diff options
author | Ondřej Surý <ondrej@sury.org> | 2012-01-30 15:38:19 +0100 |
---|---|---|
committer | Ondřej Surý <ondrej@sury.org> | 2012-01-30 15:38:19 +0100 |
commit | 4cecda6c347bd6902b960c6a35a967add7070b0d (patch) | |
tree | a462e224ff41ec9f3eb1a0b6e815806f9e8804ad /src/pkg/encoding/xml | |
parent | 6c7ca6e4d4e26e4c8cbe0d183966011b3b088a0a (diff) | |
download | golang-4cecda6c347bd6902b960c6a35a967add7070b0d.tar.gz |
Imported Upstream version 2012.01.27upstream-weekly/2012.01.27
Diffstat (limited to 'src/pkg/encoding/xml')
-rw-r--r-- | src/pkg/encoding/xml/Makefile | 15 | ||||
-rw-r--r-- | src/pkg/encoding/xml/atom_test.go | 51 | ||||
-rw-r--r-- | src/pkg/encoding/xml/marshal.go | 380 | ||||
-rw-r--r-- | src/pkg/encoding/xml/marshal_test.go | 707 | ||||
-rw-r--r-- | src/pkg/encoding/xml/read.go | 564 | ||||
-rw-r--r-- | src/pkg/encoding/xml/read_test.go | 358 | ||||
-rw-r--r-- | src/pkg/encoding/xml/typeinfo.go | 322 | ||||
-rw-r--r-- | src/pkg/encoding/xml/xml.go | 1697 | ||||
-rw-r--r-- | src/pkg/encoding/xml/xml_test.go | 578 |
9 files changed, 4672 insertions, 0 deletions
diff --git a/src/pkg/encoding/xml/Makefile b/src/pkg/encoding/xml/Makefile new file mode 100644 index 000000000..ca84c211e --- /dev/null +++ b/src/pkg/encoding/xml/Makefile @@ -0,0 +1,15 @@ +# Copyright 2009 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. + +include ../../../Make.inc + +TARG=encoding/xml + +GOFILES=\ + marshal.go\ + read.go\ + typeinfo.go\ + xml.go\ + +include ../../../Make.pkg diff --git a/src/pkg/encoding/xml/atom_test.go b/src/pkg/encoding/xml/atom_test.go new file mode 100644 index 000000000..8d003aade --- /dev/null +++ b/src/pkg/encoding/xml/atom_test.go @@ -0,0 +1,51 @@ +// Copyright 2011 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 xml + +var atomValue = &Feed{ + XMLName: Name{"http://www.w3.org/2005/Atom", "feed"}, + Title: "Example Feed", + Link: []Link{{Href: "http://example.org/"}}, + Updated: ParseTime("2003-12-13T18:30:02Z"), + Author: Person{Name: "John Doe"}, + Id: "urn:uuid:60a76c80-d399-11d9-b93C-0003939e0af6", + + Entry: []Entry{ + { + Title: "Atom-Powered Robots Run Amok", + Link: []Link{{Href: "http://example.org/2003/12/13/atom03"}}, + Id: "urn:uuid:1225c695-cfb8-4ebb-aaaa-80da344efa6a", + Updated: ParseTime("2003-12-13T18:30:02Z"), + Summary: NewText("Some text."), + }, + }, +} + +var atomXml = `` + + `<feed xmlns="http://www.w3.org/2005/Atom">` + + `<title>Example Feed</title>` + + `<id>urn:uuid:60a76c80-d399-11d9-b93C-0003939e0af6</id>` + + `<link href="http://example.org/"></link>` + + `<updated>2003-12-13T18:30:02Z</updated>` + + `<author><name>John Doe</name><uri></uri><email></email></author>` + + `<entry>` + + `<title>Atom-Powered Robots Run Amok</title>` + + `<id>urn:uuid:1225c695-cfb8-4ebb-aaaa-80da344efa6a</id>` + + `<link href="http://example.org/2003/12/13/atom03"></link>` + + `<updated>2003-12-13T18:30:02Z</updated>` + + `<author><name></name><uri></uri><email></email></author>` + + `<summary>Some text.</summary>` + + `</entry>` + + `</feed>` + +func ParseTime(str string) Time { + return Time(str) +} + +func NewText(text string) Text { + return Text{ + Body: text, + } +} diff --git a/src/pkg/encoding/xml/marshal.go b/src/pkg/encoding/xml/marshal.go new file mode 100644 index 000000000..7a05a1bb1 --- /dev/null +++ b/src/pkg/encoding/xml/marshal.go @@ -0,0 +1,380 @@ +// Copyright 2011 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 xml + +import ( + "bufio" + "bytes" + "fmt" + "io" + "reflect" + "strconv" + "strings" +) + +const ( + // A generic XML header suitable for use with the output of Marshal. + // This is not automatically added to any output of this package, + // it is provided as a convenience. + Header = `<?xml version="1.0" encoding="UTF-8"?>` + "\n" +) + +// Marshal returns the XML encoding of v. +// +// Marshal handles an array or slice by marshalling each of the elements. +// Marshal handles a pointer by marshalling the value it points at or, if the +// pointer is nil, by writing nothing. Marshal handles an interface value by +// marshalling the value it contains or, if the interface value is nil, by +// writing nothing. Marshal handles all other data by writing one or more XML +// elements containing the data. +// +// The name for the XML elements is taken from, in order of preference: +// - the tag on the XMLName field, if the data is a struct +// - the value of the XMLName field of type xml.Name +// - the tag of the struct field used to obtain the data +// - the name of the struct field used to obtain the data +// - the name of the marshalled type +// +// The XML element for a struct contains marshalled elements for each of the +// exported fields of the struct, with these exceptions: +// - the XMLName field, described above, is omitted. +// - a field with tag "-" is omitted. +// - a field with tag "name,attr" becomes an attribute with +// the given name in the XML element. +// - a field with tag ",attr" becomes an attribute with the +// field name in the in the XML element. +// - a field with tag ",chardata" is written as character data, +// not as an XML element. +// - a field with tag ",innerxml" is written verbatim, not subject +// to the usual marshalling procedure. +// - a field with tag ",comment" is written as an XML comment, not +// subject to the usual marshalling procedure. It must not contain +// the "--" string within it. +// +// If a field uses a tag "a>b>c", then the element c will be nested inside +// parent elements a and b. Fields that appear next to each other that name +// the same parent will be enclosed in one XML element. For example: +// +// type Result struct { +// XMLName xml.Name `xml:"result"` +// Id int `xml:"id,attr"` +// FirstName string `xml:"person>name>first"` +// LastName string `xml:"person>name>last"` +// Age int `xml:"person>age"` +// } +// +// xml.Marshal(&Result{Id: 13, FirstName: "John", LastName: "Doe", Age: 42}) +// +// would be marshalled as: +// +// <result> +// <person id="13"> +// <name> +// <first>John</first> +// <last>Doe</last> +// </name> +// <age>42</age> +// </person> +// </result> +// +// Marshal will return an error if asked to marshal a channel, function, or map. +func Marshal(v interface{}) ([]byte, error) { + var b bytes.Buffer + if err := NewEncoder(&b).Encode(v); err != nil { + return nil, err + } + return b.Bytes(), nil +} + +// An Encoder writes XML data to an output stream. +type Encoder struct { + printer +} + +// NewEncoder returns a new encoder that writes to w. +func NewEncoder(w io.Writer) *Encoder { + return &Encoder{printer{bufio.NewWriter(w)}} +} + +// Encode writes the XML encoding of v to the stream. +// +// See the documentation for Marshal for details about the conversion +// of Go values to XML. +func (enc *Encoder) Encode(v interface{}) error { + err := enc.marshalValue(reflect.ValueOf(v), nil) + enc.Flush() + return err +} + +type printer struct { + *bufio.Writer +} + +func (p *printer) marshalValue(val reflect.Value, finfo *fieldInfo) error { + if !val.IsValid() { + return nil + } + + kind := val.Kind() + typ := val.Type() + + // Drill into pointers/interfaces + if kind == reflect.Ptr || kind == reflect.Interface { + if val.IsNil() { + return nil + } + return p.marshalValue(val.Elem(), finfo) + } + + // Slices and arrays iterate over the elements. They do not have an enclosing tag. + if (kind == reflect.Slice || kind == reflect.Array) && typ.Elem().Kind() != reflect.Uint8 { + for i, n := 0, val.Len(); i < n; i++ { + if err := p.marshalValue(val.Index(i), finfo); err != nil { + return err + } + } + return nil + } + + tinfo, err := getTypeInfo(typ) + if err != nil { + return err + } + + // Precedence for the XML element name is: + // 1. XMLName field in underlying struct; + // 2. field name/tag in the struct field; and + // 3. type name + var xmlns, name string + if tinfo.xmlname != nil { + xmlname := tinfo.xmlname + if xmlname.name != "" { + xmlns, name = xmlname.xmlns, xmlname.name + } else if v, ok := val.FieldByIndex(xmlname.idx).Interface().(Name); ok && v.Local != "" { + xmlns, name = v.Space, v.Local + } + } + if name == "" && finfo != nil { + xmlns, name = finfo.xmlns, finfo.name + } + if name == "" { + name = typ.Name() + if name == "" { + return &UnsupportedTypeError{typ} + } + } + + p.WriteByte('<') + p.WriteString(name) + + if xmlns != "" { + p.WriteString(` xmlns="`) + // TODO: EscapeString, to avoid the allocation. + Escape(p, []byte(xmlns)) + p.WriteByte('"') + } + + // Attributes + for i := range tinfo.fields { + finfo := &tinfo.fields[i] + if finfo.flags&fAttr == 0 { + continue + } + fv := val.FieldByIndex(finfo.idx) + switch fv.Kind() { + case reflect.String, reflect.Array, reflect.Slice: + // TODO: Should we really do this once ,omitempty is in? + if fv.Len() == 0 { + continue + } + } + p.WriteByte(' ') + p.WriteString(finfo.name) + p.WriteString(`="`) + if err := p.marshalSimple(fv.Type(), fv); err != nil { + return err + } + p.WriteByte('"') + } + p.WriteByte('>') + + if val.Kind() == reflect.Struct { + err = p.marshalStruct(tinfo, val) + } else { + err = p.marshalSimple(typ, val) + } + if err != nil { + return err + } + + p.WriteByte('<') + p.WriteByte('/') + p.WriteString(name) + p.WriteByte('>') + + return nil +} + +func (p *printer) marshalSimple(typ reflect.Type, val reflect.Value) error { + switch val.Kind() { + case reflect.Int, reflect.Int8, reflect.Int16, reflect.Int32, reflect.Int64: + p.WriteString(strconv.FormatInt(val.Int(), 10)) + case reflect.Uint, reflect.Uint8, reflect.Uint16, reflect.Uint32, reflect.Uint64, reflect.Uintptr: + p.WriteString(strconv.FormatUint(val.Uint(), 10)) + case reflect.Float32, reflect.Float64: + p.WriteString(strconv.FormatFloat(val.Float(), 'g', -1, 64)) + case reflect.String: + // TODO: Add EscapeString. + Escape(p, []byte(val.String())) + case reflect.Bool: + p.WriteString(strconv.FormatBool(val.Bool())) + case reflect.Array: + // will be [...]byte + bytes := make([]byte, val.Len()) + for i := range bytes { + bytes[i] = val.Index(i).Interface().(byte) + } + Escape(p, bytes) + case reflect.Slice: + // will be []byte + Escape(p, val.Bytes()) + default: + return &UnsupportedTypeError{typ} + } + return nil +} + +var ddBytes = []byte("--") + +func (p *printer) marshalStruct(tinfo *typeInfo, val reflect.Value) error { + s := parentStack{printer: p} + for i := range tinfo.fields { + finfo := &tinfo.fields[i] + if finfo.flags&(fAttr|fAny) != 0 { + continue + } + vf := val.FieldByIndex(finfo.idx) + switch finfo.flags & fMode { + case fCharData: + switch vf.Kind() { + case reflect.String: + Escape(p, []byte(vf.String())) + case reflect.Slice: + if elem, ok := vf.Interface().([]byte); ok { + Escape(p, elem) + } + } + continue + + case fComment: + k := vf.Kind() + if !(k == reflect.String || k == reflect.Slice && vf.Type().Elem().Kind() == reflect.Uint8) { + return fmt.Errorf("xml: bad type for comment field of %s", val.Type()) + } + if vf.Len() == 0 { + continue + } + p.WriteString("<!--") + dashDash := false + dashLast := false + switch k { + case reflect.String: + s := vf.String() + dashDash = strings.Index(s, "--") >= 0 + dashLast = s[len(s)-1] == '-' + if !dashDash { + p.WriteString(s) + } + case reflect.Slice: + b := vf.Bytes() + dashDash = bytes.Index(b, ddBytes) >= 0 + dashLast = b[len(b)-1] == '-' + if !dashDash { + p.Write(b) + } + default: + panic("can't happen") + } + if dashDash { + return fmt.Errorf(`xml: comments must not contain "--"`) + } + if dashLast { + // "--->" is invalid grammar. Make it "- -->" + p.WriteByte(' ') + } + p.WriteString("-->") + continue + + case fInnerXml: + iface := vf.Interface() + switch raw := iface.(type) { + case []byte: + p.Write(raw) + continue + case string: + p.WriteString(raw) + continue + } + + case fElement: + s.trim(finfo.parents) + if len(finfo.parents) > len(s.stack) { + if vf.Kind() != reflect.Ptr && vf.Kind() != reflect.Interface || !vf.IsNil() { + s.push(finfo.parents[len(s.stack):]) + } + } + } + if err := p.marshalValue(vf, finfo); err != nil { + return err + } + } + s.trim(nil) + return nil +} + +type parentStack struct { + *printer + stack []string +} + +// trim updates the XML context to match the longest common prefix of the stack +// and the given parents. A closing tag will be written for every parent +// popped. Passing a zero slice or nil will close all the elements. +func (s *parentStack) trim(parents []string) { + split := 0 + for ; split < len(parents) && split < len(s.stack); split++ { + if parents[split] != s.stack[split] { + break + } + } + + for i := len(s.stack) - 1; i >= split; i-- { + s.WriteString("</") + s.WriteString(s.stack[i]) + s.WriteByte('>') + } + + s.stack = parents[:split] +} + +// push adds parent elements to the stack and writes open tags. +func (s *parentStack) push(parents []string) { + for i := 0; i < len(parents); i++ { + s.WriteString("<") + s.WriteString(parents[i]) + s.WriteByte('>') + } + s.stack = append(s.stack, parents...) +} + +// A MarshalXMLError is returned when Marshal encounters a type +// that cannot be converted into XML. +type UnsupportedTypeError struct { + Type reflect.Type +} + +func (e *UnsupportedTypeError) Error() string { + return "xml: unsupported type: " + e.Type.String() +} diff --git a/src/pkg/encoding/xml/marshal_test.go b/src/pkg/encoding/xml/marshal_test.go new file mode 100644 index 000000000..e0be33200 --- /dev/null +++ b/src/pkg/encoding/xml/marshal_test.go @@ -0,0 +1,707 @@ +// Copyright 2011 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 xml + +import ( + "reflect" + "strconv" + "strings" + "testing" +) + +type DriveType int + +const ( + HyperDrive DriveType = iota + ImprobabilityDrive +) + +type Passenger struct { + Name []string `xml:"name"` + Weight float32 `xml:"weight"` +} + +type Ship struct { + XMLName struct{} `xml:"spaceship"` + + Name string `xml:"name,attr"` + Pilot string `xml:"pilot,attr"` + Drive DriveType `xml:"drive"` + Age uint `xml:"age"` + Passenger []*Passenger `xml:"passenger"` + secret string +} + +type NamedType string + +type Port struct { + XMLName struct{} `xml:"port"` + Type string `xml:"type,attr"` + Comment string `xml:",comment"` + Number string `xml:",chardata"` +} + +type Domain struct { + XMLName struct{} `xml:"domain"` + Country string `xml:",attr"` + Name []byte `xml:",chardata"` + Comment []byte `xml:",comment"` +} + +type Book struct { + XMLName struct{} `xml:"book"` + Title string `xml:",chardata"` +} + +type SecretAgent struct { + XMLName struct{} `xml:"agent"` + Handle string `xml:"handle,attr"` + Identity string + Obfuscate string `xml:",innerxml"` +} + +type NestedItems struct { + XMLName struct{} `xml:"result"` + Items []string `xml:">item"` + Item1 []string `xml:"Items>item1"` +} + +type NestedOrder struct { + XMLName struct{} `xml:"result"` + Field1 string `xml:"parent>c"` + Field2 string `xml:"parent>b"` + Field3 string `xml:"parent>a"` +} + +type MixedNested struct { + XMLName struct{} `xml:"result"` + A string `xml:"parent1>a"` + B string `xml:"b"` + C string `xml:"parent1>parent2>c"` + D string `xml:"parent1>d"` +} + +type NilTest struct { + A interface{} `xml:"parent1>parent2>a"` + B interface{} `xml:"parent1>b"` + C interface{} `xml:"parent1>parent2>c"` +} + +type Service struct { + XMLName struct{} `xml:"service"` + Domain *Domain `xml:"host>domain"` + Port *Port `xml:"host>port"` + Extra1 interface{} + Extra2 interface{} `xml:"host>extra2"` +} + +var nilStruct *Ship + +type EmbedA struct { + EmbedC + EmbedB EmbedB + FieldA string +} + +type EmbedB struct { + FieldB string + EmbedC +} + +type EmbedC struct { + FieldA1 string `xml:"FieldA>A1"` + FieldA2 string `xml:"FieldA>A2"` + FieldB string + FieldC string +} + +type NameCasing struct { + XMLName struct{} `xml:"casing"` + Xy string + XY string + XyA string `xml:"Xy,attr"` + XYA string `xml:"XY,attr"` +} + +type NamePrecedence struct { + XMLName Name `xml:"Parent"` + FromTag XMLNameWithoutTag `xml:"InTag"` + FromNameVal XMLNameWithoutTag + FromNameTag XMLNameWithTag + InFieldName string +} + +type XMLNameWithTag struct { + XMLName Name `xml:"InXMLNameTag"` + Value string ",chardata" +} + +type XMLNameWithoutTag struct { + XMLName Name + Value string ",chardata" +} + +type NameInField struct { + Foo Name `xml:"ns foo"` +} + +type AttrTest struct { + Int int `xml:",attr"` + Lower int `xml:"int,attr"` + Float float64 `xml:",attr"` + Uint8 uint8 `xml:",attr"` + Bool bool `xml:",attr"` + Str string `xml:",attr"` +} + +type AnyTest struct { + XMLName struct{} `xml:"a"` + Nested string `xml:"nested>value"` + AnyField AnyHolder `xml:",any"` +} + +type AnyHolder struct { + XMLName Name + XML string `xml:",innerxml"` +} + +type RecurseA struct { + A string + B *RecurseB +} + +type RecurseB struct { + A *RecurseA + B string +} + +type PresenceTest struct { + Exists *struct{} +} + +type IgnoreTest struct { + PublicSecret string `xml:"-"` +} + +type MyBytes []byte + +type Data struct { + Bytes []byte + Attr []byte `xml:",attr"` + Custom MyBytes +} + +type Plain struct { + V interface{} +} + +// Unless explicitly stated as such (or *Plain), all of the +// tests below are two-way tests. When introducing new tests, +// please try to make them two-way as well to ensure that +// marshalling and unmarshalling are as symmetrical as feasible. +var marshalTests = []struct { + Value interface{} + ExpectXML string + MarshalOnly bool + UnmarshalOnly bool +}{ + // Test nil marshals to nothing + {Value: nil, ExpectXML: ``, MarshalOnly: true}, + {Value: nilStruct, ExpectXML: ``, MarshalOnly: true}, + + // Test value types + {Value: &Plain{true}, ExpectXML: `<Plain><V>true</V></Plain>`}, + {Value: &Plain{false}, ExpectXML: `<Plain><V>false</V></Plain>`}, + {Value: &Plain{int(42)}, ExpectXML: `<Plain><V>42</V></Plain>`}, + {Value: &Plain{int8(42)}, ExpectXML: `<Plain><V>42</V></Plain>`}, + {Value: &Plain{int16(42)}, ExpectXML: `<Plain><V>42</V></Plain>`}, + {Value: &Plain{int32(42)}, ExpectXML: `<Plain><V>42</V></Plain>`}, + {Value: &Plain{uint(42)}, ExpectXML: `<Plain><V>42</V></Plain>`}, + {Value: &Plain{uint8(42)}, ExpectXML: `<Plain><V>42</V></Plain>`}, + {Value: &Plain{uint16(42)}, ExpectXML: `<Plain><V>42</V></Plain>`}, + {Value: &Plain{uint32(42)}, ExpectXML: `<Plain><V>42</V></Plain>`}, + {Value: &Plain{float32(1.25)}, ExpectXML: `<Plain><V>1.25</V></Plain>`}, + {Value: &Plain{float64(1.25)}, ExpectXML: `<Plain><V>1.25</V></Plain>`}, + {Value: &Plain{uintptr(0xFFDD)}, ExpectXML: `<Plain><V>65501</V></Plain>`}, + {Value: &Plain{"gopher"}, ExpectXML: `<Plain><V>gopher</V></Plain>`}, + {Value: &Plain{[]byte("gopher")}, ExpectXML: `<Plain><V>gopher</V></Plain>`}, + {Value: &Plain{"</>"}, ExpectXML: `<Plain><V></></V></Plain>`}, + {Value: &Plain{[]byte("</>")}, ExpectXML: `<Plain><V></></V></Plain>`}, + {Value: &Plain{[3]byte{'<', '/', '>'}}, ExpectXML: `<Plain><V></></V></Plain>`}, + {Value: &Plain{NamedType("potato")}, ExpectXML: `<Plain><V>potato</V></Plain>`}, + {Value: &Plain{[]int{1, 2, 3}}, ExpectXML: `<Plain><V>1</V><V>2</V><V>3</V></Plain>`}, + {Value: &Plain{[3]int{1, 2, 3}}, ExpectXML: `<Plain><V>1</V><V>2</V><V>3</V></Plain>`}, + + // A pointer to struct{} may be used to test for an element's presence. + { + Value: &PresenceTest{new(struct{})}, + ExpectXML: `<PresenceTest><Exists></Exists></PresenceTest>`, + }, + { + Value: &PresenceTest{}, + ExpectXML: `<PresenceTest></PresenceTest>`, + }, + + // A pointer to struct{} may be used to test for an element's presence. + { + Value: &PresenceTest{new(struct{})}, + ExpectXML: `<PresenceTest><Exists></Exists></PresenceTest>`, + }, + { + Value: &PresenceTest{}, + ExpectXML: `<PresenceTest></PresenceTest>`, + }, + + // A []byte field is only nil if the element was not found. + { + Value: &Data{}, + ExpectXML: `<Data></Data>`, + UnmarshalOnly: true, + }, + { + Value: &Data{Bytes: []byte{}, Custom: MyBytes{}, Attr: []byte{}}, + ExpectXML: `<Data Attr=""><Bytes></Bytes><Custom></Custom></Data>`, + UnmarshalOnly: true, + }, + + // Check that []byte works, including named []byte types. + { + Value: &Data{Bytes: []byte("ab"), Custom: MyBytes("cd"), Attr: []byte{'v'}}, + ExpectXML: `<Data Attr="v"><Bytes>ab</Bytes><Custom>cd</Custom></Data>`, + }, + + // Test innerxml + { + Value: &SecretAgent{ + Handle: "007", + Identity: "James Bond", + Obfuscate: "<redacted/>", + }, + ExpectXML: `<agent handle="007"><Identity>James Bond</Identity><redacted/></agent>`, + MarshalOnly: true, + }, + { + Value: &SecretAgent{ + Handle: "007", + Identity: "James Bond", + Obfuscate: "<Identity>James Bond</Identity><redacted/>", + }, + ExpectXML: `<agent handle="007"><Identity>James Bond</Identity><redacted/></agent>`, + UnmarshalOnly: true, + }, + + // Test structs + {Value: &Port{Type: "ssl", Number: "443"}, ExpectXML: `<port type="ssl">443</port>`}, + {Value: &Port{Number: "443"}, ExpectXML: `<port>443</port>`}, + {Value: &Port{Type: "<unix>"}, ExpectXML: `<port type="<unix>"></port>`}, + {Value: &Port{Number: "443", Comment: "https"}, ExpectXML: `<port><!--https-->443</port>`}, + {Value: &Port{Number: "443", Comment: "add space-"}, ExpectXML: `<port><!--add space- -->443</port>`, MarshalOnly: true}, + {Value: &Domain{Name: []byte("google.com&friends")}, ExpectXML: `<domain>google.com&friends</domain>`}, + {Value: &Domain{Name: []byte("google.com"), Comment: []byte(" &friends ")}, ExpectXML: `<domain>google.com<!-- &friends --></domain>`}, + {Value: &Book{Title: "Pride & Prejudice"}, ExpectXML: `<book>Pride & Prejudice</book>`}, + {Value: atomValue, ExpectXML: atomXml}, + { + Value: &Ship{ + Name: "Heart of Gold", + Pilot: "Computer", + Age: 1, + Drive: ImprobabilityDrive, + Passenger: []*Passenger{ + { + Name: []string{"Zaphod", "Beeblebrox"}, + Weight: 7.25, + }, + { + Name: []string{"Trisha", "McMillen"}, + Weight: 5.5, + }, + { + Name: []string{"Ford", "Prefect"}, + Weight: 7, + }, + { + Name: []string{"Arthur", "Dent"}, + Weight: 6.75, + }, + }, + }, + ExpectXML: `<spaceship name="Heart of Gold" pilot="Computer">` + + `<drive>` + strconv.Itoa(int(ImprobabilityDrive)) + `</drive>` + + `<age>1</age>` + + `<passenger>` + + `<name>Zaphod</name>` + + `<name>Beeblebrox</name>` + + `<weight>7.25</weight>` + + `</passenger>` + + `<passenger>` + + `<name>Trisha</name>` + + `<name>McMillen</name>` + + `<weight>5.5</weight>` + + `</passenger>` + + `<passenger>` + + `<name>Ford</name>` + + `<name>Prefect</name>` + + `<weight>7</weight>` + + `</passenger>` + + `<passenger>` + + `<name>Arthur</name>` + + `<name>Dent</name>` + + `<weight>6.75</weight>` + + `</passenger>` + + `</spaceship>`, + }, + + // Test a>b + { + Value: &NestedItems{Items: nil, Item1: nil}, + ExpectXML: `<result>` + + `<Items>` + + `</Items>` + + `</result>`, + }, + { + Value: &NestedItems{Items: []string{}, Item1: []string{}}, + ExpectXML: `<result>` + + `<Items>` + + `</Items>` + + `</result>`, + MarshalOnly: true, + }, + { + Value: &NestedItems{Items: nil, Item1: []string{"A"}}, + ExpectXML: `<result>` + + `<Items>` + + `<item1>A</item1>` + + `</Items>` + + `</result>`, + }, + { + Value: &NestedItems{Items: []string{"A", "B"}, Item1: nil}, + ExpectXML: `<result>` + + `<Items>` + + `<item>A</item>` + + `<item>B</item>` + + `</Items>` + + `</result>`, + }, + { + Value: &NestedItems{Items: []string{"A", "B"}, Item1: []string{"C"}}, + ExpectXML: `<result>` + + `<Items>` + + `<item>A</item>` + + `<item>B</item>` + + `<item1>C</item1>` + + `</Items>` + + `</result>`, + }, + { + Value: &NestedOrder{Field1: "C", Field2: "B", Field3: "A"}, + ExpectXML: `<result>` + + `<parent>` + + `<c>C</c>` + + `<b>B</b>` + + `<a>A</a>` + + `</parent>` + + `</result>`, + }, + { + Value: &NilTest{A: "A", B: nil, C: "C"}, + ExpectXML: `<NilTest>` + + `<parent1>` + + `<parent2><a>A</a></parent2>` + + `<parent2><c>C</c></parent2>` + + `</parent1>` + + `</NilTest>`, + MarshalOnly: true, // Uses interface{} + }, + { + Value: &MixedNested{A: "A", B: "B", C: "C", D: "D"}, + ExpectXML: `<result>` + + `<parent1><a>A</a></parent1>` + + `<b>B</b>` + + `<parent1>` + + `<parent2><c>C</c></parent2>` + + `<d>D</d>` + + `</parent1>` + + `</result>`, + }, + { + Value: &Service{Port: &Port{Number: "80"}}, + ExpectXML: `<service><host><port>80</port></host></service>`, + }, + { + Value: &Service{}, + ExpectXML: `<service></service>`, + }, + { + Value: &Service{Port: &Port{Number: "80"}, Extra1: "A", Extra2: "B"}, + ExpectXML: `<service>` + + `<host><port>80</port></host>` + + `<Extra1>A</Extra1>` + + `<host><extra2>B</extra2></host>` + + `</service>`, + MarshalOnly: true, + }, + { + Value: &Service{Port: &Port{Number: "80"}, Extra2: "example"}, + ExpectXML: `<service>` + + `<host><port>80</port></host>` + + `<host><extra2>example</extra2></host>` + + `</service>`, + MarshalOnly: true, + }, + + // Test struct embedding + { + Value: &EmbedA{ + EmbedC: EmbedC{ + FieldA1: "", // Shadowed by A.A + FieldA2: "", // Shadowed by A.A + FieldB: "A.C.B", + FieldC: "A.C.C", + }, + EmbedB: EmbedB{ + FieldB: "A.B.B", + EmbedC: EmbedC{ + FieldA1: "A.B.C.A1", + FieldA2: "A.B.C.A2", + FieldB: "", // Shadowed by A.B.B + FieldC: "A.B.C.C", + }, + }, + FieldA: "A.A", + }, + ExpectXML: `<EmbedA>` + + `<FieldB>A.C.B</FieldB>` + + `<FieldC>A.C.C</FieldC>` + + `<EmbedB>` + + `<FieldB>A.B.B</FieldB>` + + `<FieldA>` + + `<A1>A.B.C.A1</A1>` + + `<A2>A.B.C.A2</A2>` + + `</FieldA>` + + `<FieldC>A.B.C.C</FieldC>` + + `</EmbedB>` + + `<FieldA>A.A</FieldA>` + + `</EmbedA>`, + }, + + // Test that name casing matters + { + Value: &NameCasing{Xy: "mixed", XY: "upper", XyA: "mixedA", XYA: "upperA"}, + ExpectXML: `<casing Xy="mixedA" XY="upperA"><Xy>mixed</Xy><XY>upper</XY></casing>`, + }, + + // Test the order in which the XML element name is chosen + { + Value: &NamePrecedence{ + FromTag: XMLNameWithoutTag{Value: "A"}, + FromNameVal: XMLNameWithoutTag{XMLName: Name{Local: "InXMLName"}, Value: "B"}, + FromNameTag: XMLNameWithTag{Value: "C"}, + InFieldName: "D", + }, + ExpectXML: `<Parent>` + + `<InTag><Value>A</Value></InTag>` + + `<InXMLName><Value>B</Value></InXMLName>` + + `<InXMLNameTag><Value>C</Value></InXMLNameTag>` + + `<InFieldName>D</InFieldName>` + + `</Parent>`, + MarshalOnly: true, + }, + { + Value: &NamePrecedence{ + XMLName: Name{Local: "Parent"}, + FromTag: XMLNameWithoutTag{XMLName: Name{Local: "InTag"}, Value: "A"}, + FromNameVal: XMLNameWithoutTag{XMLName: Name{Local: "FromNameVal"}, Value: "B"}, + FromNameTag: XMLNameWithTag{XMLName: Name{Local: "InXMLNameTag"}, Value: "C"}, + InFieldName: "D", + }, + ExpectXML: `<Parent>` + + `<InTag><Value>A</Value></InTag>` + + `<FromNameVal><Value>B</Value></FromNameVal>` + + `<InXMLNameTag><Value>C</Value></InXMLNameTag>` + + `<InFieldName>D</InFieldName>` + + `</Parent>`, + UnmarshalOnly: true, + }, + + // xml.Name works in a plain field as well. + { + Value: &NameInField{Name{Space: "ns", Local: "foo"}}, + ExpectXML: `<NameInField><foo xmlns="ns"></foo></NameInField>`, + }, + + // Marshaling zero xml.Name uses the tag or field name. + { + Value: &NameInField{}, + ExpectXML: `<NameInField><foo xmlns="ns"></foo></NameInField>`, + MarshalOnly: true, + }, + + // Test attributes + { + Value: &AttrTest{ + Int: 8, + Lower: 9, + Float: 23.5, + Uint8: 255, + Bool: true, + Str: "s", + }, + ExpectXML: `<AttrTest Int="8" int="9" Float="23.5" Uint8="255" Bool="true" Str="s"></AttrTest>`, + }, + + // Test ",any" + { + ExpectXML: `<a><nested><value>known</value></nested><other><sub>unknown</sub></other></a>`, + Value: &AnyTest{ + Nested: "known", + AnyField: AnyHolder{ + XMLName: Name{Local: "other"}, + XML: "<sub>unknown</sub>", + }, + }, + UnmarshalOnly: true, + }, + { + Value: &AnyTest{Nested: "known", AnyField: AnyHolder{XML: "<unknown/>"}}, + ExpectXML: `<a><nested><value>known</value></nested></a>`, + MarshalOnly: true, + }, + + // Test recursive types. + { + Value: &RecurseA{ + A: "a1", + B: &RecurseB{ + A: &RecurseA{"a2", nil}, + B: "b1", + }, + }, + ExpectXML: `<RecurseA><A>a1</A><B><A><A>a2</A></A><B>b1</B></B></RecurseA>`, + }, + + // Test ignoring fields via "-" tag + { + ExpectXML: `<IgnoreTest></IgnoreTest>`, + Value: &IgnoreTest{}, + }, + { + ExpectXML: `<IgnoreTest></IgnoreTest>`, + Value: &IgnoreTest{PublicSecret: "can't tell"}, + MarshalOnly: true, + }, + { + ExpectXML: `<IgnoreTest><PublicSecret>ignore me</PublicSecret></IgnoreTest>`, + Value: &IgnoreTest{}, + UnmarshalOnly: true, + }, +} + +func TestMarshal(t *testing.T) { + for idx, test := range marshalTests { + if test.UnmarshalOnly { + continue + } + data, err := Marshal(test.Value) + if err != nil { + t.Errorf("#%d: Error: %s", idx, err) + continue + } + if got, want := string(data), test.ExpectXML; got != want { + if strings.Contains(want, "\n") { + t.Errorf("#%d: marshal(%#v):\nHAVE:\n%s\nWANT:\n%s", idx, test.Value, got, want) + } else { + t.Errorf("#%d: marshal(%#v):\nhave %#q\nwant %#q", idx, test.Value, got, want) + } + } + } +} + +var marshalErrorTests = []struct { + Value interface{} + Err string + Kind reflect.Kind +}{ + { + Value: make(chan bool), + Err: "xml: unsupported type: chan bool", + Kind: reflect.Chan, + }, + { + Value: map[string]string{ + "question": "What do you get when you multiply six by nine?", + "answer": "42", + }, + Err: "xml: unsupported type: map[string]string", + Kind: reflect.Map, + }, + { + Value: map[*Ship]bool{nil: false}, + Err: "xml: unsupported type: map[*xml.Ship]bool", + Kind: reflect.Map, + }, + { + Value: &Domain{Comment: []byte("f--bar")}, + Err: `xml: comments must not contain "--"`, + }, +} + +func TestMarshalErrors(t *testing.T) { + for idx, test := range marshalErrorTests { + _, err := Marshal(test.Value) + if err == nil || err.Error() != test.Err { + t.Errorf("#%d: marshal(%#v) = [error] %v, want %v", idx, test.Value, err, test.Err) + } + if test.Kind != reflect.Invalid { + if kind := err.(*UnsupportedTypeError).Type.Kind(); kind != test.Kind { + t.Errorf("#%d: marshal(%#v) = [error kind] %s, want %s", idx, test.Value, kind, test.Kind) + } + } + } +} + +// Do invertibility testing on the various structures that we test +func TestUnmarshal(t *testing.T) { + for i, test := range marshalTests { + if test.MarshalOnly { + continue + } + if _, ok := test.Value.(*Plain); ok { + continue + } + + vt := reflect.TypeOf(test.Value) + dest := reflect.New(vt.Elem()).Interface() + err := Unmarshal([]byte(test.ExpectXML), dest) + + switch fix := dest.(type) { + case *Feed: + fix.Author.InnerXML = "" + for i := range fix.Entry { + fix.Entry[i].Author.InnerXML = "" + } + } + + if err != nil { + t.Errorf("#%d: unexpected error: %#v", i, err) + } else if got, want := dest, test.Value; !reflect.DeepEqual(got, want) { + t.Errorf("#%d: unmarshal(%q):\nhave %#v\nwant %#v", i, test.ExpectXML, got, want) + } + } +} + +func BenchmarkMarshal(b *testing.B) { + for i := 0; i < b.N; i++ { + Marshal(atomValue) + } +} + +func BenchmarkUnmarshal(b *testing.B) { + xml := []byte(atomXml) + for i := 0; i < b.N; i++ { + Unmarshal(xml, &Feed{}) + } +} diff --git a/src/pkg/encoding/xml/read.go b/src/pkg/encoding/xml/read.go new file mode 100644 index 000000000..871fe059c --- /dev/null +++ b/src/pkg/encoding/xml/read.go @@ -0,0 +1,564 @@ +// Copyright 2009 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 xml + +import ( + "bytes" + "errors" + "reflect" + "strconv" + "strings" +) + +// BUG(rsc): Mapping between XML elements and data structures is inherently flawed: +// an XML element is an order-dependent collection of anonymous +// values, while a data structure is an order-independent collection +// of named values. +// See package json for a textual representation more suitable +// to data structures. + +// Unmarshal parses the XML-encoded data and stores the result in +// the value pointed to by v, which must be an arbitrary struct, +// slice, or string. Well-formed data that does not fit into v is +// discarded. +// +// For example, given these definitions: +// +// type Email struct { +// Where string `xml:",attr"` +// Addr string +// } +// +// type Result struct { +// XMLName xml.Name `xml:"result"` +// Name string +// Phone string +// Email []Email +// Groups []string `xml:"group>value"` +// } +// +// result := Result{Name: "name", Phone: "phone", Email: nil} +// +// unmarshalling the XML input +// +// <result> +// <email where="home"> +// <addr>gre@example.com</addr> +// </email> +// <email where='work'> +// <addr>gre@work.com</addr> +// </email> +// <name>Grace R. Emlin</name> +// <group> +// <value>Friends</value> +// <value>Squash</value> +// </group> +// <address>123 Main Street</address> +// </result> +// +// via Unmarshal(data, &result) is equivalent to assigning +// +// r = Result{ +// xml.Name{Local: "result"}, +// "Grace R. Emlin", // name +// "phone", // no phone given +// []Email{ +// Email{"home", "gre@example.com"}, +// Email{"work", "gre@work.com"}, +// }, +// []string{"Friends", "Squash"}, +// } +// +// Note that the field r.Phone has not been modified and +// that the XML <address> element was discarded. Also, the field +// Groups was assigned considering the element path provided in the +// field tag. +// +// Because Unmarshal uses the reflect package, it can only assign +// to exported (upper case) fields. Unmarshal uses a case-sensitive +// comparison to match XML element names to tag values and struct +// field names. +// +// Unmarshal maps an XML element to a struct using the following rules. +// In the rules, the tag of a field refers to the value associated with the +// key 'xml' in the struct field's tag (see the example above). +// +// * If the struct has a field of type []byte or string with tag +// ",innerxml", Unmarshal accumulates the raw XML nested inside the +// element in that field. The rest of the rules still apply. +// +// * If the struct has a field named XMLName of type xml.Name, +// Unmarshal records the element name in that field. +// +// * If the XMLName field has an associated tag of the form +// "name" or "namespace-URL name", the XML element must have +// the given name (and, optionally, name space) or else Unmarshal +// returns an error. +// +// * If the XML element has an attribute whose name matches a +// struct field name with an associated tag containing ",attr" or +// the explicit name in a struct field tag of the form "name,attr", +// Unmarshal records the attribute value in that field. +// +// * If the XML element contains character data, that data is +// accumulated in the first struct field that has tag "chardata". +// The struct field may have type []byte or string. +// If there is no such field, the character data is discarded. +// +// * If the XML element contains comments, they are accumulated in +// the first struct field that has tag ",comments". The struct +// field may have type []byte or string. If there is no such +// field, the comments are discarded. +// +// * If the XML element contains a sub-element whose name matches +// the prefix of a tag formatted as "a" or "a>b>c", unmarshal +// will descend into the XML structure looking for elements with the +// given names, and will map the innermost elements to that struct +// field. A tag starting with ">" is equivalent to one starting +// with the field name followed by ">". +// +// * If the XML element contains a sub-element whose name matches +// a struct field's XMLName tag and the struct field has no +// explicit name tag as per the previous rule, unmarshal maps +// the sub-element to that struct field. +// +// * If the XML element contains a sub-element whose name matches a +// field without any mode flags (",attr", ",chardata", etc), Unmarshal +// maps the sub-element to that struct field. +// +// * If the XML element contains a sub-element that hasn't matched any +// of the above rules and the struct has a field with tag ",any", +// unmarshal maps the sub-element to that struct field. +// +// * A struct field with tag "-" is never unmarshalled into. +// +// Unmarshal maps an XML element to a string or []byte by saving the +// concatenation of that element's character data in the string or +// []byte. The saved []byte is never nil. +// +// Unmarshal maps an attribute value to a string or []byte by saving +// the value in the string or slice. +// +// Unmarshal maps an XML element to a slice by extending the length of +// the slice and mapping the element to the newly created value. +// +// Unmarshal maps an XML element or attribute value to a bool by +// setting it to the boolean value represented by the string. +// +// Unmarshal maps an XML element or attribute value to an integer or +// floating-point field by setting the field to the result of +// interpreting the string value in decimal. There is no check for +// overflow. +// +// Unmarshal maps an XML element to an xml.Name by recording the +// element name. +// +// Unmarshal maps an XML element to a pointer by setting the pointer +// to a freshly allocated value and then mapping the element to that value. +// +func Unmarshal(data []byte, v interface{}) error { + return NewDecoder(bytes.NewBuffer(data)).Decode(v) +} + +// Decode works like xml.Unmarshal, except it reads the decoder +// stream to find the start element. +func (d *Decoder) Decode(v interface{}) error { + return d.DecodeElement(v, nil) +} + +// DecodeElement works like xml.Unmarshal except that it takes +// a pointer to the start XML element to decode into v. +// It is useful when a client reads some raw XML tokens itself +// but also wants to defer to Unmarshal for some elements. +func (d *Decoder) DecodeElement(v interface{}, start *StartElement) error { + val := reflect.ValueOf(v) + if val.Kind() != reflect.Ptr { + return errors.New("non-pointer passed to Unmarshal") + } + return d.unmarshal(val.Elem(), start) +} + +// An UnmarshalError represents an error in the unmarshalling process. +type UnmarshalError string + +func (e UnmarshalError) Error() string { return string(e) } + +// Unmarshal a single XML element into val. +func (p *Decoder) unmarshal(val reflect.Value, start *StartElement) error { + // Find start element if we need it. + if start == nil { + for { + tok, err := p.Token() + if err != nil { + return err + } + if t, ok := tok.(StartElement); ok { + start = &t + break + } + } + } + + if pv := val; pv.Kind() == reflect.Ptr { + if pv.IsNil() { + pv.Set(reflect.New(pv.Type().Elem())) + } + val = pv.Elem() + } + + var ( + data []byte + saveData reflect.Value + comment []byte + saveComment reflect.Value + saveXML reflect.Value + saveXMLIndex int + saveXMLData []byte + saveAny reflect.Value + sv reflect.Value + tinfo *typeInfo + err error + ) + + switch v := val; v.Kind() { + default: + return errors.New("unknown type " + v.Type().String()) + + case reflect.Interface: + // TODO: For now, simply ignore the field. In the near + // future we may choose to unmarshal the start + // element on it, if not nil. + return p.Skip() + + case reflect.Slice: + typ := v.Type() + if typ.Elem().Kind() == reflect.Uint8 { + // []byte + saveData = v + break + } + + // Slice of element values. + // Grow slice. + n := v.Len() + if n >= v.Cap() { + ncap := 2 * n + if ncap < 4 { + ncap = 4 + } + new := reflect.MakeSlice(typ, n, ncap) + reflect.Copy(new, v) + v.Set(new) + } + v.SetLen(n + 1) + + // Recur to read element into slice. + if err := p.unmarshal(v.Index(n), start); err != nil { + v.SetLen(n) + return err + } + return nil + + case reflect.Bool, reflect.Float32, reflect.Float64, reflect.Int, reflect.Int8, reflect.Int16, reflect.Int32, reflect.Int64, reflect.Uint, reflect.Uint8, reflect.Uint16, reflect.Uint32, reflect.Uint64, reflect.Uintptr, reflect.String: + saveData = v + + case reflect.Struct: + sv = v + typ := sv.Type() + if typ == nameType { + v.Set(reflect.ValueOf(start.Name)) + break + } + tinfo, err = getTypeInfo(typ) + if err != nil { + return err + } + + // Validate and assign element name. + if tinfo.xmlname != nil { + finfo := tinfo.xmlname + if finfo.name != "" && finfo.name != start.Name.Local { + return UnmarshalError("expected element type <" + finfo.name + "> but have <" + start.Name.Local + ">") + } + if finfo.xmlns != "" && finfo.xmlns != start.Name.Space { + e := "expected element <" + finfo.name + "> in name space " + finfo.xmlns + " but have " + if start.Name.Space == "" { + e += "no name space" + } else { + e += start.Name.Space + } + return UnmarshalError(e) + } + fv := sv.FieldByIndex(finfo.idx) + if _, ok := fv.Interface().(Name); ok { + fv.Set(reflect.ValueOf(start.Name)) + } + } + + // Assign attributes. + // Also, determine whether we need to save character data or comments. + for i := range tinfo.fields { + finfo := &tinfo.fields[i] + switch finfo.flags & fMode { + case fAttr: + strv := sv.FieldByIndex(finfo.idx) + // Look for attribute. + for _, a := range start.Attr { + if a.Name.Local == finfo.name { + copyValue(strv, []byte(a.Value)) + break + } + } + + case fCharData: + if !saveData.IsValid() { + saveData = sv.FieldByIndex(finfo.idx) + } + + case fComment: + if !saveComment.IsValid() { + saveComment = sv.FieldByIndex(finfo.idx) + } + + case fAny: + if !saveAny.IsValid() { + saveAny = sv.FieldByIndex(finfo.idx) + } + + case fInnerXml: + if !saveXML.IsValid() { + saveXML = sv.FieldByIndex(finfo.idx) + if p.saved == nil { + saveXMLIndex = 0 + p.saved = new(bytes.Buffer) + } else { + saveXMLIndex = p.savedOffset() + } + } + } + } + } + + // Find end element. + // Process sub-elements along the way. +Loop: + for { + var savedOffset int + if saveXML.IsValid() { + savedOffset = p.savedOffset() + } + tok, err := p.Token() + if err != nil { + return err + } + switch t := tok.(type) { + case StartElement: + consumed := false + if sv.IsValid() { + consumed, err = p.unmarshalPath(tinfo, sv, nil, &t) + if err != nil { + return err + } + if !consumed && saveAny.IsValid() { + consumed = true + if err := p.unmarshal(saveAny, &t); err != nil { + return err + } + } + } + if !consumed { + if err := p.Skip(); err != nil { + return err + } + } + + case EndElement: + if saveXML.IsValid() { + saveXMLData = p.saved.Bytes()[saveXMLIndex:savedOffset] + if saveXMLIndex == 0 { + p.saved = nil + } + } + break Loop + + case CharData: + if saveData.IsValid() { + data = append(data, t...) + } + + case Comment: + if saveComment.IsValid() { + comment = append(comment, t...) + } + } + } + + if err := copyValue(saveData, data); err != nil { + return err + } + + switch t := saveComment; t.Kind() { + case reflect.String: + t.SetString(string(comment)) + case reflect.Slice: + t.Set(reflect.ValueOf(comment)) + } + + switch t := saveXML; t.Kind() { + case reflect.String: + t.SetString(string(saveXMLData)) + case reflect.Slice: + t.Set(reflect.ValueOf(saveXMLData)) + } + + return nil +} + +func copyValue(dst reflect.Value, src []byte) (err error) { + // Helper functions for integer and unsigned integer conversions + var itmp int64 + getInt64 := func() bool { + itmp, err = strconv.ParseInt(string(src), 10, 64) + // TODO: should check sizes + return err == nil + } + var utmp uint64 + getUint64 := func() bool { + utmp, err = strconv.ParseUint(string(src), 10, 64) + // TODO: check for overflow? + return err == nil + } + var ftmp float64 + getFloat64 := func() bool { + ftmp, err = strconv.ParseFloat(string(src), 64) + // TODO: check for overflow? + return err == nil + } + + // Save accumulated data. + switch t := dst; t.Kind() { + case reflect.Invalid: + // Probably a comment. + default: + return errors.New("cannot happen: unknown type " + t.Type().String()) + case reflect.Int, reflect.Int8, reflect.Int16, reflect.Int32, reflect.Int64: + if !getInt64() { + return err + } + t.SetInt(itmp) + case reflect.Uint, reflect.Uint8, reflect.Uint16, reflect.Uint32, reflect.Uint64, reflect.Uintptr: + if !getUint64() { + return err + } + t.SetUint(utmp) + case reflect.Float32, reflect.Float64: + if !getFloat64() { + return err + } + t.SetFloat(ftmp) + case reflect.Bool: + value, err := strconv.ParseBool(strings.TrimSpace(string(src))) + if err != nil { + return err + } + t.SetBool(value) + case reflect.String: + t.SetString(string(src)) + case reflect.Slice: + if len(src) == 0 { + // non-nil to flag presence + src = []byte{} + } + t.SetBytes(src) + } + return nil +} + +// unmarshalPath walks down an XML structure looking for wanted +// paths, and calls unmarshal on them. +// The consumed result tells whether XML elements have been consumed +// from the Decoder until start's matching end element, or if it's +// still untouched because start is uninteresting for sv's fields. +func (p *Decoder) unmarshalPath(tinfo *typeInfo, sv reflect.Value, parents []string, start *StartElement) (consumed bool, err error) { + recurse := false +Loop: + for i := range tinfo.fields { + finfo := &tinfo.fields[i] + if finfo.flags&fElement == 0 || len(finfo.parents) < len(parents) { + continue + } + for j := range parents { + if parents[j] != finfo.parents[j] { + continue Loop + } + } + if len(finfo.parents) == len(parents) && finfo.name == start.Name.Local { + // It's a perfect match, unmarshal the field. + return true, p.unmarshal(sv.FieldByIndex(finfo.idx), start) + } + if len(finfo.parents) > len(parents) && finfo.parents[len(parents)] == start.Name.Local { + // It's a prefix for the field. Break and recurse + // since it's not ok for one field path to be itself + // the prefix for another field path. + recurse = true + + // We can reuse the same slice as long as we + // don't try to append to it. + parents = finfo.parents[:len(parents)+1] + break + } + } + if !recurse { + // We have no business with this element. + return false, nil + } + // The element is not a perfect match for any field, but one + // or more fields have the path to this element as a parent + // prefix. Recurse and attempt to match these. + for { + var tok Token + tok, err = p.Token() + if err != nil { + return true, err + } + switch t := tok.(type) { + case StartElement: + consumed2, err := p.unmarshalPath(tinfo, sv, parents, &t) + if err != nil { + return true, err + } + if !consumed2 { + if err := p.Skip(); err != nil { + return true, err + } + } + case EndElement: + return true, nil + } + } + panic("unreachable") +} + +// Have already read a start element. +// Read tokens until we find the end element. +// Token is taking care of making sure the +// end element matches the start element we saw. +func (p *Decoder) Skip() error { + for { + tok, err := p.Token() + if err != nil { + return err + } + switch tok.(type) { + case StartElement: + if err := p.Skip(); err != nil { + return err + } + case EndElement: + return nil + } + } + panic("unreachable") +} diff --git a/src/pkg/encoding/xml/read_test.go b/src/pkg/encoding/xml/read_test.go new file mode 100644 index 000000000..833eafc9a --- /dev/null +++ b/src/pkg/encoding/xml/read_test.go @@ -0,0 +1,358 @@ +// Copyright 2009 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 xml + +import ( + "reflect" + "testing" +) + +// Stripped down Atom feed data structures. + +func TestUnmarshalFeed(t *testing.T) { + var f Feed + if err := Unmarshal([]byte(atomFeedString), &f); err != nil { + t.Fatalf("Unmarshal: %s", err) + } + if !reflect.DeepEqual(f, atomFeed) { + t.Fatalf("have %#v\nwant %#v", f, atomFeed) + } +} + +// hget http://codereview.appspot.com/rss/mine/rsc +const atomFeedString = ` +<?xml version="1.0" encoding="utf-8"?> +<feed xmlns="http://www.w3.org/2005/Atom" xml:lang="en-us"><title>Code Review - My issues</title><link href="http://codereview.appspot.com/" rel="alternate"></link><link href="http://codereview.appspot.com/rss/mine/rsc" rel="self"></link><id>http://codereview.appspot.com/</id><updated>2009-10-04T01:35:58+00:00</updated><author><name>rietveld<></name></author><entry><title>rietveld: an attempt at pubsubhubbub +</title><link href="http://codereview.appspot.com/126085" rel="alternate"></link><updated>2009-10-04T01:35:58+00:00</updated><author><name>email-address-removed</name></author><id>urn:md5:134d9179c41f806be79b3a5f7877d19a</id><summary type="html"> + An attempt at adding pubsubhubbub support to Rietveld. +http://code.google.com/p/pubsubhubbub +http://code.google.com/p/rietveld/issues/detail?id=155 + +The server side of the protocol is trivial: + 1. add a &lt;link rel=&quot;hub&quot; href=&quot;hub-server&quot;&gt; tag to all + feeds that will be pubsubhubbubbed. + 2. every time one of those feeds changes, tell the hub + with a simple POST request. + +I have tested this by adding debug prints to a local hub +server and checking that the server got the right publish +requests. + +I can&#39;t quite get the server to work, but I think the bug +is not in my code. I think that the server expects to be +able to grab the feed and see the feed&#39;s actual URL in +the link rel=&quot;self&quot;, but the default value for that drops +the :port from the URL, and I cannot for the life of me +figure out how to get the Atom generator deep inside +django not to do that, or even where it is doing that, +or even what code is running to generate the Atom feed. +(I thought I knew but I added some assert False statements +and it kept running!) + +Ignoring that particular problem, I would appreciate +feedback on the right way to get the two values at +the top of feeds.py marked NOTE(rsc). + + +</summary></entry><entry><title>rietveld: correct tab handling +</title><link href="http://codereview.appspot.com/124106" rel="alternate"></link><updated>2009-10-03T23:02:17+00:00</updated><author><name>email-address-removed</name></author><id>urn:md5:0a2a4f19bb815101f0ba2904aed7c35a</id><summary type="html"> + This fixes the buggy tab rendering that can be seen at +http://codereview.appspot.com/116075/diff/1/2 + +The fundamental problem was that the tab code was +not being told what column the text began in, so it +didn&#39;t know where to put the tab stops. Another problem +was that some of the code assumed that string byte +offsets were the same as column offsets, which is only +true if there are no tabs. + +In the process of fixing this, I cleaned up the arguments +to Fold and ExpandTabs and renamed them Break and +_ExpandTabs so that I could be sure that I found all the +call sites. I also wanted to verify that ExpandTabs was +not being used from outside intra_region_diff.py. + + +</summary></entry></feed> ` + +type Feed struct { + XMLName Name `xml:"http://www.w3.org/2005/Atom feed"` + Title string `xml:"title"` + Id string `xml:"id"` + Link []Link `xml:"link"` + Updated Time `xml:"updated"` + Author Person `xml:"author"` + Entry []Entry `xml:"entry"` +} + +type Entry struct { + Title string `xml:"title"` + Id string `xml:"id"` + Link []Link `xml:"link"` + Updated Time `xml:"updated"` + Author Person `xml:"author"` + Summary Text `xml:"summary"` +} + +type Link struct { + Rel string `xml:"rel,attr"` + Href string `xml:"href,attr"` +} + +type Person struct { + Name string `xml:"name"` + URI string `xml:"uri"` + Email string `xml:"email"` + InnerXML string `xml:",innerxml"` +} + +type Text struct { + Type string `xml:"type,attr"` + Body string `xml:",chardata"` +} + +type Time string + +var atomFeed = Feed{ + XMLName: Name{"http://www.w3.org/2005/Atom", "feed"}, + Title: "Code Review - My issues", + Link: []Link{ + {Rel: "alternate", Href: "http://codereview.appspot.com/"}, + {Rel: "self", Href: "http://codereview.appspot.com/rss/mine/rsc"}, + }, + Id: "http://codereview.appspot.com/", + Updated: "2009-10-04T01:35:58+00:00", + Author: Person{ + Name: "rietveld<>", + InnerXML: "<name>rietveld<></name>", + }, + Entry: []Entry{ + { + Title: "rietveld: an attempt at pubsubhubbub\n", + Link: []Link{ + {Rel: "alternate", Href: "http://codereview.appspot.com/126085"}, + }, + Updated: "2009-10-04T01:35:58+00:00", + Author: Person{ + Name: "email-address-removed", + InnerXML: "<name>email-address-removed</name>", + }, + Id: "urn:md5:134d9179c41f806be79b3a5f7877d19a", + Summary: Text{ + Type: "html", + Body: ` + An attempt at adding pubsubhubbub support to Rietveld. +http://code.google.com/p/pubsubhubbub +http://code.google.com/p/rietveld/issues/detail?id=155 + +The server side of the protocol is trivial: + 1. add a <link rel="hub" href="hub-server"> tag to all + feeds that will be pubsubhubbubbed. + 2. every time one of those feeds changes, tell the hub + with a simple POST request. + +I have tested this by adding debug prints to a local hub +server and checking that the server got the right publish +requests. + +I can't quite get the server to work, but I think the bug +is not in my code. I think that the server expects to be +able to grab the feed and see the feed's actual URL in +the link rel="self", but the default value for that drops +the :port from the URL, and I cannot for the life of me +figure out how to get the Atom generator deep inside +django not to do that, or even where it is doing that, +or even what code is running to generate the Atom feed. +(I thought I knew but I added some assert False statements +and it kept running!) + +Ignoring that particular problem, I would appreciate +feedback on the right way to get the two values at +the top of feeds.py marked NOTE(rsc). + + +`, + }, + }, + { + Title: "rietveld: correct tab handling\n", + Link: []Link{ + {Rel: "alternate", Href: "http://codereview.appspot.com/124106"}, + }, + Updated: "2009-10-03T23:02:17+00:00", + Author: Person{ + Name: "email-address-removed", + InnerXML: "<name>email-address-removed</name>", + }, + Id: "urn:md5:0a2a4f19bb815101f0ba2904aed7c35a", + Summary: Text{ + Type: "html", + Body: ` + This fixes the buggy tab rendering that can be seen at +http://codereview.appspot.com/116075/diff/1/2 + +The fundamental problem was that the tab code was +not being told what column the text began in, so it +didn't know where to put the tab stops. Another problem +was that some of the code assumed that string byte +offsets were the same as column offsets, which is only +true if there are no tabs. + +In the process of fixing this, I cleaned up the arguments +to Fold and ExpandTabs and renamed them Break and +_ExpandTabs so that I could be sure that I found all the +call sites. I also wanted to verify that ExpandTabs was +not being used from outside intra_region_diff.py. + + +`, + }, + }, + }, +} + +const pathTestString = ` +<Result> + <Before>1</Before> + <Items> + <Item1> + <Value>A</Value> + </Item1> + <Item2> + <Value>B</Value> + </Item2> + <Item1> + <Value>C</Value> + <Value>D</Value> + </Item1> + <_> + <Value>E</Value> + </_> + </Items> + <After>2</After> +</Result> +` + +type PathTestItem struct { + Value string +} + +type PathTestA struct { + Items []PathTestItem `xml:">Item1"` + Before, After string +} + +type PathTestB struct { + Other []PathTestItem `xml:"Items>Item1"` + Before, After string +} + +type PathTestC struct { + Values1 []string `xml:"Items>Item1>Value"` + Values2 []string `xml:"Items>Item2>Value"` + Before, After string +} + +type PathTestSet struct { + Item1 []PathTestItem +} + +type PathTestD struct { + Other PathTestSet `xml:"Items"` + Before, After string +} + +type PathTestE struct { + Underline string `xml:"Items>_>Value"` + Before, After string +} + +var pathTests = []interface{}{ + &PathTestA{Items: []PathTestItem{{"A"}, {"D"}}, Before: "1", After: "2"}, + &PathTestB{Other: []PathTestItem{{"A"}, {"D"}}, Before: "1", After: "2"}, + &PathTestC{Values1: []string{"A", "C", "D"}, Values2: []string{"B"}, Before: "1", After: "2"}, + &PathTestD{Other: PathTestSet{Item1: []PathTestItem{{"A"}, {"D"}}}, Before: "1", After: "2"}, + &PathTestE{Underline: "E", Before: "1", After: "2"}, +} + +func TestUnmarshalPaths(t *testing.T) { + for _, pt := range pathTests { + v := reflect.New(reflect.TypeOf(pt).Elem()).Interface() + if err := Unmarshal([]byte(pathTestString), v); err != nil { + t.Fatalf("Unmarshal: %s", err) + } + if !reflect.DeepEqual(v, pt) { + t.Fatalf("have %#v\nwant %#v", v, pt) + } + } +} + +type BadPathTestA struct { + First string `xml:"items>item1"` + Other string `xml:"items>item2"` + Second string `xml:"items"` +} + +type BadPathTestB struct { + Other string `xml:"items>item2>value"` + First string `xml:"items>item1"` + Second string `xml:"items>item1>value"` +} + +type BadPathTestC struct { + First string + Second string `xml:"First"` +} + +type BadPathTestD struct { + BadPathEmbeddedA + BadPathEmbeddedB +} + +type BadPathEmbeddedA struct { + First string +} + +type BadPathEmbeddedB struct { + Second string `xml:"First"` +} + +var badPathTests = []struct { + v, e interface{} +}{ + {&BadPathTestA{}, &TagPathError{reflect.TypeOf(BadPathTestA{}), "First", "items>item1", "Second", "items"}}, + {&BadPathTestB{}, &TagPathError{reflect.TypeOf(BadPathTestB{}), "First", "items>item1", "Second", "items>item1>value"}}, + {&BadPathTestC{}, &TagPathError{reflect.TypeOf(BadPathTestC{}), "First", "", "Second", "First"}}, + {&BadPathTestD{}, &TagPathError{reflect.TypeOf(BadPathTestD{}), "First", "", "Second", "First"}}, +} + +func TestUnmarshalBadPaths(t *testing.T) { + for _, tt := range badPathTests { + err := Unmarshal([]byte(pathTestString), tt.v) + if !reflect.DeepEqual(err, tt.e) { + t.Fatalf("Unmarshal with %#v didn't fail properly:\nhave %#v,\nwant %#v", tt.v, err, tt.e) + } + } +} + +const OK = "OK" +const withoutNameTypeData = ` +<?xml version="1.0" charset="utf-8"?> +<Test3 Attr="OK" />` + +type TestThree struct { + XMLName Name `xml:"Test3"` + Attr string `xml:",attr"` +} + +func TestUnmarshalWithoutNameType(t *testing.T) { + var x TestThree + if err := Unmarshal([]byte(withoutNameTypeData), &x); err != nil { + t.Fatalf("Unmarshal: %s", err) + } + if x.Attr != OK { + t.Fatalf("have %v\nwant %v", x.Attr, OK) + } +} diff --git a/src/pkg/encoding/xml/typeinfo.go b/src/pkg/encoding/xml/typeinfo.go new file mode 100644 index 000000000..2bf2c6b30 --- /dev/null +++ b/src/pkg/encoding/xml/typeinfo.go @@ -0,0 +1,322 @@ +// Copyright 2011 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 xml + +import ( + "fmt" + "reflect" + "strings" + "sync" +) + +// typeInfo holds details for the xml representation of a type. +type typeInfo struct { + xmlname *fieldInfo + fields []fieldInfo +} + +// fieldInfo holds details for the xml representation of a single field. +type fieldInfo struct { + idx []int + name string + xmlns string + flags fieldFlags + parents []string +} + +type fieldFlags int + +const ( + fElement fieldFlags = 1 << iota + fAttr + fCharData + fInnerXml + fComment + fAny + + // TODO: + //fOmitEmpty + + fMode = fElement | fAttr | fCharData | fInnerXml | fComment | fAny +) + +var tinfoMap = make(map[reflect.Type]*typeInfo) +var tinfoLock sync.RWMutex + +var nameType = reflect.TypeOf(Name{}) + +// getTypeInfo returns the typeInfo structure with details necessary +// for marshalling and unmarshalling typ. +func getTypeInfo(typ reflect.Type) (*typeInfo, error) { + tinfoLock.RLock() + tinfo, ok := tinfoMap[typ] + tinfoLock.RUnlock() + if ok { + return tinfo, nil + } + tinfo = &typeInfo{} + if typ.Kind() == reflect.Struct && typ != nameType { + n := typ.NumField() + for i := 0; i < n; i++ { + f := typ.Field(i) + if f.PkgPath != "" || f.Tag.Get("xml") == "-" { + continue // Private field + } + + // For embedded structs, embed its fields. + if f.Anonymous { + if f.Type.Kind() != reflect.Struct { + continue + } + inner, err := getTypeInfo(f.Type) + if err != nil { + return nil, err + } + for _, finfo := range inner.fields { + finfo.idx = append([]int{i}, finfo.idx...) + if err := addFieldInfo(typ, tinfo, &finfo); err != nil { + return nil, err + } + } + continue + } + + finfo, err := structFieldInfo(typ, &f) + if err != nil { + return nil, err + } + + if f.Name == "XMLName" { + tinfo.xmlname = finfo + continue + } + + // Add the field if it doesn't conflict with other fields. + if err := addFieldInfo(typ, tinfo, finfo); err != nil { + return nil, err + } + } + } + tinfoLock.Lock() + tinfoMap[typ] = tinfo + tinfoLock.Unlock() + return tinfo, nil +} + +// structFieldInfo builds and returns a fieldInfo for f. +func structFieldInfo(typ reflect.Type, f *reflect.StructField) (*fieldInfo, error) { + finfo := &fieldInfo{idx: f.Index} + + // Split the tag from the xml namespace if necessary. + tag := f.Tag.Get("xml") + if i := strings.Index(tag, " "); i >= 0 { + finfo.xmlns, tag = tag[:i], tag[i+1:] + } + + // Parse flags. + tokens := strings.Split(tag, ",") + if len(tokens) == 1 { + finfo.flags = fElement + } else { + tag = tokens[0] + for _, flag := range tokens[1:] { + switch flag { + case "attr": + finfo.flags |= fAttr + case "chardata": + finfo.flags |= fCharData + case "innerxml": + finfo.flags |= fInnerXml + case "comment": + finfo.flags |= fComment + case "any": + finfo.flags |= fAny + } + } + + // Validate the flags used. + switch mode := finfo.flags & fMode; mode { + case 0: + finfo.flags |= fElement + case fAttr, fCharData, fInnerXml, fComment, fAny: + if f.Name != "XMLName" && (tag == "" || mode == fAttr) { + break + } + fallthrough + default: + // This will also catch multiple modes in a single field. + return nil, fmt.Errorf("xml: invalid tag in field %s of type %s: %q", + f.Name, typ, f.Tag.Get("xml")) + } + } + + // Use of xmlns without a name is not allowed. + if finfo.xmlns != "" && tag == "" { + return nil, fmt.Errorf("xml: namespace without name in field %s of type %s: %q", + f.Name, typ, f.Tag.Get("xml")) + } + + if f.Name == "XMLName" { + // The XMLName field records the XML element name. Don't + // process it as usual because its name should default to + // empty rather than to the field name. + finfo.name = tag + return finfo, nil + } + + if tag == "" { + // If the name part of the tag is completely empty, get + // default from XMLName of underlying struct if feasible, + // or field name otherwise. + if xmlname := lookupXMLName(f.Type); xmlname != nil { + finfo.xmlns, finfo.name = xmlname.xmlns, xmlname.name + } else { + finfo.name = f.Name + } + return finfo, nil + } + + // Prepare field name and parents. + tokens = strings.Split(tag, ">") + if tokens[0] == "" { + tokens[0] = f.Name + } + if tokens[len(tokens)-1] == "" { + return nil, fmt.Errorf("xml: trailing '>' in field %s of type %s", f.Name, typ) + } + finfo.name = tokens[len(tokens)-1] + if len(tokens) > 1 { + finfo.parents = tokens[:len(tokens)-1] + } + + // If the field type has an XMLName field, the names must match + // so that the behavior of both marshalling and unmarshalling + // is straighforward and unambiguous. + if finfo.flags&fElement != 0 { + ftyp := f.Type + xmlname := lookupXMLName(ftyp) + if xmlname != nil && xmlname.name != finfo.name { + return nil, fmt.Errorf("xml: name %q in tag of %s.%s conflicts with name %q in %s.XMLName", + finfo.name, typ, f.Name, xmlname.name, ftyp) + } + } + return finfo, nil +} + +// lookupXMLName returns the fieldInfo for typ's XMLName field +// in case it exists and has a valid xml field tag, otherwise +// it returns nil. +func lookupXMLName(typ reflect.Type) (xmlname *fieldInfo) { + for typ.Kind() == reflect.Ptr { + typ = typ.Elem() + } + if typ.Kind() != reflect.Struct { + return nil + } + for i, n := 0, typ.NumField(); i < n; i++ { + f := typ.Field(i) + if f.Name != "XMLName" { + continue + } + finfo, err := structFieldInfo(typ, &f) + if finfo.name != "" && err == nil { + return finfo + } + // Also consider errors as a non-existent field tag + // and let getTypeInfo itself report the error. + break + } + return nil +} + +func min(a, b int) int { + if a <= b { + return a + } + return b +} + +// addFieldInfo adds finfo to tinfo.fields if there are no +// conflicts, or if conflicts arise from previous fields that were +// obtained from deeper embedded structures than finfo. In the latter +// case, the conflicting entries are dropped. +// A conflict occurs when the path (parent + name) to a field is +// itself a prefix of another path, or when two paths match exactly. +// It is okay for field paths to share a common, shorter prefix. +func addFieldInfo(typ reflect.Type, tinfo *typeInfo, newf *fieldInfo) error { + var conflicts []int +Loop: + // First, figure all conflicts. Most working code will have none. + for i := range tinfo.fields { + oldf := &tinfo.fields[i] + if oldf.flags&fMode != newf.flags&fMode { + continue + } + minl := min(len(newf.parents), len(oldf.parents)) + for p := 0; p < minl; p++ { + if oldf.parents[p] != newf.parents[p] { + continue Loop + } + } + if len(oldf.parents) > len(newf.parents) { + if oldf.parents[len(newf.parents)] == newf.name { + conflicts = append(conflicts, i) + } + } else if len(oldf.parents) < len(newf.parents) { + if newf.parents[len(oldf.parents)] == oldf.name { + conflicts = append(conflicts, i) + } + } else { + if newf.name == oldf.name { + conflicts = append(conflicts, i) + } + } + } + // Without conflicts, add the new field and return. + if conflicts == nil { + tinfo.fields = append(tinfo.fields, *newf) + return nil + } + + // If any conflict is shallower, ignore the new field. + // This matches the Go field resolution on embedding. + for _, i := range conflicts { + if len(tinfo.fields[i].idx) < len(newf.idx) { + return nil + } + } + + // Otherwise, if any of them is at the same depth level, it's an error. + for _, i := range conflicts { + oldf := &tinfo.fields[i] + if len(oldf.idx) == len(newf.idx) { + f1 := typ.FieldByIndex(oldf.idx) + f2 := typ.FieldByIndex(newf.idx) + return &TagPathError{typ, f1.Name, f1.Tag.Get("xml"), f2.Name, f2.Tag.Get("xml")} + } + } + + // Otherwise, the new field is shallower, and thus takes precedence, + // so drop the conflicting fields from tinfo and append the new one. + for c := len(conflicts) - 1; c >= 0; c-- { + i := conflicts[c] + copy(tinfo.fields[i:], tinfo.fields[i+1:]) + tinfo.fields = tinfo.fields[:len(tinfo.fields)-1] + } + tinfo.fields = append(tinfo.fields, *newf) + return nil +} + +// A TagPathError represents an error in the unmarshalling process +// caused by the use of field tags with conflicting paths. +type TagPathError struct { + Struct reflect.Type + Field1, Tag1 string + Field2, Tag2 string +} + +func (e *TagPathError) Error() string { + return fmt.Sprintf("%s field %q with tag %q conflicts with field %q with tag %q", e.Struct, e.Field1, e.Tag1, e.Field2, e.Tag2) +} diff --git a/src/pkg/encoding/xml/xml.go b/src/pkg/encoding/xml/xml.go new file mode 100644 index 000000000..5066f5c01 --- /dev/null +++ b/src/pkg/encoding/xml/xml.go @@ -0,0 +1,1697 @@ +// Copyright 2009 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 xml implements a simple XML 1.0 parser that +// understands XML name spaces. +package xml + +// References: +// Annotated XML spec: http://www.xml.com/axml/testaxml.htm +// XML name spaces: http://www.w3.org/TR/REC-xml-names/ + +// TODO(rsc): +// Test error handling. + +import ( + "bufio" + "bytes" + "fmt" + "io" + "strconv" + "strings" + "unicode" + "unicode/utf8" +) + +// A SyntaxError represents a syntax error in the XML input stream. +type SyntaxError struct { + Msg string + Line int +} + +func (e *SyntaxError) Error() string { + return "XML syntax error on line " + strconv.Itoa(e.Line) + ": " + e.Msg +} + +// A Name represents an XML name (Local) annotated +// with a name space identifier (Space). +// In tokens returned by Decoder.Token, the Space identifier +// is given as a canonical URL, not the short prefix used +// in the document being parsed. +type Name struct { + Space, Local string +} + +// An Attr represents an attribute in an XML element (Name=Value). +type Attr struct { + Name Name + Value string +} + +// A Token is an interface holding one of the token types: +// StartElement, EndElement, CharData, Comment, ProcInst, or Directive. +type Token interface{} + +// A StartElement represents an XML start element. +type StartElement struct { + Name Name + Attr []Attr +} + +func (e StartElement) Copy() StartElement { + attrs := make([]Attr, len(e.Attr)) + copy(attrs, e.Attr) + e.Attr = attrs + return e +} + +// An EndElement represents an XML end element. +type EndElement struct { + Name Name +} + +// A CharData represents XML character data (raw text), +// in which XML escape sequences have been replaced by +// the characters they represent. +type CharData []byte + +func makeCopy(b []byte) []byte { + b1 := make([]byte, len(b)) + copy(b1, b) + return b1 +} + +func (c CharData) Copy() CharData { return CharData(makeCopy(c)) } + +// A Comment represents an XML comment of the form <!--comment-->. +// The bytes do not include the <!-- and --> comment markers. +type Comment []byte + +func (c Comment) Copy() Comment { return Comment(makeCopy(c)) } + +// A ProcInst represents an XML processing instruction of the form <?target inst?> +type ProcInst struct { + Target string + Inst []byte +} + +func (p ProcInst) Copy() ProcInst { + p.Inst = makeCopy(p.Inst) + return p +} + +// A Directive represents an XML directive of the form <!text>. +// The bytes do not include the <! and > markers. +type Directive []byte + +func (d Directive) Copy() Directive { return Directive(makeCopy(d)) } + +// CopyToken returns a copy of a Token. +func CopyToken(t Token) Token { + switch v := t.(type) { + case CharData: + return v.Copy() + case Comment: + return v.Copy() + case Directive: + return v.Copy() + case ProcInst: + return v.Copy() + case StartElement: + return v.Copy() + } + return t +} + +// A Decoder represents an XML parser reading a particular input stream. +// The parser assumes that its input is encoded in UTF-8. +type Decoder struct { + // Strict defaults to true, enforcing the requirements + // of the XML specification. + // If set to false, the parser allows input containing common + // mistakes: + // * If an element is missing an end tag, the parser invents + // end tags as necessary to keep the return values from Token + // properly balanced. + // * In attribute values and character data, unknown or malformed + // character entities (sequences beginning with &) are left alone. + // + // Setting: + // + // d.Strict = false; + // d.AutoClose = HTMLAutoClose; + // d.Entity = HTMLEntity + // + // creates a parser that can handle typical HTML. + Strict bool + + // When Strict == false, AutoClose indicates a set of elements to + // consider closed immediately after they are opened, regardless + // of whether an end element is present. + AutoClose []string + + // Entity can be used to map non-standard entity names to string replacements. + // The parser behaves as if these standard mappings are present in the map, + // regardless of the actual map content: + // + // "lt": "<", + // "gt": ">", + // "amp": "&", + // "apos": "'", + // "quot": `"`, + Entity map[string]string + + // CharsetReader, if non-nil, defines a function to generate + // charset-conversion readers, converting from the provided + // non-UTF-8 charset into UTF-8. If CharsetReader is nil or + // returns an error, parsing stops with an error. One of the + // the CharsetReader's result values must be non-nil. + CharsetReader func(charset string, input io.Reader) (io.Reader, error) + + r io.ByteReader + buf bytes.Buffer + saved *bytes.Buffer + stk *stack + free *stack + needClose bool + toClose Name + nextToken Token + nextByte int + ns map[string]string + err error + line int + tmp [32]byte +} + +// NewDecoder creates a new XML parser reading from r. +func NewDecoder(r io.Reader) *Decoder { + d := &Decoder{ + ns: make(map[string]string), + nextByte: -1, + line: 1, + Strict: true, + } + d.switchToReader(r) + return d +} + +// Token returns the next XML token in the input stream. +// At the end of the input stream, Token returns nil, io.EOF. +// +// Slices of bytes in the returned token data refer to the +// parser's internal buffer and remain valid only until the next +// call to Token. To acquire a copy of the bytes, call CopyToken +// or the token's Copy method. +// +// Token expands self-closing elements such as <br/> +// into separate start and end elements returned by successive calls. +// +// Token guarantees that the StartElement and EndElement +// tokens it returns are properly nested and matched: +// if Token encounters an unexpected end element, +// it will return an error. +// +// Token implements XML name spaces as described by +// http://www.w3.org/TR/REC-xml-names/. Each of the +// Name structures contained in the Token has the Space +// set to the URL identifying its name space when known. +// If Token encounters an unrecognized name space prefix, +// it uses the prefix as the Space rather than report an error. +func (d *Decoder) Token() (t Token, err error) { + if d.nextToken != nil { + t = d.nextToken + d.nextToken = nil + } else if t, err = d.RawToken(); err != nil { + return + } + + if !d.Strict { + if t1, ok := d.autoClose(t); ok { + d.nextToken = t + t = t1 + } + } + switch t1 := t.(type) { + case StartElement: + // In XML name spaces, the translations listed in the + // attributes apply to the element name and + // to the other attribute names, so process + // the translations first. + for _, a := range t1.Attr { + if a.Name.Space == "xmlns" { + v, ok := d.ns[a.Name.Local] + d.pushNs(a.Name.Local, v, ok) + d.ns[a.Name.Local] = a.Value + } + if a.Name.Space == "" && a.Name.Local == "xmlns" { + // Default space for untagged names + v, ok := d.ns[""] + d.pushNs("", v, ok) + d.ns[""] = a.Value + } + } + + d.translate(&t1.Name, true) + for i := range t1.Attr { + d.translate(&t1.Attr[i].Name, false) + } + d.pushElement(t1.Name) + t = t1 + + case EndElement: + d.translate(&t1.Name, true) + if !d.popElement(&t1) { + return nil, d.err + } + t = t1 + } + return +} + +// Apply name space translation to name n. +// The default name space (for Space=="") +// applies only to element names, not to attribute names. +func (d *Decoder) translate(n *Name, isElementName bool) { + switch { + case n.Space == "xmlns": + return + case n.Space == "" && !isElementName: + return + case n.Space == "" && n.Local == "xmlns": + return + } + if v, ok := d.ns[n.Space]; ok { + n.Space = v + } +} + +func (d *Decoder) switchToReader(r io.Reader) { + // Get efficient byte at a time reader. + // Assume that if reader has its own + // ReadByte, it's efficient enough. + // Otherwise, use bufio. + if rb, ok := r.(io.ByteReader); ok { + d.r = rb + } else { + d.r = bufio.NewReader(r) + } +} + +// Parsing state - stack holds old name space translations +// and the current set of open elements. The translations to pop when +// ending a given tag are *below* it on the stack, which is +// more work but forced on us by XML. +type stack struct { + next *stack + kind int + name Name + ok bool +} + +const ( + stkStart = iota + stkNs +) + +func (d *Decoder) push(kind int) *stack { + s := d.free + if s != nil { + d.free = s.next + } else { + s = new(stack) + } + s.next = d.stk + s.kind = kind + d.stk = s + return s +} + +func (d *Decoder) pop() *stack { + s := d.stk + if s != nil { + d.stk = s.next + s.next = d.free + d.free = s + } + return s +} + +// Record that we are starting an element with the given name. +func (d *Decoder) pushElement(name Name) { + s := d.push(stkStart) + s.name = name +} + +// Record that we are changing the value of ns[local]. +// The old value is url, ok. +func (d *Decoder) pushNs(local string, url string, ok bool) { + s := d.push(stkNs) + s.name.Local = local + s.name.Space = url + s.ok = ok +} + +// Creates a SyntaxError with the current line number. +func (d *Decoder) syntaxError(msg string) error { + return &SyntaxError{Msg: msg, Line: d.line} +} + +// Record that we are ending an element with the given name. +// The name must match the record at the top of the stack, +// which must be a pushElement record. +// After popping the element, apply any undo records from +// the stack to restore the name translations that existed +// before we saw this element. +func (d *Decoder) popElement(t *EndElement) bool { + s := d.pop() + name := t.Name + switch { + case s == nil || s.kind != stkStart: + d.err = d.syntaxError("unexpected end element </" + name.Local + ">") + return false + case s.name.Local != name.Local: + if !d.Strict { + d.needClose = true + d.toClose = t.Name + t.Name = s.name + return true + } + d.err = d.syntaxError("element <" + s.name.Local + "> closed by </" + name.Local + ">") + return false + case s.name.Space != name.Space: + d.err = d.syntaxError("element <" + s.name.Local + "> in space " + s.name.Space + + "closed by </" + name.Local + "> in space " + name.Space) + return false + } + + // Pop stack until a Start is on the top, undoing the + // translations that were associated with the element we just closed. + for d.stk != nil && d.stk.kind != stkStart { + s := d.pop() + if s.ok { + d.ns[s.name.Local] = s.name.Space + } else { + delete(d.ns, s.name.Local) + } + } + + return true +} + +// If the top element on the stack is autoclosing and +// t is not the end tag, invent the end tag. +func (d *Decoder) autoClose(t Token) (Token, bool) { + if d.stk == nil || d.stk.kind != stkStart { + return nil, false + } + name := strings.ToLower(d.stk.name.Local) + for _, s := range d.AutoClose { + if strings.ToLower(s) == name { + // This one should be auto closed if t doesn't close it. + et, ok := t.(EndElement) + if !ok || et.Name.Local != name { + return EndElement{d.stk.name}, true + } + break + } + } + return nil, false +} + +// RawToken is like Token but does not verify that +// start and end elements match and does not translate +// name space prefixes to their corresponding URLs. +func (d *Decoder) RawToken() (Token, error) { + if d.err != nil { + return nil, d.err + } + if d.needClose { + // The last element we read was self-closing and + // we returned just the StartElement half. + // Return the EndElement half now. + d.needClose = false + return EndElement{d.toClose}, nil + } + + b, ok := d.getc() + if !ok { + return nil, d.err + } + + if b != '<' { + // Text section. + d.ungetc(b) + data := d.text(-1, false) + if data == nil { + return nil, d.err + } + return CharData(data), nil + } + + if b, ok = d.mustgetc(); !ok { + return nil, d.err + } + switch b { + case '/': + // </: End element + var name Name + if name, ok = d.nsname(); !ok { + if d.err == nil { + d.err = d.syntaxError("expected element name after </") + } + return nil, d.err + } + d.space() + if b, ok = d.mustgetc(); !ok { + return nil, d.err + } + if b != '>' { + d.err = d.syntaxError("invalid characters between </" + name.Local + " and >") + return nil, d.err + } + return EndElement{name}, nil + + case '?': + // <?: Processing instruction. + // TODO(rsc): Should parse the <?xml declaration to make sure + // the version is 1.0 and the encoding is UTF-8. + var target string + if target, ok = d.name(); !ok { + if d.err == nil { + d.err = d.syntaxError("expected target name after <?") + } + return nil, d.err + } + d.space() + d.buf.Reset() + var b0 byte + for { + if b, ok = d.mustgetc(); !ok { + return nil, d.err + } + d.buf.WriteByte(b) + if b0 == '?' && b == '>' { + break + } + b0 = b + } + data := d.buf.Bytes() + data = data[0 : len(data)-2] // chop ?> + + if target == "xml" { + enc := procInstEncoding(string(data)) + if enc != "" && enc != "utf-8" && enc != "UTF-8" { + if d.CharsetReader == nil { + d.err = fmt.Errorf("xml: encoding %q declared but Decoder.CharsetReader is nil", enc) + return nil, d.err + } + newr, err := d.CharsetReader(enc, d.r.(io.Reader)) + if err != nil { + d.err = fmt.Errorf("xml: opening charset %q: %v", enc, err) + return nil, d.err + } + if newr == nil { + panic("CharsetReader returned a nil Reader for charset " + enc) + } + d.switchToReader(newr) + } + } + return ProcInst{target, data}, nil + + case '!': + // <!: Maybe comment, maybe CDATA. + if b, ok = d.mustgetc(); !ok { + return nil, d.err + } + switch b { + case '-': // <!- + // Probably <!-- for a comment. + if b, ok = d.mustgetc(); !ok { + return nil, d.err + } + if b != '-' { + d.err = d.syntaxError("invalid sequence <!- not part of <!--") + return nil, d.err + } + // Look for terminator. + d.buf.Reset() + var b0, b1 byte + for { + if b, ok = d.mustgetc(); !ok { + return nil, d.err + } + d.buf.WriteByte(b) + if b0 == '-' && b1 == '-' && b == '>' { + break + } + b0, b1 = b1, b + } + data := d.buf.Bytes() + data = data[0 : len(data)-3] // chop --> + return Comment(data), nil + + case '[': // <![ + // Probably <![CDATA[. + for i := 0; i < 6; i++ { + if b, ok = d.mustgetc(); !ok { + return nil, d.err + } + if b != "CDATA["[i] { + d.err = d.syntaxError("invalid <![ sequence") + return nil, d.err + } + } + // Have <![CDATA[. Read text until ]]>. + data := d.text(-1, true) + if data == nil { + return nil, d.err + } + return CharData(data), nil + } + + // Probably a directive: <!DOCTYPE ...>, <!ENTITY ...>, etc. + // We don't care, but accumulate for caller. Quoted angle + // brackets do not count for nesting. + d.buf.Reset() + d.buf.WriteByte(b) + inquote := uint8(0) + depth := 0 + for { + if b, ok = d.mustgetc(); !ok { + return nil, d.err + } + if inquote == 0 && b == '>' && depth == 0 { + break + } + d.buf.WriteByte(b) + switch { + case b == inquote: + inquote = 0 + + case inquote != 0: + // in quotes, no special action + + case b == '\'' || b == '"': + inquote = b + + case b == '>' && inquote == 0: + depth-- + + case b == '<' && inquote == 0: + depth++ + } + } + return Directive(d.buf.Bytes()), nil + } + + // Must be an open element like <a href="foo"> + d.ungetc(b) + + var ( + name Name + empty bool + attr []Attr + ) + if name, ok = d.nsname(); !ok { + if d.err == nil { + d.err = d.syntaxError("expected element name after <") + } + return nil, d.err + } + + attr = make([]Attr, 0, 4) + for { + d.space() + if b, ok = d.mustgetc(); !ok { + return nil, d.err + } + if b == '/' { + empty = true + if b, ok = d.mustgetc(); !ok { + return nil, d.err + } + if b != '>' { + d.err = d.syntaxError("expected /> in element") + return nil, d.err + } + break + } + if b == '>' { + break + } + d.ungetc(b) + + n := len(attr) + if n >= cap(attr) { + nattr := make([]Attr, n, 2*cap(attr)) + copy(nattr, attr) + attr = nattr + } + attr = attr[0 : n+1] + a := &attr[n] + if a.Name, ok = d.nsname(); !ok { + if d.err == nil { + d.err = d.syntaxError("expected attribute name in element") + } + return nil, d.err + } + d.space() + if b, ok = d.mustgetc(); !ok { + return nil, d.err + } + if b != '=' { + if d.Strict { + d.err = d.syntaxError("attribute name without = in element") + return nil, d.err + } else { + d.ungetc(b) + a.Value = a.Name.Local + } + } else { + d.space() + data := d.attrval() + if data == nil { + return nil, d.err + } + a.Value = string(data) + } + } + if empty { + d.needClose = true + d.toClose = name + } + return StartElement{name, attr}, nil +} + +func (d *Decoder) attrval() []byte { + b, ok := d.mustgetc() + if !ok { + return nil + } + // Handle quoted attribute values + if b == '"' || b == '\'' { + return d.text(int(b), false) + } + // Handle unquoted attribute values for strict parsers + if d.Strict { + d.err = d.syntaxError("unquoted or missing attribute value in element") + return nil + } + // Handle unquoted attribute values for unstrict parsers + d.ungetc(b) + d.buf.Reset() + for { + b, ok = d.mustgetc() + if !ok { + return nil + } + // http://www.w3.org/TR/REC-html40/intro/sgmltut.html#h-3.2.2 + if 'a' <= b && b <= 'z' || 'A' <= b && b <= 'Z' || + '0' <= b && b <= '9' || b == '_' || b == ':' || b == '-' { + d.buf.WriteByte(b) + } else { + d.ungetc(b) + break + } + } + return d.buf.Bytes() +} + +// Skip spaces if any +func (d *Decoder) space() { + for { + b, ok := d.getc() + if !ok { + return + } + switch b { + case ' ', '\r', '\n', '\t': + default: + d.ungetc(b) + return + } + } +} + +// Read a single byte. +// If there is no byte to read, return ok==false +// and leave the error in d.err. +// Maintain line number. +func (d *Decoder) getc() (b byte, ok bool) { + if d.err != nil { + return 0, false + } + if d.nextByte >= 0 { + b = byte(d.nextByte) + d.nextByte = -1 + } else { + b, d.err = d.r.ReadByte() + if d.err != nil { + return 0, false + } + if d.saved != nil { + d.saved.WriteByte(b) + } + } + if b == '\n' { + d.line++ + } + return b, true +} + +// Return saved offset. +// If we did ungetc (nextByte >= 0), have to back up one. +func (d *Decoder) savedOffset() int { + n := d.saved.Len() + if d.nextByte >= 0 { + n-- + } + return n +} + +// Must read a single byte. +// If there is no byte to read, +// set d.err to SyntaxError("unexpected EOF") +// and return ok==false +func (d *Decoder) mustgetc() (b byte, ok bool) { + if b, ok = d.getc(); !ok { + if d.err == io.EOF { + d.err = d.syntaxError("unexpected EOF") + } + } + return +} + +// Unread a single byte. +func (d *Decoder) ungetc(b byte) { + if b == '\n' { + d.line-- + } + d.nextByte = int(b) +} + +var entity = map[string]int{ + "lt": '<', + "gt": '>', + "amp": '&', + "apos": '\'', + "quot": '"', +} + +// Read plain text section (XML calls it character data). +// If quote >= 0, we are in a quoted string and need to find the matching quote. +// If cdata == true, we are in a <![CDATA[ section and need to find ]]>. +// On failure return nil and leave the error in d.err. +func (d *Decoder) text(quote int, cdata bool) []byte { + var b0, b1 byte + var trunc int + d.buf.Reset() +Input: + for { + b, ok := d.getc() + if !ok { + if cdata { + if d.err == io.EOF { + d.err = d.syntaxError("unexpected EOF in CDATA section") + } + return nil + } + break Input + } + + // <![CDATA[ section ends with ]]>. + // It is an error for ]]> to appear in ordinary text. + if b0 == ']' && b1 == ']' && b == '>' { + if cdata { + trunc = 2 + break Input + } + d.err = d.syntaxError("unescaped ]]> not in CDATA section") + return nil + } + + // Stop reading text if we see a <. + if b == '<' && !cdata { + if quote >= 0 { + d.err = d.syntaxError("unescaped < inside quoted string") + return nil + } + d.ungetc('<') + break Input + } + if quote >= 0 && b == byte(quote) { + break Input + } + if b == '&' && !cdata { + // Read escaped character expression up to semicolon. + // XML in all its glory allows a document to define and use + // its own character names with <!ENTITY ...> directives. + // Parsers are required to recognize lt, gt, amp, apos, and quot + // even if they have not been declared. That's all we allow. + var i int + for i = 0; i < len(d.tmp); i++ { + var ok bool + d.tmp[i], ok = d.getc() + if !ok { + if d.err == io.EOF { + d.err = d.syntaxError("unexpected EOF") + } + return nil + } + c := d.tmp[i] + if c == ';' { + break + } + if 'a' <= c && c <= 'z' || + 'A' <= c && c <= 'Z' || + '0' <= c && c <= '9' || + c == '_' || c == '#' { + continue + } + d.ungetc(c) + break + } + s := string(d.tmp[0:i]) + if i >= len(d.tmp) { + if !d.Strict { + b0, b1 = 0, 0 + d.buf.WriteByte('&') + d.buf.Write(d.tmp[0:i]) + continue Input + } + d.err = d.syntaxError("character entity expression &" + s + "... too long") + return nil + } + var haveText bool + var text string + if i >= 2 && s[0] == '#' { + var n uint64 + var err error + if i >= 3 && s[1] == 'x' { + n, err = strconv.ParseUint(s[2:], 16, 64) + } else { + n, err = strconv.ParseUint(s[1:], 10, 64) + } + if err == nil && n <= unicode.MaxRune { + text = string(n) + haveText = true + } + } else { + if r, ok := entity[s]; ok { + text = string(r) + haveText = true + } else if d.Entity != nil { + text, haveText = d.Entity[s] + } + } + if !haveText { + if !d.Strict { + b0, b1 = 0, 0 + d.buf.WriteByte('&') + d.buf.Write(d.tmp[0:i]) + continue Input + } + d.err = d.syntaxError("invalid character entity &" + s + ";") + return nil + } + d.buf.Write([]byte(text)) + b0, b1 = 0, 0 + continue Input + } + d.buf.WriteByte(b) + b0, b1 = b1, b + } + data := d.buf.Bytes() + data = data[0 : len(data)-trunc] + + // Inspect each rune for being a disallowed character. + buf := data + for len(buf) > 0 { + r, size := utf8.DecodeRune(buf) + if r == utf8.RuneError && size == 1 { + d.err = d.syntaxError("invalid UTF-8") + return nil + } + buf = buf[size:] + if !isInCharacterRange(r) { + d.err = d.syntaxError(fmt.Sprintf("illegal character code %U", r)) + return nil + } + } + + // Must rewrite \r and \r\n into \n. + w := 0 + for r := 0; r < len(data); r++ { + b := data[r] + if b == '\r' { + if r+1 < len(data) && data[r+1] == '\n' { + continue + } + b = '\n' + } + data[w] = b + w++ + } + return data[0:w] +} + +// Decide whether the given rune is in the XML Character Range, per +// the Char production of http://www.xml.com/axml/testaxml.htm, +// Section 2.2 Characters. +func isInCharacterRange(r rune) (inrange bool) { + return r == 0x09 || + r == 0x0A || + r == 0x0D || + r >= 0x20 && r <= 0xDF77 || + r >= 0xE000 && r <= 0xFFFD || + r >= 0x10000 && r <= 0x10FFFF +} + +// Get name space name: name with a : stuck in the middle. +// The part before the : is the name space identifier. +func (d *Decoder) nsname() (name Name, ok bool) { + s, ok := d.name() + if !ok { + return + } + i := strings.Index(s, ":") + if i < 0 { + name.Local = s + } else { + name.Space = s[0:i] + name.Local = s[i+1:] + } + return name, true +} + +// Get name: /first(first|second)*/ +// Do not set d.err if the name is missing (unless unexpected EOF is received): +// let the caller provide better context. +func (d *Decoder) name() (s string, ok bool) { + var b byte + if b, ok = d.mustgetc(); !ok { + return + } + + // As a first approximation, we gather the bytes [A-Za-z_:.-\x80-\xFF]* + if b < utf8.RuneSelf && !isNameByte(b) { + d.ungetc(b) + return "", false + } + d.buf.Reset() + d.buf.WriteByte(b) + for { + if b, ok = d.mustgetc(); !ok { + return + } + if b < utf8.RuneSelf && !isNameByte(b) { + d.ungetc(b) + break + } + d.buf.WriteByte(b) + } + + // Then we check the characters. + s = d.buf.String() + for i, c := range s { + if !unicode.Is(first, c) && (i == 0 || !unicode.Is(second, c)) { + d.err = d.syntaxError("invalid XML name: " + s) + return "", false + } + } + return s, true +} + +func isNameByte(c byte) bool { + return 'A' <= c && c <= 'Z' || + 'a' <= c && c <= 'z' || + '0' <= c && c <= '9' || + c == '_' || c == ':' || c == '.' || c == '-' +} + +// These tables were generated by cut and paste from Appendix B of +// the XML spec at http://www.xml.com/axml/testaxml.htm +// and then reformatting. First corresponds to (Letter | '_' | ':') +// and second corresponds to NameChar. + +var first = &unicode.RangeTable{ + R16: []unicode.Range16{ + {0x003A, 0x003A, 1}, + {0x0041, 0x005A, 1}, + {0x005F, 0x005F, 1}, + {0x0061, 0x007A, 1}, + {0x00C0, 0x00D6, 1}, + {0x00D8, 0x00F6, 1}, + {0x00F8, 0x00FF, 1}, + {0x0100, 0x0131, 1}, + {0x0134, 0x013E, 1}, + {0x0141, 0x0148, 1}, + {0x014A, 0x017E, 1}, + {0x0180, 0x01C3, 1}, + {0x01CD, 0x01F0, 1}, + {0x01F4, 0x01F5, 1}, + {0x01FA, 0x0217, 1}, + {0x0250, 0x02A8, 1}, + {0x02BB, 0x02C1, 1}, + {0x0386, 0x0386, 1}, + {0x0388, 0x038A, 1}, + {0x038C, 0x038C, 1}, + {0x038E, 0x03A1, 1}, + {0x03A3, 0x03CE, 1}, + {0x03D0, 0x03D6, 1}, + {0x03DA, 0x03E0, 2}, + {0x03E2, 0x03F3, 1}, + {0x0401, 0x040C, 1}, + {0x040E, 0x044F, 1}, + {0x0451, 0x045C, 1}, + {0x045E, 0x0481, 1}, + {0x0490, 0x04C4, 1}, + {0x04C7, 0x04C8, 1}, + {0x04CB, 0x04CC, 1}, + {0x04D0, 0x04EB, 1}, + {0x04EE, 0x04F5, 1}, + {0x04F8, 0x04F9, 1}, + {0x0531, 0x0556, 1}, + {0x0559, 0x0559, 1}, + {0x0561, 0x0586, 1}, + {0x05D0, 0x05EA, 1}, + {0x05F0, 0x05F2, 1}, + {0x0621, 0x063A, 1}, + {0x0641, 0x064A, 1}, + {0x0671, 0x06B7, 1}, + {0x06BA, 0x06BE, 1}, + {0x06C0, 0x06CE, 1}, + {0x06D0, 0x06D3, 1}, + {0x06D5, 0x06D5, 1}, + {0x06E5, 0x06E6, 1}, + {0x0905, 0x0939, 1}, + {0x093D, 0x093D, 1}, + {0x0958, 0x0961, 1}, + {0x0985, 0x098C, 1}, + {0x098F, 0x0990, 1}, + {0x0993, 0x09A8, 1}, + {0x09AA, 0x09B0, 1}, + {0x09B2, 0x09B2, 1}, + {0x09B6, 0x09B9, 1}, + {0x09DC, 0x09DD, 1}, + {0x09DF, 0x09E1, 1}, + {0x09F0, 0x09F1, 1}, + {0x0A05, 0x0A0A, 1}, + {0x0A0F, 0x0A10, 1}, + {0x0A13, 0x0A28, 1}, + {0x0A2A, 0x0A30, 1}, + {0x0A32, 0x0A33, 1}, + {0x0A35, 0x0A36, 1}, + {0x0A38, 0x0A39, 1}, + {0x0A59, 0x0A5C, 1}, + {0x0A5E, 0x0A5E, 1}, + {0x0A72, 0x0A74, 1}, + {0x0A85, 0x0A8B, 1}, + {0x0A8D, 0x0A8D, 1}, + {0x0A8F, 0x0A91, 1}, + {0x0A93, 0x0AA8, 1}, + {0x0AAA, 0x0AB0, 1}, + {0x0AB2, 0x0AB3, 1}, + {0x0AB5, 0x0AB9, 1}, + {0x0ABD, 0x0AE0, 0x23}, + {0x0B05, 0x0B0C, 1}, + {0x0B0F, 0x0B10, 1}, + {0x0B13, 0x0B28, 1}, + {0x0B2A, 0x0B30, 1}, + {0x0B32, 0x0B33, 1}, + {0x0B36, 0x0B39, 1}, + {0x0B3D, 0x0B3D, 1}, + {0x0B5C, 0x0B5D, 1}, + {0x0B5F, 0x0B61, 1}, + {0x0B85, 0x0B8A, 1}, + {0x0B8E, 0x0B90, 1}, + {0x0B92, 0x0B95, 1}, + {0x0B99, 0x0B9A, 1}, + {0x0B9C, 0x0B9C, 1}, + {0x0B9E, 0x0B9F, 1}, + {0x0BA3, 0x0BA4, 1}, + {0x0BA8, 0x0BAA, 1}, + {0x0BAE, 0x0BB5, 1}, + {0x0BB7, 0x0BB9, 1}, + {0x0C05, 0x0C0C, 1}, + {0x0C0E, 0x0C10, 1}, + {0x0C12, 0x0C28, 1}, + {0x0C2A, 0x0C33, 1}, + {0x0C35, 0x0C39, 1}, + {0x0C60, 0x0C61, 1}, + {0x0C85, 0x0C8C, 1}, + {0x0C8E, 0x0C90, 1}, + {0x0C92, 0x0CA8, 1}, + {0x0CAA, 0x0CB3, 1}, + {0x0CB5, 0x0CB9, 1}, + {0x0CDE, 0x0CDE, 1}, + {0x0CE0, 0x0CE1, 1}, + {0x0D05, 0x0D0C, 1}, + {0x0D0E, 0x0D10, 1}, + {0x0D12, 0x0D28, 1}, + {0x0D2A, 0x0D39, 1}, + {0x0D60, 0x0D61, 1}, + {0x0E01, 0x0E2E, 1}, + {0x0E30, 0x0E30, 1}, + {0x0E32, 0x0E33, 1}, + {0x0E40, 0x0E45, 1}, + {0x0E81, 0x0E82, 1}, + {0x0E84, 0x0E84, 1}, + {0x0E87, 0x0E88, 1}, + {0x0E8A, 0x0E8D, 3}, + {0x0E94, 0x0E97, 1}, + {0x0E99, 0x0E9F, 1}, + {0x0EA1, 0x0EA3, 1}, + {0x0EA5, 0x0EA7, 2}, + {0x0EAA, 0x0EAB, 1}, + {0x0EAD, 0x0EAE, 1}, + {0x0EB0, 0x0EB0, 1}, + {0x0EB2, 0x0EB3, 1}, + {0x0EBD, 0x0EBD, 1}, + {0x0EC0, 0x0EC4, 1}, + {0x0F40, 0x0F47, 1}, + {0x0F49, 0x0F69, 1}, + {0x10A0, 0x10C5, 1}, + {0x10D0, 0x10F6, 1}, + {0x1100, 0x1100, 1}, + {0x1102, 0x1103, 1}, + {0x1105, 0x1107, 1}, + {0x1109, 0x1109, 1}, + {0x110B, 0x110C, 1}, + {0x110E, 0x1112, 1}, + {0x113C, 0x1140, 2}, + {0x114C, 0x1150, 2}, + {0x1154, 0x1155, 1}, + {0x1159, 0x1159, 1}, + {0x115F, 0x1161, 1}, + {0x1163, 0x1169, 2}, + {0x116D, 0x116E, 1}, + {0x1172, 0x1173, 1}, + {0x1175, 0x119E, 0x119E - 0x1175}, + {0x11A8, 0x11AB, 0x11AB - 0x11A8}, + {0x11AE, 0x11AF, 1}, + {0x11B7, 0x11B8, 1}, + {0x11BA, 0x11BA, 1}, + {0x11BC, 0x11C2, 1}, + {0x11EB, 0x11F0, 0x11F0 - 0x11EB}, + {0x11F9, 0x11F9, 1}, + {0x1E00, 0x1E9B, 1}, + {0x1EA0, 0x1EF9, 1}, + {0x1F00, 0x1F15, 1}, + {0x1F18, 0x1F1D, 1}, + {0x1F20, 0x1F45, 1}, + {0x1F48, 0x1F4D, 1}, + {0x1F50, 0x1F57, 1}, + {0x1F59, 0x1F5B, 0x1F5B - 0x1F59}, + {0x1F5D, 0x1F5D, 1}, + {0x1F5F, 0x1F7D, 1}, + {0x1F80, 0x1FB4, 1}, + {0x1FB6, 0x1FBC, 1}, + {0x1FBE, 0x1FBE, 1}, + {0x1FC2, 0x1FC4, 1}, + {0x1FC6, 0x1FCC, 1}, + {0x1FD0, 0x1FD3, 1}, + {0x1FD6, 0x1FDB, 1}, + {0x1FE0, 0x1FEC, 1}, + {0x1FF2, 0x1FF4, 1}, + {0x1FF6, 0x1FFC, 1}, + {0x2126, 0x2126, 1}, + {0x212A, 0x212B, 1}, + {0x212E, 0x212E, 1}, + {0x2180, 0x2182, 1}, + {0x3007, 0x3007, 1}, + {0x3021, 0x3029, 1}, + {0x3041, 0x3094, 1}, + {0x30A1, 0x30FA, 1}, + {0x3105, 0x312C, 1}, + {0x4E00, 0x9FA5, 1}, + {0xAC00, 0xD7A3, 1}, + }, +} + +var second = &unicode.RangeTable{ + R16: []unicode.Range16{ + {0x002D, 0x002E, 1}, + {0x0030, 0x0039, 1}, + {0x00B7, 0x00B7, 1}, + {0x02D0, 0x02D1, 1}, + {0x0300, 0x0345, 1}, + {0x0360, 0x0361, 1}, + {0x0387, 0x0387, 1}, + {0x0483, 0x0486, 1}, + {0x0591, 0x05A1, 1}, + {0x05A3, 0x05B9, 1}, + {0x05BB, 0x05BD, 1}, + {0x05BF, 0x05BF, 1}, + {0x05C1, 0x05C2, 1}, + {0x05C4, 0x0640, 0x0640 - 0x05C4}, + {0x064B, 0x0652, 1}, + {0x0660, 0x0669, 1}, + {0x0670, 0x0670, 1}, + {0x06D6, 0x06DC, 1}, + {0x06DD, 0x06DF, 1}, + {0x06E0, 0x06E4, 1}, + {0x06E7, 0x06E8, 1}, + {0x06EA, 0x06ED, 1}, + {0x06F0, 0x06F9, 1}, + {0x0901, 0x0903, 1}, + {0x093C, 0x093C, 1}, + {0x093E, 0x094C, 1}, + {0x094D, 0x094D, 1}, + {0x0951, 0x0954, 1}, + {0x0962, 0x0963, 1}, + {0x0966, 0x096F, 1}, + {0x0981, 0x0983, 1}, + {0x09BC, 0x09BC, 1}, + {0x09BE, 0x09BF, 1}, + {0x09C0, 0x09C4, 1}, + {0x09C7, 0x09C8, 1}, + {0x09CB, 0x09CD, 1}, + {0x09D7, 0x09D7, 1}, + {0x09E2, 0x09E3, 1}, + {0x09E6, 0x09EF, 1}, + {0x0A02, 0x0A3C, 0x3A}, + {0x0A3E, 0x0A3F, 1}, + {0x0A40, 0x0A42, 1}, + {0x0A47, 0x0A48, 1}, + {0x0A4B, 0x0A4D, 1}, + {0x0A66, 0x0A6F, 1}, + {0x0A70, 0x0A71, 1}, + {0x0A81, 0x0A83, 1}, + {0x0ABC, 0x0ABC, 1}, + {0x0ABE, 0x0AC5, 1}, + {0x0AC7, 0x0AC9, 1}, + {0x0ACB, 0x0ACD, 1}, + {0x0AE6, 0x0AEF, 1}, + {0x0B01, 0x0B03, 1}, + {0x0B3C, 0x0B3C, 1}, + {0x0B3E, 0x0B43, 1}, + {0x0B47, 0x0B48, 1}, + {0x0B4B, 0x0B4D, 1}, + {0x0B56, 0x0B57, 1}, + {0x0B66, 0x0B6F, 1}, + {0x0B82, 0x0B83, 1}, + {0x0BBE, 0x0BC2, 1}, + {0x0BC6, 0x0BC8, 1}, + {0x0BCA, 0x0BCD, 1}, + {0x0BD7, 0x0BD7, 1}, + {0x0BE7, 0x0BEF, 1}, + {0x0C01, 0x0C03, 1}, + {0x0C3E, 0x0C44, 1}, + {0x0C46, 0x0C48, 1}, + {0x0C4A, 0x0C4D, 1}, + {0x0C55, 0x0C56, 1}, + {0x0C66, 0x0C6F, 1}, + {0x0C82, 0x0C83, 1}, + {0x0CBE, 0x0CC4, 1}, + {0x0CC6, 0x0CC8, 1}, + {0x0CCA, 0x0CCD, 1}, + {0x0CD5, 0x0CD6, 1}, + {0x0CE6, 0x0CEF, 1}, + {0x0D02, 0x0D03, 1}, + {0x0D3E, 0x0D43, 1}, + {0x0D46, 0x0D48, 1}, + {0x0D4A, 0x0D4D, 1}, + {0x0D57, 0x0D57, 1}, + {0x0D66, 0x0D6F, 1}, + {0x0E31, 0x0E31, 1}, + {0x0E34, 0x0E3A, 1}, + {0x0E46, 0x0E46, 1}, + {0x0E47, 0x0E4E, 1}, + {0x0E50, 0x0E59, 1}, + {0x0EB1, 0x0EB1, 1}, + {0x0EB4, 0x0EB9, 1}, + {0x0EBB, 0x0EBC, 1}, + {0x0EC6, 0x0EC6, 1}, + {0x0EC8, 0x0ECD, 1}, + {0x0ED0, 0x0ED9, 1}, + {0x0F18, 0x0F19, 1}, + {0x0F20, 0x0F29, 1}, + {0x0F35, 0x0F39, 2}, + {0x0F3E, 0x0F3F, 1}, + {0x0F71, 0x0F84, 1}, + {0x0F86, 0x0F8B, 1}, + {0x0F90, 0x0F95, 1}, + {0x0F97, 0x0F97, 1}, + {0x0F99, 0x0FAD, 1}, + {0x0FB1, 0x0FB7, 1}, + {0x0FB9, 0x0FB9, 1}, + {0x20D0, 0x20DC, 1}, + {0x20E1, 0x3005, 0x3005 - 0x20E1}, + {0x302A, 0x302F, 1}, + {0x3031, 0x3035, 1}, + {0x3099, 0x309A, 1}, + {0x309D, 0x309E, 1}, + {0x30FC, 0x30FE, 1}, + }, +} + +// HTMLEntity is an entity map containing translations for the +// standard HTML entity characters. +var HTMLEntity = htmlEntity + +var htmlEntity = map[string]string{ + /* + hget http://www.w3.org/TR/html4/sgml/entities.html | + ssam ' + ,y /\>/ x/\<(.|\n)+/ s/\n/ /g + ,x v/^\<!ENTITY/d + ,s/\<!ENTITY ([^ ]+) .*U\+([0-9A-F][0-9A-F][0-9A-F][0-9A-F]) .+/ "\1": "\\u\2",/g + ' + */ + "nbsp": "\u00A0", + "iexcl": "\u00A1", + "cent": "\u00A2", + "pound": "\u00A3", + "curren": "\u00A4", + "yen": "\u00A5", + "brvbar": "\u00A6", + "sect": "\u00A7", + "uml": "\u00A8", + "copy": "\u00A9", + "ordf": "\u00AA", + "laquo": "\u00AB", + "not": "\u00AC", + "shy": "\u00AD", + "reg": "\u00AE", + "macr": "\u00AF", + "deg": "\u00B0", + "plusmn": "\u00B1", + "sup2": "\u00B2", + "sup3": "\u00B3", + "acute": "\u00B4", + "micro": "\u00B5", + "para": "\u00B6", + "middot": "\u00B7", + "cedil": "\u00B8", + "sup1": "\u00B9", + "ordm": "\u00BA", + "raquo": "\u00BB", + "frac14": "\u00BC", + "frac12": "\u00BD", + "frac34": "\u00BE", + "iquest": "\u00BF", + "Agrave": "\u00C0", + "Aacute": "\u00C1", + "Acirc": "\u00C2", + "Atilde": "\u00C3", + "Auml": "\u00C4", + "Aring": "\u00C5", + "AElig": "\u00C6", + "Ccedil": "\u00C7", + "Egrave": "\u00C8", + "Eacute": "\u00C9", + "Ecirc": "\u00CA", + "Euml": "\u00CB", + "Igrave": "\u00CC", + "Iacute": "\u00CD", + "Icirc": "\u00CE", + "Iuml": "\u00CF", + "ETH": "\u00D0", + "Ntilde": "\u00D1", + "Ograve": "\u00D2", + "Oacute": "\u00D3", + "Ocirc": "\u00D4", + "Otilde": "\u00D5", + "Ouml": "\u00D6", + "times": "\u00D7", + "Oslash": "\u00D8", + "Ugrave": "\u00D9", + "Uacute": "\u00DA", + "Ucirc": "\u00DB", + "Uuml": "\u00DC", + "Yacute": "\u00DD", + "THORN": "\u00DE", + "szlig": "\u00DF", + "agrave": "\u00E0", + "aacute": "\u00E1", + "acirc": "\u00E2", + "atilde": "\u00E3", + "auml": "\u00E4", + "aring": "\u00E5", + "aelig": "\u00E6", + "ccedil": "\u00E7", + "egrave": "\u00E8", + "eacute": "\u00E9", + "ecirc": "\u00EA", + "euml": "\u00EB", + "igrave": "\u00EC", + "iacute": "\u00ED", + "icirc": "\u00EE", + "iuml": "\u00EF", + "eth": "\u00F0", + "ntilde": "\u00F1", + "ograve": "\u00F2", + "oacute": "\u00F3", + "ocirc": "\u00F4", + "otilde": "\u00F5", + "ouml": "\u00F6", + "divide": "\u00F7", + "oslash": "\u00F8", + "ugrave": "\u00F9", + "uacute": "\u00FA", + "ucirc": "\u00FB", + "uuml": "\u00FC", + "yacute": "\u00FD", + "thorn": "\u00FE", + "yuml": "\u00FF", + "fnof": "\u0192", + "Alpha": "\u0391", + "Beta": "\u0392", + "Gamma": "\u0393", + "Delta": "\u0394", + "Epsilon": "\u0395", + "Zeta": "\u0396", + "Eta": "\u0397", + "Theta": "\u0398", + "Iota": "\u0399", + "Kappa": "\u039A", + "Lambda": "\u039B", + "Mu": "\u039C", + "Nu": "\u039D", + "Xi": "\u039E", + "Omicron": "\u039F", + "Pi": "\u03A0", + "Rho": "\u03A1", + "Sigma": "\u03A3", + "Tau": "\u03A4", + "Upsilon": "\u03A5", + "Phi": "\u03A6", + "Chi": "\u03A7", + "Psi": "\u03A8", + "Omega": "\u03A9", + "alpha": "\u03B1", + "beta": "\u03B2", + "gamma": "\u03B3", + "delta": "\u03B4", + "epsilon": "\u03B5", + "zeta": "\u03B6", + "eta": "\u03B7", + "theta": "\u03B8", + "iota": "\u03B9", + "kappa": "\u03BA", + "lambda": "\u03BB", + "mu": "\u03BC", + "nu": "\u03BD", + "xi": "\u03BE", + "omicron": "\u03BF", + "pi": "\u03C0", + "rho": "\u03C1", + "sigmaf": "\u03C2", + "sigma": "\u03C3", + "tau": "\u03C4", + "upsilon": "\u03C5", + "phi": "\u03C6", + "chi": "\u03C7", + "psi": "\u03C8", + "omega": "\u03C9", + "thetasym": "\u03D1", + "upsih": "\u03D2", + "piv": "\u03D6", + "bull": "\u2022", + "hellip": "\u2026", + "prime": "\u2032", + "Prime": "\u2033", + "oline": "\u203E", + "frasl": "\u2044", + "weierp": "\u2118", + "image": "\u2111", + "real": "\u211C", + "trade": "\u2122", + "alefsym": "\u2135", + "larr": "\u2190", + "uarr": "\u2191", + "rarr": "\u2192", + "darr": "\u2193", + "harr": "\u2194", + "crarr": "\u21B5", + "lArr": "\u21D0", + "uArr": "\u21D1", + "rArr": "\u21D2", + "dArr": "\u21D3", + "hArr": "\u21D4", + "forall": "\u2200", + "part": "\u2202", + "exist": "\u2203", + "empty": "\u2205", + "nabla": "\u2207", + "isin": "\u2208", + "notin": "\u2209", + "ni": "\u220B", + "prod": "\u220F", + "sum": "\u2211", + "minus": "\u2212", + "lowast": "\u2217", + "radic": "\u221A", + "prop": "\u221D", + "infin": "\u221E", + "ang": "\u2220", + "and": "\u2227", + "or": "\u2228", + "cap": "\u2229", + "cup": "\u222A", + "int": "\u222B", + "there4": "\u2234", + "sim": "\u223C", + "cong": "\u2245", + "asymp": "\u2248", + "ne": "\u2260", + "equiv": "\u2261", + "le": "\u2264", + "ge": "\u2265", + "sub": "\u2282", + "sup": "\u2283", + "nsub": "\u2284", + "sube": "\u2286", + "supe": "\u2287", + "oplus": "\u2295", + "otimes": "\u2297", + "perp": "\u22A5", + "sdot": "\u22C5", + "lceil": "\u2308", + "rceil": "\u2309", + "lfloor": "\u230A", + "rfloor": "\u230B", + "lang": "\u2329", + "rang": "\u232A", + "loz": "\u25CA", + "spades": "\u2660", + "clubs": "\u2663", + "hearts": "\u2665", + "diams": "\u2666", + "quot": "\u0022", + "amp": "\u0026", + "lt": "\u003C", + "gt": "\u003E", + "OElig": "\u0152", + "oelig": "\u0153", + "Scaron": "\u0160", + "scaron": "\u0161", + "Yuml": "\u0178", + "circ": "\u02C6", + "tilde": "\u02DC", + "ensp": "\u2002", + "emsp": "\u2003", + "thinsp": "\u2009", + "zwnj": "\u200C", + "zwj": "\u200D", + "lrm": "\u200E", + "rlm": "\u200F", + "ndash": "\u2013", + "mdash": "\u2014", + "lsquo": "\u2018", + "rsquo": "\u2019", + "sbquo": "\u201A", + "ldquo": "\u201C", + "rdquo": "\u201D", + "bdquo": "\u201E", + "dagger": "\u2020", + "Dagger": "\u2021", + "permil": "\u2030", + "lsaquo": "\u2039", + "rsaquo": "\u203A", + "euro": "\u20AC", +} + +// HTMLAutoClose is the set of HTML elements that +// should be considered to close automatically. +var HTMLAutoClose = htmlAutoClose + +var htmlAutoClose = []string{ + /* + hget http://www.w3.org/TR/html4/loose.dtd | + 9 sed -n 's/<!ELEMENT (.*) - O EMPTY.+/ "\1",/p' | tr A-Z a-z + */ + "basefont", + "br", + "area", + "link", + "img", + "param", + "hr", + "input", + "col ", + "frame", + "isindex", + "base", + "meta", +} + +var ( + esc_quot = []byte(""") // shorter than """ + esc_apos = []byte("'") // shorter than "'" + esc_amp = []byte("&") + esc_lt = []byte("<") + esc_gt = []byte(">") +) + +// Escape writes to w the properly escaped XML equivalent +// of the plain text data s. +func Escape(w io.Writer, s []byte) { + var esc []byte + last := 0 + for i, c := range s { + switch c { + case '"': + esc = esc_quot + case '\'': + esc = esc_apos + case '&': + esc = esc_amp + case '<': + esc = esc_lt + case '>': + esc = esc_gt + default: + continue + } + w.Write(s[last:i]) + w.Write(esc) + last = i + 1 + } + w.Write(s[last:]) +} + +// procInstEncoding parses the `encoding="..."` or `encoding='...'` +// value out of the provided string, returning "" if not found. +func procInstEncoding(s string) string { + // TODO: this parsing is somewhat lame and not exact. + // It works for all actual cases, though. + idx := strings.Index(s, "encoding=") + if idx == -1 { + return "" + } + v := s[idx+len("encoding="):] + if v == "" { + return "" + } + if v[0] != '\'' && v[0] != '"' { + return "" + } + idx = strings.IndexRune(v[1:], rune(v[0])) + if idx == -1 { + return "" + } + return v[1 : idx+1] +} diff --git a/src/pkg/encoding/xml/xml_test.go b/src/pkg/encoding/xml/xml_test.go new file mode 100644 index 000000000..1d0696ce0 --- /dev/null +++ b/src/pkg/encoding/xml/xml_test.go @@ -0,0 +1,578 @@ +// Copyright 2009 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 xml + +import ( + "io" + "reflect" + "strings" + "testing" +) + +const testInput = ` +<?xml version="1.0" encoding="UTF-8"?> +<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" + "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> +<body xmlns:foo="ns1" xmlns="ns2" xmlns:tag="ns3" ` + + "\r\n\t" + ` > + <hello lang="en">World <>'" 白鵬翔</hello> + <goodbye /> + <outer foo:attr="value" xmlns:tag="ns4"> + <inner/> + </outer> + <tag:name> + <![CDATA[Some text here.]]> + </tag:name> +</body><!-- missing final newline -->` + +var rawTokens = []Token{ + CharData("\n"), + ProcInst{"xml", []byte(`version="1.0" encoding="UTF-8"`)}, + CharData("\n"), + Directive(`DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" + "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"`), + CharData("\n"), + StartElement{Name{"", "body"}, []Attr{{Name{"xmlns", "foo"}, "ns1"}, {Name{"", "xmlns"}, "ns2"}, {Name{"xmlns", "tag"}, "ns3"}}}, + CharData("\n "), + StartElement{Name{"", "hello"}, []Attr{{Name{"", "lang"}, "en"}}}, + CharData("World <>'\" 白鵬翔"), + EndElement{Name{"", "hello"}}, + CharData("\n "), + StartElement{Name{"", "goodbye"}, []Attr{}}, + EndElement{Name{"", "goodbye"}}, + CharData("\n "), + StartElement{Name{"", "outer"}, []Attr{{Name{"foo", "attr"}, "value"}, {Name{"xmlns", "tag"}, "ns4"}}}, + CharData("\n "), + StartElement{Name{"", "inner"}, []Attr{}}, + EndElement{Name{"", "inner"}}, + CharData("\n "), + EndElement{Name{"", "outer"}}, + CharData("\n "), + StartElement{Name{"tag", "name"}, []Attr{}}, + CharData("\n "), + CharData("Some text here."), + CharData("\n "), + EndElement{Name{"tag", "name"}}, + CharData("\n"), + EndElement{Name{"", "body"}}, + Comment(" missing final newline "), +} + +var cookedTokens = []Token{ + CharData("\n"), + ProcInst{"xml", []byte(`version="1.0" encoding="UTF-8"`)}, + CharData("\n"), + Directive(`DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" + "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"`), + CharData("\n"), + StartElement{Name{"ns2", "body"}, []Attr{{Name{"xmlns", "foo"}, "ns1"}, {Name{"", "xmlns"}, "ns2"}, {Name{"xmlns", "tag"}, "ns3"}}}, + CharData("\n "), + StartElement{Name{"ns2", "hello"}, []Attr{{Name{"", "lang"}, "en"}}}, + CharData("World <>'\" 白鵬翔"), + EndElement{Name{"ns2", "hello"}}, + CharData("\n "), + StartElement{Name{"ns2", "goodbye"}, []Attr{}}, + EndElement{Name{"ns2", "goodbye"}}, + CharData("\n "), + StartElement{Name{"ns2", "outer"}, []Attr{{Name{"ns1", "attr"}, "value"}, {Name{"xmlns", "tag"}, "ns4"}}}, + CharData("\n "), + StartElement{Name{"ns2", "inner"}, []Attr{}}, + EndElement{Name{"ns2", "inner"}}, + CharData("\n "), + EndElement{Name{"ns2", "outer"}}, + CharData("\n "), + StartElement{Name{"ns3", "name"}, []Attr{}}, + CharData("\n "), + CharData("Some text here."), + CharData("\n "), + EndElement{Name{"ns3", "name"}}, + CharData("\n"), + EndElement{Name{"ns2", "body"}}, + Comment(" missing final newline "), +} + +const testInputAltEncoding = ` +<?xml version="1.0" encoding="x-testing-uppercase"?> +<TAG>VALUE</TAG>` + +var rawTokensAltEncoding = []Token{ + CharData("\n"), + ProcInst{"xml", []byte(`version="1.0" encoding="x-testing-uppercase"`)}, + CharData("\n"), + StartElement{Name{"", "tag"}, []Attr{}}, + CharData("value"), + EndElement{Name{"", "tag"}}, +} + +var xmlInput = []string{ + // unexpected EOF cases + "<", + "<t", + "<t ", + "<t/", + "<!", + "<!-", + "<!--", + "<!--c-", + "<!--c--", + "<!d", + "<t></", + "<t></t", + "<?", + "<?p", + "<t a", + "<t a=", + "<t a='", + "<t a=''", + "<t/><![", + "<t/><![C", + "<t/><![CDATA[d", + "<t/><![CDATA[d]", + "<t/><![CDATA[d]]", + + // other Syntax errors + "<>", + "<t/a", + "<0 />", + "<?0 >", + // "<!0 >", // let the Token() caller handle + "</0>", + "<t 0=''>", + "<t a='&'>", + "<t a='<'>", + "<t> c;</t>", + "<t a>", + "<t a=>", + "<t a=v>", + // "<![CDATA[d]]>", // let the Token() caller handle + "<t></e>", + "<t></>", + "<t></t!", + "<t>cdata]]></t>", +} + +func TestRawToken(t *testing.T) { + d := NewDecoder(strings.NewReader(testInput)) + testRawToken(t, d, rawTokens) +} + +type downCaser struct { + t *testing.T + r io.ByteReader +} + +func (d *downCaser) ReadByte() (c byte, err error) { + c, err = d.r.ReadByte() + if c >= 'A' && c <= 'Z' { + c += 'a' - 'A' + } + return +} + +func (d *downCaser) Read(p []byte) (int, error) { + d.t.Fatalf("unexpected Read call on downCaser reader") + panic("unreachable") +} + +func TestRawTokenAltEncoding(t *testing.T) { + sawEncoding := "" + d := NewDecoder(strings.NewReader(testInputAltEncoding)) + d.CharsetReader = func(charset string, input io.Reader) (io.Reader, error) { + sawEncoding = charset + if charset != "x-testing-uppercase" { + t.Fatalf("unexpected charset %q", charset) + } + return &downCaser{t, input.(io.ByteReader)}, nil + } + testRawToken(t, d, rawTokensAltEncoding) +} + +func TestRawTokenAltEncodingNoConverter(t *testing.T) { + d := NewDecoder(strings.NewReader(testInputAltEncoding)) + token, err := d.RawToken() + if token == nil { + t.Fatalf("expected a token on first RawToken call") + } + if err != nil { + t.Fatal(err) + } + token, err = d.RawToken() + if token != nil { + t.Errorf("expected a nil token; got %#v", token) + } + if err == nil { + t.Fatalf("expected an error on second RawToken call") + } + const encoding = "x-testing-uppercase" + if !strings.Contains(err.Error(), encoding) { + t.Errorf("expected error to contain %q; got error: %v", + encoding, err) + } +} + +func testRawToken(t *testing.T, d *Decoder, rawTokens []Token) { + for i, want := range rawTokens { + have, err := d.RawToken() + if err != nil { + t.Fatalf("token %d: unexpected error: %s", i, err) + } + if !reflect.DeepEqual(have, want) { + t.Errorf("token %d = %#v want %#v", i, have, want) + } + } +} + +// Ensure that directives (specifically !DOCTYPE) include the complete +// text of any nested directives, noting that < and > do not change +// nesting depth if they are in single or double quotes. + +var nestedDirectivesInput = ` +<!DOCTYPE [<!ENTITY rdf "http://www.w3.org/1999/02/22-rdf-syntax-ns#">]> +<!DOCTYPE [<!ENTITY xlt ">">]> +<!DOCTYPE [<!ENTITY xlt "<">]> +<!DOCTYPE [<!ENTITY xlt '>'>]> +<!DOCTYPE [<!ENTITY xlt '<'>]> +<!DOCTYPE [<!ENTITY xlt '">'>]> +<!DOCTYPE [<!ENTITY xlt "'<">]> +` + +var nestedDirectivesTokens = []Token{ + CharData("\n"), + Directive(`DOCTYPE [<!ENTITY rdf "http://www.w3.org/1999/02/22-rdf-syntax-ns#">]`), + CharData("\n"), + Directive(`DOCTYPE [<!ENTITY xlt ">">]`), + CharData("\n"), + Directive(`DOCTYPE [<!ENTITY xlt "<">]`), + CharData("\n"), + Directive(`DOCTYPE [<!ENTITY xlt '>'>]`), + CharData("\n"), + Directive(`DOCTYPE [<!ENTITY xlt '<'>]`), + CharData("\n"), + Directive(`DOCTYPE [<!ENTITY xlt '">'>]`), + CharData("\n"), + Directive(`DOCTYPE [<!ENTITY xlt "'<">]`), + CharData("\n"), +} + +func TestNestedDirectives(t *testing.T) { + d := NewDecoder(strings.NewReader(nestedDirectivesInput)) + + for i, want := range nestedDirectivesTokens { + have, err := d.Token() + if err != nil { + t.Fatalf("token %d: unexpected error: %s", i, err) + } + if !reflect.DeepEqual(have, want) { + t.Errorf("token %d = %#v want %#v", i, have, want) + } + } +} + +func TestToken(t *testing.T) { + d := NewDecoder(strings.NewReader(testInput)) + + for i, want := range cookedTokens { + have, err := d.Token() + if err != nil { + t.Fatalf("token %d: unexpected error: %s", i, err) + } + if !reflect.DeepEqual(have, want) { + t.Errorf("token %d = %#v want %#v", i, have, want) + } + } +} + +func TestSyntax(t *testing.T) { + for i := range xmlInput { + d := NewDecoder(strings.NewReader(xmlInput[i])) + var err error + for _, err = d.Token(); err == nil; _, err = d.Token() { + } + if _, ok := err.(*SyntaxError); !ok { + t.Fatalf(`xmlInput "%s": expected SyntaxError not received`, xmlInput[i]) + } + } +} + +type allScalars struct { + True1 bool + True2 bool + False1 bool + False2 bool + Int int + Int8 int8 + Int16 int16 + Int32 int32 + Int64 int64 + Uint int + Uint8 uint8 + Uint16 uint16 + Uint32 uint32 + Uint64 uint64 + Uintptr uintptr + Float32 float32 + Float64 float64 + String string + PtrString *string +} + +var all = allScalars{ + True1: true, + True2: true, + False1: false, + False2: false, + Int: 1, + Int8: -2, + Int16: 3, + Int32: -4, + Int64: 5, + Uint: 6, + Uint8: 7, + Uint16: 8, + Uint32: 9, + Uint64: 10, + Uintptr: 11, + Float32: 13.0, + Float64: 14.0, + String: "15", + PtrString: &sixteen, +} + +var sixteen = "16" + +const testScalarsInput = `<allscalars> + <True1>true</True1> + <True2>1</True2> + <False1>false</False1> + <False2>0</False2> + <Int>1</Int> + <Int8>-2</Int8> + <Int16>3</Int16> + <Int32>-4</Int32> + <Int64>5</Int64> + <Uint>6</Uint> + <Uint8>7</Uint8> + <Uint16>8</Uint16> + <Uint32>9</Uint32> + <Uint64>10</Uint64> + <Uintptr>11</Uintptr> + <Float>12.0</Float> + <Float32>13.0</Float32> + <Float64>14.0</Float64> + <String>15</String> + <PtrString>16</PtrString> +</allscalars>` + +func TestAllScalars(t *testing.T) { + var a allScalars + err := Unmarshal([]byte(testScalarsInput), &a) + + if err != nil { + t.Fatal(err) + } + if !reflect.DeepEqual(a, all) { + t.Errorf("have %+v want %+v", a, all) + } +} + +type item struct { + Field_a string +} + +func TestIssue569(t *testing.T) { + data := `<item><Field_a>abcd</Field_a></item>` + var i item + err := Unmarshal([]byte(data), &i) + + if err != nil || i.Field_a != "abcd" { + t.Fatal("Expecting abcd") + } +} + +func TestUnquotedAttrs(t *testing.T) { + data := "<tag attr=azAZ09:-_\t>" + d := NewDecoder(strings.NewReader(data)) + d.Strict = false + token, err := d.Token() + if _, ok := err.(*SyntaxError); ok { + t.Errorf("Unexpected error: %v", err) + } + if token.(StartElement).Name.Local != "tag" { + t.Errorf("Unexpected tag name: %v", token.(StartElement).Name.Local) + } + attr := token.(StartElement).Attr[0] + if attr.Value != "azAZ09:-_" { + t.Errorf("Unexpected attribute value: %v", attr.Value) + } + if attr.Name.Local != "attr" { + t.Errorf("Unexpected attribute name: %v", attr.Name.Local) + } +} + +func TestValuelessAttrs(t *testing.T) { + tests := [][3]string{ + {"<p nowrap>", "p", "nowrap"}, + {"<p nowrap >", "p", "nowrap"}, + {"<input checked/>", "input", "checked"}, + {"<input checked />", "input", "checked"}, + } + for _, test := range tests { + d := NewDecoder(strings.NewReader(test[0])) + d.Strict = false + token, err := d.Token() + if _, ok := err.(*SyntaxError); ok { + t.Errorf("Unexpected error: %v", err) + } + if token.(StartElement).Name.Local != test[1] { + t.Errorf("Unexpected tag name: %v", token.(StartElement).Name.Local) + } + attr := token.(StartElement).Attr[0] + if attr.Value != test[2] { + t.Errorf("Unexpected attribute value: %v", attr.Value) + } + if attr.Name.Local != test[2] { + t.Errorf("Unexpected attribute name: %v", attr.Name.Local) + } + } +} + +func TestCopyTokenCharData(t *testing.T) { + data := []byte("same data") + var tok1 Token = CharData(data) + tok2 := CopyToken(tok1) + if !reflect.DeepEqual(tok1, tok2) { + t.Error("CopyToken(CharData) != CharData") + } + data[1] = 'o' + if reflect.DeepEqual(tok1, tok2) { + t.Error("CopyToken(CharData) uses same buffer.") + } +} + +func TestCopyTokenStartElement(t *testing.T) { + elt := StartElement{Name{"", "hello"}, []Attr{{Name{"", "lang"}, "en"}}} + var tok1 Token = elt + tok2 := CopyToken(tok1) + if tok1.(StartElement).Attr[0].Value != "en" { + t.Error("CopyToken overwrote Attr[0]") + } + if !reflect.DeepEqual(tok1, tok2) { + t.Error("CopyToken(StartElement) != StartElement") + } + tok1.(StartElement).Attr[0] = Attr{Name{"", "lang"}, "de"} + if reflect.DeepEqual(tok1, tok2) { + t.Error("CopyToken(CharData) uses same buffer.") + } +} + +func TestSyntaxErrorLineNum(t *testing.T) { + testInput := "<P>Foo<P>\n\n<P>Bar</>\n" + d := NewDecoder(strings.NewReader(testInput)) + var err error + for _, err = d.Token(); err == nil; _, err = d.Token() { + } + synerr, ok := err.(*SyntaxError) + if !ok { + t.Error("Expected SyntaxError.") + } + if synerr.Line != 3 { + t.Error("SyntaxError didn't have correct line number.") + } +} + +func TestTrailingRawToken(t *testing.T) { + input := `<FOO></FOO> ` + d := NewDecoder(strings.NewReader(input)) + var err error + for _, err = d.RawToken(); err == nil; _, err = d.RawToken() { + } + if err != io.EOF { + t.Fatalf("d.RawToken() = _, %v, want _, io.EOF", err) + } +} + +func TestTrailingToken(t *testing.T) { + input := `<FOO></FOO> ` + d := NewDecoder(strings.NewReader(input)) + var err error + for _, err = d.Token(); err == nil; _, err = d.Token() { + } + if err != io.EOF { + t.Fatalf("d.Token() = _, %v, want _, io.EOF", err) + } +} + +func TestEntityInsideCDATA(t *testing.T) { + input := `<test><![CDATA[ &val=foo ]]></test>` + d := NewDecoder(strings.NewReader(input)) + var err error + for _, err = d.Token(); err == nil; _, err = d.Token() { + } + if err != io.EOF { + t.Fatalf("d.Token() = _, %v, want _, io.EOF", err) + } +} + +// The last three tests (respectively one for characters in attribute +// names and two for character entities) pass not because of code +// changed for issue 1259, but instead pass with the given messages +// from other parts of xml.Decoder. I provide these to note the +// current behavior of situations where one might think that character +// range checking would detect the error, but it does not in fact. + +var characterTests = []struct { + in string + err string +}{ + {"\x12<doc/>", "illegal character code U+0012"}, + {"<?xml version=\"1.0\"?>\x0b<doc/>", "illegal character code U+000B"}, + {"\xef\xbf\xbe<doc/>", "illegal character code U+FFFE"}, + {"<?xml version=\"1.0\"?><doc>\r\n<hiya/>\x07<toots/></doc>", "illegal character code U+0007"}, + {"<?xml version=\"1.0\"?><doc \x12='value'>what's up</doc>", "expected attribute name in element"}, + {"<doc>&\x01;</doc>", "invalid character entity &;"}, + {"<doc>&\xef\xbf\xbe;</doc>", "invalid character entity &;"}, +} + +func TestDisallowedCharacters(t *testing.T) { + + for i, tt := range characterTests { + d := NewDecoder(strings.NewReader(tt.in)) + var err error + + for err == nil { + _, err = d.Token() + } + synerr, ok := err.(*SyntaxError) + if !ok { + t.Fatalf("input %d d.Token() = _, %v, want _, *SyntaxError", i, err) + } + if synerr.Msg != tt.err { + t.Fatalf("input %d synerr.Msg wrong: want '%s', got '%s'", i, tt.err, synerr.Msg) + } + } +} + +type procInstEncodingTest struct { + expect, got string +} + +var procInstTests = []struct { + input, expect string +}{ + {`version="1.0" encoding="utf-8"`, "utf-8"}, + {`version="1.0" encoding='utf-8'`, "utf-8"}, + {`version="1.0" encoding='utf-8' `, "utf-8"}, + {`version="1.0" encoding=utf-8`, ""}, + {`encoding="FOO" `, "FOO"}, +} + +func TestProcInstEncoding(t *testing.T) { + for _, test := range procInstTests { + got := procInstEncoding(test.input) + if got != test.expect { + t.Errorf("procInstEncoding(%q) = %q; want %q", test.input, got, test.expect) + } + } +} |