1 | import Tokenizer from "./Tokenizer";
|
---|
2 | export interface ParserOptions {
|
---|
3 | /**
|
---|
4 | * Indicates whether special tags (`<script>`, `<style>`, and `<title>`) should get special treatment
|
---|
5 | * and if "empty" tags (eg. `<br>`) can have children. If `false`, the content of special tags
|
---|
6 | * will be text only. For feeds and other XML content (documents that don't consist of HTML),
|
---|
7 | * set this to `true`.
|
---|
8 | *
|
---|
9 | * @default false
|
---|
10 | */
|
---|
11 | xmlMode?: boolean;
|
---|
12 | /**
|
---|
13 | * Decode entities within the document.
|
---|
14 | *
|
---|
15 | * @default true
|
---|
16 | */
|
---|
17 | decodeEntities?: boolean;
|
---|
18 | /**
|
---|
19 | * If set to true, all tags will be lowercased.
|
---|
20 | *
|
---|
21 | * @default !xmlMode
|
---|
22 | */
|
---|
23 | lowerCaseTags?: boolean;
|
---|
24 | /**
|
---|
25 | * If set to `true`, all attribute names will be lowercased. This has noticeable impact on speed.
|
---|
26 | *
|
---|
27 | * @default !xmlMode
|
---|
28 | */
|
---|
29 | lowerCaseAttributeNames?: boolean;
|
---|
30 | /**
|
---|
31 | * If set to true, CDATA sections will be recognized as text even if the xmlMode option is not enabled.
|
---|
32 | * NOTE: If xmlMode is set to `true` then CDATA sections will always be recognized as text.
|
---|
33 | *
|
---|
34 | * @default xmlMode
|
---|
35 | */
|
---|
36 | recognizeCDATA?: boolean;
|
---|
37 | /**
|
---|
38 | * If set to `true`, self-closing tags will trigger the onclosetag event even if xmlMode is not set to `true`.
|
---|
39 | * NOTE: If xmlMode is set to `true` then self-closing tags will always be recognized.
|
---|
40 | *
|
---|
41 | * @default xmlMode
|
---|
42 | */
|
---|
43 | recognizeSelfClosing?: boolean;
|
---|
44 | /**
|
---|
45 | * Allows the default tokenizer to be overwritten.
|
---|
46 | */
|
---|
47 | Tokenizer?: typeof Tokenizer;
|
---|
48 | }
|
---|
49 | export interface Handler {
|
---|
50 | onparserinit(parser: Parser): void;
|
---|
51 | /**
|
---|
52 | * Resets the handler back to starting state
|
---|
53 | */
|
---|
54 | onreset(): void;
|
---|
55 | /**
|
---|
56 | * Signals the handler that parsing is done
|
---|
57 | */
|
---|
58 | onend(): void;
|
---|
59 | onerror(error: Error): void;
|
---|
60 | onclosetag(name: string): void;
|
---|
61 | onopentagname(name: string): void;
|
---|
62 | /**
|
---|
63 | *
|
---|
64 | * @param name Name of the attribute
|
---|
65 | * @param value Value of the attribute.
|
---|
66 | * @param quote Quotes used around the attribute. `null` if the attribute has no quotes around the value, `undefined` if the attribute has no value.
|
---|
67 | */
|
---|
68 | onattribute(name: string, value: string, quote?: string | undefined | null): void;
|
---|
69 | onopentag(name: string, attribs: {
|
---|
70 | [s: string]: string;
|
---|
71 | }): void;
|
---|
72 | ontext(data: string): void;
|
---|
73 | oncomment(data: string): void;
|
---|
74 | oncdatastart(): void;
|
---|
75 | oncdataend(): void;
|
---|
76 | oncommentend(): void;
|
---|
77 | onprocessinginstruction(name: string, data: string): void;
|
---|
78 | }
|
---|
79 | export declare class Parser {
|
---|
80 | /** The start index of the last event. */
|
---|
81 | startIndex: number;
|
---|
82 | /** The end index of the last event. */
|
---|
83 | endIndex: number | null;
|
---|
84 | private tagname;
|
---|
85 | private attribname;
|
---|
86 | private attribvalue;
|
---|
87 | private attribs;
|
---|
88 | private stack;
|
---|
89 | private readonly foreignContext;
|
---|
90 | private readonly cbs;
|
---|
91 | private readonly options;
|
---|
92 | private readonly lowerCaseTagNames;
|
---|
93 | private readonly lowerCaseAttributeNames;
|
---|
94 | private readonly tokenizer;
|
---|
95 | constructor(cbs: Partial<Handler> | null, options?: ParserOptions);
|
---|
96 | private updatePosition;
|
---|
97 | ontext(data: string): void;
|
---|
98 | onopentagname(name: string): void;
|
---|
99 | onopentagend(): void;
|
---|
100 | onclosetag(name: string): void;
|
---|
101 | onselfclosingtag(): void;
|
---|
102 | private closeCurrentTag;
|
---|
103 | onattribname(name: string): void;
|
---|
104 | onattribdata(value: string): void;
|
---|
105 | onattribend(quote: string | undefined | null): void;
|
---|
106 | private getInstructionName;
|
---|
107 | ondeclaration(value: string): void;
|
---|
108 | onprocessinginstruction(value: string): void;
|
---|
109 | oncomment(value: string): void;
|
---|
110 | oncdata(value: string): void;
|
---|
111 | onerror(err: Error): void;
|
---|
112 | onend(): void;
|
---|
113 | /**
|
---|
114 | * Resets the parser to a blank state, ready to parse a new HTML document
|
---|
115 | */
|
---|
116 | reset(): void;
|
---|
117 | /**
|
---|
118 | * Resets the parser, then parses a complete document and
|
---|
119 | * pushes it to the handler.
|
---|
120 | *
|
---|
121 | * @param data Document to parse.
|
---|
122 | */
|
---|
123 | parseComplete(data: string): void;
|
---|
124 | /**
|
---|
125 | * Parses a chunk of data and calls the corresponding callbacks.
|
---|
126 | *
|
---|
127 | * @param chunk Chunk to parse.
|
---|
128 | */
|
---|
129 | write(chunk: string): void;
|
---|
130 | /**
|
---|
131 | * Parses the end of the buffer and clears the stack, calls onend.
|
---|
132 | *
|
---|
133 | * @param chunk Optional final chunk to parse.
|
---|
134 | */
|
---|
135 | end(chunk?: string): void;
|
---|
136 | /**
|
---|
137 | * Pauses parsing. The parser won't emit events until `resume` is called.
|
---|
138 | */
|
---|
139 | pause(): void;
|
---|
140 | /**
|
---|
141 | * Resumes parsing after `pause` was called.
|
---|
142 | */
|
---|
143 | resume(): void;
|
---|
144 | /**
|
---|
145 | * Alias of `write`, for backwards compatibility.
|
---|
146 | *
|
---|
147 | * @param chunk Chunk to parse.
|
---|
148 | * @deprecated
|
---|
149 | */
|
---|
150 | parseChunk(chunk: string): void;
|
---|
151 | /**
|
---|
152 | * Alias of `end`, for backwards compatibility.
|
---|
153 | *
|
---|
154 | * @param chunk Optional final chunk to parse.
|
---|
155 | * @deprecated
|
---|
156 | */
|
---|
157 | done(chunk?: string): void;
|
---|
158 | }
|
---|
159 | //# sourceMappingURL=Parser.d.ts.map |
---|