// Copyright 2013 The Go Authors. All rights reserved. // Use of this source code is governed by a BSD-style // license that can be found in the LICENSE file. // Package godoc is a work-in-progress (2013-07-17) package to // begin splitting up the godoc binary into multiple pieces. // // This package comment will evolve over time as this package splits // into smaller pieces. package godoc // import "golang.org/x/tools/godoc" import ( "bufio" "bytes" "fmt" "go/ast" "go/doc" "go/format" "go/printer" "go/token" htmltemplate "html/template" "io" "log" "os" pathpkg "path" "regexp" "strconv" "strings" "text/template" "time" "unicode" "unicode/utf8" ) // Fake relative package path for built-ins. Documentation for all globals // (not just exported ones) will be shown for packages in this directory, // and there will be no association of consts, vars, and factory functions // with types (see issue 6645). const builtinPkgPath = "builtin" // FuncMap defines template functions used in godoc templates. // // Convention: template function names ending in "_html" or "_url" produce // HTML- or URL-escaped strings; all other function results may // require explicit escaping in the template. func (p *Presentation) FuncMap() template.FuncMap { p.initFuncMapOnce.Do(p.initFuncMap) return p.funcMap } func (p *Presentation) TemplateFuncs() template.FuncMap { p.initFuncMapOnce.Do(p.initFuncMap) return p.templateFuncs } func (p *Presentation) initFuncMap() { if p.Corpus == nil { panic("nil Presentation.Corpus") } p.templateFuncs = template.FuncMap{ "code": p.code, } p.funcMap = template.FuncMap{ // various helpers "filename": filenameFunc, "repeat": strings.Repeat, "since": p.Corpus.pkgAPIInfo.sinceVersionFunc, // access to FileInfos (directory listings) "fileInfoName": fileInfoNameFunc, "fileInfoTime": fileInfoTimeFunc, // access to search result information "infoKind_html": infoKind_htmlFunc, "infoLine": p.infoLineFunc, "infoSnippet_html": p.infoSnippet_htmlFunc, // formatting of AST nodes "node": p.nodeFunc, "node_html": p.node_htmlFunc, "comment_html": comment_htmlFunc, "sanitize": sanitizeFunc, // support for URL attributes "pkgLink": pkgLinkFunc, "srcLink": srcLinkFunc, "posLink_url": newPosLink_urlFunc(srcPosLinkFunc), "docLink": docLinkFunc, "queryLink": queryLinkFunc, "srcBreadcrumb": srcBreadcrumbFunc, "srcToPkgLink": srcToPkgLinkFunc, // formatting of Examples "example_html": p.example_htmlFunc, "example_name": p.example_nameFunc, "example_suffix": p.example_suffixFunc, // formatting of analysis information "callgraph_html": p.callgraph_htmlFunc, "implements_html": p.implements_htmlFunc, "methodset_html": p.methodset_htmlFunc, // formatting of Notes "noteTitle": noteTitle, // Number operation "multiply": multiply, // formatting of PageInfoMode query string "modeQueryString": modeQueryString, // check whether to display third party section or not "hasThirdParty": hasThirdParty, // get the no. of columns to split the toc in search page "tocColCount": tocColCount, } if p.URLForSrc != nil { p.funcMap["srcLink"] = p.URLForSrc } if p.URLForSrcPos != nil { p.funcMap["posLink_url"] = newPosLink_urlFunc(p.URLForSrcPos) } if p.URLForSrcQuery != nil { p.funcMap["queryLink"] = p.URLForSrcQuery } } func multiply(a, b int) int { return a * b } func filenameFunc(path string) string { _, localname := pathpkg.Split(path) return localname } func fileInfoNameFunc(fi os.FileInfo) string { name := fi.Name() if fi.IsDir() { name += "/" } return name } func fileInfoTimeFunc(fi os.FileInfo) string { if t := fi.ModTime(); t.Unix() != 0 { return t.Local().String() } return "" // don't return epoch if time is obviously not set } // The strings in infoKinds must be properly html-escaped. var infoKinds = [nKinds]string{ PackageClause: "package clause", ImportDecl: "import decl", ConstDecl: "const decl", TypeDecl: "type decl", VarDecl: "var decl", FuncDecl: "func decl", MethodDecl: "method decl", Use: "use", } func infoKind_htmlFunc(info SpotInfo) string { return infoKinds[info.Kind()] // infoKind entries are html-escaped } func (p *Presentation) infoLineFunc(info SpotInfo) int { line := info.Lori() if info.IsIndex() { index, _ := p.Corpus.searchIndex.Get() if index != nil { line = index.(*Index).Snippet(line).Line } else { // no line information available because // we don't have an index - this should // never happen; be conservative and don't // crash line = 0 } } return line } func (p *Presentation) infoSnippet_htmlFunc(info SpotInfo) string { if info.IsIndex() { index, _ := p.Corpus.searchIndex.Get() // Snippet.Text was HTML-escaped when it was generated return index.(*Index).Snippet(info.Lori()).Text } return `no snippet text available` } func (p *Presentation) nodeFunc(info *PageInfo, node interface{}) string { var buf bytes.Buffer p.writeNode(&buf, info, info.FSet, node) return buf.String() } func (p *Presentation) node_htmlFunc(info *PageInfo, node interface{}, linkify bool) string { var buf1 bytes.Buffer p.writeNode(&buf1, info, info.FSet, node) var buf2 bytes.Buffer if n, _ := node.(ast.Node); n != nil && linkify && p.DeclLinks { LinkifyText(&buf2, buf1.Bytes(), n) if st, name := isStructTypeDecl(n); st != nil { addStructFieldIDAttributes(&buf2, name, st) } } else { FormatText(&buf2, buf1.Bytes(), -1, true, "", nil) } return buf2.String() } // isStructTypeDecl checks whether n is a struct declaration. // It either returns a non-nil StructType and its name, or zero values. func isStructTypeDecl(n ast.Node) (st *ast.StructType, name string) { gd, ok := n.(*ast.GenDecl) if !ok || gd.Tok != token.TYPE { return nil, "" } if gd.Lparen > 0 { // Parenthesized type. Who does that, anyway? // TODO: Reportedly gri does. Fix this to handle that too. return nil, "" } if len(gd.Specs) != 1 { return nil, "" } ts, ok := gd.Specs[0].(*ast.TypeSpec) if !ok { return nil, "" } st, ok = ts.Type.(*ast.StructType) if !ok { return nil, "" } return st, ts.Name.Name } // addStructFieldIDAttributes modifies the contents of buf such that // all struct fields of the named struct have // in them, so people can link to /#Struct.Field. func addStructFieldIDAttributes(buf *bytes.Buffer, name string, st *ast.StructType) { if st.Fields == nil { return } // needsLink is a set of identifiers that still need to be // linked, where value == key, to avoid an allocation in func // linkedField. needsLink := make(map[string]string) for _, f := range st.Fields.List { if len(f.Names) == 0 { continue } fieldName := f.Names[0].Name needsLink[fieldName] = fieldName } var newBuf bytes.Buffer foreachLine(buf.Bytes(), func(line []byte) { if fieldName := linkedField(line, needsLink); fieldName != "" { fmt.Fprintf(&newBuf, ``, name, fieldName) delete(needsLink, fieldName) } newBuf.Write(line) }) buf.Reset() buf.Write(newBuf.Bytes()) } // foreachLine calls fn for each line of in, where a line includes // the trailing "\n", except on the last line, if it doesn't exist. func foreachLine(in []byte, fn func(line []byte)) { for len(in) > 0 { nl := bytes.IndexByte(in, '\n') if nl == -1 { fn(in) return } fn(in[:nl+1]) in = in[nl+1:] } } // commentPrefix is the line prefix for comments after they've been HTMLified. var commentPrefix = []byte(`// `) // linkedField determines whether the given line starts with an // identifier in the provided ids map (mapping from identifier to the // same identifier). The line can start with either an identifier or // an identifier in a comment. If one matches, it returns the // identifier that matched. Otherwise it returns the empty string. func linkedField(line []byte, ids map[string]string) string { line = bytes.TrimSpace(line) // For fields with a doc string of the // conventional form, we put the new span into // the comment instead of the field. // The "conventional" form is a complete sentence // per https://golang.org/s/style#comment-sentences like: // // // Foo is an optional Fooer to foo the foos. // Foo Fooer // // In this case, we want the #StructName.Foo // link to make the browser go to the comment // line "Foo is an optional Fooer" instead of // the "Foo Fooer" line, which could otherwise // obscure the docs above the browser's "fold". // // TODO: do this better, so it works for all // comments, including unconventional ones. line = bytes.TrimPrefix(line, commentPrefix) id := scanIdentifier(line) if len(id) == 0 { // No leading identifier. Avoid map lookup for // somewhat common case. return "" } return ids[string(id)] } // scanIdentifier scans a valid Go identifier off the front of v and // either returns a subslice of v if there's a valid identifier, or // returns a zero-length slice. func scanIdentifier(v []byte) []byte { var n int // number of leading bytes of v belonging to an identifier for { r, width := utf8.DecodeRune(v[n:]) if !(isLetter(r) || n > 0 && isDigit(r)) { break } n += width } return v[:n] } func isLetter(ch rune) bool { return 'a' <= ch && ch <= 'z' || 'A' <= ch && ch <= 'Z' || ch == '_' || ch >= utf8.RuneSelf && unicode.IsLetter(ch) } func isDigit(ch rune) bool { return '0' <= ch && ch <= '9' || ch >= utf8.RuneSelf && unicode.IsDigit(ch) } func comment_htmlFunc(comment string) string { var buf bytes.Buffer // TODO(gri) Provide list of words (e.g. function parameters) // to be emphasized by ToHTML. doc.ToHTML(&buf, comment, nil) // does html-escaping return buf.String() } // sanitizeFunc sanitizes the argument src by replacing newlines with // blanks, removing extra blanks, and by removing trailing whitespace // and commas before closing parentheses. func sanitizeFunc(src string) string { buf := make([]byte, len(src)) j := 0 // buf index comma := -1 // comma index if >= 0 for i := 0; i < len(src); i++ { ch := src[i] switch ch { case '\t', '\n', ' ': // ignore whitespace at the beginning, after a blank, or after opening parentheses if j == 0 { continue } if p := buf[j-1]; p == ' ' || p == '(' || p == '{' || p == '[' { continue } // replace all whitespace with blanks ch = ' ' case ',': comma = j case ')', '}', ']': // remove any trailing comma if comma >= 0 { j = comma } // remove any trailing whitespace if j > 0 && buf[j-1] == ' ' { j-- } default: comma = -1 } buf[j] = ch j++ } // remove trailing blank, if any if j > 0 && buf[j-1] == ' ' { j-- } return string(buf[:j]) } type PageInfo struct { Dirname string // directory containing the package Err error // error or nil GoogleCN bool // page is being served from golang.google.cn Mode PageInfoMode // display metadata from query string // package info FSet *token.FileSet // nil if no package documentation PDoc *doc.Package // nil if no package documentation Examples []*doc.Example // nil if no example code Notes map[string][]*doc.Note // nil if no package Notes PAst map[string]*ast.File // nil if no AST with package exports IsMain bool // true for package main IsFiltered bool // true if results were filtered // analysis info TypeInfoIndex map[string]int // index of JSON datum for type T (if -analysis=type) AnalysisData htmltemplate.JS // array of TypeInfoJSON values CallGraph htmltemplate.JS // array of PCGNodeJSON values (if -analysis=pointer) CallGraphIndex map[string]int // maps func name to index in CallGraph // directory info Dirs *DirList // nil if no directory information DirTime time.Time // directory time stamp DirFlat bool // if set, show directory in a flat (non-indented) manner } func (info *PageInfo) IsEmpty() bool { return info.Err != nil || info.PAst == nil && info.PDoc == nil && info.Dirs == nil } func pkgLinkFunc(path string) string { // because of the irregular mapping under goroot // we need to correct certain relative paths path = strings.TrimPrefix(path, "/") path = strings.TrimPrefix(path, "src/") path = strings.TrimPrefix(path, "pkg/") return "pkg/" + path } // srcToPkgLinkFunc builds an tag linking to the package // documentation of relpath. func srcToPkgLinkFunc(relpath string) string { relpath = pkgLinkFunc(relpath) relpath = pathpkg.Dir(relpath) if relpath == "pkg" { return `Index` } return fmt.Sprintf(`%s`, relpath, relpath[len("pkg/"):]) } // srcBreadcrumbFun converts each segment of relpath to a HTML . // Each segment links to its corresponding src directories. func srcBreadcrumbFunc(relpath string) string { segments := strings.Split(relpath, "/") var buf bytes.Buffer var selectedSegment string var selectedIndex int if strings.HasSuffix(relpath, "/") { // relpath is a directory ending with a "/". // Selected segment is the segment before the last slash. selectedIndex = len(segments) - 2 selectedSegment = segments[selectedIndex] + "/" } else { selectedIndex = len(segments) - 1 selectedSegment = segments[selectedIndex] } for i := range segments[:selectedIndex] { buf.WriteString(fmt.Sprintf(`%s/`, strings.Join(segments[:i+1], "/"), segments[i], )) } buf.WriteString(``) buf.WriteString(selectedSegment) buf.WriteString(``) return buf.String() } func newPosLink_urlFunc(srcPosLinkFunc func(s string, line, low, high int) string) func(info *PageInfo, n interface{}) string { // n must be an ast.Node or a *doc.Note return func(info *PageInfo, n interface{}) string { var pos, end token.Pos switch n := n.(type) { case ast.Node: pos = n.Pos() end = n.End() case *doc.Note: pos = n.Pos end = n.End default: panic(fmt.Sprintf("wrong type for posLink_url template formatter: %T", n)) } var relpath string var line int var low, high int // selection offset range if pos.IsValid() { p := info.FSet.Position(pos) relpath = p.Filename line = p.Line low = p.Offset } if end.IsValid() { high = info.FSet.Position(end).Offset } return srcPosLinkFunc(relpath, line, low, high) } } func srcPosLinkFunc(s string, line, low, high int) string { s = srcLinkFunc(s) var buf bytes.Buffer template.HTMLEscape(&buf, []byte(s)) // selection ranges are of form "s=low:high" if low < high { fmt.Fprintf(&buf, "?s=%d:%d", low, high) // no need for URL escaping // if we have a selection, position the page // such that the selection is a bit below the top line -= 10 if line < 1 { line = 1 } } // line id's in html-printed source are of the // form "L%d" where %d stands for the line number if line > 0 { fmt.Fprintf(&buf, "#L%d", line) // no need for URL escaping } return buf.String() } func srcLinkFunc(s string) string { s = pathpkg.Clean("/" + s) if !strings.HasPrefix(s, "/src/") { s = "/src" + s } return s } // queryLinkFunc returns a URL for a line in a source file with a highlighted // query term. // s is expected to be a path to a source file. // query is expected to be a string that has already been appropriately escaped // for use in a URL query. func queryLinkFunc(s, query string, line int) string { url := pathpkg.Clean("/"+s) + "?h=" + query if line > 0 { url += "#L" + strconv.Itoa(line) } return url } func docLinkFunc(s string, ident string) string { return pathpkg.Clean("/pkg/"+s) + "/#" + ident } func (p *Presentation) example_htmlFunc(info *PageInfo, funcName string) string { var buf bytes.Buffer for _, eg := range info.Examples { name := stripExampleSuffix(eg.Name) if name != funcName { continue } // print code cnode := &printer.CommentedNode{Node: eg.Code, Comments: eg.Comments} code := p.node_htmlFunc(info, cnode, true) out := eg.Output wholeFile := true // Additional formatting if this is a function body. if n := len(code); n >= 2 && code[0] == '{' && code[n-1] == '}' { wholeFile = false // remove surrounding braces code = code[1 : n-1] // unindent code = replaceLeadingIndentation(code, strings.Repeat(" ", p.TabWidth), "") // remove output comment if loc := exampleOutputRx.FindStringIndex(code); loc != nil { code = strings.TrimSpace(code[:loc[0]]) } } // Write out the playground code in standard Go style // (use tabs, no comment highlight, etc). play := "" if eg.Play != nil && p.ShowPlayground { var buf bytes.Buffer eg.Play.Comments = filterOutBuildAnnotations(eg.Play.Comments) if err := format.Node(&buf, info.FSet, eg.Play); err != nil { log.Print(err) } else { play = buf.String() } } // Drop output, as the output comment will appear in the code. if wholeFile && play == "" { out = "" } if p.ExampleHTML == nil { out = "" return "" } err := p.ExampleHTML.Execute(&buf, struct { Name, Doc, Code, Play, Output string GoogleCN bool }{eg.Name, eg.Doc, code, play, out, info.GoogleCN}) if err != nil { log.Print(err) } } return buf.String() } func filterOutBuildAnnotations(cg []*ast.CommentGroup) []*ast.CommentGroup { if len(cg) == 0 { return cg } for i := range cg { if !strings.HasPrefix(cg[i].Text(), "+build ") { // Found the first non-build tag, return from here until the end // of the slice. return cg[i:] } } // There weren't any non-build tags, return an empty slice. return []*ast.CommentGroup{} } // example_nameFunc takes an example function name and returns its display // name. For example, "Foo_Bar_quux" becomes "Foo.Bar (Quux)". func (p *Presentation) example_nameFunc(s string) string { name, suffix := splitExampleName(s) // replace _ with . for method names name = strings.Replace(name, "_", ".", 1) // use "Package" if no name provided if name == "" { name = "Package" } return name + suffix } // example_suffixFunc takes an example function name and returns its suffix in // parenthesized form. For example, "Foo_Bar_quux" becomes " (Quux)". func (p *Presentation) example_suffixFunc(name string) string { _, suffix := splitExampleName(name) return suffix } // implements_html returns the "> Implements" toggle for a package-level named type. // Its contents are populated from JSON data by client-side JS at load time. func (p *Presentation) implements_htmlFunc(info *PageInfo, typeName string) string { if p.ImplementsHTML == nil { return "" } index, ok := info.TypeInfoIndex[typeName] if !ok { return "" } var buf bytes.Buffer err := p.ImplementsHTML.Execute(&buf, struct{ Index int }{index}) if err != nil { log.Print(err) } return buf.String() } // methodset_html returns the "> Method set" toggle for a package-level named type. // Its contents are populated from JSON data by client-side JS at load time. func (p *Presentation) methodset_htmlFunc(info *PageInfo, typeName string) string { if p.MethodSetHTML == nil { return "" } index, ok := info.TypeInfoIndex[typeName] if !ok { return "" } var buf bytes.Buffer err := p.MethodSetHTML.Execute(&buf, struct{ Index int }{index}) if err != nil { log.Print(err) } return buf.String() } // callgraph_html returns the "> Call graph" toggle for a package-level func. // Its contents are populated from JSON data by client-side JS at load time. func (p *Presentation) callgraph_htmlFunc(info *PageInfo, recv, name string) string { if p.CallGraphHTML == nil { return "" } if recv != "" { // Format must match (*ssa.Function).RelString(). name = fmt.Sprintf("(%s).%s", recv, name) } index, ok := info.CallGraphIndex[name] if !ok { return "" } var buf bytes.Buffer err := p.CallGraphHTML.Execute(&buf, struct{ Index int }{index}) if err != nil { log.Print(err) } return buf.String() } func noteTitle(note string) string { return strings.Title(strings.ToLower(note)) } func startsWithUppercase(s string) bool { r, _ := utf8.DecodeRuneInString(s) return unicode.IsUpper(r) } var exampleOutputRx = regexp.MustCompile(`(?i)//[[:space:]]*(unordered )?output:`) // stripExampleSuffix strips lowercase braz in Foo_braz or Foo_Bar_braz from name // while keeping uppercase Braz in Foo_Braz. func stripExampleSuffix(name string) string { if i := strings.LastIndex(name, "_"); i != -1 { if i < len(name)-1 && !startsWithUppercase(name[i+1:]) { name = name[:i] } } return name } func splitExampleName(s string) (name, suffix string) { i := strings.LastIndex(s, "_") if 0 <= i && i < len(s)-1 && !startsWithUppercase(s[i+1:]) { name = s[:i] suffix = " (" + strings.Title(s[i+1:]) + ")" return } name = s return } // replaceLeadingIndentation replaces oldIndent at the beginning of each line // with newIndent. This is used for formatting examples. Raw strings that // span multiple lines are handled specially: oldIndent is not removed (since // go/printer will not add any indentation there), but newIndent is added // (since we may still want leading indentation). func replaceLeadingIndentation(body, oldIndent, newIndent string) string { // Handle indent at the beginning of the first line. After this, we handle // indentation only after a newline. var buf bytes.Buffer if strings.HasPrefix(body, oldIndent) { buf.WriteString(newIndent) body = body[len(oldIndent):] } // Use a state machine to keep track of whether we're in a string or // rune literal while we process the rest of the code. const ( codeState = iota runeState interpretedStringState rawStringState ) searchChars := []string{ "'\"`\n", // codeState `\'`, // runeState `\"`, // interpretedStringState "`\n", // rawStringState // newlineState does not need to search } state := codeState for { i := strings.IndexAny(body, searchChars[state]) if i < 0 { buf.WriteString(body) break } c := body[i] buf.WriteString(body[:i+1]) body = body[i+1:] switch state { case codeState: switch c { case '\'': state = runeState case '"': state = interpretedStringState case '`': state = rawStringState case '\n': if strings.HasPrefix(body, oldIndent) { buf.WriteString(newIndent) body = body[len(oldIndent):] } } case runeState: switch c { case '\\': r, size := utf8.DecodeRuneInString(body) buf.WriteRune(r) body = body[size:] case '\'': state = codeState } case interpretedStringState: switch c { case '\\': r, size := utf8.DecodeRuneInString(body) buf.WriteRune(r) body = body[size:] case '"': state = codeState } case rawStringState: switch c { case '`': state = codeState case '\n': buf.WriteString(newIndent) } } } return buf.String() } // writeNode writes the AST node x to w. // // The provided fset must be non-nil. The pageInfo is optional. If // present, the pageInfo is used to add comments to struct fields to // say which version of Go introduced them. func (p *Presentation) writeNode(w io.Writer, pageInfo *PageInfo, fset *token.FileSet, x interface{}) { // convert trailing tabs into spaces using a tconv filter // to ensure a good outcome in most browsers (there may still // be tabs in comments and strings, but converting those into // the right number of spaces is much harder) // // TODO(gri) rethink printer flags - perhaps tconv can be eliminated // with an another printer mode (which is more efficiently // implemented in the printer than here with another layer) var pkgName, structName string var apiInfo pkgAPIVersions if gd, ok := x.(*ast.GenDecl); ok && pageInfo != nil && pageInfo.PDoc != nil && p.Corpus != nil && gd.Tok == token.TYPE && len(gd.Specs) != 0 { pkgName = pageInfo.PDoc.ImportPath if ts, ok := gd.Specs[0].(*ast.TypeSpec); ok { if _, ok := ts.Type.(*ast.StructType); ok { structName = ts.Name.Name } } apiInfo = p.Corpus.pkgAPIInfo[pkgName] } var out = w var buf bytes.Buffer if structName != "" { out = &buf } mode := printer.TabIndent | printer.UseSpaces err := (&printer.Config{Mode: mode, Tabwidth: p.TabWidth}).Fprint(&tconv{p: p, output: out}, fset, x) if err != nil { log.Print(err) } // Add comments to struct fields saying which Go version introduced them. if structName != "" { fieldSince := apiInfo.fieldSince[structName] typeSince := apiInfo.typeSince[structName] // Add/rewrite comments on struct fields to note which Go version added them. var buf2 bytes.Buffer buf2.Grow(buf.Len() + len(" // Added in Go 1.n")*10) bs := bufio.NewScanner(&buf) for bs.Scan() { line := bs.Bytes() field := firstIdent(line) var since string if field != "" { since = fieldSince[field] if since != "" && since == typeSince { // Don't highlight field versions if they were the // same as the struct itself. since = "" } } if since == "" { buf2.Write(line) } else { if bytes.Contains(line, slashSlash) { line = bytes.TrimRight(line, " \t.") buf2.Write(line) buf2.WriteString("; added in Go ") } else { buf2.Write(line) buf2.WriteString(" // Go ") } buf2.WriteString(since) } buf2.WriteByte('\n') } w.Write(buf2.Bytes()) } } var slashSlash = []byte("//") // WriteNode writes x to w. // TODO(bgarcia) Is this method needed? It's just a wrapper for p.writeNode. func (p *Presentation) WriteNode(w io.Writer, fset *token.FileSet, x interface{}) { p.writeNode(w, nil, fset, x) } // firstIdent returns the first identifier in x. // This actually parses "identifiers" that begin with numbers too, but we // never feed it such input, so it's fine. func firstIdent(x []byte) string { x = bytes.TrimSpace(x) i := bytes.IndexFunc(x, func(r rune) bool { return !unicode.IsLetter(r) && !unicode.IsNumber(r) }) if i == -1 { return string(x) } return string(x[:i]) }