summaryrefslogtreecommitdiff
path: root/src/pkg/xml
diff options
context:
space:
mode:
Diffstat (limited to 'src/pkg/xml')
-rw-r--r--src/pkg/xml/Makefile14
-rw-r--r--src/pkg/xml/atom_test.go50
-rw-r--r--src/pkg/xml/embed_test.go124
-rw-r--r--src/pkg/xml/marshal.go228
-rw-r--r--src/pkg/xml/marshal_test.go305
-rw-r--r--src/pkg/xml/read.go631
-rw-r--r--src/pkg/xml/read_test.go371
-rw-r--r--src/pkg/xml/xml.go1694
-rw-r--r--src/pkg/xml/xml_test.go609
9 files changed, 4026 insertions, 0 deletions
diff --git a/src/pkg/xml/Makefile b/src/pkg/xml/Makefile
new file mode 100644
index 000000000..d66c4988a
--- /dev/null
+++ b/src/pkg/xml/Makefile
@@ -0,0 +1,14 @@
+# 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=xml
+
+GOFILES=\
+ marshal.go\
+ read.go\
+ xml.go\
+
+include ../../Make.pkg
diff --git a/src/pkg/xml/atom_test.go b/src/pkg/xml/atom_test.go
new file mode 100644
index 000000000..d365510bf
--- /dev/null
+++ b/src/pkg/xml/atom_test.go
@@ -0,0 +1,50 @@
+// 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{
+ 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/xml/embed_test.go b/src/pkg/xml/embed_test.go
new file mode 100644
index 000000000..ec7f478be
--- /dev/null
+++ b/src/pkg/xml/embed_test.go
@@ -0,0 +1,124 @@
+// Copyright 2010 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 "testing"
+
+type C struct {
+ Name string
+ Open bool
+}
+
+type A struct {
+ XMLName Name `xml:"http://domain a"`
+ C
+ B B
+ FieldA string
+}
+
+type B struct {
+ XMLName Name `xml:"b"`
+ C
+ FieldB string
+}
+
+const _1a = `
+<?xml version="1.0" encoding="UTF-8"?>
+<a xmlns="http://domain">
+ <name>KmlFile</name>
+ <open>1</open>
+ <b>
+ <name>Absolute</name>
+ <open>0</open>
+ <fieldb>bar</fieldb>
+ </b>
+ <fielda>foo</fielda>
+</a>
+`
+
+// Tests that embedded structs are marshalled.
+func TestEmbedded1(t *testing.T) {
+ var a A
+ if e := Unmarshal(StringReader(_1a), &a); e != nil {
+ t.Fatalf("Unmarshal: %s", e)
+ }
+ if a.FieldA != "foo" {
+ t.Fatalf("Unmarshal: expected 'foo' but found '%s'", a.FieldA)
+ }
+ if a.Name != "KmlFile" {
+ t.Fatalf("Unmarshal: expected 'KmlFile' but found '%s'", a.Name)
+ }
+ if !a.Open {
+ t.Fatal("Unmarshal: expected 'true' but found otherwise")
+ }
+ if a.B.FieldB != "bar" {
+ t.Fatalf("Unmarshal: expected 'bar' but found '%s'", a.B.FieldB)
+ }
+ if a.B.Name != "Absolute" {
+ t.Fatalf("Unmarshal: expected 'Absolute' but found '%s'", a.B.Name)
+ }
+ if a.B.Open {
+ t.Fatal("Unmarshal: expected 'false' but found otherwise")
+ }
+}
+
+type A2 struct {
+ XMLName Name `xml:"http://domain a"`
+ XY string
+ Xy string
+}
+
+const _2a = `
+<?xml version="1.0" encoding="UTF-8"?>
+<a xmlns="http://domain">
+ <xy>foo</xy>
+</a>
+`
+
+// Tests that conflicting field names get excluded.
+func TestEmbedded2(t *testing.T) {
+ var a A2
+ if e := Unmarshal(StringReader(_2a), &a); e != nil {
+ t.Fatalf("Unmarshal: %s", e)
+ }
+ if a.XY != "" {
+ t.Fatalf("Unmarshal: expected empty string but found '%s'", a.XY)
+ }
+ if a.Xy != "" {
+ t.Fatalf("Unmarshal: expected empty string but found '%s'", a.Xy)
+ }
+}
+
+type A3 struct {
+ XMLName Name `xml:"http://domain a"`
+ xy string
+}
+
+// Tests that private fields are not set.
+func TestEmbedded3(t *testing.T) {
+ var a A3
+ if e := Unmarshal(StringReader(_2a), &a); e != nil {
+ t.Fatalf("Unmarshal: %s", e)
+ }
+ if a.xy != "" {
+ t.Fatalf("Unmarshal: expected empty string but found '%s'", a.xy)
+ }
+}
+
+type A4 struct {
+ XMLName Name `xml:"http://domain a"`
+ Any string
+}
+
+// Tests that private fields are not set.
+func TestEmbedded4(t *testing.T) {
+ var a A4
+ if e := Unmarshal(StringReader(_2a), &a); e != nil {
+ t.Fatalf("Unmarshal: %s", e)
+ }
+ if a.Any != "foo" {
+ t.Fatalf("Unmarshal: expected 'foo' but found '%s'", a.Any)
+ }
+}
diff --git a/src/pkg/xml/marshal.go b/src/pkg/xml/marshal.go
new file mode 100644
index 000000000..ea421c1b1
--- /dev/null
+++ b/src/pkg/xml/marshal.go
@@ -0,0 +1,228 @@
+// 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"
+ "io"
+ "os"
+ "reflect"
+ "strconv"
+ "strings"
+)
+
+const (
+ // A generic XML header suitable for use with the output of Marshal and
+ // MarshalIndent. 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"
+)
+
+// A Marshaler can produce well-formatted XML representing its internal state.
+// It is used by both Marshal and MarshalIndent.
+type Marshaler interface {
+ MarshalXML() ([]byte, os.Error)
+}
+
+type printer struct {
+ *bufio.Writer
+}
+
+// Marshal writes an XML-formatted representation of v to w.
+//
+// If v implements Marshaler, then Marshal calls its MarshalXML method.
+// Otherwise, Marshal uses the following procedure to create the XML.
+//
+// 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 a single XML
+// element containing the data.
+//
+// The name of that XML element is taken from, in order of preference:
+// - the tag on an XMLName field, if the data is a struct
+// - the value of an 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 '???'.
+//
+// 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 "attr" becomes an attribute 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.
+//
+// Marshal will return an error if asked to marshal a channel, function, or map.
+func Marshal(w io.Writer, v interface{}) (err os.Error) {
+ p := &printer{bufio.NewWriter(w)}
+ err = p.marshalValue(reflect.ValueOf(v), "???")
+ p.Flush()
+ return err
+}
+
+func (p *printer) marshalValue(val reflect.Value, name string) os.Error {
+ if !val.IsValid() {
+ return nil
+ }
+
+ kind := val.Kind()
+ typ := val.Type()
+
+ // Try Marshaler
+ if typ.NumMethod() > 0 {
+ if marshaler, ok := val.Interface().(Marshaler); ok {
+ bytes, err := marshaler.MarshalXML()
+ if err != nil {
+ return err
+ }
+ p.Write(bytes)
+ return nil
+ }
+ }
+
+ // Drill into pointers/interfaces
+ if kind == reflect.Ptr || kind == reflect.Interface {
+ if val.IsNil() {
+ return nil
+ }
+ return p.marshalValue(val.Elem(), name)
+ }
+
+ // 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), name); err != nil {
+ return err
+ }
+ }
+ return nil
+ }
+
+ // Find XML name
+ xmlns := ""
+ if kind == reflect.Struct {
+ if f, ok := typ.FieldByName("XMLName"); ok {
+ if tag := f.Tag.Get("xml"); tag != "" {
+ if i := strings.Index(tag, " "); i >= 0 {
+ xmlns, name = tag[:i], tag[i+1:]
+ } else {
+ name = tag
+ }
+ } else if v, ok := val.FieldByIndex(f.Index).Interface().(Name); ok && v.Local != "" {
+ xmlns, name = v.Space, v.Local
+ }
+ }
+ }
+
+ p.WriteByte('<')
+ p.WriteString(name)
+
+ // Attributes
+ if kind == reflect.Struct {
+ if len(xmlns) > 0 {
+ p.WriteString(` xmlns="`)
+ Escape(p, []byte(xmlns))
+ p.WriteByte('"')
+ }
+
+ for i, n := 0, typ.NumField(); i < n; i++ {
+ if f := typ.Field(i); f.PkgPath == "" && f.Tag.Get("xml") == "attr" {
+ if f.Type.Kind() == reflect.String {
+ if str := val.Field(i).String(); str != "" {
+ p.WriteByte(' ')
+ p.WriteString(strings.ToLower(f.Name))
+ p.WriteString(`="`)
+ Escape(p, []byte(str))
+ p.WriteByte('"')
+ }
+ }
+ }
+ }
+ }
+ p.WriteByte('>')
+
+ switch k := val.Kind(); k {
+ case reflect.Int, reflect.Int8, reflect.Int16, reflect.Int32, reflect.Int64:
+ p.WriteString(strconv.Itoa64(val.Int()))
+ case reflect.Uint, reflect.Uint8, reflect.Uint16, reflect.Uint32, reflect.Uint64, reflect.Uintptr:
+ p.WriteString(strconv.Uitoa64(val.Uint()))
+ case reflect.Float32, reflect.Float64:
+ p.WriteString(strconv.Ftoa64(val.Float(), 'g', -1))
+ case reflect.String:
+ Escape(p, []byte(val.String()))
+ case reflect.Bool:
+ p.WriteString(strconv.Btoa(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
+ bytes := val.Interface().([]byte)
+ Escape(p, bytes)
+ case reflect.Struct:
+ for i, n := 0, val.NumField(); i < n; i++ {
+ if f := typ.Field(i); f.Name != "XMLName" && f.PkgPath == "" {
+ name := f.Name
+ switch tag := f.Tag.Get("xml"); tag {
+ case "":
+ case "chardata":
+ if tk := f.Type.Kind(); tk == reflect.String {
+ Escape(p, []byte(val.Field(i).String()))
+ } else if tk == reflect.Slice {
+ if elem, ok := val.Field(i).Interface().([]byte); ok {
+ Escape(p, elem)
+ }
+ }
+ continue
+ case "innerxml":
+ iface := val.Field(i).Interface()
+ switch raw := iface.(type) {
+ case []byte:
+ p.Write(raw)
+ continue
+ case string:
+ p.WriteString(raw)
+ continue
+ }
+ case "attr":
+ continue
+ default:
+ name = tag
+ }
+
+ if err := p.marshalValue(val.Field(i), name); err != nil {
+ return err
+ }
+ }
+ }
+ default:
+ return &UnsupportedTypeError{typ}
+ }
+
+ p.WriteByte('<')
+ p.WriteByte('/')
+ p.WriteString(name)
+ p.WriteByte('>')
+
+ return nil
+}
+
+// A MarshalXMLError is returned when Marshal or MarshalIndent encounter a type
+// that cannot be converted into XML.
+type UnsupportedTypeError struct {
+ Type reflect.Type
+}
+
+func (e *UnsupportedTypeError) String() string {
+ return "xml: unsupported type: " + e.Type.String()
+}
diff --git a/src/pkg/xml/marshal_test.go b/src/pkg/xml/marshal_test.go
new file mode 100644
index 000000000..5b972fafe
--- /dev/null
+++ b/src/pkg/xml/marshal_test.go
@@ -0,0 +1,305 @@
+// 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"
+ "testing"
+
+ "os"
+ "bytes"
+ "strings"
+ "strconv"
+)
+
+type DriveType int
+
+const (
+ HyperDrive DriveType = iota
+ ImprobabilityDrive
+)
+
+type Passenger struct {
+ Name []string `xml:"name"`
+ Weight float32 `xml:"weight"`
+}
+
+type Ship struct {
+ XMLName Name `xml:"spaceship"`
+
+ Name string `xml:"attr"`
+ Pilot string `xml:"attr"`
+ Drive DriveType `xml:"drive"`
+ Age uint `xml:"age"`
+ Passenger []*Passenger `xml:"passenger"`
+ secret string
+}
+
+type RawXML string
+
+func (rx RawXML) MarshalXML() ([]byte, os.Error) {
+ return []byte(rx), nil
+}
+
+type NamedType string
+
+type Port struct {
+ XMLName Name `xml:"port"`
+ Type string `xml:"attr"`
+ Number string `xml:"chardata"`
+}
+
+type Domain struct {
+ XMLName Name `xml:"domain"`
+ Country string `xml:"attr"`
+ Name []byte `xml:"chardata"`
+}
+
+type Book struct {
+ XMLName Name `xml:"book"`
+ Title string `xml:"chardata"`
+}
+
+type SecretAgent struct {
+ XMLName Name `xml:"agent"`
+ Handle string `xml:"attr"`
+ Identity string
+ Obfuscate string `xml:"innerxml"`
+}
+
+var nilStruct *Ship
+
+var marshalTests = []struct {
+ Value interface{}
+ ExpectXML string
+}{
+ // Test nil marshals to nothing
+ {Value: nil, ExpectXML: ``},
+ {Value: nilStruct, ExpectXML: ``},
+
+ // Test value types (no tag name, so ???)
+ {Value: true, ExpectXML: `<???>true</???>`},
+ {Value: int(42), ExpectXML: `<???>42</???>`},
+ {Value: int8(42), ExpectXML: `<???>42</???>`},
+ {Value: int16(42), ExpectXML: `<???>42</???>`},
+ {Value: int32(42), ExpectXML: `<???>42</???>`},
+ {Value: uint(42), ExpectXML: `<???>42</???>`},
+ {Value: uint8(42), ExpectXML: `<???>42</???>`},
+ {Value: uint16(42), ExpectXML: `<???>42</???>`},
+ {Value: uint32(42), ExpectXML: `<???>42</???>`},
+ {Value: float32(1.25), ExpectXML: `<???>1.25</???>`},
+ {Value: float64(1.25), ExpectXML: `<???>1.25</???>`},
+ {Value: uintptr(0xFFDD), ExpectXML: `<???>65501</???>`},
+ {Value: "gopher", ExpectXML: `<???>gopher</???>`},
+ {Value: []byte("gopher"), ExpectXML: `<???>gopher</???>`},
+ {Value: "</>", ExpectXML: `<???>&lt;/&gt;</???>`},
+ {Value: []byte("</>"), ExpectXML: `<???>&lt;/&gt;</???>`},
+ {Value: [3]byte{'<', '/', '>'}, ExpectXML: `<???>&lt;/&gt;</???>`},
+ {Value: NamedType("potato"), ExpectXML: `<???>potato</???>`},
+ {Value: []int{1, 2, 3}, ExpectXML: `<???>1</???><???>2</???><???>3</???>`},
+ {Value: [3]int{1, 2, 3}, ExpectXML: `<???>1</???><???>2</???><???>3</???>`},
+
+ // Test innerxml
+ {Value: RawXML("</>"), ExpectXML: `</>`},
+ {
+ Value: &SecretAgent{
+ Handle: "007",
+ Identity: "James Bond",
+ Obfuscate: "<redacted/>",
+ },
+ //ExpectXML: `<agent handle="007"><redacted/></agent>`,
+ ExpectXML: `<agent handle="007"><Identity>James Bond</Identity><redacted/></agent>`,
+ },
+
+ // 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="&lt;unix&gt;"></port>`},
+ {Value: &Domain{Name: []byte("google.com&friends")}, ExpectXML: `<domain>google.com&amp;friends</domain>`},
+ {Value: &Book{Title: "Pride & Prejudice"}, ExpectXML: `<book>Pride &amp; Prejudice</book>`},
+ {Value: atomValue, ExpectXML: atomXml},
+ {
+ Value: &Ship{
+ Name: "Heart of Gold",
+ Pilot: "Computer",
+ Age: 1,
+ Drive: ImprobabilityDrive,
+ Passenger: []*Passenger{
+ &Passenger{
+ Name: []string{"Zaphod", "Beeblebrox"},
+ Weight: 7.25,
+ },
+ &Passenger{
+ Name: []string{"Trisha", "McMillen"},
+ Weight: 5.5,
+ },
+ &Passenger{
+ Name: []string{"Ford", "Prefect"},
+ Weight: 7,
+ },
+ &Passenger{
+ 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>`,
+ },
+}
+
+func TestMarshal(t *testing.T) {
+ for idx, test := range marshalTests {
+ buf := bytes.NewBuffer(nil)
+ err := Marshal(buf, test.Value)
+ if err != nil {
+ t.Errorf("#%d: Error: %s", idx, err)
+ continue
+ }
+ if got, want := buf.String(), test.ExpectXML; got != want {
+ if strings.Contains(want, "\n") {
+ t.Errorf("#%d: marshal(%#v) - GOT:\n%s\nWANT:\n%s", idx, test.Value, got, want)
+ } else {
+ t.Errorf("#%d: marshal(%#v) = %#q want %#q", idx, test.Value, got, want)
+ }
+ }
+ }
+}
+
+var marshalErrorTests = []struct {
+ Value interface{}
+ ExpectErr string
+ ExpectKind reflect.Kind
+}{
+ {
+ Value: make(chan bool),
+ ExpectErr: "xml: unsupported type: chan bool",
+ ExpectKind: reflect.Chan,
+ },
+ {
+ Value: map[string]string{
+ "question": "What do you get when you multiply six by nine?",
+ "answer": "42",
+ },
+ ExpectErr: "xml: unsupported type: map[string] string",
+ ExpectKind: reflect.Map,
+ },
+ {
+ Value: map[*Ship]bool{nil: false},
+ ExpectErr: "xml: unsupported type: map[*xml.Ship] bool",
+ ExpectKind: reflect.Map,
+ },
+}
+
+func TestMarshalErrors(t *testing.T) {
+ for idx, test := range marshalErrorTests {
+ buf := bytes.NewBuffer(nil)
+ err := Marshal(buf, test.Value)
+ if got, want := err, test.ExpectErr; got == nil {
+ t.Errorf("#%d: want error %s", idx, want)
+ continue
+ } else if got.String() != want {
+ t.Errorf("#%d: marshal(%#v) = [error] %q, want %q", idx, test.Value, got, want)
+ }
+ if got, want := err.(*UnsupportedTypeError).Type.Kind(), test.ExpectKind; got != want {
+ t.Errorf("#%d: marshal(%#v) = [error kind] %s, want %s", idx, test.Value, got, want)
+ }
+ }
+}
+
+// Do invertibility testing on the various structures that we test
+func TestUnmarshal(t *testing.T) {
+ for i, test := range marshalTests {
+ // Skip the nil pointers
+ if i <= 1 {
+ continue
+ }
+
+ var dest interface{}
+
+ switch test.Value.(type) {
+ case *Ship, Ship:
+ dest = &Ship{}
+ case *Port, Port:
+ dest = &Port{}
+ case *Domain, Domain:
+ dest = &Domain{}
+ case *Feed, Feed:
+ dest = &Feed{}
+ default:
+ continue
+ }
+
+ buffer := bytes.NewBufferString(test.ExpectXML)
+ err := Unmarshal(buffer, dest)
+
+ // Don't compare XMLNames
+ switch fix := dest.(type) {
+ case *Ship:
+ fix.XMLName = Name{}
+ case *Port:
+ fix.XMLName = Name{}
+ case *Domain:
+ fix.XMLName = Name{}
+ case *Feed:
+ fix.XMLName = Name{}
+ 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(%#s) = %#v, want %#v", i, test.ExpectXML, got, want)
+ }
+ }
+}
+
+func BenchmarkMarshal(b *testing.B) {
+ idx := len(marshalTests) - 1
+ test := marshalTests[idx]
+
+ buf := bytes.NewBuffer(nil)
+ for i := 0; i < b.N; i++ {
+ Marshal(buf, test.Value)
+ buf.Truncate(0)
+ }
+}
+
+func BenchmarkUnmarshal(b *testing.B) {
+ idx := len(marshalTests) - 1
+ test := marshalTests[idx]
+ sm := &Ship{}
+ xml := []byte(test.ExpectXML)
+
+ for i := 0; i < b.N; i++ {
+ buffer := bytes.NewBuffer(xml)
+ Unmarshal(buffer, sm)
+ }
+}
diff --git a/src/pkg/xml/read.go b/src/pkg/xml/read.go
new file mode 100644
index 000000000..786b69f5a
--- /dev/null
+++ b/src/pkg/xml/read.go
@@ -0,0 +1,631 @@
+// 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"
+ "fmt"
+ "io"
+ "os"
+ "reflect"
+ "strconv"
+ "strings"
+ "unicode"
+ "utf8"
+)
+
+// 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 an XML element from r and uses the
+// reflect library to fill in an arbitrary struct, slice, or string
+// pointed at by val. Well-formed data that does not fit
+// into val 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(r, &result) is equivalent to assigning
+//
+// r = Result{xml.Name{"", "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-insensitive
+// comparison to match XML element names to 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 of type string with tag "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>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 field whose tag is neither "attr" nor "chardata",
+// Unmarshal maps the sub-element to that struct field.
+// Otherwise, if the struct has a field named Any, unmarshal
+// maps the sub-element to that struct field.
+//
+// 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.
+//
+// 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(r io.Reader, val interface{}) os.Error {
+ v := reflect.ValueOf(val)
+ if v.Kind() != reflect.Ptr {
+ return os.NewError("non-pointer passed to Unmarshal")
+ }
+ p := NewParser(r)
+ elem := v.Elem()
+ err := p.unmarshal(elem, nil)
+ if err != nil {
+ return err
+ }
+ return nil
+}
+
+// An UnmarshalError represents an error in the unmarshalling process.
+type UnmarshalError string
+
+func (e UnmarshalError) String() string { return string(e) }
+
+// 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) String() 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)
+}
+
+// The Parser's Unmarshal method is like xml.Unmarshal
+// except that it can be passed a pointer to the initial start element,
+// useful when a client reads some raw XML tokens itself
+// but also defers to Unmarshal for some elements.
+// Passing a nil start element indicates that Unmarshal should
+// read the token stream to find the start element.
+func (p *Parser) Unmarshal(val interface{}, start *StartElement) os.Error {
+ v := reflect.ValueOf(val)
+ if v.Kind() != reflect.Ptr {
+ return os.NewError("non-pointer passed to Unmarshal")
+ }
+ return p.unmarshal(v.Elem(), start)
+}
+
+// fieldName strips invalid characters from an XML name
+// to create a valid Go struct name. It also converts the
+// name to lower case letters.
+func fieldName(original string) string {
+
+ var i int
+ //remove leading underscores
+ for i = 0; i < len(original) && original[i] == '_'; i++ {
+ }
+
+ return strings.Map(
+ func(x int) int {
+ if x == '_' || unicode.IsDigit(x) || unicode.IsLetter(x) {
+ return unicode.ToLower(x)
+ }
+ return -1
+ },
+ original[i:])
+}
+
+// Unmarshal a single XML element into val.
+func (p *Parser) unmarshal(val reflect.Value, start *StartElement) os.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
+ sv reflect.Value
+ styp reflect.Type
+ fieldPaths map[string]pathInfo
+ )
+
+ switch v := val; v.Kind() {
+ default:
+ return os.NewError("unknown type " + v.Type().String())
+
+ 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:
+ if _, ok := v.Interface().(Name); ok {
+ v.Set(reflect.ValueOf(start.Name))
+ break
+ }
+
+ sv = v
+ typ := sv.Type()
+ styp = typ
+ // Assign name.
+ if f, ok := typ.FieldByName("XMLName"); ok {
+ // Validate element name.
+ if tag := f.Tag.Get("xml"); tag != "" {
+ ns := ""
+ i := strings.LastIndex(tag, " ")
+ if i >= 0 {
+ ns, tag = tag[0:i], tag[i+1:]
+ }
+ if tag != start.Name.Local {
+ return UnmarshalError("expected element type <" + tag + "> but have <" + start.Name.Local + ">")
+ }
+ if ns != "" && ns != start.Name.Space {
+ e := "expected element <" + tag + "> in name space " + ns + " but have "
+ if start.Name.Space == "" {
+ e += "no name space"
+ } else {
+ e += start.Name.Space
+ }
+ return UnmarshalError(e)
+ }
+ }
+
+ // Save
+ v := sv.FieldByIndex(f.Index)
+ if _, ok := v.Interface().(Name); !ok {
+ return UnmarshalError(sv.Type().String() + " field XMLName does not have type xml.Name")
+ }
+ v.Set(reflect.ValueOf(start.Name))
+ }
+
+ // Assign attributes.
+ // Also, determine whether we need to save character data or comments.
+ for i, n := 0, typ.NumField(); i < n; i++ {
+ f := typ.Field(i)
+ switch f.Tag.Get("xml") {
+ case "attr":
+ strv := sv.FieldByIndex(f.Index)
+ // Look for attribute.
+ val := ""
+ k := strings.ToLower(f.Name)
+ for _, a := range start.Attr {
+ if fieldName(a.Name.Local) == k {
+ val = a.Value
+ break
+ }
+ }
+ copyValue(strv, []byte(val))
+
+ case "comment":
+ if !saveComment.IsValid() {
+ saveComment = sv.FieldByIndex(f.Index)
+ }
+
+ case "chardata":
+ if !saveData.IsValid() {
+ saveData = sv.FieldByIndex(f.Index)
+ }
+
+ case "innerxml":
+ if !saveXML.IsValid() {
+ saveXML = sv.FieldByIndex(f.Index)
+ if p.saved == nil {
+ saveXMLIndex = 0
+ p.saved = new(bytes.Buffer)
+ } else {
+ saveXMLIndex = p.savedOffset()
+ }
+ }
+
+ default:
+ if tag := f.Tag.Get("xml"); strings.Contains(tag, ">") {
+ if fieldPaths == nil {
+ fieldPaths = make(map[string]pathInfo)
+ }
+ path := strings.ToLower(tag)
+ if strings.HasPrefix(tag, ">") {
+ path = strings.ToLower(f.Name) + path
+ }
+ if strings.HasSuffix(tag, ">") {
+ path = path[:len(path)-1]
+ }
+ err := addFieldPath(sv, fieldPaths, path, f.Index)
+ if err != nil {
+ return err
+ }
+ }
+ }
+ }
+ }
+
+ // 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:
+ // Sub-element.
+ // Look up by tag name.
+ if sv.IsValid() {
+ k := fieldName(t.Name.Local)
+
+ if fieldPaths != nil {
+ if _, found := fieldPaths[k]; found {
+ if err := p.unmarshalPaths(sv, fieldPaths, k, &t); err != nil {
+ return err
+ }
+ continue Loop
+ }
+ }
+
+ match := func(s string) bool {
+ // check if the name matches ignoring case
+ if strings.ToLower(s) != k {
+ return false
+ }
+ // now check that it's public
+ c, _ := utf8.DecodeRuneInString(s)
+ return unicode.IsUpper(c)
+ }
+
+ f, found := styp.FieldByNameFunc(match)
+ if !found { // fall back to mop-up field named "Any"
+ f, found = styp.FieldByName("Any")
+ }
+ if found {
+ if err := p.unmarshal(sv.FieldByIndex(f.Index), &t); err != nil {
+ return err
+ }
+ continue Loop
+ }
+ }
+ // Not saving sub-element but still have to skip over it.
+ 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 os.Error) {
+ // Helper functions for integer and unsigned integer conversions
+ var itmp int64
+ getInt64 := func() bool {
+ itmp, err = strconv.Atoi64(string(src))
+ // TODO: should check sizes
+ return err == nil
+ }
+ var utmp uint64
+ getUint64 := func() bool {
+ utmp, err = strconv.Atoui64(string(src))
+ // TODO: check for overflow?
+ return err == nil
+ }
+ var ftmp float64
+ getFloat64 := func() bool {
+ ftmp, err = strconv.Atof64(string(src))
+ // TODO: check for overflow?
+ return err == nil
+ }
+
+ // Save accumulated data and comments
+ switch t := dst; t.Kind() {
+ case reflect.Invalid:
+ // Probably a comment, handled below
+ default:
+ return os.NewError("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.Atob(strings.TrimSpace(string(src)))
+ if err != nil {
+ return err
+ }
+ t.SetBool(value)
+ case reflect.String:
+ t.SetString(string(src))
+ case reflect.Slice:
+ t.Set(reflect.ValueOf(src))
+ }
+ return nil
+}
+
+type pathInfo struct {
+ fieldIdx []int
+ complete bool
+}
+
+// addFieldPath takes an element path such as "a>b>c" and fills the
+// paths map with all paths leading to it ("a", "a>b", and "a>b>c").
+// It is okay for paths to share a common, shorter prefix but not ok
+// for one path to itself be a prefix of another.
+func addFieldPath(sv reflect.Value, paths map[string]pathInfo, path string, fieldIdx []int) os.Error {
+ if info, found := paths[path]; found {
+ return tagError(sv, info.fieldIdx, fieldIdx)
+ }
+ paths[path] = pathInfo{fieldIdx, true}
+ for {
+ i := strings.LastIndex(path, ">")
+ if i < 0 {
+ break
+ }
+ path = path[:i]
+ if info, found := paths[path]; found {
+ if info.complete {
+ return tagError(sv, info.fieldIdx, fieldIdx)
+ }
+ } else {
+ paths[path] = pathInfo{fieldIdx, false}
+ }
+ }
+ return nil
+
+}
+
+func tagError(sv reflect.Value, idx1 []int, idx2 []int) os.Error {
+ t := sv.Type()
+ f1 := t.FieldByIndex(idx1)
+ f2 := t.FieldByIndex(idx2)
+ return &TagPathError{t, f1.Name, f1.Tag.Get("xml"), f2.Name, f2.Tag.Get("xml")}
+}
+
+// unmarshalPaths walks down an XML structure looking for
+// wanted paths, and calls unmarshal on them.
+func (p *Parser) unmarshalPaths(sv reflect.Value, paths map[string]pathInfo, path string, start *StartElement) os.Error {
+ if info, _ := paths[path]; info.complete {
+ return p.unmarshal(sv.FieldByIndex(info.fieldIdx), start)
+ }
+ for {
+ tok, err := p.Token()
+ if err != nil {
+ return err
+ }
+ switch t := tok.(type) {
+ case StartElement:
+ k := path + ">" + fieldName(t.Name.Local)
+ if _, found := paths[k]; found {
+ if err := p.unmarshalPaths(sv, paths, k, &t); err != nil {
+ return err
+ }
+ continue
+ }
+ if err := p.Skip(); err != nil {
+ return err
+ }
+ case EndElement:
+ return 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 *Parser) Skip() os.Error {
+ for {
+ tok, err := p.Token()
+ if err != nil {
+ return err
+ }
+ switch t := tok.(type) {
+ case StartElement:
+ if err := p.Skip(); err != nil {
+ return err
+ }
+ case EndElement:
+ return nil
+ }
+ }
+ panic("unreachable")
+}
diff --git a/src/pkg/xml/read_test.go b/src/pkg/xml/read_test.go
new file mode 100644
index 000000000..2126da3c7
--- /dev/null
+++ b/src/pkg/xml/read_test.go
@@ -0,0 +1,371 @@
+// 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(StringReader(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><li-nk href="http://codereview.appspot.com/rss/mine/rsc" rel="self"></li-nk><id>http://codereview.appspot.com/</id><updated>2009-10-04T01:35:58+00:00</updated><author><name>rietveld&lt;&gt;</name></author><entry><title>rietveld: an attempt at pubsubhubbub
+</title><link hre-f="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 &amp;lt;link rel=&amp;quot;hub&amp;quot; href=&amp;quot;hub-server&amp;quot;&amp;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&amp;#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&amp;#39;s actual URL in
+the link rel=&amp;quot;self&amp;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&amp;#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
+ Id string
+ Link []Link
+ Updated Time
+ Author Person
+ Entry []Entry
+}
+
+type Entry struct {
+ Title string
+ Id string
+ Link []Link
+ Updated Time
+ Author Person
+ Summary Text
+}
+
+type Link struct {
+ Rel string `xml:"attr"`
+ Href string `xml:"attr"`
+}
+
+type Person struct {
+ Name string
+ URI string
+ Email string
+ InnerXML string `xml:"innerxml"`
+}
+
+type Text struct {
+ Type string `xml:"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&lt;&gt;</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 &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).
+
+
+`,
+ },
+ },
+ {
+ 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&#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.
+
+
+`,
+ },
+ },
+ },
+}
+
+type FieldNameTest struct {
+ in, out string
+}
+
+var FieldNameTests = []FieldNameTest{
+ {"Profile-Image", "profileimage"},
+ {"_score", "score"},
+}
+
+func TestFieldName(t *testing.T) {
+ for _, tt := range FieldNameTests {
+ a := fieldName(tt.in)
+ if a != tt.out {
+ t.Fatalf("have %#v\nwant %#v\n\n", a, tt.out)
+ }
+ }
+}
+
+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>
+ </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
+}
+
+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"},
+}
+
+func TestUnmarshalPaths(t *testing.T) {
+ for _, pt := range pathTests {
+ v := reflect.New(reflect.TypeOf(pt).Elem()).Interface()
+ if err := Unmarshal(StringReader(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"`
+}
+
+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"}},
+}
+
+func TestUnmarshalBadPaths(t *testing.T) {
+ for _, tt := range badPathTests {
+ err := Unmarshal(StringReader(pathTestString), tt.v)
+ if !reflect.DeepEqual(err, tt.e) {
+ t.Fatalf("Unmarshal with %#v didn't fail properly: %#v", tt.v, err)
+ }
+ }
+}
+
+func TestUnmarshalAttrs(t *testing.T) {
+ var f AttrTest
+ if err := Unmarshal(StringReader(attrString), &f); err != nil {
+ t.Fatalf("Unmarshal: %s", err)
+ }
+ if !reflect.DeepEqual(f, attrStruct) {
+ t.Fatalf("have %#v\nwant %#v", f, attrStruct)
+ }
+}
+
+type AttrTest struct {
+ Test1 Test1
+ Test2 Test2
+}
+
+type Test1 struct {
+ Int int `xml:"attr"`
+ Float float64 `xml:"attr"`
+ Uint8 uint8 `xml:"attr"`
+}
+
+type Test2 struct {
+ Bool bool `xml:"attr"`
+}
+
+const attrString = `
+<?xml version="1.0" charset="utf-8"?>
+<attrtest>
+ <test1 int="8" float="23.5" uint8="255"/>
+ <test2 bool="true"/>
+</attrtest>
+`
+
+var attrStruct = AttrTest{
+ Test1: Test1{
+ Int: 8,
+ Float: 23.5,
+ Uint8: 255,
+ },
+ Test2: Test2{
+ Bool: true,
+ },
+}
diff --git a/src/pkg/xml/xml.go b/src/pkg/xml/xml.go
new file mode 100644
index 000000000..e7ba44e4a
--- /dev/null
+++ b/src/pkg/xml/xml.go
@@ -0,0 +1,1694 @@
+// 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"
+ "os"
+ "strconv"
+ "strings"
+ "unicode"
+ "utf8"
+)
+
+// A SyntaxError represents a syntax error in the XML input stream.
+type SyntaxError struct {
+ Msg string
+ Line int
+}
+
+func (e *SyntaxError) String() 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 Parser.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(e.Attr, attrs)
+ 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 Parser represents an XML parser reading a particular input stream.
+// The parser assumes that its input is encoded in UTF-8.
+type Parser 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:
+ //
+ // p.Strict = false;
+ // p.AutoClose = HTMLAutoClose;
+ // p.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, os.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 os.Error
+ line int
+ tmp [32]byte
+}
+
+// NewParser creates a new XML parser reading from r.
+func NewParser(r io.Reader) *Parser {
+ p := &Parser{
+ ns: make(map[string]string),
+ nextByte: -1,
+ line: 1,
+ Strict: true,
+ }
+ p.switchToReader(r)
+ return p
+}
+
+// Token returns the next XML token in the input stream.
+// At the end of the input stream, Token returns nil, os.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 (p *Parser) Token() (t Token, err os.Error) {
+ if p.nextToken != nil {
+ t = p.nextToken
+ p.nextToken = nil
+ } else if t, err = p.RawToken(); err != nil {
+ return
+ }
+
+ if !p.Strict {
+ if t1, ok := p.autoClose(t); ok {
+ p.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 := p.ns[a.Name.Local]
+ p.pushNs(a.Name.Local, v, ok)
+ p.ns[a.Name.Local] = a.Value
+ }
+ if a.Name.Space == "" && a.Name.Local == "xmlns" {
+ // Default space for untagged names
+ v, ok := p.ns[""]
+ p.pushNs("", v, ok)
+ p.ns[""] = a.Value
+ }
+ }
+
+ p.translate(&t1.Name, true)
+ for i := range t1.Attr {
+ p.translate(&t1.Attr[i].Name, false)
+ }
+ p.pushElement(t1.Name)
+ t = t1
+
+ case EndElement:
+ p.translate(&t1.Name, true)
+ if !p.popElement(&t1) {
+ return nil, p.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 (p *Parser) 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 := p.ns[n.Space]; ok {
+ n.Space = v
+ }
+}
+
+func (p *Parser) 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 {
+ p.r = rb
+ } else {
+ p.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 (p *Parser) push(kind int) *stack {
+ s := p.free
+ if s != nil {
+ p.free = s.next
+ } else {
+ s = new(stack)
+ }
+ s.next = p.stk
+ s.kind = kind
+ p.stk = s
+ return s
+}
+
+func (p *Parser) pop() *stack {
+ s := p.stk
+ if s != nil {
+ p.stk = s.next
+ s.next = p.free
+ p.free = s
+ }
+ return s
+}
+
+// Record that we are starting an element with the given name.
+func (p *Parser) pushElement(name Name) {
+ s := p.push(stkStart)
+ s.name = name
+}
+
+// Record that we are changing the value of ns[local].
+// The old value is url, ok.
+func (p *Parser) pushNs(local string, url string, ok bool) {
+ s := p.push(stkNs)
+ s.name.Local = local
+ s.name.Space = url
+ s.ok = ok
+}
+
+// Creates a SyntaxError with the current line number.
+func (p *Parser) syntaxError(msg string) os.Error {
+ return &SyntaxError{Msg: msg, Line: p.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 (p *Parser) popElement(t *EndElement) bool {
+ s := p.pop()
+ name := t.Name
+ switch {
+ case s == nil || s.kind != stkStart:
+ p.err = p.syntaxError("unexpected end element </" + name.Local + ">")
+ return false
+ case s.name.Local != name.Local:
+ if !p.Strict {
+ p.needClose = true
+ p.toClose = t.Name
+ t.Name = s.name
+ return true
+ }
+ p.err = p.syntaxError("element <" + s.name.Local + "> closed by </" + name.Local + ">")
+ return false
+ case s.name.Space != name.Space:
+ p.err = p.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 p.stk != nil && p.stk.kind != stkStart {
+ s := p.pop()
+ p.ns[s.name.Local] = s.name.Space, s.ok
+ }
+
+ return true
+}
+
+// If the top element on the stack is autoclosing and
+// t is not the end tag, invent the end tag.
+func (p *Parser) autoClose(t Token) (Token, bool) {
+ if p.stk == nil || p.stk.kind != stkStart {
+ return nil, false
+ }
+ name := strings.ToLower(p.stk.name.Local)
+ for _, s := range p.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{p.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 (p *Parser) RawToken() (Token, os.Error) {
+ if p.err != nil {
+ return nil, p.err
+ }
+ if p.needClose {
+ // The last element we read was self-closing and
+ // we returned just the StartElement half.
+ // Return the EndElement half now.
+ p.needClose = false
+ return EndElement{p.toClose}, nil
+ }
+
+ b, ok := p.getc()
+ if !ok {
+ return nil, p.err
+ }
+
+ if b != '<' {
+ // Text section.
+ p.ungetc(b)
+ data := p.text(-1, false)
+ if data == nil {
+ return nil, p.err
+ }
+ return CharData(data), nil
+ }
+
+ if b, ok = p.mustgetc(); !ok {
+ return nil, p.err
+ }
+ switch b {
+ case '/':
+ // </: End element
+ var name Name
+ if name, ok = p.nsname(); !ok {
+ if p.err == nil {
+ p.err = p.syntaxError("expected element name after </")
+ }
+ return nil, p.err
+ }
+ p.space()
+ if b, ok = p.mustgetc(); !ok {
+ return nil, p.err
+ }
+ if b != '>' {
+ p.err = p.syntaxError("invalid characters between </" + name.Local + " and >")
+ return nil, p.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 = p.name(); !ok {
+ if p.err == nil {
+ p.err = p.syntaxError("expected target name after <?")
+ }
+ return nil, p.err
+ }
+ p.space()
+ p.buf.Reset()
+ var b0 byte
+ for {
+ if b, ok = p.mustgetc(); !ok {
+ return nil, p.err
+ }
+ p.buf.WriteByte(b)
+ if b0 == '?' && b == '>' {
+ break
+ }
+ b0 = b
+ }
+ data := p.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 p.CharsetReader == nil {
+ p.err = fmt.Errorf("xml: encoding %q declared but Parser.CharsetReader is nil", enc)
+ return nil, p.err
+ }
+ newr, err := p.CharsetReader(enc, p.r.(io.Reader))
+ if err != nil {
+ p.err = fmt.Errorf("xml: opening charset %q: %v", enc, err)
+ return nil, p.err
+ }
+ if newr == nil {
+ panic("CharsetReader returned a nil Reader for charset " + enc)
+ }
+ p.switchToReader(newr)
+ }
+ }
+ return ProcInst{target, data}, nil
+
+ case '!':
+ // <!: Maybe comment, maybe CDATA.
+ if b, ok = p.mustgetc(); !ok {
+ return nil, p.err
+ }
+ switch b {
+ case '-': // <!-
+ // Probably <!-- for a comment.
+ if b, ok = p.mustgetc(); !ok {
+ return nil, p.err
+ }
+ if b != '-' {
+ p.err = p.syntaxError("invalid sequence <!- not part of <!--")
+ return nil, p.err
+ }
+ // Look for terminator.
+ p.buf.Reset()
+ var b0, b1 byte
+ for {
+ if b, ok = p.mustgetc(); !ok {
+ return nil, p.err
+ }
+ p.buf.WriteByte(b)
+ if b0 == '-' && b1 == '-' && b == '>' {
+ break
+ }
+ b0, b1 = b1, b
+ }
+ data := p.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 = p.mustgetc(); !ok {
+ return nil, p.err
+ }
+ if b != "CDATA["[i] {
+ p.err = p.syntaxError("invalid <![ sequence")
+ return nil, p.err
+ }
+ }
+ // Have <![CDATA[. Read text until ]]>.
+ data := p.text(-1, true)
+ if data == nil {
+ return nil, p.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.
+ p.buf.Reset()
+ p.buf.WriteByte(b)
+ inquote := uint8(0)
+ depth := 0
+ for {
+ if b, ok = p.mustgetc(); !ok {
+ return nil, p.err
+ }
+ if inquote == 0 && b == '>' && depth == 0 {
+ break
+ }
+ p.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(p.buf.Bytes()), nil
+ }
+
+ // Must be an open element like <a href="foo">
+ p.ungetc(b)
+
+ var (
+ name Name
+ empty bool
+ attr []Attr
+ )
+ if name, ok = p.nsname(); !ok {
+ if p.err == nil {
+ p.err = p.syntaxError("expected element name after <")
+ }
+ return nil, p.err
+ }
+
+ attr = make([]Attr, 0, 4)
+ for {
+ p.space()
+ if b, ok = p.mustgetc(); !ok {
+ return nil, p.err
+ }
+ if b == '/' {
+ empty = true
+ if b, ok = p.mustgetc(); !ok {
+ return nil, p.err
+ }
+ if b != '>' {
+ p.err = p.syntaxError("expected /> in element")
+ return nil, p.err
+ }
+ break
+ }
+ if b == '>' {
+ break
+ }
+ p.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 = p.nsname(); !ok {
+ if p.err == nil {
+ p.err = p.syntaxError("expected attribute name in element")
+ }
+ return nil, p.err
+ }
+ p.space()
+ if b, ok = p.mustgetc(); !ok {
+ return nil, p.err
+ }
+ if b != '=' {
+ if p.Strict {
+ p.err = p.syntaxError("attribute name without = in element")
+ return nil, p.err
+ } else {
+ p.ungetc(b)
+ a.Value = a.Name.Local
+ }
+ } else {
+ p.space()
+ data := p.attrval()
+ if data == nil {
+ return nil, p.err
+ }
+ a.Value = string(data)
+ }
+ }
+ if empty {
+ p.needClose = true
+ p.toClose = name
+ }
+ return StartElement{name, attr}, nil
+}
+
+func (p *Parser) attrval() []byte {
+ b, ok := p.mustgetc()
+ if !ok {
+ return nil
+ }
+ // Handle quoted attribute values
+ if b == '"' || b == '\'' {
+ return p.text(int(b), false)
+ }
+ // Handle unquoted attribute values for strict parsers
+ if p.Strict {
+ p.err = p.syntaxError("unquoted or missing attribute value in element")
+ return nil
+ }
+ // Handle unquoted attribute values for unstrict parsers
+ p.ungetc(b)
+ p.buf.Reset()
+ for {
+ b, ok = p.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 == '-' {
+ p.buf.WriteByte(b)
+ } else {
+ p.ungetc(b)
+ break
+ }
+ }
+ return p.buf.Bytes()
+}
+
+// Skip spaces if any
+func (p *Parser) space() {
+ for {
+ b, ok := p.getc()
+ if !ok {
+ return
+ }
+ switch b {
+ case ' ', '\r', '\n', '\t':
+ default:
+ p.ungetc(b)
+ return
+ }
+ }
+}
+
+// Read a single byte.
+// If there is no byte to read, return ok==false
+// and leave the error in p.err.
+// Maintain line number.
+func (p *Parser) getc() (b byte, ok bool) {
+ if p.err != nil {
+ return 0, false
+ }
+ if p.nextByte >= 0 {
+ b = byte(p.nextByte)
+ p.nextByte = -1
+ } else {
+ b, p.err = p.r.ReadByte()
+ if p.err != nil {
+ return 0, false
+ }
+ if p.saved != nil {
+ p.saved.WriteByte(b)
+ }
+ }
+ if b == '\n' {
+ p.line++
+ }
+ return b, true
+}
+
+// Return saved offset.
+// If we did ungetc (nextByte >= 0), have to back up one.
+func (p *Parser) savedOffset() int {
+ n := p.saved.Len()
+ if p.nextByte >= 0 {
+ n--
+ }
+ return n
+}
+
+// Must read a single byte.
+// If there is no byte to read,
+// set p.err to SyntaxError("unexpected EOF")
+// and return ok==false
+func (p *Parser) mustgetc() (b byte, ok bool) {
+ if b, ok = p.getc(); !ok {
+ if p.err == os.EOF {
+ p.err = p.syntaxError("unexpected EOF")
+ }
+ }
+ return
+}
+
+// Unread a single byte.
+func (p *Parser) ungetc(b byte) {
+ if b == '\n' {
+ p.line--
+ }
+ p.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 p.err.
+func (p *Parser) text(quote int, cdata bool) []byte {
+ var b0, b1 byte
+ var trunc int
+ p.buf.Reset()
+Input:
+ for {
+ b, ok := p.getc()
+ if !ok {
+ if cdata {
+ if p.err == os.EOF {
+ p.err = p.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
+ }
+ p.err = p.syntaxError("unescaped ]]> not in CDATA section")
+ return nil
+ }
+
+ // Stop reading text if we see a <.
+ if b == '<' && !cdata {
+ if quote >= 0 {
+ p.err = p.syntaxError("unescaped < inside quoted string")
+ return nil
+ }
+ p.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(p.tmp); i++ {
+ var ok bool
+ p.tmp[i], ok = p.getc()
+ if !ok {
+ if p.err == os.EOF {
+ p.err = p.syntaxError("unexpected EOF")
+ }
+ return nil
+ }
+ c := p.tmp[i]
+ if c == ';' {
+ break
+ }
+ if 'a' <= c && c <= 'z' ||
+ 'A' <= c && c <= 'Z' ||
+ '0' <= c && c <= '9' ||
+ c == '_' || c == '#' {
+ continue
+ }
+ p.ungetc(c)
+ break
+ }
+ s := string(p.tmp[0:i])
+ if i >= len(p.tmp) {
+ if !p.Strict {
+ b0, b1 = 0, 0
+ p.buf.WriteByte('&')
+ p.buf.Write(p.tmp[0:i])
+ continue Input
+ }
+ p.err = p.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 os.Error
+ if i >= 3 && s[1] == 'x' {
+ n, err = strconv.Btoui64(s[2:], 16)
+ } else {
+ n, err = strconv.Btoui64(s[1:], 10)
+ }
+ 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 p.Entity != nil {
+ text, haveText = p.Entity[s]
+ }
+ }
+ if !haveText {
+ if !p.Strict {
+ b0, b1 = 0, 0
+ p.buf.WriteByte('&')
+ p.buf.Write(p.tmp[0:i])
+ continue Input
+ }
+ p.err = p.syntaxError("invalid character entity &" + s + ";")
+ return nil
+ }
+ p.buf.Write([]byte(text))
+ b0, b1 = 0, 0
+ continue Input
+ }
+ p.buf.WriteByte(b)
+ b0, b1 = b1, b
+ }
+ data := p.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 {
+ p.err = p.syntaxError("invalid UTF-8")
+ return nil
+ }
+ buf = buf[size:]
+ if !isInCharacterRange(r) {
+ p.err = p.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(rune int) (inrange bool) {
+ return rune == 0x09 ||
+ rune == 0x0A ||
+ rune == 0x0D ||
+ rune >= 0x20 && rune <= 0xDF77 ||
+ rune >= 0xE000 && rune <= 0xFFFD ||
+ rune >= 0x10000 && rune <= 0x10FFFF
+}
+
+// Get name space name: name with a : stuck in the middle.
+// The part before the : is the name space identifier.
+func (p *Parser) nsname() (name Name, ok bool) {
+ s, ok := p.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 p.err if the name is missing (unless unexpected EOF is received):
+// let the caller provide better context.
+func (p *Parser) name() (s string, ok bool) {
+ var b byte
+ if b, ok = p.mustgetc(); !ok {
+ return
+ }
+
+ // As a first approximation, we gather the bytes [A-Za-z_:.-\x80-\xFF]*
+ if b < utf8.RuneSelf && !isNameByte(b) {
+ p.ungetc(b)
+ return "", false
+ }
+ p.buf.Reset()
+ p.buf.WriteByte(b)
+ for {
+ if b, ok = p.mustgetc(); !ok {
+ return
+ }
+ if b < utf8.RuneSelf && !isNameByte(b) {
+ p.ungetc(b)
+ break
+ }
+ p.buf.WriteByte(b)
+ }
+
+ // Then we check the characters.
+ s = p.buf.String()
+ for i, c := range s {
+ if !unicode.Is(first, c) && (i == 0 || !unicode.Is(second, c)) {
+ p.err = p.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 /\&gt;/ x/\&lt;(.|\n)+/ s/\n/ /g
+ ,x v/^\&lt;!ENTITY/d
+ ,s/\&lt;!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("&#34;") // shorter than "&quot;"
+ esc_apos = []byte("&#39;") // shorter than "&apos;"
+ esc_amp = []byte("&amp;")
+ esc_lt = []byte("&lt;")
+ esc_gt = []byte("&gt;")
+)
+
+// 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:], int(v[0]))
+ if idx == -1 {
+ return ""
+ }
+ return v[1 : idx+1]
+}
diff --git a/src/pkg/xml/xml_test.go b/src/pkg/xml/xml_test.go
new file mode 100644
index 000000000..640762405
--- /dev/null
+++ b/src/pkg/xml/xml_test.go
@@ -0,0 +1,609 @@
+// 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"
+ "io"
+ "os"
+ "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 &lt;&gt;&apos;&quot; &#x767d;&#40300;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([]byte("\n")),
+ ProcInst{"xml", []byte(`version="1.0" encoding="UTF-8"`)},
+ CharData([]byte("\n")),
+ Directive([]byte(`DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
+ "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"`),
+ ),
+ CharData([]byte("\n")),
+ StartElement{Name{"", "body"}, []Attr{{Name{"xmlns", "foo"}, "ns1"}, {Name{"", "xmlns"}, "ns2"}, {Name{"xmlns", "tag"}, "ns3"}}},
+ CharData([]byte("\n ")),
+ StartElement{Name{"", "hello"}, []Attr{{Name{"", "lang"}, "en"}}},
+ CharData([]byte("World <>'\" 白鵬翔")),
+ EndElement{Name{"", "hello"}},
+ CharData([]byte("\n ")),
+ StartElement{Name{"", "goodbye"}, nil},
+ EndElement{Name{"", "goodbye"}},
+ CharData([]byte("\n ")),
+ StartElement{Name{"", "outer"}, []Attr{{Name{"foo", "attr"}, "value"}, {Name{"xmlns", "tag"}, "ns4"}}},
+ CharData([]byte("\n ")),
+ StartElement{Name{"", "inner"}, nil},
+ EndElement{Name{"", "inner"}},
+ CharData([]byte("\n ")),
+ EndElement{Name{"", "outer"}},
+ CharData([]byte("\n ")),
+ StartElement{Name{"tag", "name"}, nil},
+ CharData([]byte("\n ")),
+ CharData([]byte("Some text here.")),
+ CharData([]byte("\n ")),
+ EndElement{Name{"tag", "name"}},
+ CharData([]byte("\n")),
+ EndElement{Name{"", "body"}},
+ Comment([]byte(" missing final newline ")),
+}
+
+var cookedTokens = []Token{
+ CharData([]byte("\n")),
+ ProcInst{"xml", []byte(`version="1.0" encoding="UTF-8"`)},
+ CharData([]byte("\n")),
+ Directive([]byte(`DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
+ "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"`),
+ ),
+ CharData([]byte("\n")),
+ StartElement{Name{"ns2", "body"}, []Attr{{Name{"xmlns", "foo"}, "ns1"}, {Name{"", "xmlns"}, "ns2"}, {Name{"xmlns", "tag"}, "ns3"}}},
+ CharData([]byte("\n ")),
+ StartElement{Name{"ns2", "hello"}, []Attr{{Name{"", "lang"}, "en"}}},
+ CharData([]byte("World <>'\" 白鵬翔")),
+ EndElement{Name{"ns2", "hello"}},
+ CharData([]byte("\n ")),
+ StartElement{Name{"ns2", "goodbye"}, nil},
+ EndElement{Name{"ns2", "goodbye"}},
+ CharData([]byte("\n ")),
+ StartElement{Name{"ns2", "outer"}, []Attr{{Name{"ns1", "attr"}, "value"}, {Name{"xmlns", "tag"}, "ns4"}}},
+ CharData([]byte("\n ")),
+ StartElement{Name{"ns2", "inner"}, nil},
+ EndElement{Name{"ns2", "inner"}},
+ CharData([]byte("\n ")),
+ EndElement{Name{"ns2", "outer"}},
+ CharData([]byte("\n ")),
+ StartElement{Name{"ns3", "name"}, nil},
+ CharData([]byte("\n ")),
+ CharData([]byte("Some text here.")),
+ CharData([]byte("\n ")),
+ EndElement{Name{"ns3", "name"}},
+ CharData([]byte("\n")),
+ EndElement{Name{"ns2", "body"}},
+ Comment([]byte(" missing final newline ")),
+}
+
+const testInputAltEncoding = `
+<?xml version="1.0" encoding="x-testing-uppercase"?>
+<TAG>VALUE</TAG>`
+
+var rawTokensAltEncoding = []Token{
+ CharData([]byte("\n")),
+ ProcInst{"xml", []byte(`version="1.0" encoding="x-testing-uppercase"`)},
+ CharData([]byte("\n")),
+ StartElement{Name{"", "tag"}, nil},
+ CharData([]byte("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>&nbspc;</t>",
+ "<t a>",
+ "<t a=>",
+ "<t a=v>",
+ // "<![CDATA[d]]>", // let the Token() caller handle
+ "<t></e>",
+ "<t></>",
+ "<t></t!",
+ "<t>cdata]]></t>",
+}
+
+type stringReader struct {
+ s string
+ off int
+}
+
+func (r *stringReader) Read(b []byte) (n int, err os.Error) {
+ if r.off >= len(r.s) {
+ return 0, os.EOF
+ }
+ for r.off < len(r.s) && n < len(b) {
+ b[n] = r.s[r.off]
+ n++
+ r.off++
+ }
+ return
+}
+
+func (r *stringReader) ReadByte() (b byte, err os.Error) {
+ if r.off >= len(r.s) {
+ return 0, os.EOF
+ }
+ b = r.s[r.off]
+ r.off++
+ return
+}
+
+func StringReader(s string) io.Reader { return &stringReader{s, 0} }
+
+func TestRawToken(t *testing.T) {
+ p := NewParser(StringReader(testInput))
+ testRawToken(t, p, rawTokens)
+}
+
+type downCaser struct {
+ t *testing.T
+ r io.ByteReader
+}
+
+func (d *downCaser) ReadByte() (c byte, err os.Error) {
+ c, err = d.r.ReadByte()
+ if c >= 'A' && c <= 'Z' {
+ c += 'a' - 'A'
+ }
+ return
+}
+
+func (d *downCaser) Read(p []byte) (int, os.Error) {
+ d.t.Fatalf("unexpected Read call on downCaser reader")
+ return 0, os.EINVAL
+}
+
+func TestRawTokenAltEncoding(t *testing.T) {
+ sawEncoding := ""
+ p := NewParser(StringReader(testInputAltEncoding))
+ p.CharsetReader = func(charset string, input io.Reader) (io.Reader, os.Error) {
+ sawEncoding = charset
+ if charset != "x-testing-uppercase" {
+ t.Fatalf("unexpected charset %q", charset)
+ }
+ return &downCaser{t, input.(io.ByteReader)}, nil
+ }
+ testRawToken(t, p, rawTokensAltEncoding)
+}
+
+func TestRawTokenAltEncodingNoConverter(t *testing.T) {
+ p := NewParser(StringReader(testInputAltEncoding))
+ token, err := p.RawToken()
+ if token == nil {
+ t.Fatalf("expected a token on first RawToken call")
+ }
+ if err != nil {
+ t.Fatal(err)
+ }
+ token, err = p.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.String(), encoding) {
+ t.Errorf("expected error to contain %q; got error: %v",
+ encoding, err)
+ }
+}
+
+func testRawToken(t *testing.T, p *Parser, rawTokens []Token) {
+ for i, want := range rawTokens {
+ have, err := p.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([]byte("\n")),
+ Directive([]byte(`DOCTYPE [<!ENTITY rdf "http://www.w3.org/1999/02/22-rdf-syntax-ns#">]`)),
+ CharData([]byte("\n")),
+ Directive([]byte(`DOCTYPE [<!ENTITY xlt ">">]`)),
+ CharData([]byte("\n")),
+ Directive([]byte(`DOCTYPE [<!ENTITY xlt "<">]`)),
+ CharData([]byte("\n")),
+ Directive([]byte(`DOCTYPE [<!ENTITY xlt '>'>]`)),
+ CharData([]byte("\n")),
+ Directive([]byte(`DOCTYPE [<!ENTITY xlt '<'>]`)),
+ CharData([]byte("\n")),
+ Directive([]byte(`DOCTYPE [<!ENTITY xlt '">'>]`)),
+ CharData([]byte("\n")),
+ Directive([]byte(`DOCTYPE [<!ENTITY xlt "'<">]`)),
+ CharData([]byte("\n")),
+}
+
+func TestNestedDirectives(t *testing.T) {
+ p := NewParser(StringReader(nestedDirectivesInput))
+
+ for i, want := range nestedDirectivesTokens {
+ have, err := p.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) {
+ p := NewParser(StringReader(testInput))
+
+ for i, want := range cookedTokens {
+ have, err := p.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 {
+ p := NewParser(StringReader(xmlInput[i]))
+ var err os.Error
+ for _, err = p.Token(); err == nil; _, err = p.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
+ buf := bytes.NewBufferString(testScalarsInput)
+ err := Unmarshal(buf, &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
+ buf := bytes.NewBufferString(data)
+ err := Unmarshal(buf, &i)
+
+ if err != nil || i.Field_a != "abcd" {
+ t.Fatal("Expecting abcd")
+ }
+}
+
+func TestUnquotedAttrs(t *testing.T) {
+ data := "<tag attr=azAZ09:-_\t>"
+ p := NewParser(StringReader(data))
+ p.Strict = false
+ token, err := p.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 {
+ p := NewParser(StringReader(test[0]))
+ p.Strict = false
+ token, err := p.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 !reflect.DeepEqual(tok1, tok2) {
+ t.Error("CopyToken(StartElement) != StartElement")
+ }
+ elt.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"
+ p := NewParser(StringReader(testInput))
+ var err os.Error
+ for _, err = p.Token(); err == nil; _, err = p.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> `
+ p := NewParser(StringReader(input))
+ var err os.Error
+ for _, err = p.RawToken(); err == nil; _, err = p.RawToken() {
+ }
+ if err != os.EOF {
+ t.Fatalf("p.RawToken() = _, %v, want _, os.EOF", err)
+ }
+}
+
+func TestTrailingToken(t *testing.T) {
+ input := `<FOO></FOO> `
+ p := NewParser(StringReader(input))
+ var err os.Error
+ for _, err = p.Token(); err == nil; _, err = p.Token() {
+ }
+ if err != os.EOF {
+ t.Fatalf("p.Token() = _, %v, want _, os.EOF", err)
+ }
+}
+
+func TestEntityInsideCDATA(t *testing.T) {
+ input := `<test><![CDATA[ &val=foo ]]></test>`
+ p := NewParser(StringReader(input))
+ var err os.Error
+ for _, err = p.Token(); err == nil; _, err = p.Token() {
+ }
+ if err != os.EOF {
+ t.Fatalf("p.Token() = _, %v, want _, os.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.Parser. 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 {
+ p := NewParser(StringReader(tt.in))
+ var err os.Error
+
+ for err == nil {
+ _, err = p.Token()
+ }
+ synerr, ok := err.(*SyntaxError)
+ if !ok {
+ t.Fatalf("input %d p.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)
+ }
+ }
+}
generated by cgit v1.2.3 (git 2.39.1) at 2025-07-22 22:41:46 +0000