 +407,9 @@ func sameBindingFont(source, target []byte) bool {
 // 返回: FontFile 独立字体, error 匹配或解析错误
 func (e *Editor) resolveBindingFont(scan *editorFontDocument, declaration *editorFontDeclaration) (FontFile, error) {
 	reader := &Reader{Zip: scan.reader.Zip, files: maps.Clone(scan.reader.files), documentIndex: scan.index}
+	if scan.index == scan.reader.documentIndex {
+		reader.selectedVersion = scan.reader.selectedVersion
+	}
 	if err := reader.initRoot(); err != nil {
 		return FontFile{}, err
 	}
@@ -501,7 +504,7 @@ func (e *Editor) commitFontSnapshot(ctx context.Context, reader *Reader) error {
 	if err := ctx.Err(); err != nil {
 		return err
 	}
-	clean := &Reader{Zip: reader.Zip, files: maps.Clone(reader.files), documentIndex: reader.DocumentIndex(), encryption: e.encryption}
+	clean := &Reader{Zip: reader.Zip, files: maps.Clone(reader.files), documentIndex: reader.DocumentIndex(), encryption: e.encryption, selectedVersion: reader.selectedVersion}
 	if err := clean.initRoot(); err != nil {
 		return err
 	}
@@ -544,7 +547,7 @@ func pruneExternalFontFiles(ctx context.Context, reader *Reader, candidates ...s
 	}
 	removed := make(map[string]bool, len(candidates))
 	for _, candidate := range candidates {
-		if key := strings.ToLower(cleanPackagePath(candidate)); key != "" {
+		if key := cleanPackagePath(candidate); key != "" {
 			removed[key] = true
 		}
 	}
@@ -566,7 +569,7 @@ func pruneExternalFontFiles(ctx context.Context, reader *Reader, candidates ...s
 			return nil
 		}
 	}
-	for _, name := range reader.fileNamesFold {
+	for name := range reader.fileNames {
 		if err := ctx.Err(); err != nil {
 			return err
 		}
@@ -597,7 +600,7 @@ func pruneExternalFontFiles(ctx context.Context, reader *Reader, candidates ...s
 			}
 			if node.matchesOFD("FontFile") {
 				location := resolveResourcePath(name, base, strings.TrimSpace(editorImportText(data, node)))
-				delete(removed, strings.ToLower(cleanPackagePath(location)))
+				delete(removed, cleanPackagePath(location))
 			}
 			return slices.ContainsFunc(node.children, retains)
 		}
@@ -609,12 +612,12 @@ func pruneExternalFontFiles(ctx context.Context, reader *Reader, candidates ...s
 		return err
 	}
 	for name := range reader.files {
-		if removed[strings.ToLower(cleanPackagePath(name))] {
+		if removed[cleanPackagePath(name)] {
 			delete(reader.files, name)
 		}
 	}
 	if reader.Zip != nil {
-		reader.Zip = &zip.Reader{Comment: reader.Zip.Comment, File: slices.DeleteFunc(slices.Clone(reader.Zip.File), func(file *zip.File) bool { return removed[strings.ToLower(cleanPackagePath(file.Name))] })}
+		reader.Zip = &zip.Reader{Comment: reader.Zip.Comment, File: slices.DeleteFunc(slices.Clone(reader.Zip.File), func(file *zip.File) bool { return removed[cleanPackagePath(file.Name)] })}
 	}
 	return reader.initRoot()
 }

ofdgo_editor_font_scan.go

@@ -62,16 +62,17 @@ type editorFontDeclaration struct {
 // 返回: *editorFontDocument 字体引用, error 读取或结构错误
 func scanEditorFonts(ctx context.Context, reader *Reader, index int) (*editorFontDocument, error) {
 	scan := &editorFontDocument{reader: reader, index: index, files: make(map[string]*editorFontXML), fonts: make(map[string]*editorFontDeclaration)}
-	queue := []string{cleanPackagePath(reader.OFD.DocBody[index].DocRoot)}
+	entry, err := reader.documentEntry(index)
+	if err != nil {
+		return nil, err
+	}
+	queue := []string{entry}
 	for len(queue) != 0 {
 		if err := ctx.Err(); err != nil {
 			return nil, err
 		}
 		name := queue[0]
 		queue = queue[1:]
-		if actual := reader.fileNamesFold[strings.ToLower(name)]; actual != "" {
-			name = actual
-		}
 		if scan.files[name] != nil {
 			continue
 		}
@@ -148,9 +149,6 @@ func scanEditorFonts(ctx context.Context, reader *Reader, index int) (*editorFon
 			}
 			if value != "" {
 				target := resolveResourcePath(name, base, value)
-				if actual := reader.fileNamesFold[strings.ToLower(target)]; actual != "" {
-					target = actual
-				}
 				file.links = append(file.links, editorFontLink{node: node, attribute: attribute, target: target})
 				queue = append(queue, target)
 			}
@@ -172,6 +170,29 @@ func scanEditorFonts(ctx context.Context, reader *Reader, index int) (*editorFon
 // 入参: ctx 取消上下文, changes 按原文件路径保存的修改
 // 返回: error 引用扫描或XML修改错误
 func (s *editorFontDocument) apply(ctx context.Context, changes map[string][]byte) error {
+	if s.reader.OFD.DocBody[s.index].versioned {
+		view := *s.reader
+		if s.index != view.documentIndex {
+			view.selectedVersion = nil
+		}
+		view.documentIndex = s.index
+		view.documentRoot, view.versionInfo = "", nil
+		parts, err := view.versionChanges(ctx, changes)
+		if err != nil {
+			return err
+		}
+		next := *s.reader
+		next.files = maps.Clone(next.files)
+		if next.files == nil {
+			next.files = make(map[string][]byte)
+		}
+		maps.Copy(next.files, parts)
+		if err := next.initRoot(); err != nil {
+			return err
+		}
+		*s.reader = next
+		return nil
+	}
 	shared := make(map[string]bool)
 	if s.reader.DocumentCount() > 1 {
 		for index := range s.reader.DocumentCount() {
@@ -187,14 +208,14 @@ func (s *editorFontDocument) apply(ctx context.Context, changes map[string][]byt
 	}
 	paths := make(map[string]string)
 	for name := range changes {
-		if shared[strings.ToLower(name)] {
+		if shared[name] {
 			paths[name] = packageAvailableName(s.reader, changes, name+".font.xml")
 		}
 	}
 	for changed := true; changed; {
 		changed = false
 		for name, file := range s.files {
-			if paths[name] != "" || !shared[strings.ToLower(name)] {
+			if paths[name] != "" || !shared[name] {
 				continue
 			}
 			for _, link := range file.links {
@@ -236,7 +257,11 @@ func (s *editorFontDocument) apply(ctx context.Context, changes map[string][]byt
 					if err != nil {
 						return err
 					}
+					if node.open == node.end {
+						patches = append(patches, editorXMLPatch{node.start, node.end, fragment})
+					} else {
 						patches = append(patches, editorXMLPatch{node.start, node.open, fragment[:header.open]})
+					}
 					break
 				}
 				if link.attribute == "" && editorImportText(data, node) == editorImportText(file.data, link.node) {
@@ -261,9 +286,9 @@ func (s *editorFontDocument) apply(ctx context.Context, changes map[string][]byt
 			changes[name] = data
 		}
 	}
-	rootName := cleanPackagePath(s.reader.OFD.DocBody[s.index].DocRoot)
+	rootName, err := s.reader.documentEntry(s.index)
-	if actual := s.reader.fileNamesFold[strings.ToLower(rootName)]; actual != "" {
+	if err != nil {
-		rootName = actual
+		return err
 	}
 	if target := paths[rootName]; target != "" {
 		data, err := s.reader.readFile("OFD.xml")
@@ -275,7 +300,7 @@ func (s *editorFontDocument) apply(ctx context.Context, changes map[string][]byt
 			return err
 		}
 		node := root.childAt("DocBody", s.index).child("DocRoot")
-		changes[s.reader.fileNamesFold["ofd.xml"]] = editorPatchXML(data, []editorXMLPatch{editorXMLContent(data, node, editorFontXMLText("/"+target))})
+		changes["OFD.xml"] = editorPatchXML(data, []editorXMLPatch{editorXMLContent(data, node, editorFontXMLText("/"+target))})
 	}
 	for name, data := range changes {
 		if target := paths[name]; target != "" {

ofdgo_editor_import.go

@@ -16,6 +16,7 @@ package ofdgo
 
 import (
 	"bytes"
+	"context"
 	"encoding/xml"
 	"fmt"
 	"io"
@@ -229,6 +230,7 @@ func (e *Editor) ImportObjects(source *Editor, sourcePage int, ids []string, pag
 // ImportPages 将来源当前文档中的指定页面插入目标位置，保持输入顺序并作为一次撤销操作
 // 页面索引从0开始，at可等于当前页数；复制关联模板、注释、签章外观和实际引用的资源
 // 不导入来源元数据和目录，指向未选页面的跳转被移除；签名数据仅保留原始凭据，不代表合并后文档有效
+// 目标文档含版本列表时暂不支持导入带签章页面，返回错误且不修改目标
 // 返回后可关闭来源Reader，目标原Reader的生命周期要求不变
 // 入参: source 来源阅读器, indexes 来源页面索引，不可重复, at 目标插入位置
 // 返回: []string 新页面标识, error 错误信息
@@ -455,7 +457,7 @@ func (e *Editor) prepareImport(source *Reader, doc *Document, selected map[strin
 	for _, resource := range e.resources {
 		m.files[resource.name] = nil
 	}
-	m.documentXML, err = m.readFile(source.ResPath(source.OFD.DocBody[source.documentIndex].DocRoot))
+	m.documentXML, err = m.readFile(source.documentRoot)
 	if err != nil {
 		return nil, nil, err
 	}
@@ -535,7 +537,7 @@ func (m *editorPageImport) readFile(name string) ([]byte, error) {
 // outlines 筛选指向导入页面的目录树并复用标准动作迁移
 // 返回: []byte 目录片段, error 错误信息
 func (m *editorPageImport) outlines() ([]byte, error) {
-	name := m.reader.ResPath(m.reader.OFD.DocBody[m.reader.documentIndex].DocRoot)
+	name := m.reader.documentRoot
 	data, err := m.readFile(name)
 	if err != nil {
 		return nil, err
@@ -838,7 +840,7 @@ func (m *editorPageImport) copyData(name, relative string, evidence bool) (strin
 		paths = m.evidence
 		relative = path.Join("Signs", "Signed", path.Base(name))
 	}
-	key := strings.ToLower(cleanPackagePath(name))
+	key := cleanPackagePath(name)
 	if target := paths[key]; target != "" {
 		return "/" + target, nil
 	}
@@ -858,7 +860,7 @@ func (m *editorPageImport) copyData(name, relative string, evidence bool) (strin
 // 入参: name 来源路径, candidate 目标候选路径
 // 返回: string 新路径, error 错误信息
 func (m *editorPageImport) copyXML(name, candidate string) (string, error) {
-	key := strings.ToLower(cleanPackagePath(name))
+	key := cleanPackagePath(name)
 	if target := m.paths[key]; target != "" {
 		return "/" + target, nil
 	}
@@ -1188,7 +1190,7 @@ func (m *editorPageImport) attachment(id string) (string, error) {
 	if _, exists := m.attachments[id]; exists {
 		return m.attachmentID(id)
 	}
-	name := m.reader.ResPath(m.reader.OFD.DocBody[m.reader.documentIndex].DocRoot)
+	name := m.reader.documentRoot
 	if m.doc.Attachments.Path != "" {
 		name = m.reader.ResPath(m.doc.Attachments.Path)
 	}
@@ -1350,7 +1352,7 @@ func (m *editorPageImport) merge(base *editorSource, refs []Page, annotations, s
 		files = make(map[string][]byte)
 	}
 	maps.Copy(files, m.files)
-	name := base.reader.ResPath(base.reader.OFD.DocBody[base.reader.documentIndex].DocRoot)
+	name := base.reader.documentRoot
 	if file, ok := base.reader.packageFile(name); ok {
 		name = cleanPackagePath(file.Name)
 	}
@@ -1454,6 +1456,9 @@ func (m *editorPageImport) merge(base *editorSource, refs []Page, annotations, s
 	}
 	files[name] = data
 	if len(signatures) != 0 {
+		if base.reader.OFD.DocBody[base.reader.documentIndex].versioned {
+			return nil, fmt.Errorf("signature import into a versioned document is not supported")
+		}
 		loc, err := mergeImportIndex(base.reader, files, base.document.Signatures, path.Join(m.directory, "Signs", "Signatures.xml"), "Signatures", signatures)
 		if err != nil {
 			return nil, err
@@ -1490,7 +1495,16 @@ func (m *editorPageImport) merge(base *editorSource, refs []Page, annotations, s
 			}
 		}
 	}
-	reader := &Reader{Zip: base.reader.Zip, files: files, encryption: base.reader.encryption, documentIndex: base.reader.documentIndex}
+	parts, err := base.reader.versionChanges(context.Background(), files)
+	if err != nil {
+		return nil, err
+	}
+	files = maps.Clone(base.reader.files)
+	if files == nil {
+		files = make(map[string][]byte)
+	}
+	maps.Copy(files, parts)
+	reader := &Reader{Zip: base.reader.Zip, files: files, encryption: base.reader.encryption, documentIndex: base.reader.documentIndex, selectedVersion: base.reader.selectedVersion}
 	if err := reader.initRoot(); err != nil {
 		return nil, err
 	}

ofdgo_editor_open.go

@@ -113,6 +113,7 @@ type PageCapabilities struct {
 // Editor 打开当前文档进行保真编辑，不修改输入包，不以权限声明或签名限制编辑
 // 原权限和签名数据保留，修改受保护内容会使原签名失效，不重新签名
 // 页面按需解析，未修改条目直接保留；编辑器及其Reader快照使用期间不得关闭输入Reader
+// 版本共享文件在提交修改时隔离，保留其他版本及其文件清单
 // 返回: *Editor 编辑器, error 错误信息
 func (r *Reader) Editor() (*Editor, error) {
 	_, err := r.Doc()
@@ -419,7 +420,7 @@ func (e *Editor) repairSourceObjectIDs(data []byte, root *editorXML) ([]byte, bo
 // 返回: error 引用歧义或读取错误
 func (e *Editor) checkDuplicateObjectReferences(duplicates map[string]bool) error {
 	reader := e.source.reader
-	for _, name := range reader.fileNamesFold {
+	for name := range reader.fileNames {
 		if !strings.EqualFold(path.Ext(name), ".xml") {
 			continue
 		}

ofdgo_editor_outline.go

@@ -319,7 +319,7 @@ func (e *Editor) outlineXML() ([]byte, error) {
 	}
 	if e.source != nil {
 		reader := e.source.reader
-		data, err := reader.readFile(reader.ResPath(reader.OFD.DocBody[reader.documentIndex].DocRoot))
+		data, err := reader.readFile(reader.documentRoot)
 		if err != nil {
 			return nil, err
 		}

ofdgo_editor_package.go

@@ -17,6 +17,7 @@ package ofdgo
 import (
 	"archive/zip"
 	"bytes"
+	"context"
 	"encoding/xml"
 	"fmt"
 	"io"
@@ -31,7 +32,11 @@ import (
 // 入参: progress 保存进度回调，预览时为nil
 // 返回: map[string][]byte 替换及新增条目, error 错误信息
 func (e *Editor) sourceParts(progress editorProgress) (map[string][]byte, error) {
-	return e.sourcePartsReferences(progress, nil)
+	parts, err := e.sourcePartsReferences(progress, nil)
+	if err != nil {
+		return nil, err
+	}
+	return e.source.reader.versionChanges(context.Background(), parts)
 }
 
 // sourcePartsReferences 生成保存快照，同步收集完全自产页面的实际编码引用
@@ -49,6 +54,7 @@ func (e *Editor) sourcePartsPrepared(progress editorProgress, generated map[stri
 	var stager editorPageStager
 	source := e.source
 	reader := source.reader
+	stage = stage && !reader.OFD.DocBody[reader.documentIndex].versioned
 	pageRefs := make([]Page, len(e.pages))
 	originalCount := 0
 	for i, page := range e.pages {
@@ -111,7 +117,7 @@ func (e *Editor) sourcePartsPrepared(progress editorProgress, generated map[stri
 	commonData := source.document.CommonData
 	for _, name := range append(slices.Clone(commonData.DocumentRes), commonData.PublicRes...) {
 		resourceFiles = slices.DeleteFunc(resourceFiles, func(file string) bool {
-			return strings.EqualFold(file, reader.ResPath(name))
+			return file == reader.ResPath(name)
 		})
 	}
 	if len(fonts)+len(images)+len(spaces) != 0 {
@@ -134,7 +140,7 @@ func (e *Editor) sourcePartsPrepared(progress editorProgress, generated map[stri
 	}
 	pagesChanged := !slices.Equal(pageRefs, source.document.Pages.Page)
 	if pagesChanged || len(resourceFiles) != 0 || len(parts) != 0 && e.maxID != source.document.CommonData.MaxUnitID {
-		name := reader.ResPath(reader.OFD.DocBody[reader.documentIndex].DocRoot)
+		name := reader.documentRoot
 		data, err := reader.readFile(name)
 		if err != nil {
 			return nil, err
@@ -232,7 +238,7 @@ func (e *Editor) sourcePartsPrepared(progress editorProgress, generated map[stri
 		parts["OFD.xml"] = data
 	}
 	if e.outlines != nil {
-		name := reader.ResPath(reader.OFD.DocBody[reader.documentIndex].DocRoot)
+		name := reader.documentRoot
 		data, ok := parts[name]
 		if !ok {
 			var err error
@@ -586,6 +592,14 @@ func (e *Editor) writeSource(writer io.Writer, fonts map[string][]byte, progress
 	if err := e.compressResourceReferences(parts, reader, removed, generated); err != nil {
 		return 0, err
 	}
+	ctx := context.Background()
+	if e.output != nil {
+		ctx = e.output.ctx
+	}
+	parts, err = reader.versionChanges(ctx, parts)
+	if err != nil {
+		return 0, err
+	}
 	if len(parts) != 0 || len(removed) != 0 || e.output != nil && e.output.options.Mode != CompressionUnchanged && !e.output.protected {
 		name := "OFD.xml"
 		if file, ok := reader.packageFile(name); ok {
@@ -610,7 +624,7 @@ func (e *Editor) writeSource(writer io.Writer, fonts map[string][]byte, progress
 	}
 	maps.Copy(remaining, parts)
 	for name := range remaining {
-		if removed[strings.ToLower(cleanPackagePath(name))] {
+		if removed[cleanPackagePath(name)] {
 			delete(remaining, name)
 		}
 	}
@@ -631,7 +645,7 @@ func (e *Editor) writeSource(writer io.Writer, fonts map[string][]byte, progress
 				continue
 			}
 			name := cleanPackagePath(file.Name)
-			if removed[strings.ToLower(name)] {
+			if removed[name] {
 				continue
 			}
 			if data, ok := remaining[name]; ok {
@@ -710,8 +724,7 @@ func (e *Editor) sourceReader(progress editorProgress) (*Reader, error) {
 		maps.Copy(reader.files, parts)
 		reader.encryption = e.encryption
 		for name := range reader.pageHeaderCache {
-			actual := reader.fileNamesFold[strings.ToLower(name)]
+			if _, changed := parts[name]; changed {
-			if _, changed := parts[actual]; changed {
 				delete(reader.pageHeaderCache, name)
 			}
 		}
@@ -725,7 +738,7 @@ func (e *Editor) sourceReader(progress editorProgress) (*Reader, error) {
 		files = make(map[string][]byte)
 	}
 	maps.Copy(files, parts)
-	reader := &Reader{Zip: e.source.reader.Zip, files: files, encryption: e.encryption, documentIndex: e.source.reader.documentIndex}
+	reader := &Reader{Zip: e.source.reader.Zip, files: files, encryption: e.encryption, documentIndex: e.source.reader.documentIndex, selectedVersion: e.source.reader.selectedVersion}
 	if err := reader.initRoot(); err != nil {
 		return nil, err
 	}
@@ -733,8 +746,7 @@ func (e *Editor) sourceReader(progress editorProgress) (*Reader, error) {
 		return nil, err
 	}
 	for name, header := range e.source.reader.pageHeaderCache {
-		actual := e.source.reader.fileNamesFold[strings.ToLower(name)]
+		if _, changed := parts[name]; !changed {
-		if _, changed := parts[actual]; !changed {
 			reader.pageHeaderCache[name] = header
 		}
 	}

ofdgo_editor_pages.go

@@ -23,6 +23,7 @@ import (
 
 // prunePageReferences 在输出快照中删除已移除页面的标准引用，不修改历史快照或其他文档
 // 沿当前文档、页面、模板、资源、注释和签名索引检查，不进入附件、私有扩展和签名原始凭据
+// 版本包保留原签名及其页面引用，避免改写其他版本共用的签名数据
 // 入参: parts 输出覆盖条目, generated 本次自产页面及引用，可为nil
 // 返回: error 读取或解析错误
 func (e *Editor) prunePageReferences(parts map[string][]byte, generated map[string]editorGeneratedReferences) error {
@@ -50,8 +51,8 @@ func (e *Editor) prunePageReferences(parts map[string][]byte, generated map[stri
 			bookmarks[bookmark.Name] = true
 		}
 	}
-	queue := []string{reader.ResPath(reader.OFD.DocBody[reader.documentIndex].DocRoot)}
+	queue := []string{reader.documentRoot}
-	if name := e.source.document.Signatures; name != "" {
+	if name := e.source.document.Signatures; name != "" && !reader.OFD.DocBody[reader.documentIndex].versioned {
 		queue = append(queue, reader.ResPath(name))
 	}
 	seen := make(map[string]bool)
@@ -63,10 +64,10 @@ func (e *Editor) prunePageReferences(parts map[string][]byte, generated map[stri
 		}
 		name := queue[0]
 		queue = queue[1:]
-		if seen[strings.ToLower(name)] {
+		if seen[name] {
 			continue
 		}
-		seen[strings.ToLower(name)] = true
+		seen[name] = true
 		data, ok := parts[name]
 		if !ok {
 			var err error

ofdgo_editor_reference.go

@@ -164,7 +164,11 @@ func (s *editorReferenceScan) accept(token xml.Token) bool {
 				s.refs.reference(s.name, s.base, attr.Name.Local, attr.Value, s.resource)
 			}
 			if s.objects != nil && editorObjectReference(attr.Name.Local) {
-				s.objects[attr.Value] = true
+				id := attr.Value
+				if key := editorResourceID(id); key != "" {
+					id = key
+				}
+				s.objects[id] = true
 			}
 			if attr.Name.Local == "Font" {
 				usage = s.refs.fontUsage(attr.Value)
@@ -189,6 +193,14 @@ func (s *editorReferenceScan) accept(token xml.Token) bool {
 			return false
 		}
 		s.refs.reference(s.name, s.base, token.Name.Local, s.text.String(), s.resource)
+		if len(s.stack) == 3 && s.stack[0] == "DocVersion" && s.stack[1] == "FileList" && token.Name.Local == "File" {
+			if name := editorResourceLocation(s.name, "", s.text.String()); name != "" {
+				if s.refs.versions == nil {
+					s.refs.versions = make(map[string]bool)
+				}
+				s.refs.versions[name] = true
+			}
+		}
 		if s.objects != nil && (token.Name.Local == "Thumbnail" || token.Name.Local == "Substitution") {
 			if id := editorResourceID(s.text.String()); id != "" {
 				s.objects[id] = true
@@ -219,6 +231,12 @@ func (s *editorReferenceScan) accept(token xml.Token) bool {
 func (r *editorResourceRefs) merge(other *editorResourceRefs) {
 	maps.Copy(r.ids, other.ids)
 	maps.Copy(r.files, other.files)
+	if len(other.versions) != 0 {
+		if r.versions == nil {
+			r.versions = make(map[string]bool)
+		}
+		maps.Copy(r.versions, other.versions)
+	}
 	for id, usage := range other.fonts {
 		r.fontUsage(id).merge(usage, false)
 	}

ofdgo_editor_resources.go

@@ -39,6 +39,7 @@ type editorResourceRefs struct {
 	ids      map[string]bool
 	files    map[string]bool
 	fonts    map[string]*editorFontUsage
+	versions map[string]bool
 }
 
 // compactSourceReferences 合并自产页面的已知引用，其他条目仍完整扫描
@@ -135,6 +136,7 @@ func (e *Editor) compactSourceReferences(parts map[string][]byte, progress edito
 		if err != nil {
 			return nil, nil
 		}
+		preserved := refs.versions[cleanPackagePath(name)]
 		var patches []editorXMLPatch
 		for _, group := range root.children {
 			var members []editorXMLPatch
@@ -149,18 +151,18 @@ func (e *Editor) compactSourceReferences(parts map[string][]byte, progress edito
 					continue
 				}
 				id := editorResourceID(node.attr("ID"))
-				keep := id == "" || refs.ids[id] || field == "MediaFile" && node.attr("Type") != "Image"
+				keep := preserved || id == "" || refs.ids[id] || field == "MediaFile" && node.attr("Type") != "Image"
 				if file := node.child(field); file != nil {
 					var value string
 					if err := xml.Unmarshal(data[file.start:file.end], &value); err != nil {
 						return nil, nil
 					}
 					if location := editorResourceLocation(name, root.attr("BaseLoc"), value); location != "" {
-						alternate := strings.ToLower(cleanPackagePath(resolveResourcePath(name, root.attr("BaseLoc"), value)))
+						alternate := cleanPackagePath(resolveResourcePath(name, root.attr("BaseLoc"), value))
 						if field == "FontFile" {
 							usage := refs.fonts[id]
 							charset := node.attr("Charset")
-							unsafe := id == "" || charset != "" && !strings.EqualFold(charset, "unicode") || location != alternate
+							unsafe := preserved || id == "" || charset != "" && !strings.EqualFold(charset, "unicode") || location != alternate
 							for _, file := range []string{location, alternate} {
 								if fontFiles[file] == nil {
 									fontFiles[file] = newEditorFontUsage()
@@ -200,7 +202,7 @@ func (e *Editor) compactSourceReferences(parts map[string][]byte, progress edito
 		if err := progress.report("fonts", i, len(ordered)); err != nil {
 			return nil, err
 		}
-		key := strings.ToLower(cleanPackagePath(name))
+		key := cleanPackagePath(name)
 		usage := fontFiles[key]
 		if e.backends.FontResources == nil || usage == nil || usage.unsafe || refs.files[key] || removed[key] || parts[name] != nil {
 			continue
@@ -222,7 +224,7 @@ func (e *Editor) compactSourceReferences(parts map[string][]byte, progress edito
 	}
 	maps.Copy(parts, updates)
 	for name := range parts {
-		if removed[strings.ToLower(cleanPackagePath(name))] {
+		if removed[cleanPackagePath(name)] {
 			delete(parts, name)
 		}
 	}
@@ -279,7 +281,7 @@ func (r *editorResourceRefs) reference(name, base, key, value string, resource b
 		return
 	}
 	r.files[editorResourceLocation(name, base, value)] = true
-	r.files[strings.ToLower(cleanPackagePath(resolveResourcePath(name, base, value)))] = true
+	r.files[cleanPackagePath(resolveResourcePath(name, base, value))] = true
 }
 
 // editorResourceReferenceKind 区分标准标识和文件引用，其他文本不参与资源扫描
@@ -340,7 +342,7 @@ func editorResourceNode(root *editorXML, group, kind, id string) *editorXML {
 	return node
 }
 
-// editorResourceLocation 按ST_Loc与Res.BaseLoc解析路径，大小写折叠用于保守保留共享文件
+// editorResourceLocation 按ST_Loc与Res.BaseLoc解析路径，保留路径大小写
 // 入参: name 所在XML, base 资源目录, value 路径值
 // 返回: string 包内路径
 func editorResourceLocation(name, base, value string) string {
@@ -349,7 +351,7 @@ func editorResourceLocation(name, base, value string) string {
 		return ""
 	}
 	if strings.HasPrefix(value, "/") {
-		return strings.ToLower(cleanPackagePath(value))
+		return cleanPackagePath(value)
 	}
 	base = strings.ReplaceAll(base, "\\", "/")
 	dir := path.Dir(name)
@@ -358,5 +360,5 @@ func editorResourceLocation(name, base, value string) string {
 	} else {
 		dir = path.Join(dir, base)
 	}
-	return strings.ToLower(cleanPackagePath(path.Join(dir, value)))
+	return cleanPackagePath(path.Join(dir, value))
 }

ofdgo_format.go

@@ -49,7 +49,7 @@ func DetectFormat(source io.ReaderAt, size int64) (string, error) {
 		if err := reader.indexPackage(); err != nil {
 			return "", err
 		}
-		if reader.fileIndexFold["ofd.xml"] != nil || reader.fileIndexFold["encryptions.xml"] != nil {
+		if reader.fileNames["OFD.xml"] || reader.fileNames["Encryptions.xml"] {
 			return "ofd", nil
 		}
 	} else {

ofdgo_meta.go

@@ -34,6 +34,24 @@ type DocBody struct {
 	DocInfo    DocInfo `xml:"DocInfo"`
 	DocRoot    string  `xml:"DocRoot"`
 	Signatures string  `xml:"Signatures"`
+	versioned  bool
+}
+
+// UnmarshalXML 读取文档入口并记录版本列表是否存在，版本描述在选择时读取
+// 入参: decoder XML解码器, start 起始节点
+// 返回: error 解析错误
+func (body *DocBody) UnmarshalXML(decoder *xml.Decoder, start xml.StartElement) error {
+	type docBody DocBody
+	var value struct {
+		docBody
+		Versions *struct{} `xml:"Versions"`
+	}
+	if err := decoder.DecodeElement(&value, &start); err != nil {
+		return err
+	}
+	*body = DocBody(value.docBody)
+	body.versioned = value.Versions != nil
+	return nil
 }
 
 // DocInfo 文档元数据

ofdgo_package.go

@@ -53,7 +53,7 @@ func (e *Editor) packageName(relative string) string {
 	return packageAvailableName(reader, files, path.Join(e.packageDirectory(), relative))
 }
 
-// packageAvailableName 仅在包内同名条目冲突时添加序号，忽略大小写差异
+// packageAvailableName 仅在包内同名条目冲突时添加序号，路径区分大小写
 // 入参: reader 原包，可为nil, files 待写入条目, name 期望路径
 // 返回: string 未占用路径
 func packageAvailableName(reader *Reader, files map[string][]byte, name string) string {
@@ -66,13 +66,11 @@ func packageAvailableName(reader *Reader, files map[string][]byte, name string)
 		occupied := false
 		if reader != nil {
 			_, occupied = reader.packageFile(candidate)
-			for file := range reader.files {
+			_, cached := reader.files[candidate]
-				occupied = occupied || strings.EqualFold(cleanPackagePath(file), candidate)
+			occupied = occupied || cached
-			}
-		}
-		for file := range files {
-			occupied = occupied || strings.EqualFold(cleanPackagePath(file), candidate)
 		}
+		_, pending := files[candidate]
+		occupied = occupied || pending
 		if !occupied {
 			return candidate
 		}

ofdgo_reader.go

@@ -48,13 +48,15 @@ type Reader struct {
 	doc                       *Document
 	relatedPending            bool
 	documentIndex             int
+	documentRoot              string
+	selectedVersion           *string
+	versionInfo               *DocumentVersionInfo
 	Stamps                    map[string][]Stamp
 	Annots                    map[string][]Annotation
 	annotationFiles           map[string][]string
 	fileIndex                 map[string]*zip.File
-	fileIndexFold             map[string]*zip.File
 	files                     map[string][]byte
-	fileNamesFold             map[string]string
+	fileNames                 map[string]bool
 	encryption                *encryptionState
 }
 
@@ -94,7 +96,11 @@ func (r *Reader) DocumentPageCount(index int) (int, error) {
 	if index == r.documentIndex && r.doc != nil {
 		return len(r.doc.Pages.Page), nil
 	}
-	data, err := r.readFile(r.OFD.DocBody[index].DocRoot)
+	root, err := r.documentEntry(index)
+	if err != nil {
+		return 0, err
+	}
+	data, err := r.readFile(root)
 	if err != nil {
 		return 0, err
 	}
@@ -151,7 +157,10 @@ func (r *Reader) docStructure() (*Document, error) {
 		return nil, fmt.Errorf("no docbody found")
 	}
 	docAttr := r.OFD.DocBody[r.documentIndex]
-	docRootPath := docAttr.DocRoot
+	docRootPath, err := r.DocumentRoot()
+	if err != nil {
+		return nil, err
+	}
 	r.RootDir = path.Dir(docRootPath)
 	data, err := r.readFile(docRootPath)
 	if err != nil {
@@ -162,7 +171,7 @@ func (r *Reader) docStructure() (*Document, error) {
 		return nil, fmt.Errorf("failed to unmarshal document.xml: %w", err)
 	}
 	if doc.Signatures == "" {
-		doc.Signatures = docAttr.Signatures
+		doc.Signatures = r.versionSignatures(docAttr.Signatures)
 	}
 	for _, res := range doc.CommonData.DocumentRes {
 		r.loadRes(res)
@@ -459,6 +468,11 @@ func (r *Reader) initRoot() error {
 		return fmt.Errorf("failed to unmarshal ofd.xml: %w", err)
 	}
 	r.OFD = &ofd
+	r.doc, r.versionInfo = nil, nil
+	r.documentRoot, r.RootDir = "", ""
+	r.relatedPending, r.fontResourcesRead = false, false
+	r.imageCatalog = nil
+	r.Stamps, r.Annots, r.annotationFiles = nil, nil, nil
 	r.ResMap = make(map[string]string)
 	r.resourcesRead = make(map[string]bool)
 	r.mediaCache = make(map[string]MultiMedia)
@@ -472,7 +486,7 @@ func (r *Reader) initRoot() error {
 	return nil
 }
 
-// indexPackage 建立包路径索引，拒绝会使显示与验签产生歧义的同名条目
+// indexPackage 按大小写敏感路径建立包索引，拒绝规范化后重复的ZIP条目
 // 返回: error 错误信息
 func (r *Reader) indexPackage() error {
 	count := 0
@@ -480,8 +494,7 @@ func (r *Reader) indexPackage() error {
 		count = len(r.Zip.File)
 	}
 	r.fileIndex = make(map[string]*zip.File, count)
-	r.fileIndexFold = make(map[string]*zip.File, count)
+	r.fileNames = make(map[string]bool, count+len(r.files))
-	r.fileNamesFold = make(map[string]string, count+len(r.files))
 	if r.Zip != nil {
 		for _, f := range r.Zip.File {
 			if f.FileInfo().IsDir() {
@@ -491,24 +504,18 @@ func (r *Reader) indexPackage() error {
 			if !validPackagePath(name) {
 				return fmt.Errorf("invalid package path: %q", f.Name)
 			}
-			fold := strings.ToLower(name)
+			if previous, ok := r.fileIndex[name]; ok {
-			if previous, ok := r.fileIndexFold[fold]; ok {
 				return fmt.Errorf("ambiguous package paths: %q and %q", previous.Name, f.Name)
 			}
 			r.fileIndex[name] = f
-			r.fileIndexFold[fold] = f
+			r.fileNames[name] = true
-			r.fileNamesFold[fold] = name
 		}
 	}
 	for name := range r.files {
 		if name != cleanPackagePath(name) || !validPackagePath(name) {
 			return fmt.Errorf("invalid package path: %q", name)
 		}
-		fold := strings.ToLower(name)
+		r.fileNames[name] = true
-		if previous, ok := r.fileNamesFold[fold]; ok && previous != name {
-			return fmt.Errorf("ambiguous package paths: %q and %q", previous, name)
-		}
-		r.fileNamesFold[fold] = name
 	}
 	return nil
 }
@@ -544,16 +551,10 @@ func (r *Reader) readFileView(name string) ([]byte, error) {
 // 返回: []byte 文件内容, bool 是否借用缓存, error 错误信息
 func (r *Reader) readFileData(name string) ([]byte, bool, error) {
 	name = cleanPackagePath(name)
-	if actual, ok := r.fileNamesFold[strings.ToLower(name)]; ok {
-		name = actual
-	}
 	if data, ok := r.files[name]; ok {
 		return data, true, nil
 	}
 	if f, ok := r.packageFile(name); ok {
-		if data, ok := r.files[cleanPackagePath(f.Name)]; ok {
-			return data, true, nil
-		}
 		data, err := readZipFile(f)
 		return data, false, err
 	}
@@ -565,16 +566,10 @@ func (r *Reader) readFileData(name string) ([]byte, bool, error) {
 // 返回: io.ReadCloser 文件流, error 错误信息
 func (r *Reader) openFile(name string) (io.ReadCloser, error) {
 	name = cleanPackagePath(name)
-	if actual, ok := r.fileNamesFold[strings.ToLower(name)]; ok {
-		name = actual
-	}
 	if data, ok := r.files[name]; ok {
 		return io.NopCloser(bytes.NewReader(data)), nil
 	}
 	if f, ok := r.packageFile(name); ok {
-		if data, ok := r.files[cleanPackagePath(f.Name)]; ok {
-			return io.NopCloser(bytes.NewReader(data)), nil
-		}
 		return f.Open()
 	}
 	return nil, fmt.Errorf("file not found: %s", name)
@@ -593,13 +588,8 @@ func cleanPackagePath(name string) string {
 // 入参: name 文件路径
 // 返回: *zip.File 压缩包文件, bool 是否存在
 func (r *Reader) packageFile(name string) (*zip.File, bool) {
-	if f, ok := r.fileIndex[name]; ok {
+	f, ok := r.fileIndex[cleanPackagePath(name)]
-		return f, true
+	return f, ok
-	}
-	if f, ok := r.fileIndexFold[strings.ToLower(name)]; ok {
-		return f, true
-	}
-	return nil, false
 }
 
 // readZipFile 读取zip文件内容

ofdgo_reader_dependencies.go

@@ -26,7 +26,7 @@ import (
 
 // documentFiles 按标准路径遍历独立文档依赖，签名摘要引用不作为内容依赖
 // 入参: ctx 取消上下文, index 文档索引
-// 返回: map[string]bool 大小写折叠的文件路径, error 读取或取消错误
+// 返回: map[string]bool 包内文件路径, error 读取或取消错误
 func (r *Reader) documentFiles(ctx context.Context, index int) (map[string]bool, error) {
 	if err := ctx.Err(); err != nil {
 		return nil, err
@@ -67,15 +67,13 @@ func (r *Reader) documentFiles(ctx context.Context, index int) (map[string]bool,
 		if err := ctx.Err(); err != nil {
 			return nil, err
 		}
-		name := queue[0]
+		actual := queue[0]
 		queue = queue[1:]
-		key := strings.ToLower(name)
+		if seen[actual] {
-		if seen[key] {
 			continue
 		}
-		seen[key] = true
+		seen[actual] = true
-		actual, exists := r.fileNamesFold[key]
+		if !r.fileNames[actual] {
-		if !exists {
 			continue
 		}
 		input, err := r.openFile(actual)

ofdgo_sign_policy.go

@@ -219,7 +219,7 @@ func (r *Reader) applySignatureCoverage(report *SignatureVerifyReport, listPath
 	covered := make(map[string]bool)
 	for _, ref := range report.References {
 		if ref.Checked && ref.OK {
-			covered[r.signatureCoveragePath(ref.Path)] = true
+			covered[cleanPackagePath(ref.Path)] = true
 		}
 	}
 	for name := range covered {
@@ -232,12 +232,12 @@ func (r *Reader) applySignatureCoverage(report *SignatureVerifyReport, listPath
 	report.CoverageChecked, report.CoverageOK = true, true
 	required := make(map[string]bool)
 	for _, name := range o.Policy.RequiredFiles {
-		required[r.signatureCoveragePath(name)] = true
+		required[cleanPackagePath(name)] = true
 	}
 	if o.Policy.RequireDocumentCoverage {
-		required[r.signatureCoveragePath("OFD.xml")] = true
+		required["OFD.xml"] = true
-		excluded := map[string]bool{r.signatureCoveragePath(listPath): true}
+		excluded := map[string]bool{cleanPackagePath(listPath): true}
-		name := r.signatureCoveragePath(signatureRefPath(listPath, report.BaseLoc))
+		name := cleanPackagePath(signatureRefPath(listPath, report.BaseLoc))
 		data, err := r.readFile(name)
 		if err != nil {
 			report.CoverageOK = false
@@ -252,9 +252,9 @@ func (r *Reader) applySignatureCoverage(report *SignatureVerifyReport, listPath
 		}
 		excluded[name] = true
 		if file.SignedValue != "" {
-			excluded[r.signatureCoveragePath(signatureRefPath(name, file.SignedValue))] = true
+			excluded[cleanPackagePath(signatureRefPath(name, file.SignedValue))] = true
 		}
-		root := path.Dir(r.signatureCoveragePath(o.DocRoot))
+		root := path.Dir(cleanPackagePath(o.DocRoot))
 		include := func(name string) {
 			if (root == "." || strings.HasPrefix(name, root+"/")) && !excluded[name] && !strings.HasSuffix(name, "/") {
 				required[name] = true
@@ -281,22 +281,11 @@ func (r *Reader) applySignatureCoverage(report *SignatureVerifyReport, listPath
 	report.CoverageOK = report.CoverageOK && len(report.UncoveredFiles) == 0
 }
 
-// signatureCoveragePath 复用阅读器大小写索引获取同一文件身份
-// 入参: name 包内路径
-// 返回: string 实际包内路径
-func (r *Reader) signatureCoveragePath(name string) string {
-	name = cleanPackagePath(name)
-	if actual, ok := r.fileNamesFold[strings.ToLower(name)]; ok {
-		return actual
-	}
-	return name
-}
-
 // signatureCoverageReferences 收集标准OFD引用的传递闭包且包含跨文档资源
 // 入参: root 文档入口, required 待补充的必需文件集合
 // 返回: error 引用文件无法读取或解析时的错误
 func (r *Reader) signatureCoverageReferences(root string, required map[string]bool) error {
-	queue := []string{r.signatureCoveragePath(root)}
+	queue := []string{cleanPackagePath(root)}
 	seen := make(map[string]bool)
 	for len(queue) != 0 {
 		name := queue[0]
@@ -332,7 +321,7 @@ func (r *Reader) signatureCoverageReferences(root string, required map[string]bo
 			if resource {
 				resBase = base
 			}
-			resolved := r.signatureCoveragePath(resolveResourcePath(name, resBase, loc))
+			resolved := cleanPackagePath(resolveResourcePath(name, resBase, loc))
 			required[resolved] = true
 			if !seen[resolved] {
 				queue = append(queue, resolved)

ofdgo_sign_verify.go

@@ -299,11 +299,22 @@ func (r *Reader) VerifyDocumentSignatures(index int, opts ...SignatureVerifyOpti
 		}
 	}
 	body := r.OFD.DocBody[index]
-	options.DocIndex, options.DocRoot = index, r.signatureCoveragePath(body.DocRoot)
 	if index == r.documentIndex && r.doc != nil && r.doc.Signatures == "" {
 		return nil, nil
 	}
-	data, err := r.readFile(body.DocRoot)
+	view := &Reader{Zip: r.Zip, files: r.files, encryption: r.encryption, documentIndex: index}
+	if index == r.documentIndex {
+		view.selectedVersion = r.selectedVersion
+	}
+	if err := view.initRoot(); err != nil {
+		return nil, err
+	}
+	docRoot, err := view.DocumentRoot()
+	if err != nil {
+		return nil, err
+	}
+	options.DocIndex, options.DocRoot = index, docRoot
+	data, err := view.readFile(docRoot)
 	if err != nil {
 		return nil, err
 	}
@@ -312,16 +323,12 @@ func (r *Reader) VerifyDocumentSignatures(index int, opts ...SignatureVerifyOpti
 		return nil, err
 	}
 	if doc.Signatures == "" {
-		doc.Signatures = body.Signatures
+		doc.Signatures = view.versionSignatures(body.Signatures)
 	}
 	if doc.Signatures == "" {
 		return nil, nil
 	}
-	view := &Reader{Zip: r.Zip, files: r.files, encryption: r.encryption, documentIndex: index}
+	view.doc, view.RootDir = &doc, path.Dir(docRoot)
-	if err := view.initRoot(); err != nil {
-		return nil, err
-	}
-	view.doc, view.RootDir = &doc, path.Dir(body.DocRoot)
 	sigListPath := view.ResPath(doc.Signatures)
 	data, err = view.readFile(sigListPath)
 	if err != nil {

ofdgo_sign_write.go

@@ -78,6 +78,7 @@ type signatureWriteInfo struct {
 }
 
 // SignPackage 签署最终OFD包，失败时返回nil且不修改输入
+// 含版本列表的文档需先通过ExtractVersion提取后签署
 // 入参: data 最终包字节, options 签署选项
 // 返回: []byte 签署后的完整包, error 错误信息
 func SignPackage(data []byte, options SignatureWriteOptions) ([]byte, error) {
@@ -89,6 +90,7 @@ func SignPackage(data []byte, options SignatureWriteOptions) ([]byte, error) {
 }
 
 // SignEditorTo 将编辑器序列化为最终文件后签署，保留或显式覆盖加密策略
+// 含版本列表的文档需先通过ExtractVersion提取后签署
 // 不修改编辑器状态；OnWriteProgress可在准备、签署和加密阶段取消，准备失败不写入目标
 // 输出目标不得覆盖源文件；目标自身写入失败时调用方应丢弃部分输出
 // 入参: writer 输出目标, editor 编辑器, options 签署选项
@@ -121,6 +123,7 @@ func SignEditorTo(writer io.Writer, editor *Editor, options SignatureWriteOption
 }
 
 // SignTo 从阅读器包内容创建独立快照，完成签署和回验后写出
+// 含版本列表的文档需先通过ExtractVersion提取后签署
 // 不采用可变的Doc/OFD模型；调用方应先将Editor写出为最终包
 // 准备失败不写入writer；writer自身失败可能已写入部分输出，不保证目标原子替换
 // 入参: writer 独立的输出目标，不得覆盖源文件, options 签署选项
@@ -147,6 +150,9 @@ func signatureWriteOutput(r *Reader, options SignatureWriteOptions) ([]byte, err
 	if r == nil {
 		return nil, fmt.Errorf("nil signature reader")
 	}
+	if r.documentIndex >= 0 && r.documentIndex < r.DocumentCount() && r.OFD.DocBody[r.documentIndex].versioned {
+		return nil, fmt.Errorf("in-package version signing is not supported; extract the version before signing")
+	}
 	encryption := options.Encryption
 	if encryption == nil && r.encryption != nil {
 		encryption = r.encryption.options
@@ -275,7 +281,7 @@ func signatureWritePackage(parts map[string][]byte, options SignatureWriteOption
 			return nil, err
 		}
 		for name := range parts {
-			if removed[strings.ToLower(cleanPackagePath(name))] {
+			if removed[cleanPackagePath(name)] {
 				delete(parts, name)
 			}
 		}
@@ -364,7 +370,7 @@ func signatureWritePackage(parts map[string][]byte, options SignatureWriteOption
 		directory = path.Join(path.Dir(listPath), "Sign_"+strconv.Itoa(n))
 		occupied := false
 		for name := range parts {
-			if strings.EqualFold(name, directory) || strings.HasPrefix(strings.ToLower(name), strings.ToLower(directory)+"/") {
+			if name == directory || strings.HasPrefix(name, directory+"/") {
 				occupied = true
 				break
 			}
@@ -474,7 +480,7 @@ func signatureWriteIDs(list *Signatures, listPath string, parts map[string][]byt
 	var sequence signatureIDSequence
 	sequence.observe(list.MaxSignID)
 	add := func(id string) error {
-		if !signatureWriteIDValid(id) || used[id] {
+		if !ofdXMLIDValid(id) || used[id] {
 			return fmt.Errorf("invalid or duplicate signature ID: %s", id)
 		}
 		used[id] = true
@@ -534,14 +540,7 @@ func signatureWriteXML(name string, value any) ([]byte, error) {
 func signatureWriteAvailable(parts map[string][]byte, name string) string {
 	original := name
 	for i := 1; ; i++ {
-		found := false
+		if _, found := parts[name]; !found {
-		for existing := range parts {
-			if strings.EqualFold(existing, name) {
-				found = true
-				break
-			}
-		}
-		if !found {
 			return name
 		}
 		name = path.Join(path.Dir(original), strconv.Itoa(i)+"_"+path.Base(original))

ofdgo_sign_write_stamp.go

@@ -18,7 +18,6 @@ import (
 	"fmt"
 	"math"
 	"strconv"
-	"unicode/utf8"
 )
 
 // SignaturePageStamps 生成所选页面相同毫米坐标的印章位置并检查页内范围
@@ -110,7 +109,7 @@ func signatureWriteStamps(r *Reader, stamps []SignatureStamp, seal bool) ([]Sign
 		if stamp.ID == "" {
 			continue
 		}
-		if !signatureWriteIDValid(stamp.ID) || ids[stamp.ID] {
+		if !ofdXMLIDValid(stamp.ID) || ids[stamp.ID] {
 			return nil, fmt.Errorf("invalid or duplicate stamp ID: %s", stamp.ID)
 		}
 		ids[stamp.ID] = true
@@ -142,31 +141,6 @@ func signatureWriteStamps(r *Reader, stamps []SignatureStamp, seal bool) ([]Sign
 	return result, nil
 }
 
-// signatureWriteIDValid 按XML 1.0的NCName规则校验签章标识
-// 入参: id 标识
-// 返回: bool 是否有效
-func signatureWriteIDValid(id string) bool {
-	if id == "" || !utf8.ValidString(id) {
-		return false
-	}
-	for i, c := range id {
-		if c == '_' || c >= 'A' && c <= 'Z' || c >= 'a' && c <= 'z' ||
-			c >= 0xC0 && c <= 0xD6 || c >= 0xD8 && c <= 0xF6 ||
-			c >= 0xF8 && c <= 0x2FF || c >= 0x370 && c <= 0x37D ||
-			c >= 0x37F && c <= 0x1FFF || c >= 0x200C && c <= 0x200D ||
-			c >= 0x2070 && c <= 0x218F || c >= 0x2C00 && c <= 0x2FEF ||
-			c >= 0x3001 && c <= 0xD7FF || c >= 0xF900 && c <= 0xFDCF ||
-			c >= 0xFDF0 && c <= 0xFFFD || c >= 0x10000 && c <= 0xEFFFF {
-			continue
-		}
-		if i == 0 || !(c == '-' || c == '.' || c >= '0' && c <= '9' ||
-			c == 0xB7 || c >= 0x300 && c <= 0x36F || c >= 0x203F && c <= 0x2040) {
-			return false
-		}
-	}
-	return true
-}
-
 // signatureWriteBoxValid 判断印章区域是否有限且尺寸为正
 // 入参: box 区域
 // 返回: bool 是否有效

ofdgo_version.go

@@ -0,0 +1,185 @@
+// Copyright 2025-2026 肖其顿 (XIAO QI DUN)
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+// http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+package ofdgo
+
+import (
+	"fmt"
+	"strconv"
+	"strings"
+)
+
+// DocumentVersionInfo 文档版本信息，Location、DocRoot及Files中的路径均相对于包根目录
+// Index为版本号，Version为该版本适用的格式版本，CreationDate保留原始日期文本
+type DocumentVersionInfo struct {
+	ID           string
+	Index        int32
+	Current      bool
+	Location     string
+	Version      string
+	Name         string
+	CreationDate string
+	DocRoot      string
+	Files        []VersionFile
+}
+
+// VersionFile 文档版本中的文件标识与包内路径
+type VersionFile struct {
+	ID       string
+	Location string
+}
+
+// DocumentVersions 按入口顺序读取当前文档的版本信息，不切换文档或读取文件清单中的内容
+// 返回独立副本；没有版本列表时返回nil，不以版本号推测默认版本
+// 返回: []DocumentVersionInfo 版本信息, error 读取或描述错误
+func (r *Reader) DocumentVersions() ([]DocumentVersionInfo, error) {
+	entries, err := r.documentVersionEntries()
+	if err != nil {
+		return nil, err
+	}
+	for i, entry := range entries {
+		info, err := r.readDocumentVersion(entry.ID, entry.Location)
+		if err != nil {
+			return nil, fmt.Errorf("version %s: %w", entry.ID, err)
+		}
+		info.Index, info.Current = entry.Index, entry.Current
+		entries[i] = info
+	}
+	return entries, nil
+}
+
+// documentVersionEntries 读取当前文档的版本入口，不提前加载各版本描述
+// 返回: []DocumentVersionInfo 入口属性, error 读取或描述错误
+func (r *Reader) documentVersionEntries() ([]DocumentVersionInfo, error) {
+	if r.documentIndex < 0 || r.documentIndex >= r.DocumentCount() {
+		return nil, fmt.Errorf("no docbody found")
+	}
+	data, err := r.readFile("OFD.xml")
+	if err != nil {
+		return nil, err
+	}
+	root, err := parseEditorXML(data)
+	if err != nil {
+		return nil, err
+	}
+	body := root.childAt("DocBody", r.documentIndex)
+	if body == nil {
+		return nil, fmt.Errorf("document body not found: %d", r.documentIndex)
+	}
+	versions := body.child("Versions")
+	if versions == nil {
+		return nil, nil
+	}
+	if body.childAt("Versions", 1) != nil {
+		return nil, fmt.Errorf("multiple version lists")
+	}
+	var result []DocumentVersionInfo
+	ids := make(map[string]bool)
+	for _, node := range versions.children {
+		if !node.matchesOFD("Version") {
+			continue
+		}
+		id := strings.TrimSpace(node.attr("ID"))
+		if !ofdXMLIDValid(id) || ids[id] {
+			return nil, fmt.Errorf("invalid or duplicate version ID %q", id)
+		}
+		ids[id] = true
+		index, err := strconv.ParseInt(strings.TrimSpace(node.attr("Index")), 10, 32)
+		if err != nil {
+			return nil, fmt.Errorf("invalid version Index for %s: %w", id, err)
+		}
+		current := false
+		if node.hasAttr("Current") {
+			switch strings.TrimSpace(node.attr("Current")) {
+			case "true", "1":
+				current = true
+			case "false", "0":
+			default:
+				return nil, fmt.Errorf("invalid version Current for %s", id)
+			}
+		}
+		location, err := versionFileLocation("OFD.xml", node.attr("BaseLoc"))
+		if err != nil {
+			return nil, fmt.Errorf("version %s: %w", id, err)
+		}
+		result = append(result, DocumentVersionInfo{ID: id, Index: int32(index), Current: current, Location: location})
+	}
+	if len(result) == 0 {
+		return nil, fmt.Errorf("empty version list")
+	}
+	return result, nil
+}
+
+// readDocumentVersion 读取版本描述，按描述文件位置解析文件清单和文档入口
+// 入参: id 入口中的版本标识, location 描述文件路径
+// 返回: DocumentVersionInfo 版本信息, error 读取或描述错误
+func (r *Reader) readDocumentVersion(id, location string) (DocumentVersionInfo, error) {
+	var info DocumentVersionInfo
+	data, err := r.readFile(location)
+	if err != nil {
+		return info, err
+	}
+	root, err := parseEditorXML(data)
+	if err != nil {
+		return info, err
+	}
+	if root.name.Local != "DocVersion" || strings.TrimSpace(root.attr("ID")) != id {
+		return info, fmt.Errorf("invalid version descriptor %s", location)
+	}
+	list, docRoot := root.child("FileList"), root.child("DocRoot")
+	if list == nil || docRoot == nil {
+		return info, fmt.Errorf("missing version FileList or DocRoot in %s", location)
+	}
+	if root.childAt("FileList", 1) != nil || root.childAt("DocRoot", 1) != nil {
+		return info, fmt.Errorf("duplicate version FileList or DocRoot in %s", location)
+	}
+	entry, err := versionFileLocation(location, editorImportText(data, docRoot))
+	if err != nil {
+		return info, err
+	}
+	info = DocumentVersionInfo{ID: id, Location: location, Version: root.attr("Version"), Name: root.attr("Name"), CreationDate: root.attr("CreationDate"), DocRoot: entry}
+	ids := map[string]bool{id: true}
+	found := false
+	for _, node := range list.children {
+		if !node.matchesOFD("File") {
+			continue
+		}
+		fileID := strings.TrimSpace(node.attr("ID"))
+		if !ofdXMLIDValid(fileID) || ids[fileID] {
+			return DocumentVersionInfo{}, fmt.Errorf("invalid or duplicate version file ID %q", fileID)
+		}
+		ids[fileID] = true
+		name, err := versionFileLocation(location, editorImportText(data, node))
+		if err != nil {
+			return DocumentVersionInfo{}, err
+		}
+		found = found || name == entry
+		info.Files = append(info.Files, VersionFile{ID: fileID, Location: name})
+	}
+	if !found {
+		return DocumentVersionInfo{}, fmt.Errorf("version DocRoot absent from FileList: %s", entry)
+	}
+	return info, nil
+}
+
+// versionFileLocation 解析版本描述中的ST_Loc，拒绝空路径和包外路径
+// 入参: owner 所在XML路径, value 引用路径
+// 返回: string 包内路径, error 路径错误
+func versionFileLocation(owner, value string) (string, error) {
+	name := resolveResourcePath(owner, "", value)
+	if name == "" || !validPackagePath(name) {
+		return "", fmt.Errorf("invalid version file location %q", value)
+	}
+	return name, nil
+}

ofdgo_version_edit.go

@@ -0,0 +1,422 @@
+// Copyright 2025-2026 肖其顿 (XIAO QI DUN)
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+// http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+package ofdgo
+
+import (
+	"bytes"
+	"context"
+	"encoding/xml"
+	"fmt"
+	"maps"
+	"path"
+	"slices"
+	"strconv"
+	"strings"
+)
+
+// versionEditFile 保存版本内容XML及标准文件引用，不解析字体、图片或附件数据
+type versionEditFile struct {
+	data  []byte
+	links []versionEditLink
+}
+
+// versionEditLink 记录文件引用的位置与包内目标
+type versionEditLink struct {
+	node      *editorXML
+	attribute string
+	target    string
+}
+
+// versionEditGraph 保存一个入口的结构文件与全部直接、间接文件依赖
+type versionEditGraph struct {
+	files map[string]*versionEditFile
+	used  map[string]bool
+}
+
+// versionEditGraph 沿标准文件引用建立版本依赖图，二进制资源只记录路径
+// 入参: ctx 取消上下文, entry 文档入口, parts 待修改条目
+// 返回: *versionEditGraph 依赖图, error 读取或结构错误
+func (r *Reader) versionEditGraph(ctx context.Context, entry string, parts map[string][]byte) (*versionEditGraph, error) {
+	graph := &versionEditGraph{files: make(map[string]*versionEditFile), used: make(map[string]bool)}
+	queue := []string{entry}
+	for len(queue) != 0 {
+		if err := ctx.Err(); err != nil {
+			return nil, err
+		}
+		name := queue[0]
+		queue = queue[1:]
+		if graph.files[name] != nil {
+			continue
+		}
+		data, ok := parts[name]
+		if !ok {
+			var err error
+			data, err = r.readFile(name)
+			if err != nil {
+				return nil, err
+			}
+		}
+		root, err := parseEditorXML(data)
+		if err != nil {
+			return nil, fmt.Errorf("version file %s: %w", name, err)
+		}
+		file := &versionEditFile{data: data}
+		graph.files[name], graph.used[name] = file, true
+		base := ""
+		if root.name.Local == "Res" {
+			base = root.attr("BaseLoc")
+		}
+		var visit func(*editorXML)
+		visit = func(node *editorXML) {
+			if node != root && !node.matchesOFD(node.name.Local) {
+				return
+			}
+			add := func(attribute, value string, structure bool) {
+				if strings.TrimSpace(value) == "" {
+					return
+				}
+				target := resolveResourcePath(name, base, value)
+				file.links = append(file.links, versionEditLink{node: node, attribute: attribute, target: target})
+				graph.used[target] = true
+				if structure {
+					queue = append(queue, target)
+				}
+			}
+			switch node.name.Local {
+			case "PublicRes", "DocumentRes", "PageRes", "Annotations", "Attachments", "Signatures", "CustomTags", "Extensions":
+				if len(node.children) == 0 {
+					add("", editorImportText(data, node), true)
+				}
+			case "Page", "TemplatePage", "CompositeGraphicUnit", "Signature":
+				add("BaseLoc", node.attr("BaseLoc"), true)
+			case "PageAnnot":
+				add("FileLoc", node.attr("FileLoc"), true)
+			case "DrawParam":
+				add("Link", node.attr("Link"), true)
+			case "ColorSpace":
+				add("Profile", node.attr("Profile"), false)
+			case "FontFile", "MediaFile", "Profile", "SignedValue", "SchemaLoc", "ExtendData":
+				add("", editorImportText(data, node), false)
+			case "FileLoc":
+				add("", editorImportText(data, node), root.name.Local == "Annotations")
+			}
+			for _, child := range node.children {
+				visit(child)
+			}
+		}
+		visit(root)
+	}
+	return graph, nil
+}
+
+// versionSharedFiles 汇总其他版本、主入口及其他文档依赖，保护共享条目不被覆盖
+// 入参: ctx 取消上下文
+// 返回: map[string]bool 受保护路径, error 读取或结构错误
+func (r *Reader) versionSharedFiles(ctx context.Context) (map[string]bool, error) {
+	shared := make(map[string]bool)
+	data, err := r.readFile("OFD.xml")
+	if err != nil {
+		return nil, err
+	}
+	root, err := parseEditorXML(data)
+	if err != nil {
+		return nil, err
+	}
+	body := root.childAt("DocBody", r.documentIndex)
+	if info := body.child("DocInfo"); info != nil {
+		if cover := info.child("Cover"); cover != nil {
+			name, err := versionFileLocation("OFD.xml", editorImportText(data, cover))
+			if err != nil {
+				return nil, err
+			}
+			shared[name] = true
+		}
+	}
+	for index := range r.DocumentCount() {
+		if index == r.documentIndex {
+			continue
+		}
+		files, err := r.documentFiles(ctx, index)
+		if err != nil {
+			return nil, err
+		}
+		maps.Copy(shared, files)
+	}
+	versions, err := r.DocumentVersions()
+	if err != nil {
+		return nil, err
+	}
+	for _, version := range versions {
+		if r.versionInfo != nil && version.ID == r.versionInfo.ID {
+			continue
+		}
+		shared[version.Location] = true
+		for _, file := range version.Files {
+			shared[file.Location] = true
+		}
+	}
+	if r.versionInfo != nil {
+		entry, err := versionFileLocation("OFD.xml", r.OFD.DocBody[r.documentIndex].DocRoot)
+		if err != nil {
+			return nil, err
+		}
+		graph, err := r.versionEditGraph(ctx, entry, nil)
+		if err != nil {
+			return nil, err
+		}
+		maps.Copy(shared, graph.used)
+	}
+	return shared, nil
+}
+
+// versionChanges 隔离当前入口修改涉及的共享文件，同步版本清单，不修改输入包或修改集
+// 入参: ctx 取消上下文, changes 待写入条目
+// 返回: map[string][]byte 隔离后的修改集, error 读取、引用或结构错误
+func (r *Reader) versionChanges(ctx context.Context, changes map[string][]byte) (map[string][]byte, error) {
+	if err := ctx.Err(); err != nil {
+		return nil, err
+	}
+	entry, err := r.DocumentRoot()
+	if err != nil {
+		return nil, err
+	}
+	if !r.OFD.DocBody[r.documentIndex].versioned || len(changes) == 0 {
+		return changes, ctx.Err()
+	}
+	parts := maps.Clone(changes)
+	for name, data := range parts {
+		if original, err := r.readFileView(name); err == nil && bytes.Equal(original, data) {
+			delete(parts, name)
+		}
+	}
+	if len(parts) == 0 {
+		return parts, ctx.Err()
+	}
+	shared, err := r.versionSharedFiles(ctx)
+	if err != nil {
+		return nil, err
+	}
+	graph, err := r.versionEditGraph(ctx, entry, parts)
+	if err != nil {
+		return nil, err
+	}
+	owned := maps.Clone(graph.used)
+	if r.versionInfo != nil {
+		for _, file := range r.versionInfo.Files {
+			owned[file.Location] = true
+		}
+	}
+	paths := make(map[string]string)
+	reserved := maps.Clone(parts)
+	var pending []string
+	allocate := func(name string) {
+		extension := path.Ext(name)
+		target := packageAvailableName(r, reserved, strings.TrimSuffix(name, extension)+".edit"+extension)
+		paths[name], reserved[target] = target, nil
+		pending = append(pending, name)
+	}
+	for _, name := range slices.Sorted(maps.Keys(parts)) {
+		if name != "OFD.xml" && shared[name] && owned[name] {
+			allocate(name)
+		}
+	}
+	parents := make(map[string][]string)
+	for _, name := range slices.Sorted(maps.Keys(graph.files)) {
+		for _, link := range graph.files[name].links {
+			parents[link.target] = append(parents[link.target], name)
+		}
+	}
+	for i := 0; i < len(pending); i++ {
+		for _, name := range parents[pending[i]] {
+			if shared[name] && paths[name] == "" {
+				allocate(name)
+			}
+		}
+	}
+	for name, file := range graph.files {
+		data, err := file.rewrite(paths)
+		if err != nil {
+			return nil, err
+		}
+		if !bytes.Equal(data, file.data) {
+			parts[name] = data
+		}
+	}
+	for source, target := range paths {
+		data, ok := parts[source]
+		if !ok {
+			data, err = r.readFile(source)
+			if err != nil {
+				return nil, err
+			}
+		}
+		parts[target] = data
+		delete(parts, source)
+	}
+	if r.versionInfo != nil {
+		if err := r.versionManifest(parts, paths, graph.used, shared); err != nil {
+			return nil, err
+		}
+	} else if target := paths[entry]; target != "" {
+		data, ok := parts["OFD.xml"]
+		if !ok {
+			data, err = r.readFile("OFD.xml")
+			if err != nil {
+				return nil, err
+			}
+		}
+		root, err := parseEditorXML(data)
+		if err != nil {
+			return nil, err
+		}
+		node := root.childAt("DocBody", r.documentIndex).child("DocRoot")
+		parts["OFD.xml"] = editorPatchXML(data, []editorXMLPatch{editorXMLContent(data, node, editorFontXMLText("/"+target))})
+	}
+	return parts, ctx.Err()
+}
+
+// rewrite 更新已隔离文件的引用，保留其余XML及相对路径基准
+// 入参: paths 原路径到新路径的映射
+// 返回: []byte 更新后的XML, error 属性编码错误
+func (file *versionEditFile) rewrite(paths map[string]string) ([]byte, error) {
+	var patches []editorXMLPatch
+	headers := make(map[*editorXML][]versionEditLink)
+	for _, link := range file.links {
+		if paths[link.target] == "" {
+			continue
+		}
+		if link.attribute == "" {
+			patches = append(patches, editorXMLContent(file.data, link.node, editorFontXMLText("/"+paths[link.target])))
+		} else {
+			headers[link.node] = append(headers[link.node], link)
+		}
+	}
+	for node, links := range headers {
+		data := file.data[node.start:node.end]
+		for _, link := range links {
+			root, err := parseEditorXML(data)
+			if err != nil {
+				return nil, err
+			}
+			data, err = editorXMLAttribute(data, root, link.attribute, "/"+paths[link.target])
+			if err != nil {
+				return nil, err
+			}
+		}
+		root, err := parseEditorXML(data)
+		if err != nil {
+			return nil, err
+		}
+		if node.open == node.end {
+			patches = append(patches, editorXMLPatch{node.start, node.end, data})
+		} else {
+			patches = append(patches, editorXMLPatch{node.start, node.open, data[:root.open]})
+		}
+	}
+	return editorPatchXML(file.data, patches), nil
+}
+
+// versionManifest 更新所选版本的入口与文件清单，保留已有文件标识及扩展字段
+// 入参: parts 输出条目, paths 隔离路径, used 当前内容依赖, shared 其他入口依赖
+// 返回: error 描述读取或XML编码错误
+func (r *Reader) versionManifest(parts map[string][]byte, paths map[string]string, used, shared map[string]bool) error {
+	version := r.versionInfo
+	data, err := r.readFile(version.Location)
+	if err != nil {
+		return err
+	}
+	root, err := parseEditorXML(data)
+	if err != nil {
+		return err
+	}
+	list := root.child("FileList")
+	var patches []editorXMLPatch
+	listed := make(map[string]bool)
+	ids := map[string]bool{version.ID: true}
+	for _, node := range list.children {
+		if !node.matchesOFD("File") {
+			continue
+		}
+		ids[strings.TrimSpace(node.attr("ID"))] = true
+		name, err := versionFileLocation(version.Location, editorImportText(data, node))
+		if err != nil {
+			return err
+		}
+		if target := paths[name]; target != "" {
+			name = target
+			patches = append(patches, editorXMLContent(data, node, editorFontXMLText("/"+target)))
+		}
+		listed[name] = true
+	}
+	var added []byte
+	serial := 0
+	for _, name := range slices.Sorted(maps.Keys(used)) {
+		if target := paths[name]; target != "" {
+			name = target
+		}
+		if listed[name] {
+			continue
+		}
+		id := ""
+		for id == "" || ids[id] {
+			serial++
+			id = "file" + strconv.Itoa(serial)
+		}
+		ids[id], listed[name] = true, true
+		item, err := list.containerXML("File", ofdAttrs{{Name: xml.Name{Local: "ID"}, Value: id}}, editorFontXMLText("/"+name))
+		if err != nil {
+			return err
+		}
+		added = append(added, bytes.TrimPrefix(item, []byte(xml.Header))...)
+	}
+	if len(added) != 0 {
+		patches = append(patches, editorXMLPatch{list.close, list.close, added})
+	}
+	if target := paths[version.DocRoot]; target != "" {
+		patches = append(patches, editorXMLContent(data, root.child("DocRoot"), editorFontXMLText("/"+target)))
+	}
+	updated := editorPatchXML(data, patches)
+	if bytes.Equal(updated, data) {
+		return nil
+	}
+	name := version.Location
+	if shared[name] {
+		extension := path.Ext(name)
+		name = packageAvailableName(r, parts, strings.TrimSuffix(name, extension)+".edit"+extension)
+		entry, ok := parts["OFD.xml"]
+		if !ok {
+			entry, err = r.readFile("OFD.xml")
+			if err != nil {
+				return err
+			}
+		}
+		root, err := parseEditorXML(entry)
+		if err != nil {
+			return err
+		}
+		for _, node := range root.childAt("DocBody", r.documentIndex).child("Versions").children {
+			if node.matchesOFD("Version") && node.attr("ID") == version.ID {
+				fragment, err := editorXMLAttribute(entry, node, "BaseLoc", "/"+name)
+				if err != nil {
+					return err
+				}
+				parts["OFD.xml"] = editorPatchXML(entry, []editorXMLPatch{{node.start, node.end, fragment}})
+				break
+			}
+		}
+	}
+	parts[name] = updated
+	return nil
+}

ofdgo_version_extract.go

@@ -0,0 +1,140 @@
+// Copyright 2025-2026 肖其顿 (XIAO QI DUN)
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+// http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+package ofdgo
+
+import (
+	"archive/zip"
+	"fmt"
+)
+
+// ExtractVersion 将当前文档的指定版本提取为独立阅读器，可继续编辑或另存，不改变原包
+// 仅保留版本文件清单中的文件并重建单文档入口，不保留版本列表；原签名可能因入口变化而失效
+// 文件内容按需读取，使用期间不得关闭持有输入文件的原阅读器；返回值不关闭输入文件
+// 入参: id 版本标识，空值选择唯一的Current版本，没有默认版本或存在歧义时返回错误
+// 返回: *Reader 版本阅读器, error 选择、读取或文件缺失错误
+func (r *Reader) ExtractVersion(id string) (*Reader, error) {
+	entries, err := r.documentVersionEntries()
+	if err != nil {
+		return nil, err
+	}
+	var selected *DocumentVersionInfo
+	for i := range entries {
+		entry := &entries[i]
+		if id != "" && entry.ID == id || id == "" && entry.Current {
+			if selected != nil {
+				return nil, fmt.Errorf("multiple current document versions")
+			}
+			selected = entry
+		}
+	}
+	if selected == nil {
+		if id == "" {
+			return nil, fmt.Errorf("no current document version")
+		}
+		return nil, fmt.Errorf("document version not found: %s", id)
+	}
+	version, err := r.readDocumentVersion(selected.ID, selected.Location)
+	if err != nil {
+		return nil, fmt.Errorf("version %s: %w", selected.ID, err)
+	}
+	files := make(map[string]bool, len(version.Files))
+	for _, file := range version.Files {
+		if !r.fileNames[file.Location] {
+			return nil, fmt.Errorf("version %s file not found: %s", selected.ID, file.Location)
+		}
+		files[file.Location] = true
+	}
+	root, err := r.extractedVersionRoot(version, files)
+	if err != nil {
+		return nil, err
+	}
+	next := &Reader{Path: r.Path, files: map[string][]byte{"OFD.xml": root}, encryption: r.encryption}
+	for name := range files {
+		if name != "OFD.xml" {
+			if data, ok := r.files[name]; ok {
+				next.files[name] = data
+			}
+		}
+	}
+	if r.Zip != nil {
+		next.Zip = &zip.Reader{Comment: r.Zip.Comment}
+		for _, file := range r.Zip.File {
+			name := cleanPackagePath(file.Name)
+			if name != "OFD.xml" && files[name] && !file.FileInfo().IsDir() {
+				next.Zip.File = append(next.Zip.File, file)
+			}
+		}
+	}
+	if err := next.initRoot(); err != nil {
+		return nil, err
+	}
+	if _, err := next.docStructure(); err != nil {
+		return nil, err
+	}
+	return next, nil
+}
+
+// extractedVersionRoot 重建版本副本入口，保留元数据及命名空间，移除其他文档与版本列表
+// 入参: version 版本描述, files 版本包含的文件
+// 返回: []byte 主入口XML, error 读取或结构错误
+func (r *Reader) extractedVersionRoot(version DocumentVersionInfo, files map[string]bool) ([]byte, error) {
+	data, err := r.readFile("OFD.xml")
+	if err != nil {
+		return nil, err
+	}
+	root, err := parseEditorXML(data)
+	if err != nil {
+		return nil, err
+	}
+	body := root.childAt("DocBody", r.documentIndex)
+	if body == nil || body.child("DocRoot") == nil {
+		return nil, fmt.Errorf("document entry not found: %d", r.documentIndex)
+	}
+	var patches []editorXMLPatch
+	for _, child := range root.children {
+		if child.matchesOFD("DocBody") && child != body {
+			patches = append(patches, editorXMLPatch{start: child.start, end: child.end})
+		}
+	}
+	entry := body.child("DocRoot")
+	patches = append(patches, editorXMLContent(data, entry, editorFontXMLText(version.DocRoot)))
+	if versions := body.child("Versions"); versions != nil {
+		patches = append(patches, editorXMLPatch{start: versions.start, end: versions.end})
+	}
+	references := []*editorXML{body.child("Signatures")}
+	if info := body.child("DocInfo"); info != nil {
+		references = append(references, info.child("Cover"))
+	}
+	for _, node := range references {
+		if node == nil {
+			continue
+		}
+		name := resolveResourcePath("OFD.xml", "", editorImportText(data, node))
+		if files[name] {
+			patches = append(patches, editorXMLContent(data, node, editorFontXMLText("/"+name)))
+		} else {
+			patches = append(patches, editorXMLPatch{start: node.start, end: node.end})
+		}
+	}
+	data = editorPatchXML(data, patches)
+	if version.Version != "" {
+		root, err = parseEditorXML(data)
+		if err != nil {
+			return nil, err
+		}
+		data, err = editorXMLAttribute(data, root, "Version", version.Version)
+	}
+	return data, err
+}

ofdgo_version_reader.go

@@ -0,0 +1,127 @@
+// Copyright 2025-2026 肖其顿 (XIAO QI DUN)
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+// http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+package ofdgo
+
+import "fmt"
+
+// DocumentRoot 获取当前阅读入口的包内路径，不改写OFD中的主入口
+// 未显式选择时使用唯一的Current版本；没有默认版本时使用主入口，多个默认版本返回错误
+// 返回: string 阅读入口, error 版本选择或描述错误
+func (r *Reader) DocumentRoot() (string, error) {
+	if r.documentRoot != "" {
+		return r.documentRoot, nil
+	}
+	if r.documentIndex < 0 || r.documentIndex >= r.DocumentCount() {
+		return "", fmt.Errorf("no docbody found")
+	}
+	body := r.OFD.DocBody[r.documentIndex]
+	if r.selectedVersion != nil && *r.selectedVersion == "" || !body.versioned && r.selectedVersion == nil {
+		root, err := versionFileLocation("OFD.xml", body.DocRoot)
+		if err == nil {
+			r.documentRoot = root
+		}
+		return root, err
+	}
+	entries, err := r.documentVersionEntries()
+	if err != nil {
+		return "", err
+	}
+	var selected *DocumentVersionInfo
+	for i := range entries {
+		entry := &entries[i]
+		if r.selectedVersion != nil && entry.ID == *r.selectedVersion || r.selectedVersion == nil && entry.Current {
+			if selected != nil {
+				return "", fmt.Errorf("multiple current document versions")
+			}
+			selected = entry
+		}
+	}
+	if selected == nil {
+		if r.selectedVersion != nil {
+			return "", fmt.Errorf("document version not found: %s", *r.selectedVersion)
+		}
+		root, err := versionFileLocation("OFD.xml", body.DocRoot)
+		if err == nil {
+			r.documentRoot = root
+		}
+		return root, err
+	}
+	info, err := r.readDocumentVersion(selected.ID, selected.Location)
+	if err != nil {
+		return "", fmt.Errorf("version %s: %w", selected.ID, err)
+	}
+	info.Index, info.Current = selected.Index, selected.Current
+	r.documentRoot, r.versionInfo = info.DocRoot, &info
+	return r.documentRoot, nil
+}
+
+// DocumentVersionID 获取当前阅读的版本标识，主入口返回空字符串
+// 返回: string 版本标识, error 版本选择或描述错误
+func (r *Reader) DocumentVersionID() (string, error) {
+	if _, err := r.DocumentRoot(); err != nil {
+		return "", err
+	}
+	if r.versionInfo == nil {
+		return "", nil
+	}
+	return r.versionInfo.ID, nil
+}
+
+// DocumentVersion 创建指定版本的独立阅读器，保留整个包，不改变当前阅读器或共享可变缓存
+// 返回值不关闭输入文件，使用期间不得关闭持有输入文件的原阅读器
+// 入参: id 版本标识，空值显式选择主入口；Document可重新按Current选择默认版本
+// 返回: *Reader 版本阅读器, error 选择或读取错误
+func (r *Reader) DocumentVersion(id string) (*Reader, error) {
+	next := &Reader{Path: r.Path, Zip: r.Zip, files: r.files, encryption: r.encryption, documentIndex: r.documentIndex, selectedVersion: &id}
+	if err := next.initRoot(); err != nil {
+		return nil, err
+	}
+	if _, err := next.Doc(); err != nil {
+		return nil, err
+	}
+	return next, nil
+}
+
+// documentEntry 获取指定文档的阅读入口，当前文档保留显式选择，其他文档使用默认版本
+// 入参: index 文档索引
+// 返回: string 包内路径, error 选择错误
+func (r *Reader) documentEntry(index int) (string, error) {
+	if index == r.documentIndex {
+		return r.DocumentRoot()
+	}
+	view := *r
+	view.documentIndex, view.documentRoot = index, ""
+	view.selectedVersion, view.versionInfo = nil, nil
+	return view.DocumentRoot()
+}
+
+// versionSignatures 按OFD入口解析包级签名路径，选中版本仅使用文件清单内的签名入口
+// 入参: location OFD入口中的签名路径
+// 返回: string 签名路径，不属于该版本时为空
+func (r *Reader) versionSignatures(location string) string {
+	if location == "" {
+		return ""
+	}
+	name := resolveResourcePath("OFD.xml", "", location)
+	if r.versionInfo == nil {
+		return "/" + name
+	}
+	for _, file := range r.versionInfo.Files {
+		if file.Location == name {
+			return "/" + name
+		}
+	}
+	return ""
+}

ofdgo_writer.go

@@ -190,7 +190,7 @@ func (e *Editor) writePlaintext(writer io.Writer) (int64, error) {
 			return 0, err
 		}
 		for _, name := range slices.Sorted(maps.Keys(parts)) {
-			if removed[strings.ToLower(cleanPackagePath(name))] {
+			if removed[cleanPackagePath(name)] {
 				continue
 			}
 			if err := e.writeOutputEntry(archive, zip.FileHeader{Name: name, Method: zip.Deflate}, parts[name]); err != nil {

ofdgo_xml.go

@@ -14,6 +14,8 @@
 
 package ofdgo
 
+import "unicode/utf8"
+
 // 旧版与2016版OFD命名空间
 const (
 	ofdNamespaceLegacy = "http://www.ofdspec.org"
@@ -46,3 +48,28 @@ func classifyOFDNamespace(namespace string) ofdXMLNamespace {
 		return ofdXMLUnknown
 	}
 }
+
+// ofdXMLIDValid 按XML 1.0的NCName规则校验xs:ID，不用于数值型ST_ID
+// 入参: id 标识
+// 返回: bool 是否有效
+func ofdXMLIDValid(id string) bool {
+	if id == "" || !utf8.ValidString(id) {
+		return false
+	}
+	for i, c := range id {
+		if c == '_' || c >= 'A' && c <= 'Z' || c >= 'a' && c <= 'z' ||
+			c >= 0xC0 && c <= 0xD6 || c >= 0xD8 && c <= 0xF6 ||
+			c >= 0xF8 && c <= 0x2FF || c >= 0x370 && c <= 0x37D ||
+			c >= 0x37F && c <= 0x1FFF || c >= 0x200C && c <= 0x200D ||
+			c >= 0x2070 && c <= 0x218F || c >= 0x2C00 && c <= 0x2FEF ||
+			c >= 0x3001 && c <= 0xD7FF || c >= 0xF900 && c <= 0xFDCF ||
+			c >= 0xFDF0 && c <= 0xFFFD || c >= 0x10000 && c <= 0xEFFFF {
+			continue
+		}
+		if i == 0 || !(c == '-' || c == '.' || c >= '0' && c <= '9' ||
+			c == 0xB7 || c >= 0x300 && c <= 0x36F || c >= 0x203F && c <= 0x2040) {
+			return false
+		}
+	}
+	return true
+}