diff --git a/go.mod b/go.mod index 8c46b8ab..56e0800a 100644 --- a/go.mod +++ b/go.mod @@ -3,7 +3,7 @@ module github.com/lightstep/saml go 1.17 require ( - github.com/beevik/etree v1.1.0 + github.com/beevik/etree v1.1.4 github.com/crewjam/httperr v0.2.0 github.com/dchest/uniuri v0.0.0-20200228104902-7aecb25e1fe5 github.com/form3tech-oss/jwt-go v3.2.2+incompatible diff --git a/go.sum b/go.sum index 600aad56..164fd703 100644 --- a/go.sum +++ b/go.sum @@ -1,5 +1,6 @@ -github.com/beevik/etree v1.1.0 h1:T0xke/WvNtMoCqgzPhkX2r4rjY3GDZFi+FjpRZY2Jbs= github.com/beevik/etree v1.1.0/go.mod h1:r8Aw8JqVegEf0w2fDnATrX9VpkMcyFeM0FhwO62wh+A= +github.com/beevik/etree v1.1.4 h1:34PFKrJczQ1qXVC4QCqvY0Iz7m3xu89OShTjYRl4Nbk= +github.com/beevik/etree v1.1.4/go.mod h1:aiPf89g/1k3AShMVAzriilpcE4R/Vuor90y83zVZWFc= github.com/creack/pty v1.1.9/go.mod h1:oKZEueFk5CKHvIhNR5MUki03XCEU+Q6VDXinZuGJ33E= github.com/crewjam/httperr v0.2.0 h1:b2BfXR8U3AlIHwNeFFvZ+BV1LFvKLlzMjzaTnZMybNo= github.com/crewjam/httperr v0.2.0/go.mod h1:Jlz+Sg/XqBQhyMjdDiC+GNNRzZTD7x39Gu3pglZ5oH4= diff --git a/vendor/github.com/beevik/etree/.travis.yml b/vendor/github.com/beevik/etree/.travis.yml deleted file mode 100644 index f4cb25d4..00000000 --- a/vendor/github.com/beevik/etree/.travis.yml +++ /dev/null @@ -1,14 +0,0 @@ -language: go -sudo: false - -go: - - 1.11.x - - tip - -matrix: - allow_failures: - - go: tip - -script: - - go vet ./... - - go test -v ./... diff --git a/vendor/github.com/beevik/etree/CONTRIBUTORS b/vendor/github.com/beevik/etree/CONTRIBUTORS index 03211a85..2cf09949 100644 --- a/vendor/github.com/beevik/etree/CONTRIBUTORS +++ b/vendor/github.com/beevik/etree/CONTRIBUTORS @@ -8,3 +8,4 @@ Nicolas Piganeau (npiganeau) Chris Brown (ccbrown) Earncef Sequeira (earncef) Gabriel de Labachelerie (wuzuf) +Martin Dosch (mdosch) diff --git a/vendor/github.com/beevik/etree/LICENSE b/vendor/github.com/beevik/etree/LICENSE index 26f1f775..ef7b286a 100644 --- a/vendor/github.com/beevik/etree/LICENSE +++ b/vendor/github.com/beevik/etree/LICENSE @@ -1,4 +1,4 @@ -Copyright 2015-2019 Brett Vickers. All rights reserved. +Copyright 2015-2023 Brett Vickers. All rights reserved. Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions diff --git a/vendor/github.com/beevik/etree/README.md b/vendor/github.com/beevik/etree/README.md index 08ec26b0..62f0babc 100644 --- a/vendor/github.com/beevik/etree/README.md +++ b/vendor/github.com/beevik/etree/README.md @@ -1,4 +1,3 @@ -[![Build Status](https://travis-ci.org/beevik/etree.svg?branch=master)](https://travis-ci.org/beevik/etree) [![GoDoc](https://godoc.org/github.com/beevik/etree?status.svg)](https://godoc.org/github.com/beevik/etree) etree diff --git a/vendor/github.com/beevik/etree/RELEASE_NOTES.md b/vendor/github.com/beevik/etree/RELEASE_NOTES.md index ee59d7ab..e29bf7cf 100644 --- a/vendor/github.com/beevik/etree/RELEASE_NOTES.md +++ b/vendor/github.com/beevik/etree/RELEASE_NOTES.md @@ -1,3 +1,36 @@ +Release v1.1.4 +============== + +**New Features** + +* Add the ability to preserve whitespace in leaf elements during indent. +* Add the ability to suppress a document-trailing newline during indent. +* Add choice of XML attribute quoting style (single-quote or double-quote). + +**Removed Features** + +* Removed the CDATA preservation change introduced in v1.1.3. It was + implemented in a way that broke the ability to process XML documents + encoded using non-UTF8 character sets. + +Release v1.1.3 +============== + +* XML reads now preserve CDATA sections instead of converting them to + standard character data. + +Release v1.1.2 +============== + +* Fixed a path parsing bug. +* The `Element.Text` function now handles comments embedded between + character data spans. + +Release v1.1.1 +============== + +* Updated go version in `go.mod` to 1.20 + Release v1.1.0 ============== diff --git a/vendor/github.com/beevik/etree/etree.go b/vendor/github.com/beevik/etree/etree.go index 9e24f901..ac9dc74e 100644 --- a/vendor/github.com/beevik/etree/etree.go +++ b/vendor/github.com/beevik/etree/etree.go @@ -18,14 +18,15 @@ import ( ) const ( - // NoIndent is used with Indent to disable all indenting. + // NoIndent is used with the IndentSettings record to remove all + // indenting. NoIndent = -1 ) // ErrXML is returned when XML parsing fails due to incorrect formatting. var ErrXML = errors.New("etree: invalid XML format") -// ReadSettings allow for changing the default behavior of the ReadFrom* +// ReadSettings determine the default behavior of the Document's ReadFrom* // methods. type ReadSettings struct { // CharsetReader to be passed to standard xml.Decoder. Default: nil. @@ -49,8 +50,23 @@ func newReadSettings() ReadSettings { } } -// WriteSettings allow for changing the serialization behavior of the WriteTo* -// methods. +// dup creates a duplicate of the ReadSettings object. +func (s *ReadSettings) dup() ReadSettings { + var entityCopy map[string]string + if s.Entity != nil { + entityCopy = make(map[string]string) + for k, v := range s.Entity { + entityCopy[k] = v + } + } + return ReadSettings{ + CharsetReader: s.CharsetReader, + Permissive: s.Permissive, + Entity: entityCopy, + } +} + +// WriteSettings determine the behavior of the Document's WriteTo* methods. type WriteSettings struct { // CanonicalEndTags forces the production of XML end tags, even for // elements that have no child elements. Default: false. @@ -66,9 +82,16 @@ type WriteSettings struct { // references are also produced for > and '. Default: false. CanonicalAttrVal bool - // When outputting indented XML, use a carriage return and linefeed - // ("\r\n") as a new-line delimiter instead of just a linefeed ("\n"). - // This is useful on Windows-based systems. + // AttrSingleQuote causes attributes to use single quotes (attr='example') + // instead of double quotes (attr = "example") when set to true. Default: + // false. + AttrSingleQuote bool + + // UseCRLF causes the document's Indent* methods to use a carriage return + // followed by a linefeed ("\r\n") when outputting a newline. If false, + // only a linefeed is used ("\n"). Default: false. + // + // Deprecated: UseCRLF is deprecated. Use IndentSettings.UseCRLF instead. UseCRLF bool } @@ -78,12 +101,56 @@ func newWriteSettings() WriteSettings { CanonicalEndTags: false, CanonicalText: false, CanonicalAttrVal: false, + AttrSingleQuote: false, UseCRLF: false, } } -// A Token is an empty interface that represents an Element, CharData, -// Comment, Directive, or ProcInst. +// dup creates a duplicate of the WriteSettings object. +func (s *WriteSettings) dup() WriteSettings { + return *s +} + +// IndentSettings determine the behavior of the Document's Indent* methods. +type IndentSettings struct { + // Spaces indicates the number of spaces to insert for each level of + // indentation. Set to etree.NoIndent to remove all indentation. Ignored + // when UseTabs is true. Default: 4. + Spaces int + + // UseTabs causes tabs to be used instead of spaces when indenting. + // Default: false. + UseTabs bool + + // UseCRLF causes newlines to be written as a carriage return followed by + // a linefeed ("\r\n"). If false, only a linefeed character is output + // for a newline ("\n"). Default: false. + UseCRLF bool + + // PreserveLeafWhitespace causes indent methods to preserve whitespace + // within XML elements containing only non-CDATA character data. Default: + // false. + PreserveLeafWhitespace bool + + // SuppressTrailingNewline suppresses the generation of a trailing newline + // character at the end of the indented document. Default: false. + SuppressTrailingNewline bool +} + +// NewIndentSettings creates a default IndentSettings record. +func NewIndentSettings() IndentSettings { + return IndentSettings{ + Spaces: 4, + UseTabs: false, + UseCRLF: false, + PreserveLeafWhitespace: false, + SuppressTrailingNewline: false, + } +} + +// A Token is an interface type used to represent XML elements, character +// data, CDATA sections, XML comments, XML directives, and XML processing +// instructions. type Token interface { Parent() *Element Index() int @@ -93,10 +160,16 @@ type Token interface { writeTo(w *bufio.Writer, s *WriteSettings) } -// A Document is a container holding a complete XML hierarchy. Its embedded -// element contains zero or more children, one of which is usually the root -// element. The embedded element may include other children such as -// processing instructions or BOM CharData tokens. +// A Document is a container holding a complete XML tree. +// +// A document has a single embedded element, which contains zero or more child +// tokens, one of which is usually the root element. The embedded element may +// include other children such as processing instruction tokens or character +// data tokens. The document's embedded element is never directly serialized; +// only its children are. +// +// A document also contains read and write settings, which influence the way +// the document is deserialized, serialized, and indented. type Document struct { Element ReadSettings ReadSettings @@ -112,7 +185,7 @@ type Element struct { index int // token index in parent's children } -// An Attr represents a key-value attribute of an XML element. +// An Attr represents a key-value attribute within an XML element. type Attr struct { Space, Key string // The attribute's namespace prefix and key Value string // The attribute value string @@ -123,17 +196,18 @@ type Attr struct { type charDataFlags uint8 const ( - // The CharData was created by an indent function as whitespace. + // The CharData contains only whitespace. whitespaceFlag charDataFlags = 1 << iota // The CharData contains a CDATA section. cdataFlag ) -// CharData can be used to represent character data or a CDATA section within -// an XML document. +// CharData may be used to represent simple text data or a CDATA section +// within an XML document. The Data property should never be modified +// directly; use the SetData method instead. type CharData struct { - Data string + Data string // the simple text or CDATA section content parent *Element index int flags charDataFlags @@ -141,22 +215,22 @@ type CharData struct { // A Comment represents an XML comment. type Comment struct { - Data string + Data string // the comment's text parent *Element index int } // A Directive represents an XML directive. type Directive struct { - Data string + Data string // the directive string parent *Element index int } // A ProcInst represents an XML processing instruction. type ProcInst struct { - Target string - Inst string + Target string // the processing instruction target + Inst string // the processing instruction value parent *Element index int } @@ -164,19 +238,32 @@ type ProcInst struct { // NewDocument creates an XML document without a root element. func NewDocument() *Document { return &Document{ - Element{Child: make([]Token, 0)}, - newReadSettings(), - newWriteSettings(), + Element: Element{Child: make([]Token, 0)}, + ReadSettings: newReadSettings(), + WriteSettings: newWriteSettings(), } } +// NewDocumentWithRoot creates an XML document and sets the element 'e' as its +// root element. If the element 'e' is already part of another document, it is +// first removed from its existing document. +func NewDocumentWithRoot(e *Element) *Document { + d := NewDocument() + d.SetRoot(e) + return d +} + // Copy returns a recursive, deep copy of the document. func (d *Document) Copy() *Document { - return &Document{*(d.dup(nil).(*Element)), d.ReadSettings, d.WriteSettings} + return &Document{ + Element: *(d.Element.dup(nil).(*Element)), + ReadSettings: d.ReadSettings.dup(), + WriteSettings: d.WriteSettings.dup(), + } } -// Root returns the root element of the document, or nil if there is no root -// element. +// Root returns the root element of the document. It returns nil if there is +// no root element. func (d *Document) Root() *Element { for _, t := range d.Child { if c, ok := t.(*Element); ok { @@ -186,25 +273,23 @@ func (d *Document) Root() *Element { return nil } -// SetRoot replaces the document's root element with e. If the document -// already has a root when this function is called, then the document's -// original root is unbound first. If the element e is bound to another -// document (or to another element within a document), then it is unbound -// first. +// SetRoot replaces the document's root element with the element 'e'. If the +// document already has a root element when this function is called, then the +// existing root element is unbound from the document. If the element 'e' is +// part of another document, then it is unbound from the other document. func (d *Document) SetRoot(e *Element) { if e.parent != nil { e.parent.RemoveChild(e) } - p := &d.Element - e.setParent(p) - // If there is already a root element, replace it. + p := &d.Element for i, t := range p.Child { if _, ok := t.(*Element); ok { t.setParent(nil) t.setIndex(-1) p.Child[i] = e + e.setParent(p) e.setIndex(i) return } @@ -214,15 +299,16 @@ func (d *Document) SetRoot(e *Element) { p.addChild(e) } -// ReadFrom reads XML from the reader r into the document d. It returns the -// number of bytes read and any error encountered. +// ReadFrom reads XML from the reader 'r' into this document. The function +// returns the number of bytes read and any error encountered. func (d *Document) ReadFrom(r io.Reader) (n int64, err error) { return d.Element.readFrom(r, d.ReadSettings) } -// ReadFromFile reads XML from the string s into the document d. -func (d *Document) ReadFromFile(filename string) error { - f, err := os.Open(filename) +// ReadFromFile reads XML from a local file at path 'filepath' into this +// document. +func (d *Document) ReadFromFile(filepath string) error { + f, err := os.Open(filepath) if err != nil { return err } @@ -231,34 +317,33 @@ func (d *Document) ReadFromFile(filename string) error { return err } -// ReadFromBytes reads XML from the byte slice b into the document d. +// ReadFromBytes reads XML from the byte slice 'b' into the this document. func (d *Document) ReadFromBytes(b []byte) error { _, err := d.ReadFrom(bytes.NewReader(b)) return err } -// ReadFromString reads XML from the string s into the document d. +// ReadFromString reads XML from the string 's' into this document. func (d *Document) ReadFromString(s string) error { _, err := d.ReadFrom(strings.NewReader(s)) return err } -// WriteTo serializes an XML document into the writer w. It -// returns the number of bytes written and any error encountered. +// WriteTo serializes the document out to the writer 'w'. The function returns +// the number of bytes written and any error encountered. func (d *Document) WriteTo(w io.Writer) (n int64, err error) { - cw := newCountWriter(w) - b := bufio.NewWriter(cw) + xw := newXmlWriter(w) + b := bufio.NewWriter(xw) for _, c := range d.Child { c.writeTo(b, &d.WriteSettings) } - err, n = b.Flush(), cw.bytes + err, n = b.Flush(), xw.bytes return } -// WriteToFile serializes an XML document into the file named -// filename. -func (d *Document) WriteToFile(filename string) error { - f, err := os.Create(filename) +// WriteToFile serializes the document out to the file at path 'filepath'. +func (d *Document) WriteToFile(filepath string) error { + f, err := os.Create(filepath) if err != nil { return err } @@ -267,8 +352,7 @@ func (d *Document) WriteToFile(filename string) error { return err } -// WriteToBytes serializes the XML document into a slice of -// bytes. +// WriteToBytes serializes this document into a slice of bytes. func (d *Document) WriteToBytes() (b []byte, err error) { var buf bytes.Buffer if _, err = d.WriteTo(&buf); err != nil { @@ -277,7 +361,7 @@ func (d *Document) WriteToBytes() (b []byte, err error) { return buf.Bytes(), nil } -// WriteToString serializes the XML document into a string. +// WriteToString serializes this document into a string. func (d *Document) WriteToString() (s string, err error) { var b []byte if b, err = d.WriteToBytes(); err != nil { @@ -289,38 +373,73 @@ func (d *Document) WriteToString() (s string, err error) { type indentFunc func(depth int) string // Indent modifies the document's element tree by inserting character data -// tokens containing newlines and indentation. The amount of indentation per -// depth level is given as spaces. Pass etree.NoIndent for spaces if you want -// no indentation at all. +// tokens containing newlines and spaces for indentation. The amount of +// indentation per depth level is given by the 'spaces' parameter. Other than +// the number of spaces, default IndentSettings are used. func (d *Document) Indent(spaces int) { - var indent indentFunc - switch { - case spaces < 0: - indent = func(depth int) string { return "" } - case d.WriteSettings.UseCRLF == true: - indent = func(depth int) string { return indentCRLF(depth*spaces, indentSpaces) } - default: - indent = func(depth int) string { return indentLF(depth*spaces, indentSpaces) } - } - d.Element.indent(0, indent) + s := NewIndentSettings() + s.Spaces = spaces + d.IndentWithSettings(s) } // IndentTabs modifies the document's element tree by inserting CharData -// tokens containing newlines and tabs for indentation. One tab is used per -// indentation level. +// tokens containing newlines and tabs for indentation. One tab is used per +// indentation level. Other than the use of tabs, default IndentSettings +// are used. func (d *Document) IndentTabs() { + s := NewIndentSettings() + s.UseTabs = true + d.IndentWithSettings(s) +} + +// IndentWithSettings modifies the document's element tree by inserting +// character data tokens containing newlines and indentation. The behavior +// of the indentation algorithm is configured by the indent settings. +func (d *Document) IndentWithSettings(s IndentSettings) { + // WriteSettings.UseCRLF is deprecated. Until removed from the package, it + // overrides IndentSettings.UseCRLF when true. + if d.WriteSettings.UseCRLF { + s.UseCRLF = true + } + var indent indentFunc - switch d.WriteSettings.UseCRLF { - case true: - indent = func(depth int) string { return indentCRLF(depth, indentTabs) } - default: - indent = func(depth int) string { return indentLF(depth, indentTabs) } + if s.UseTabs { + if s.UseCRLF { + indent = func(depth int) string { return indentCRLF(depth, indentTabs) } + } else { + indent = func(depth int) string { return indentLF(depth, indentTabs) } + } + } else { + if s.Spaces < 0 { + indent = func(depth int) string { return "" } + } else if s.UseCRLF { + indent = func(depth int) string { return indentCRLF(depth*s.Spaces, indentSpaces) } + } else { + indent = func(depth int) string { return indentLF(depth*s.Spaces, indentSpaces) } + } } - d.Element.indent(0, indent) + + d.Element.indent(0, indent, &s) + + if s.SuppressTrailingNewline && len(d.Element.Child) > 0 { + n := len(d.Element.Child) - 1 + if cd, ok := d.Element.Child[n].(*CharData); ok && (cd.flags&whitespaceFlag) != 0 { + d.Element.Child = d.Element.Child[:n] + } + } +} + +// Unindent modifies the document's element tree by removing character data +// tokens containing only whitespace. Other than the removal of indentation, +// default IndentSettings are used. +func (d *Document) Unindent() { + s := NewIndentSettings() + s.Spaces = NoIndent + d.IndentWithSettings(s) } -// NewElement creates an unparented element with the specified tag. The tag -// may be prefixed by a namespace prefix and a colon. +// NewElement creates an unparented element with the specified tag (i.e., +// name). The tag may include a namespace prefix followed by a colon. func NewElement(tag string) *Element { space, stag := spaceDecompose(tag) return newElement(space, stag, nil) @@ -345,7 +464,8 @@ func newElement(space, tag string, parent *Element) *Element { // Copy creates a recursive, deep copy of the element and all its attributes // and children. The returned element has no parent but can be parented to a -// another element using AddElement, or to a document using SetRoot. +// another element using AddChild, or added to a document with SetRoot or +// NewDocumentWithRoot. func (e *Element) Copy() *Element { return e.dup(nil).(*Element) } @@ -400,16 +520,6 @@ func (e *Element) findDefaultNamespaceURI() string { return e.parent.findDefaultNamespaceURI() } -// hasText returns true if the element has character data immediately -// folllowing the element's opening tag. -func (e *Element) hasText() bool { - if len(e.Child) == 0 { - return false - } - _, ok := e.Child[0].(*CharData) - return ok -} - // namespacePrefix returns the namespace prefix associated with the element. func (e *Element) namespacePrefix() string { return e.Space @@ -433,8 +543,10 @@ func (e *Element) Text() string { if text == "" { text = cd.Data } else { - text = text + cd.Data + text += cd.Data } + } else if _, ok := ch.(*Comment); ok { + // ignore } else { break } @@ -470,7 +582,7 @@ func (e *Element) Tail() string { if text == "" { text = cd.Data } else { - text = text + cd.Data + text += cd.Data } } else { break @@ -548,30 +660,30 @@ func (e *Element) findTermCharDataIndex(start int) int { return len(e.Child) } -// CreateElement creates an element with the specified tag and adds it as the -// last child element of the element e. The tag may be prefixed by a namespace -// prefix and a colon. +// CreateElement creates a new element with the specified tag (i.e., name) and +// adds it as the last child token of this element. The tag may include a +// prefix followed by a colon. func (e *Element) CreateElement(tag string) *Element { space, stag := spaceDecompose(tag) return newElement(space, stag, e) } -// AddChild adds the token t as the last child of element e. If token t was -// already the child of another element, it is first removed from its current +// AddChild adds the token 't' as the last child of the element. If token 't' +// was already the child of another element, it is first removed from its // parent element. func (e *Element) AddChild(t Token) { if t.Parent() != nil { t.Parent().RemoveChild(t) } - - t.setParent(e) e.addChild(t) } -// InsertChild inserts the token t before e's existing child token ex. If ex -// is nil or ex is not a child of e, then t is added to the end of e's child -// token list. If token t was already the child of another element, it is -// first removed from its current parent element. +// InsertChild inserts the token 't' into this element's list of children just +// before the element's existing child token 'ex'. If the existing element +// 'ex' does not appear in this element's list of child tokens, then 't' is +// added to the end of this element's list of child tokens. If token 't' is +// already the child of another element, it is first removed from the other +// element's list of child tokens. // // Deprecated: InsertChild is deprecated. Use InsertChildAt instead. func (e *Element) InsertChild(ex Token, t Token) { @@ -596,10 +708,10 @@ func (e *Element) InsertChild(ex Token, t Token) { } } -// InsertChildAt inserts the token t into the element e's list of child tokens -// just before the requested index. If the index is greater than or equal to -// the length of the list of child tokens, the token t is added to the end of -// the list. +// InsertChildAt inserts the token 't' into this element's list of child +// tokens just before the requested 'index'. If the index is greater than or +// equal to the length of the list of child tokens, then the token 't' is +// added to the end of the list of child tokens. func (e *Element) InsertChildAt(index int, t Token) { if index >= len(e.Child) { e.AddChild(t) @@ -624,9 +736,9 @@ func (e *Element) InsertChildAt(index int, t Token) { } } -// RemoveChild attempts to remove the token t from element e's list of -// children. If the token t is a child of e, then it is returned. Otherwise, -// nil is returned. +// RemoveChild attempts to remove the token 't' from this element's list of +// child tokens. If the token 't' was a child of this element, then it is +// removed and returned. Otherwise, nil is returned. func (e *Element) RemoveChild(t Token) Token { if t.Parent() != e { return nil @@ -634,9 +746,9 @@ func (e *Element) RemoveChild(t Token) Token { return e.RemoveChildAt(t.Index()) } -// RemoveChildAt removes the index-th child token from the element e. The -// removed child token is returned. If the index is out of bounds, no child is -// removed and nil is returned. +// RemoveChildAt removes the child token appearing in slot 'index' of this +// element's list of child tokens. The removed child token is then returned. +// If the index is out of bounds, no child is removed and nil is returned. func (e *Element) RemoveChildAt(index int) Token { if index >= len(e.Child) { return nil @@ -652,11 +764,11 @@ func (e *Element) RemoveChildAt(index int) Token { return t } -// ReadFrom reads XML from the reader r and stores the result as a new child -// of element e. +// ReadFrom reads XML from the reader 'ri' and stores the result as a new +// child of this element. func (e *Element) readFrom(ri io.Reader, settings ReadSettings) (n int64, err error) { - r := newCountReader(ri) - dec := xml.NewDecoder(r) + xr := newXmlReader(ri) + dec := xml.NewDecoder(xr) dec.CharsetReader = settings.CharsetReader dec.Strict = !settings.Permissive dec.Entity = settings.Entity @@ -666,11 +778,14 @@ func (e *Element) readFrom(ri io.Reader, settings ReadSettings) (n int64, err er t, err := dec.RawToken() switch { case err == io.EOF: - return r.bytes, nil + if len(stack.data) != 1 { + return xr.bytes, ErrXML + } + return xr.bytes, nil case err != nil: - return r.bytes, err + return xr.bytes, err case stack.empty(): - return r.bytes, ErrXML + return xr.bytes, ErrXML } top := stack.peek().(*Element) @@ -683,6 +798,9 @@ func (e *Element) readFrom(ri io.Reader, settings ReadSettings) (n int64, err er } stack.push(e) case xml.EndElement: + if top.Tag != t.Name.Local || top.Space != t.Name.Space { + return xr.bytes, ErrXML + } stack.pop() case xml.CharData: data := string(t) @@ -701,9 +819,10 @@ func (e *Element) readFrom(ri io.Reader, settings ReadSettings) (n int64, err er } } -// SelectAttr finds an element attribute matching the requested key and -// returns it if found. Returns nil if no matching attribute is found. The key -// may be prefixed by a namespace prefix and a colon. +// SelectAttr finds an element attribute matching the requested 'key' and, if +// found, returns a pointer to the matching attribute. The function returns +// nil if no matching attribute is found. The key may include a namespace +// prefix followed by a colon. func (e *Element) SelectAttr(key string) *Attr { space, skey := spaceDecompose(key) for i, a := range e.Attr { @@ -714,9 +833,10 @@ func (e *Element) SelectAttr(key string) *Attr { return nil } -// SelectAttrValue finds an element attribute matching the requested key and -// returns its value if found. The key may be prefixed by a namespace prefix -// and a colon. If the key is not found, the dflt value is returned instead. +// SelectAttrValue finds an element attribute matching the requested 'key' and +// returns its value if found. If no matching attribute is found, the function +// returns the 'dflt' value instead. The key may include a namespace prefix +// followed by a colon. func (e *Element) SelectAttrValue(key, dflt string) string { space, skey := spaceDecompose(key) for _, a := range e.Attr { @@ -727,7 +847,7 @@ func (e *Element) SelectAttrValue(key, dflt string) string { return dflt } -// ChildElements returns all elements that are children of element e. +// ChildElements returns all elements that are children of this element. func (e *Element) ChildElements() []*Element { var elements []*Element for _, t := range e.Child { @@ -738,9 +858,9 @@ func (e *Element) ChildElements() []*Element { return elements } -// SelectElement returns the first child element with the given tag. The tag -// may be prefixed by a namespace prefix and a colon. Returns nil if no -// element with a matching tag was found. +// SelectElement returns the first child element with the given 'tag' (i.e., +// name). The function returns nil if no child element matching the tag is +// found. The tag may include a namespace prefix followed by a colon. func (e *Element) SelectElement(tag string) *Element { space, stag := spaceDecompose(tag) for _, t := range e.Child { @@ -751,8 +871,8 @@ func (e *Element) SelectElement(tag string) *Element { return nil } -// SelectElements returns a slice of all child elements with the given tag. -// The tag may be prefixed by a namespace prefix and a colon. +// SelectElements returns a slice of all child elements with the given 'tag' +// (i.e., name). The tag may include a namespace prefix followed by a colon. func (e *Element) SelectElements(tag string) []*Element { space, stag := spaceDecompose(tag) var elements []*Element @@ -764,39 +884,39 @@ func (e *Element) SelectElements(tag string) []*Element { return elements } -// FindElement returns the first element matched by the XPath-like path -// string. Returns nil if no element is found using the path. Panics if an -// invalid path string is supplied. +// FindElement returns the first element matched by the XPath-like 'path' +// string. The function returns nil if no child element is found using the +// path. It panics if an invalid path string is supplied. func (e *Element) FindElement(path string) *Element { return e.FindElementPath(MustCompilePath(path)) } -// FindElementPath returns the first element matched by the XPath-like path -// string. Returns nil if no element is found using the path. +// FindElementPath returns the first element matched by the 'path' object. The +// function returns nil if no element is found using the path. func (e *Element) FindElementPath(path Path) *Element { p := newPather() elements := p.traverse(e, path) - switch { - case len(elements) > 0: + if len(elements) > 0 { return elements[0] - default: - return nil } + return nil } -// FindElements returns a slice of elements matched by the XPath-like path -// string. Panics if an invalid path string is supplied. +// FindElements returns a slice of elements matched by the XPath-like 'path' +// string. The function returns nil if no child element is found using the +// path. It panics if an invalid path string is supplied. func (e *Element) FindElements(path string) []*Element { return e.FindElementsPath(MustCompilePath(path)) } -// FindElementsPath returns a slice of elements matched by the Path object. +// FindElementsPath returns a slice of elements matched by the 'path' object. func (e *Element) FindElementsPath(path Path) []*Element { p := newPather() return p.traverse(e, path) } -// GetPath returns the absolute path of the element. +// GetPath returns the absolute path of the element. The absolute path is the +// full path from the document's root. func (e *Element) GetPath() string { path := []string{} for seg := e; seg != nil; seg = seg.Parent() { @@ -813,9 +933,9 @@ func (e *Element) GetPath() string { return "/" + strings.Join(path, "/") } -// GetRelativePath returns the path of the element relative to the source +// GetRelativePath returns the path of this element relative to the 'source' // element. If the two elements are not part of the same element tree, then -// GetRelativePath returns the empty string. +// the function returns the empty string. func (e *Element) GetRelativePath(source *Element) string { var path []*Element @@ -884,10 +1004,10 @@ func (e *Element) GetRelativePath(source *Element) string { return strings.Join(parts, "/") } -// indent recursively inserts proper indentation between an -// XML element's child tokens. -func (e *Element) indent(depth int, indent indentFunc) { - e.stripIndent() +// indent recursively inserts proper indentation between an XML element's +// child tokens. +func (e *Element) indent(depth int, indent indentFunc, s *IndentSettings) { + e.stripIndent(s) n := len(e.Child) if n == 0 { return @@ -915,7 +1035,7 @@ func (e *Element) indent(depth int, indent indentFunc) { // Recursively process child elements. if ce, ok := c.(*Element); ok { - ce.indent(depth+1, indent) + ce.indent(depth+1, indent, s) } } @@ -931,7 +1051,7 @@ func (e *Element) indent(depth int, indent indentFunc) { } // stripIndent removes any previously inserted indentation. -func (e *Element) stripIndent() { +func (e *Element) stripIndent(s *IndentSettings) { // Count the number of non-indent child tokens n := len(e.Child) for _, c := range e.Child { @@ -942,6 +1062,9 @@ func (e *Element) stripIndent() { if n == len(e.Child) { return } + if n == 0 && len(e.Child) == 1 && s.PreserveLeafWhitespace { + return + } // Strip out indent CharData newChild := make([]Token, n) @@ -970,31 +1093,29 @@ func (e *Element) dup(parent *Element) Token { for i, t := range e.Child { ne.Child[i] = t.dup(ne) } - for i, a := range e.Attr { - ne.Attr[i] = a - } + copy(ne.Attr, e.Attr) return ne } -// Parent returns the element token's parent element, or nil if it has no -// parent. +// Parent returns this element's parent element. It returns nil if this +// element has no parent. func (e *Element) Parent() *Element { return e.parent } // Index returns the index of this element within its parent element's -// list of child tokens. If this element has no parent element, the index -// is -1. +// list of child tokens. If this element has no parent, then the function +// returns -1. func (e *Element) Index() int { return e.index } -// setParent replaces the element token's parent. +// setParent replaces this element token's parent. func (e *Element) setParent(parent *Element) { e.parent = parent } -// setIndex sets the element token's index within its parent's Child slice. +// setIndex sets this element token's index within its parent's Child slice. func (e *Element) setIndex(index int) { e.index = index } @@ -1008,7 +1129,7 @@ func (e *Element) writeTo(w *bufio.Writer, s *WriteSettings) { a.writeTo(w, s) } if len(e.Child) > 0 { - w.WriteString(">") + w.WriteByte('>') for _, c := range e.Child { c.writeTo(w, s) } @@ -1028,13 +1149,15 @@ func (e *Element) writeTo(w *bufio.Writer, s *WriteSettings) { // addChild adds a child token to the element e. func (e *Element) addChild(t Token) { + t.setParent(e) t.setIndex(len(e.Child)) e.Child = append(e.Child, t) } -// CreateAttr creates an attribute and adds it to element e. The key may be -// prefixed by a namespace prefix and a colon. If an attribute with the key -// already exists, its value is replaced. +// CreateAttr creates an attribute with the specified 'key' and 'value' and +// adds it to this element. If an attribute with same key already exists on +// this element, then its value is replaced. The key may include a namespace +// prefix followed by a colon. func (e *Element) CreateAttr(key, value string) *Attr { space, skey := spaceDecompose(key) return e.createAttr(space, skey, value, e) @@ -1058,10 +1181,10 @@ func (e *Element) createAttr(space, key, value string, parent *Element) *Attr { return &e.Attr[len(e.Attr)-1] } -// RemoveAttr removes and returns a copy of the first attribute of the element -// whose key matches the given key. The key may be prefixed by a namespace -// prefix and a colon. If a matching attribute does not exist, nil is -// returned. +// RemoveAttr removes the first attribute of this element whose key matches +// 'key'. It returns a copy of the removed attribute if a match is found. If +// no match is found, it returns nil. The key may include a namespace prefix +// followed by a colon. func (e *Element) RemoveAttr(key string) *Attr { space, skey := spaceDecompose(key) for i, a := range e.Attr { @@ -1078,7 +1201,7 @@ func (e *Element) RemoveAttr(key string) *Attr { return nil } -// SortAttrs sorts the element's attributes lexicographically by key. +// SortAttrs sorts this element's attributes lexicographically by key. func (e *Element) SortAttrs() { sort.Sort(byAttr(e.Attr)) } @@ -1101,7 +1224,7 @@ func (a byAttr) Less(i, j int) bool { return sp < 0 } -// FullKey returns the attribute a's complete key, including namespace prefix +// FullKey returns this attribute's complete key, including namespace prefix // if present. func (a *Attr) FullKey() string { if a.Space == "" { @@ -1110,22 +1233,29 @@ func (a *Attr) FullKey() string { return a.Space + ":" + a.Key } -// Element returns the element containing the attribute. +// Element returns a pointer to the element containing this attribute. func (a *Attr) Element() *Element { return a.element } -// NamespaceURI returns the XML namespace URI associated with the attribute. -// If the element is part of the XML default namespace, NamespaceURI returns -// the empty string. +// NamespaceURI returns the XML namespace URI associated with this attribute. +// The function returns the empty string if the attribute is unprefixed or +// if the attribute is part of the XML default namespace. func (a *Attr) NamespaceURI() string { - return a.element.NamespaceURI() + if a.Space == "" { + return "" + } + return a.element.findLocalNamespaceURI(a.Space) } // writeTo serializes the attribute to the writer. func (a *Attr) writeTo(w *bufio.Writer, s *WriteSettings) { w.WriteString(a.FullKey()) - w.WriteString(`="`) + if s.AttrSingleQuote { + w.WriteString(`='`) + } else { + w.WriteString(`="`) + } var m escapeMode if s.CanonicalAttrVal { m = escapeCanonicalAttr @@ -1133,20 +1263,26 @@ func (a *Attr) writeTo(w *bufio.Writer, s *WriteSettings) { m = escapeNormal } escapeString(w, a.Value, m) - w.WriteByte('"') + if s.AttrSingleQuote { + w.WriteByte('\'') + } else { + w.WriteByte('"') + } } -// NewText creates a parentless CharData token containing character data. +// NewText creates an unparented CharData token containing simple text data. func NewText(text string) *CharData { return newCharData(text, 0, nil) } -// NewCData creates a parentless XML character CDATA section. +// NewCData creates an unparented XML character CDATA section with 'data' as +// its content. func NewCData(data string) *CharData { return newCharData(data, cdataFlag, nil) } -// NewCharData creates a parentless CharData token containing character data. +// NewCharData creates an unparented CharData token containing simple text +// data. // // Deprecated: NewCharData is deprecated. Instead, use NewText, which does the // same thing. @@ -1159,7 +1295,7 @@ func NewCharData(data string) *CharData { func newCharData(data string, flags charDataFlags, parent *Element) *CharData { c := &CharData{ Data: data, - parent: parent, + parent: nil, index: -1, flags: flags, } @@ -1169,62 +1305,75 @@ func newCharData(data string, flags charDataFlags, parent *Element) *CharData { return c } -// CreateText creates a CharData token containing character data and adds it -// as a child of element e. +// CreateText creates a CharData token containing simple text data and adds it +// to the end of this element's list of child tokens. func (e *Element) CreateText(text string) *CharData { return newCharData(text, 0, e) } -// CreateCData creates a CharData token containing a CDATA section and adds it -// as a child of element e. +// CreateCData creates a CharData token containing a CDATA section with 'data' +// as its content and adds it to the end of this element's list of child +// tokens. func (e *Element) CreateCData(data string) *CharData { return newCharData(data, cdataFlag, e) } -// CreateCharData creates a CharData token containing character data and adds -// it as a child of element e. +// CreateCharData creates a CharData token containing simple text data and +// adds it to the end of this element's list of child tokens. // // Deprecated: CreateCharData is deprecated. Instead, use CreateText, which // does the same thing. func (e *Element) CreateCharData(data string) *CharData { - return newCharData(data, 0, e) + return e.CreateText(data) } -// dup duplicates the character data. -func (c *CharData) dup(parent *Element) Token { - return &CharData{ - Data: c.Data, - flags: c.flags, - parent: parent, - index: c.index, +// SetData modifies the content of the CharData token. In the case of a +// CharData token containing simple text, the simple text is modified. In the +// case of a CharData token containing a CDATA section, the CDATA section's +// content is modified. +func (c *CharData) SetData(text string) { + c.Data = text + if isWhitespace(text) { + c.flags |= whitespaceFlag + } else { + c.flags &= ^whitespaceFlag } } -// IsCData returns true if the character data token is to be encoded as a -// CDATA section. +// IsCData returns true if this CharData token is contains a CDATA section. It +// returns false if the CharData token contains simple text. func (c *CharData) IsCData() bool { return (c.flags & cdataFlag) != 0 } -// IsWhitespace returns true if the character data token was created by one of -// the document Indent methods to contain only whitespace. +// IsWhitespace returns true if this CharData token contains only whitespace. func (c *CharData) IsWhitespace() bool { return (c.flags & whitespaceFlag) != 0 } -// Parent returns the character data token's parent element, or nil if it has -// no parent. +// Parent returns this CharData token's parent element, or nil if it has no +// parent. func (c *CharData) Parent() *Element { return c.parent } // Index returns the index of this CharData token within its parent element's -// list of child tokens. If this CharData token has no parent element, the -// index is -1. +// list of child tokens. If this CharData token has no parent, then the +// function returns -1. func (c *CharData) Index() int { return c.index } +// dup duplicates the character data. +func (c *CharData) dup(parent *Element) Token { + return &CharData{ + Data: c.Data, + flags: c.flags, + parent: parent, + index: c.index, + } +} + // setParent replaces the character data token's parent. func (c *CharData) setParent(parent *Element) { c.parent = parent @@ -1253,17 +1402,16 @@ func (c *CharData) writeTo(w *bufio.Writer, s *WriteSettings) { } } -// NewComment creates a parentless XML comment. +// NewComment creates an unparented comment token. func NewComment(comment string) *Comment { return newComment(comment, nil) } -// NewComment creates an XML comment and binds it to a parent element. If -// parent is nil, the Comment remains unbound. +// NewComment creates a comment token and sets its parent element to 'parent'. func newComment(comment string, parent *Element) *Comment { c := &Comment{ Data: comment, - parent: parent, + parent: nil, index: -1, } if parent != nil { @@ -1272,7 +1420,8 @@ func newComment(comment string, parent *Element) *Comment { return c } -// CreateComment creates an XML comment and adds it as a child of element e. +// CreateComment creates a comment token using the specified 'comment' string +// and adds it as the last child token of this element. func (e *Element) CreateComment(comment string) *Comment { return newComment(comment, e) } @@ -1292,8 +1441,8 @@ func (c *Comment) Parent() *Element { } // Index returns the index of this Comment token within its parent element's -// list of child tokens. If this Comment token has no parent element, the -// index is -1. +// list of child tokens. If this Comment token has no parent, then the +// function returns -1. func (c *Comment) Index() int { return c.index } @@ -1316,7 +1465,7 @@ func (c *Comment) writeTo(w *bufio.Writer, s *WriteSettings) { w.WriteString("-->") } -// NewDirective creates a parentless XML directive. +// NewDirective creates an unparented XML directive token. func NewDirective(data string) *Directive { return newDirective(data, nil) } @@ -1326,7 +1475,7 @@ func NewDirective(data string) *Directive { func newDirective(data string, parent *Element) *Directive { d := &Directive{ Data: data, - parent: parent, + parent: nil, index: -1, } if parent != nil { @@ -1335,8 +1484,8 @@ func newDirective(data string, parent *Element) *Directive { return d } -// CreateDirective creates an XML directive and adds it as the last child of -// element e. +// CreateDirective creates an XML directive token with the specified 'data' +// value and adds it as the last child token of this element. func (e *Element) CreateDirective(data string) *Directive { return newDirective(data, e) } @@ -1357,8 +1506,8 @@ func (d *Directive) Parent() *Element { } // Index returns the index of this Directive token within its parent element's -// list of child tokens. If this Directive token has no parent element, the -// index is -1. +// list of child tokens. If this Directive token has no parent, then the +// function returns -1. func (d *Directive) Index() int { return d.index } @@ -1381,7 +1530,7 @@ func (d *Directive) writeTo(w *bufio.Writer, s *WriteSettings) { w.WriteString(">") } -// NewProcInst creates a parentless XML processing instruction. +// NewProcInst creates an unparented XML processing instruction. func NewProcInst(target, inst string) *ProcInst { return newProcInst(target, inst, nil) } @@ -1392,7 +1541,7 @@ func newProcInst(target, inst string, parent *Element) *ProcInst { p := &ProcInst{ Target: target, Inst: inst, - parent: parent, + parent: nil, index: -1, } if parent != nil { @@ -1401,8 +1550,9 @@ func newProcInst(target, inst string, parent *Element) *ProcInst { return p } -// CreateProcInst creates a processing instruction and adds it as a child of -// element e. +// CreateProcInst creates an XML processing instruction token with the +// specified 'target' and instruction 'inst'. It is then added as the last +// child token of this element. func (e *Element) CreateProcInst(target, inst string) *ProcInst { return newProcInst(target, inst, e) } @@ -1424,8 +1574,8 @@ func (p *ProcInst) Parent() *Element { } // Index returns the index of this ProcInst token within its parent element's -// list of child tokens. If this ProcInst token has no parent element, the -// index is -1. +// list of child tokens. If this ProcInst token has no parent, then the +// function returns -1. func (p *ProcInst) Index() int { return p.index } diff --git a/vendor/github.com/beevik/etree/helpers.go b/vendor/github.com/beevik/etree/helpers.go index 825e14e9..39a31ee6 100644 --- a/vendor/github.com/beevik/etree/helpers.go +++ b/vendor/github.com/beevik/etree/helpers.go @@ -83,38 +83,41 @@ func (f *fifo) grow() { f.data, f.head, f.tail = buf, 0, count } -// countReader implements a proxy reader that counts the number of -// bytes read from its encapsulated reader. -type countReader struct { +// xmlReader implements a proxy reader that counts the number of +// bytes read from its encapsulated reader and detects when a CDATA +// prefix has been parsed. +type xmlReader struct { r io.Reader bytes int64 } -func newCountReader(r io.Reader) *countReader { - return &countReader{r: r} +var cdataPrefix = []byte("