preprocessor.js
4.38 KB
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
'use strict';
const unicode = require('../common/unicode');
const ERR = require('../common/error-codes');
//Aliases
const $ = unicode.CODE_POINTS;
//Const
const DEFAULT_BUFFER_WATERLINE = 1 << 16;
//Preprocessor
//NOTE: HTML input preprocessing
//(see: http://www.whatwg.org/specs/web-apps/current-work/multipage/parsing.html#preprocessing-the-input-stream)
class Preprocessor {
constructor() {
this.html = null;
this.pos = -1;
this.lastGapPos = -1;
this.lastCharPos = -1;
this.gapStack = [];
this.skipNextNewLine = false;
this.lastChunkWritten = false;
this.endOfChunkHit = false;
this.bufferWaterline = DEFAULT_BUFFER_WATERLINE;
}
_err() {
// NOTE: err reporting is noop by default. Enabled by mixin.
}
_addGap() {
this.gapStack.push(this.lastGapPos);
this.lastGapPos = this.pos;
}
_processSurrogate(cp) {
//NOTE: try to peek a surrogate pair
if (this.pos !== this.lastCharPos) {
const nextCp = this.html.charCodeAt(this.pos + 1);
if (unicode.isSurrogatePair(nextCp)) {
//NOTE: we have a surrogate pair. Peek pair character and recalculate code point.
this.pos++;
//NOTE: add gap that should be avoided during retreat
this._addGap();
return unicode.getSurrogatePairCodePoint(cp, nextCp);
}
}
//NOTE: we are at the end of a chunk, therefore we can't infer surrogate pair yet.
else if (!this.lastChunkWritten) {
this.endOfChunkHit = true;
return $.EOF;
}
//NOTE: isolated surrogate
this._err(ERR.surrogateInInputStream);
return cp;
}
dropParsedChunk() {
if (this.pos > this.bufferWaterline) {
this.lastCharPos -= this.pos;
this.html = this.html.substring(this.pos);
this.pos = 0;
this.lastGapPos = -1;
this.gapStack = [];
}
}
write(chunk, isLastChunk) {
if (this.html) {
this.html += chunk;
} else {
this.html = chunk;
}
this.lastCharPos = this.html.length - 1;
this.endOfChunkHit = false;
this.lastChunkWritten = isLastChunk;
}
insertHtmlAtCurrentPos(chunk) {
this.html = this.html.substring(0, this.pos + 1) + chunk + this.html.substring(this.pos + 1, this.html.length);
this.lastCharPos = this.html.length - 1;
this.endOfChunkHit = false;
}
advance() {
this.pos++;
if (this.pos > this.lastCharPos) {
this.endOfChunkHit = !this.lastChunkWritten;
return $.EOF;
}
let cp = this.html.charCodeAt(this.pos);
//NOTE: any U+000A LINE FEED (LF) characters that immediately follow a U+000D CARRIAGE RETURN (CR) character
//must be ignored.
if (this.skipNextNewLine && cp === $.LINE_FEED) {
this.skipNextNewLine = false;
this._addGap();
return this.advance();
}
//NOTE: all U+000D CARRIAGE RETURN (CR) characters must be converted to U+000A LINE FEED (LF) characters
if (cp === $.CARRIAGE_RETURN) {
this.skipNextNewLine = true;
return $.LINE_FEED;
}
this.skipNextNewLine = false;
if (unicode.isSurrogate(cp)) {
cp = this._processSurrogate(cp);
}
//OPTIMIZATION: first check if code point is in the common allowed
//range (ASCII alphanumeric, whitespaces, big chunk of BMP)
//before going into detailed performance cost validation.
const isCommonValidRange =
(cp > 0x1f && cp < 0x7f) || cp === $.LINE_FEED || cp === $.CARRIAGE_RETURN || (cp > 0x9f && cp < 0xfdd0);
if (!isCommonValidRange) {
this._checkForProblematicCharacters(cp);
}
return cp;
}
_checkForProblematicCharacters(cp) {
if (unicode.isControlCodePoint(cp)) {
this._err(ERR.controlCharacterInInputStream);
} else if (unicode.isUndefinedCodePoint(cp)) {
this._err(ERR.noncharacterInInputStream);
}
}
retreat() {
if (this.pos === this.lastGapPos) {
this.lastGapPos = this.gapStack.pop();
this.pos--;
}
this.pos--;
}
}
module.exports = Preprocessor;