1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
|
// 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 html
import (
"bufio"
"bytes"
"fmt"
"io"
"io/ioutil"
"os"
"strings"
"testing"
)
func pipeErr(err os.Error) io.Reader {
pr, pw := io.Pipe()
pw.CloseWithError(err)
return pr
}
func readDat(filename string, c chan io.Reader) {
f, err := os.Open("testdata/webkit/" + filename)
if err != nil {
c <- pipeErr(err)
return
}
defer f.Close()
// Loop through the lines of the file. Each line beginning with "#" denotes
// a new section, which is returned as a separate io.Reader.
r := bufio.NewReader(f)
var pw *io.PipeWriter
for {
line, err := r.ReadSlice('\n')
if err != nil {
if pw != nil {
pw.CloseWithError(err)
pw = nil
} else {
c <- pipeErr(err)
}
return
}
if len(line) == 0 {
continue
}
if line[0] == '#' {
if pw != nil {
pw.Close()
}
var pr *io.PipeReader
pr, pw = io.Pipe()
c <- pr
continue
}
if line[0] != '|' {
// Strip the trailing '\n'.
line = line[:len(line)-1]
}
if pw != nil {
if _, err := pw.Write(line); err != nil {
pw.CloseWithError(err)
pw = nil
}
}
}
}
func dumpIndent(w io.Writer, level int) {
io.WriteString(w, "| ")
for i := 0; i < level; i++ {
io.WriteString(w, " ")
}
}
func dumpLevel(w io.Writer, n *Node, level int) os.Error {
dumpIndent(w, level)
switch n.Type {
case ErrorNode:
return os.NewError("unexpected ErrorNode")
case DocumentNode:
return os.NewError("unexpected DocumentNode")
case ElementNode:
fmt.Fprintf(w, "<%s>", n.Data)
for _, a := range n.Attr {
io.WriteString(w, "\n")
dumpIndent(w, level+1)
fmt.Fprintf(w, `%s="%s"`, a.Key, a.Val)
}
case TextNode:
fmt.Fprintf(w, "%q", n.Data)
case CommentNode:
fmt.Fprintf(w, "<!-- %s -->", n.Data)
case DoctypeNode:
fmt.Fprintf(w, "<!DOCTYPE %s>", n.Data)
case scopeMarkerNode:
return os.NewError("unexpected scopeMarkerNode")
default:
return os.NewError("unknown node type")
}
io.WriteString(w, "\n")
for _, c := range n.Child {
if err := dumpLevel(w, c, level+1); err != nil {
return err
}
}
return nil
}
func dump(n *Node) (string, os.Error) {
if n == nil || len(n.Child) == 0 {
return "", nil
}
b := bytes.NewBuffer(nil)
for _, child := range n.Child {
if err := dumpLevel(b, child, 0); err != nil {
return "", err
}
}
return b.String(), nil
}
func TestParser(t *testing.T) {
// TODO(nigeltao): Process all the .dat files, not just the first one.
filenames := []string{
"tests1.dat",
}
for _, filename := range filenames {
rc := make(chan io.Reader)
go readDat(filename, rc)
// TODO(nigeltao): Process all test cases, not just a subset.
for i := 0; i < 80; i++ {
// Parse the #data section.
b, err := ioutil.ReadAll(<-rc)
if err != nil {
t.Fatal(err)
}
text := string(b)
doc, err := Parse(strings.NewReader(text))
if err != nil {
t.Fatal(err)
}
got, err := dump(doc)
if err != nil {
t.Fatal(err)
}
// Skip the #error section.
if _, err := io.Copy(ioutil.Discard, <-rc); err != nil {
t.Fatal(err)
}
// Compare the parsed tree to the #document section.
b, err = ioutil.ReadAll(<-rc)
if err != nil {
t.Fatal(err)
}
if want := string(b); got != want {
t.Errorf("%s test #%d %q, got vs want:\n----\n%s----\n%s----", filename, i, text, got, want)
continue
}
if renderTestBlacklist[text] {
continue
}
// Check that rendering and re-parsing results in an identical tree.
pr, pw := io.Pipe()
go func() {
pw.CloseWithError(Render(pw, doc))
}()
doc1, err := Parse(pr)
if err != nil {
t.Fatal(err)
}
got1, err := dump(doc1)
if err != nil {
t.Fatal(err)
}
if got != got1 {
t.Errorf("%s test #%d %q, got vs got1:\n----\n%s----\n%s----", filename, i, text, got, got1)
continue
}
}
}
}
// Some test input result in parse trees are not 'well-formed' despite
// following the HTML5 recovery algorithms. Rendering and re-parsing such a
// tree will not result in an exact clone of that tree. We blacklist such
// inputs from the render test.
var renderTestBlacklist = map[string]bool{
// The second <a> will be reparented to the first <table>'s parent. This
// results in an <a> whose parent is an <a>, which is not 'well-formed'.
`<a><table><td><a><table></table><a></tr><a></table><b>X</b>C<a>Y`: true,
// The second <a> will be reparented, similar to the case above.
`<a href="blah">aba<table><a href="foo">br<tr><td></td></tr>x</table>aoe`: true,
}
|
