// 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" "reflect" "strings" "testing" "unicode/utf8" ) const testInput = ` World <>'" 白鵬翔 &何; &is-it; ` var testEntity = map[string]string{"何": "What", "is-it": "is it?"} var rawTokens = []Token{ CharData("\n"), ProcInst{"xml", []byte(`version="1.0" encoding="UTF-8"`)}, CharData("\n"), Directive(`DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"`), CharData("\n"), StartElement{Name{"", "body"}, []Attr{{Name{"xmlns", "foo"}, "ns1"}, {Name{"", "xmlns"}, "ns2"}, {Name{"xmlns", "tag"}, "ns3"}}}, CharData("\n "), StartElement{Name{"", "hello"}, []Attr{{Name{"", "lang"}, "en"}}}, CharData("World <>'\" 白鵬翔"), EndElement{Name{"", "hello"}}, CharData("\n "), StartElement{Name{"", "query"}, []Attr{}}, CharData("What is it?"), EndElement{Name{"", "query"}}, CharData("\n "), StartElement{Name{"", "goodbye"}, []Attr{}}, EndElement{Name{"", "goodbye"}}, CharData("\n "), StartElement{Name{"", "outer"}, []Attr{{Name{"foo", "attr"}, "value"}, {Name{"xmlns", "tag"}, "ns4"}}}, CharData("\n "), StartElement{Name{"", "inner"}, []Attr{}}, EndElement{Name{"", "inner"}}, CharData("\n "), EndElement{Name{"", "outer"}}, CharData("\n "), StartElement{Name{"tag", "name"}, []Attr{}}, CharData("\n "), CharData("Some text here."), CharData("\n "), EndElement{Name{"tag", "name"}}, CharData("\n"), EndElement{Name{"", "body"}}, Comment(" missing final newline "), } var cookedTokens = []Token{ CharData("\n"), ProcInst{"xml", []byte(`version="1.0" encoding="UTF-8"`)}, CharData("\n"), Directive(`DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"`), CharData("\n"), StartElement{Name{"ns2", "body"}, []Attr{{Name{"xmlns", "foo"}, "ns1"}, {Name{"", "xmlns"}, "ns2"}, {Name{"xmlns", "tag"}, "ns3"}}}, CharData("\n "), StartElement{Name{"ns2", "hello"}, []Attr{{Name{"", "lang"}, "en"}}}, CharData("World <>'\" 白鵬翔"), EndElement{Name{"ns2", "hello"}}, CharData("\n "), StartElement{Name{"ns2", "query"}, []Attr{}}, CharData("What is it?"), EndElement{Name{"ns2", "query"}}, CharData("\n "), StartElement{Name{"ns2", "goodbye"}, []Attr{}}, EndElement{Name{"ns2", "goodbye"}}, CharData("\n "), StartElement{Name{"ns2", "outer"}, []Attr{{Name{"ns1", "attr"}, "value"}, {Name{"xmlns", "tag"}, "ns4"}}}, CharData("\n "), StartElement{Name{"ns2", "inner"}, []Attr{}}, EndElement{Name{"ns2", "inner"}}, CharData("\n "), EndElement{Name{"ns2", "outer"}}, CharData("\n "), StartElement{Name{"ns3", "name"}, []Attr{}}, CharData("\n "), CharData("Some text here."), CharData("\n "), EndElement{Name{"ns3", "name"}}, CharData("\n"), EndElement{Name{"ns2", "body"}}, Comment(" missing final newline "), } const testInputAltEncoding = ` VALUE` var rawTokensAltEncoding = []Token{ CharData("\n"), ProcInst{"xml", []byte(`version="1.0" encoding="x-testing-uppercase"`)}, CharData("\n"), StartElement{Name{"", "tag"}, []Attr{}}, CharData("value"), EndElement{Name{"", "tag"}}, } var xmlInput = []string{ // unexpected EOF cases "<", "", "", "", // "", // let the Token() caller handle "", "", "", "", " c;", "", "", "", // "", // let the Token() caller handle "", "", "cdata]]>", } func TestRawToken(t *testing.T) { d := NewDecoder(strings.NewReader(testInput)) d.Entity = testEntity testRawToken(t, d, rawTokens) } const nonStrictInput = ` non&entity &unknown;entity { &#zzz; &なまえ3; <-gt; &; &0a; ` var nonStringEntity = map[string]string{"": "oops!", "0a": "oops!"} var nonStrictTokens = []Token{ CharData("\n"), StartElement{Name{"", "tag"}, []Attr{}}, CharData("non&entity"), EndElement{Name{"", "tag"}}, CharData("\n"), StartElement{Name{"", "tag"}, []Attr{}}, CharData("&unknown;entity"), EndElement{Name{"", "tag"}}, CharData("\n"), StartElement{Name{"", "tag"}, []Attr{}}, CharData("{"), EndElement{Name{"", "tag"}}, CharData("\n"), StartElement{Name{"", "tag"}, []Attr{}}, CharData("&#zzz;"), EndElement{Name{"", "tag"}}, CharData("\n"), StartElement{Name{"", "tag"}, []Attr{}}, CharData("&なまえ3;"), EndElement{Name{"", "tag"}}, CharData("\n"), StartElement{Name{"", "tag"}, []Attr{}}, CharData("<-gt;"), EndElement{Name{"", "tag"}}, CharData("\n"), StartElement{Name{"", "tag"}, []Attr{}}, CharData("&;"), EndElement{Name{"", "tag"}}, CharData("\n"), StartElement{Name{"", "tag"}, []Attr{}}, CharData("&0a;"), EndElement{Name{"", "tag"}}, CharData("\n"), } func TestNonStrictRawToken(t *testing.T) { d := NewDecoder(strings.NewReader(nonStrictInput)) d.Strict = false testRawToken(t, d, nonStrictTokens) } type downCaser struct { t *testing.T r io.ByteReader } func (d *downCaser) ReadByte() (c byte, err error) { c, err = d.r.ReadByte() if c >= 'A' && c <= 'Z' { c += 'a' - 'A' } return } func (d *downCaser) Read(p []byte) (int, error) { d.t.Fatalf("unexpected Read call on downCaser reader") panic("unreachable") } func TestRawTokenAltEncoding(t *testing.T) { d := NewDecoder(strings.NewReader(testInputAltEncoding)) d.CharsetReader = func(charset string, input io.Reader) (io.Reader, error) { if charset != "x-testing-uppercase" { t.Fatalf("unexpected charset %q", charset) } return &downCaser{t, input.(io.ByteReader)}, nil } testRawToken(t, d, rawTokensAltEncoding) } func TestRawTokenAltEncodingNoConverter(t *testing.T) { d := NewDecoder(strings.NewReader(testInputAltEncoding)) token, err := d.RawToken() if token == nil { t.Fatalf("expected a token on first RawToken call") } if err != nil { t.Fatal(err) } token, err = d.RawToken() if token != nil { t.Errorf("expected a nil token; got %#v", token) } if err == nil { t.Fatalf("expected an error on second RawToken call") } const encoding = "x-testing-uppercase" if !strings.Contains(err.Error(), encoding) { t.Errorf("expected error to contain %q; got error: %v", encoding, err) } } func testRawToken(t *testing.T, d *Decoder, rawTokens []Token) { for i, want := range rawTokens { have, err := d.RawToken() if err != nil { t.Fatalf("token %d: unexpected error: %s", i, err) } if !reflect.DeepEqual(have, want) { var shave, swant string if _, ok := have.(CharData); ok { shave = fmt.Sprintf("CharData(%q)", have) } else { shave = fmt.Sprintf("%#v", have) } if _, ok := want.(CharData); ok { swant = fmt.Sprintf("CharData(%q)", want) } else { swant = fmt.Sprintf("%#v", want) } t.Errorf("token %d = %s, want %s", i, shave, swant) } } } // 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 = ` ]> ">]> ]> '>]> ]> '>]> ]> ` var nestedDirectivesTokens = []Token{ CharData("\n"), Directive(`DOCTYPE []`), CharData("\n"), Directive(`DOCTYPE [">]`), CharData("\n"), Directive(`DOCTYPE []`), CharData("\n"), Directive(`DOCTYPE ['>]`), CharData("\n"), Directive(`DOCTYPE []`), CharData("\n"), Directive(`DOCTYPE ['>]`), CharData("\n"), Directive(`DOCTYPE []`), CharData("\n"), } func TestNestedDirectives(t *testing.T) { d := NewDecoder(strings.NewReader(nestedDirectivesInput)) for i, want := range nestedDirectivesTokens { have, err := d.Token() if err != nil { t.Fatalf("token %d: unexpected error: %s", i, err) } if !reflect.DeepEqual(have, want) { t.Errorf("token %d = %#v want %#v", i, have, want) } } } func TestToken(t *testing.T) { d := NewDecoder(strings.NewReader(testInput)) d.Entity = testEntity for i, want := range cookedTokens { have, err := d.Token() if err != nil { t.Fatalf("token %d: unexpected error: %s", i, err) } if !reflect.DeepEqual(have, want) { t.Errorf("token %d = %#v want %#v", i, have, want) } } } func TestSyntax(t *testing.T) { for i := range xmlInput { d := NewDecoder(strings.NewReader(xmlInput[i])) var err error for _, err = d.Token(); err == nil; _, err = d.Token() { } if _, ok := err.(*SyntaxError); !ok { t.Fatalf(`xmlInput "%s": expected SyntaxError not received`, xmlInput[i]) } } } type allScalars struct { True1 bool True2 bool False1 bool False2 bool Int int Int8 int8 Int16 int16 Int32 int32 Int64 int64 Uint int Uint8 uint8 Uint16 uint16 Uint32 uint32 Uint64 uint64 Uintptr uintptr Float32 float32 Float64 float64 String string PtrString *string } var all = allScalars{ True1: true, True2: true, False1: false, False2: false, Int: 1, Int8: -2, Int16: 3, Int32: -4, Int64: 5, Uint: 6, Uint8: 7, Uint16: 8, Uint32: 9, Uint64: 10, Uintptr: 11, Float32: 13.0, Float64: 14.0, String: "15", PtrString: &sixteen, } var sixteen = "16" const testScalarsInput = ` true 1 false 0 1 -2 3 -4 5 6 7 8 9 10 11 12.0 13.0 14.0 15 16 ` func TestAllScalars(t *testing.T) { var a allScalars err := Unmarshal([]byte(testScalarsInput), &a) if err != nil { t.Fatal(err) } if !reflect.DeepEqual(a, all) { t.Errorf("have %+v want %+v", a, all) } } type item struct { Field_a string } func TestIssue569(t *testing.T) { data := `abcd` var i item err := Unmarshal([]byte(data), &i) if err != nil || i.Field_a != "abcd" { t.Fatal("Expecting abcd") } } func TestUnquotedAttrs(t *testing.T) { data := "" d := NewDecoder(strings.NewReader(data)) d.Strict = false token, err := d.Token() if _, ok := err.(*SyntaxError); ok { t.Errorf("Unexpected error: %v", err) } if token.(StartElement).Name.Local != "tag" { t.Errorf("Unexpected tag name: %v", token.(StartElement).Name.Local) } attr := token.(StartElement).Attr[0] if attr.Value != "azAZ09:-_" { t.Errorf("Unexpected attribute value: %v", attr.Value) } if attr.Name.Local != "attr" { t.Errorf("Unexpected attribute name: %v", attr.Name.Local) } } func TestValuelessAttrs(t *testing.T) { tests := [][3]string{ {"

", "p", "nowrap"}, {"

", "p", "nowrap"}, {"", "input", "checked"}, {"", "input", "checked"}, } for _, test := range tests { d := NewDecoder(strings.NewReader(test[0])) d.Strict = false token, err := d.Token() if _, ok := err.(*SyntaxError); ok { t.Errorf("Unexpected error: %v", err) } if token.(StartElement).Name.Local != test[1] { t.Errorf("Unexpected tag name: %v", token.(StartElement).Name.Local) } attr := token.(StartElement).Attr[0] if attr.Value != test[2] { t.Errorf("Unexpected attribute value: %v", attr.Value) } if attr.Name.Local != test[2] { t.Errorf("Unexpected attribute name: %v", attr.Name.Local) } } } func TestCopyTokenCharData(t *testing.T) { data := []byte("same data") var tok1 Token = CharData(data) tok2 := CopyToken(tok1) if !reflect.DeepEqual(tok1, tok2) { t.Error("CopyToken(CharData) != CharData") } data[1] = 'o' if reflect.DeepEqual(tok1, tok2) { t.Error("CopyToken(CharData) uses same buffer.") } } func TestCopyTokenStartElement(t *testing.T) { elt := StartElement{Name{"", "hello"}, []Attr{{Name{"", "lang"}, "en"}}} var tok1 Token = elt tok2 := CopyToken(tok1) if tok1.(StartElement).Attr[0].Value != "en" { t.Error("CopyToken overwrote Attr[0]") } if !reflect.DeepEqual(tok1, tok2) { t.Error("CopyToken(StartElement) != StartElement") } tok1.(StartElement).Attr[0] = Attr{Name{"", "lang"}, "de"} if reflect.DeepEqual(tok1, tok2) { t.Error("CopyToken(CharData) uses same buffer.") } } func TestSyntaxErrorLineNum(t *testing.T) { testInput := "

Foo

\n\n

Bar\n" d := NewDecoder(strings.NewReader(testInput)) var err error for _, err = d.Token(); err == nil; _, err = d.Token() { } synerr, ok := err.(*SyntaxError) if !ok { t.Error("Expected SyntaxError.") } if synerr.Line != 3 { t.Error("SyntaxError didn't have correct line number.") } } func TestTrailingRawToken(t *testing.T) { input := ` ` d := NewDecoder(strings.NewReader(input)) var err error for _, err = d.RawToken(); err == nil; _, err = d.RawToken() { } if err != io.EOF { t.Fatalf("d.RawToken() = _, %v, want _, io.EOF", err) } } func TestTrailingToken(t *testing.T) { input := ` ` d := NewDecoder(strings.NewReader(input)) var err error for _, err = d.Token(); err == nil; _, err = d.Token() { } if err != io.EOF { t.Fatalf("d.Token() = _, %v, want _, io.EOF", err) } } func TestEntityInsideCDATA(t *testing.T) { input := `` d := NewDecoder(strings.NewReader(input)) var err error for _, err = d.Token(); err == nil; _, err = d.Token() { } if err != io.EOF { t.Fatalf("d.Token() = _, %v, want _, io.EOF", err) } } var characterTests = []struct { in string err string }{ {"\x12", "illegal character code U+0012"}, {"\x0b", "illegal character code U+000B"}, {"\xef\xbf\xbe", "illegal character code U+FFFE"}, {"\r\n\x07", "illegal character code U+0007"}, {"what's up", "expected attribute name in element"}, {"&abc\x01;", "invalid character entity &abc (no semicolon)"}, {"&\x01;", "invalid character entity & (no semicolon)"}, {"&\xef\xbf\xbe;", "invalid character entity &\uFFFE;"}, {"&hello;", "invalid character entity &hello;"}, } func TestDisallowedCharacters(t *testing.T) { for i, tt := range characterTests { d := NewDecoder(strings.NewReader(tt.in)) var err error for err == nil { _, err = d.Token() } synerr, ok := err.(*SyntaxError) if !ok { t.Fatalf("input %d d.Token() = _, %v, want _, *SyntaxError", i, err) } if synerr.Msg != tt.err { t.Fatalf("input %d synerr.Msg wrong: want %q, got %q", 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) } } } // Ensure that directives with comments include the complete // text of any nested directives. var directivesWithCommentsInput = ` ]> ]> --> --> []> ` var directivesWithCommentsTokens = []Token{ CharData("\n"), Directive(`DOCTYPE []`), CharData("\n"), Directive(`DOCTYPE []`), CharData("\n"), Directive(`DOCTYPE []`), CharData("\n"), } func TestDirectivesWithComments(t *testing.T) { d := NewDecoder(strings.NewReader(directivesWithCommentsInput)) for i, want := range directivesWithCommentsTokens { have, err := d.Token() if err != nil { t.Fatalf("token %d: unexpected error: %s", i, err) } if !reflect.DeepEqual(have, want) { t.Errorf("token %d = %#v want %#v", i, have, want) } } } // Writer whose Write method always returns an error. type errWriter struct{} func (errWriter) Write(p []byte) (n int, err error) { return 0, fmt.Errorf("unwritable") } func TestEscapeTextIOErrors(t *testing.T) { expectErr := "unwritable" err := EscapeText(errWriter{}, []byte{'A'}) if err == nil || err.Error() != expectErr { t.Errorf("have %v, want %v", err, expectErr) } } func TestEscapeTextInvalidChar(t *testing.T) { input := []byte("A \x00 terminated string.") expected := "A \uFFFD terminated string." buff := new(bytes.Buffer) if err := EscapeText(buff, input); err != nil { t.Fatalf("have %v, want nil", err) } text := buff.String() if text != expected { t.Errorf("have %v, want %v", text, expected) } } func TestIssue5880(t *testing.T) { type T []byte data, err := Marshal(T{192, 168, 0, 1}) if err != nil { t.Errorf("Marshal error: %v", err) } if !utf8.Valid(data) { t.Errorf("Marshal generated invalid UTF-8: %x", data) } }