Function TestHTML5LibTests

html/token_test.go:1106–1172

(t *testing.T)

}

func TestHTML5LibTests(t *testing.T) {
	skipTests := map[string]bool{
		// We emit a comment token here, instead of no token. This is a specification
		// divergence that we may want to fix.
		"test1.test/Empty end tag":                           true,
		"test2.test/Empty end tag with following characters": true,
		"test2.test/Empty end tag with following tag":        true,
		"test2.test/Empty end tag with following comment":    true,
		"test2.test/Empty end tag with following end tag":    true,
		"test3.test/</>":                                 true,
		"test4.test/CR EOF after doctype name":           true,
		"test4.test/Doctype public case-sensitivity (1)": true,
		"test4.test/Doctype public case-sensitivity (2)": true,
		"test4.test/Doctype system case-sensitivity (1)": true,
		"test4.test/Doctype system case-sensitivity (2)": true,
	}

	var tests struct {
		Tests []h5libTest
	}
	testFiles, err := filepath.Glob("testdata/html5lib-tests/tokenizer/*.test")
	if err != nil {
		t.Fatal(err)
	}
	for _, testFile := range testFiles {
		data, err := os.ReadFile(testFile)
		if err != nil {
			t.Fatal(err)
		}
		if err := json.Unmarshal(data, &tests); err != nil {
			t.Fatal(err)
		}

		base := filepath.Base(testFile)

		for _, tc := range tests.Tests {
			name := fmt.Sprintf("%s/%s", base, tc.Description)
			t.Run(name, func(t *testing.T) {
				if skipTests[name] {
					t.Skip("skipping, known failure")
				}
				if len(tc.InitialStates) > 0 {
					t.Skip("Initial states not supported yet")
				}
				if strings.Contains(tc.Input, "<!DOCTYPE") {
					t.Skip("Skipping DOCTYPE")
				}
				z := NewTokenizer(strings.NewReader(tc.Input))
				var tokens []Token
				for {
					if z.Next() == ErrorToken {
						if z.Err() == io.EOF {
							break
						}
						t.Fatalf("Error: %v", z.Err())
					}
					tokens = append(tokens, z.Token())
				}