Updated the files.

This commit is contained in:
Batuhan Berk Başoğlu 2024-02-08 19:38:41 -05:00
parent 1553e6b971
commit 753967d4f5
23418 changed files with 3784666 additions and 0 deletions

View file

@ -0,0 +1,90 @@
import { SAXParser, type EndTag, type StartTag, type Doctype, type Text, type Comment, type SaxToken } from 'parse5-sax-parser';
/**
* Streaming [SAX](https://en.wikipedia.org/wiki/Simple_API_for_XML)-style HTML rewriter.
* A [transform stream](https://nodejs.org/api/stream.html#stream_class_stream_transform) (which means you can pipe _through_ it, see example).
*
* The rewriter uses the raw source representation of tokens if they are not modified by the user. Therefore, the resulting
* HTML is not affected by parser error-recovery mechanisms as in a classical parsing-serialization roundtrip.
*
* @example
*
* ```js
* const RewritingStream = require('parse5-html-rewriting-stream');
* const http = require('http');
* const fs = require('fs');
*
* const file = fs.createWriteStream('/home/google.com.html');
* const rewriter = new RewritingStream();
*
* // Replace divs with spans
* rewriter.on('startTag', startTag => {
* if (startTag.tagName === 'span') {
* startTag.tagName = 'div';
* }
*
* rewriter.emitStartTag(startTag);
* });
*
* rewriter.on('endTag', endTag => {
* if (endTag.tagName === 'span') {
* endTag.tagName = 'div';
* }
*
* rewriter.emitEndTag(endTag);
* });
*
* // Wrap all text nodes with an <i> tag
* rewriter.on('text', (_, raw) => {
* // Use the raw representation of text without HTML entities decoding
* rewriter.emitRaw(`<i>${raw}</i>`);
* });
*
* http.get('http://google.com', res => {
* // Assumes response is UTF-8.
* res.setEncoding('utf8');
* // `RewritingStream` is a `Transform` stream, which means you can pipe
* // through it.
* res.pipe(rewriter).pipe(file);
* });
* ```
*/
export declare class RewritingStream extends SAXParser {
/** Note: `sourceCodeLocationInfo` is always enabled. */
constructor();
_transformChunk(chunk: string): string;
private _getRawHtml;
protected emitIfListenerExists(eventName: string, token: SaxToken): boolean;
protected _emitToken(eventName: string, token: SaxToken): void;
/** Emits a serialized document type token into the output stream. */
emitDoctype(token: Doctype): void;
/** Emits a serialized start tag token into the output stream. */
emitStartTag(token: StartTag): void;
/** Emits a serialized end tag token into the output stream. */
emitEndTag(token: EndTag): void;
/** Emits a serialized text token into the output stream. */
emitText({ text }: Text): void;
/** Emits a serialized comment token into the output stream. */
emitComment(token: Comment): void;
/** Emits a raw HTML string into the output stream. */
emitRaw(html: string): void;
}
export interface RewritingStream {
/** Raised when the rewriter encounters a start tag. */
on(event: 'startTag', listener: (startTag: StartTag, rawHtml: string) => void): this;
/** Raised when rewriter encounters an end tag. */
on(event: 'endTag', listener: (endTag: EndTag, rawHtml: string) => void): this;
/** Raised when rewriter encounters a comment. */
on(event: 'comment', listener: (comment: Comment, rawHtml: string) => void): this;
/** Raised when rewriter encounters text content. */
on(event: 'text', listener: (text: Text, rawHtml: string) => void): this;
/** Raised when rewriter encounters a [document type declaration](https://en.wikipedia.org/wiki/Document_type_declaration). */
on(event: 'doctype', listener: (doctype: Doctype, rawHtml: string) => void): this;
/**
* Base event handler.
*
* @param event Name of the event
* @param handler Event handler
*/
on(event: string, handler: (...args: any[]) => void): this;
}
//# sourceMappingURL=index.d.ts.map

View file

@ -0,0 +1 @@
{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../lib/index.ts"],"names":[],"mappings":"AACA,OAAO,EACH,SAAS,EACT,KAAK,MAAM,EACX,KAAK,QAAQ,EACb,KAAK,OAAO,EACZ,KAAK,IAAI,EACT,KAAK,OAAO,EACZ,KAAK,QAAQ,EAChB,MAAM,mBAAmB,CAAC;AAG3B;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;GAgDG;AACH,qBAAa,eAAgB,SAAQ,SAAS;IAC1C,wDAAwD;;IAK/C,eAAe,CAAC,KAAK,EAAE,MAAM,GAAG,MAAM;IAO/C,OAAO,CAAC,WAAW;cASA,oBAAoB,CAAC,SAAS,EAAE,MAAM,EAAE,KAAK,EAAE,QAAQ,GAAG,OAAO;cAYjE,UAAU,CAAC,SAAS,EAAE,MAAM,EAAE,KAAK,EAAE,QAAQ,GAAG,IAAI;IAIvE,qEAAqE;IAC9D,WAAW,CAAC,KAAK,EAAE,OAAO,GAAG,IAAI;IAkBxC,iEAAiE;IAC1D,YAAY,CAAC,KAAK,EAAE,QAAQ,GAAG,IAAI;IAY1C,+DAA+D;IACxD,UAAU,CAAC,KAAK,EAAE,MAAM,GAAG,IAAI;IAItC,4DAA4D;IACrD,QAAQ,CAAC,EAAE,IAAI,EAAE,EAAE,IAAI,GAAG,IAAI;IASrC,+DAA+D;IACxD,WAAW,CAAC,KAAK,EAAE,OAAO,GAAG,IAAI;IAIxC,sDAAsD;IAC/C,OAAO,CAAC,IAAI,EAAE,MAAM,GAAG,IAAI;CAGrC;AAED,MAAM,WAAW,eAAe;IAC5B,uDAAuD;IACvD,EAAE,CAAC,KAAK,EAAE,UAAU,EAAE,QAAQ,EAAE,CAAC,QAAQ,EAAE,QAAQ,EAAE,OAAO,EAAE,MAAM,KAAK,IAAI,GAAG,IAAI,CAAC;IACrF,kDAAkD;IAClD,EAAE,CAAC,KAAK,EAAE,QAAQ,EAAE,QAAQ,EAAE,CAAC,MAAM,EAAE,MAAM,EAAE,OAAO,EAAE,MAAM,KAAK,IAAI,GAAG,IAAI,CAAC;IAC/E,iDAAiD;IACjD,EAAE,CAAC,KAAK,EAAE,SAAS,EAAE,QAAQ,EAAE,CAAC,OAAO,EAAE,OAAO,EAAE,OAAO,EAAE,MAAM,KAAK,IAAI,GAAG,IAAI,CAAC;IAClF,oDAAoD;IACpD,EAAE,CAAC,KAAK,EAAE,MAAM,EAAE,QAAQ,EAAE,CAAC,IAAI,EAAE,IAAI,EAAE,OAAO,EAAE,MAAM,KAAK,IAAI,GAAG,IAAI,CAAC;IACzE,8HAA8H;IAC9H,EAAE,CAAC,KAAK,EAAE,SAAS,EAAE,QAAQ,EAAE,CAAC,OAAO,EAAE,OAAO,EAAE,OAAO,EAAE,MAAM,KAAK,IAAI,GAAG,IAAI,CAAC;IAElF;;;;;OAKG;IACH,EAAE,CAAC,KAAK,EAAE,MAAM,EAAE,OAAO,EAAE,CAAC,GAAG,IAAI,EAAE,GAAG,EAAE,KAAK,IAAI,GAAG,IAAI,CAAC;CAC9D"}

View file

@ -0,0 +1,128 @@
import { html } from 'parse5';
import { SAXParser, } from 'parse5-sax-parser';
import { escapeText, escapeAttribute } from 'entities/lib/escape.js';
/**
* Streaming [SAX](https://en.wikipedia.org/wiki/Simple_API_for_XML)-style HTML rewriter.
* A [transform stream](https://nodejs.org/api/stream.html#stream_class_stream_transform) (which means you can pipe _through_ it, see example).
*
* The rewriter uses the raw source representation of tokens if they are not modified by the user. Therefore, the resulting
* HTML is not affected by parser error-recovery mechanisms as in a classical parsing-serialization roundtrip.
*
* @example
*
* ```js
* const RewritingStream = require('parse5-html-rewriting-stream');
* const http = require('http');
* const fs = require('fs');
*
* const file = fs.createWriteStream('/home/google.com.html');
* const rewriter = new RewritingStream();
*
* // Replace divs with spans
* rewriter.on('startTag', startTag => {
* if (startTag.tagName === 'span') {
* startTag.tagName = 'div';
* }
*
* rewriter.emitStartTag(startTag);
* });
*
* rewriter.on('endTag', endTag => {
* if (endTag.tagName === 'span') {
* endTag.tagName = 'div';
* }
*
* rewriter.emitEndTag(endTag);
* });
*
* // Wrap all text nodes with an <i> tag
* rewriter.on('text', (_, raw) => {
* // Use the raw representation of text without HTML entities decoding
* rewriter.emitRaw(`<i>${raw}</i>`);
* });
*
* http.get('http://google.com', res => {
* // Assumes response is UTF-8.
* res.setEncoding('utf8');
* // `RewritingStream` is a `Transform` stream, which means you can pipe
* // through it.
* res.pipe(rewriter).pipe(file);
* });
* ```
*/
export class RewritingStream extends SAXParser {
/** Note: `sourceCodeLocationInfo` is always enabled. */
constructor() {
super({ sourceCodeLocationInfo: true });
}
_transformChunk(chunk) {
// NOTE: ignore upstream return values as we want to push to
// the `Writable` part of the `Transform` stream ourselves.
super._transformChunk(chunk);
return '';
}
_getRawHtml(location) {
const { droppedBufferSize, html } = this.tokenizer.preprocessor;
const start = location.startOffset - droppedBufferSize;
const end = location.endOffset - droppedBufferSize;
return html.slice(start, end);
}
// Events
emitIfListenerExists(eventName, token) {
if (!super.emitIfListenerExists(eventName, token)) {
this.emitRaw(this._getRawHtml(token.sourceCodeLocation));
}
// NOTE: don't skip new lines after `<pre>` and other tags,
// otherwise we'll have incorrect raw data.
this.parserFeedbackSimulator.skipNextNewLine = false;
return true;
}
// Emitter API
_emitToken(eventName, token) {
this.emit(eventName, token, this._getRawHtml(token.sourceCodeLocation));
}
/** Emits a serialized document type token into the output stream. */
emitDoctype(token) {
let res = `<!DOCTYPE ${token.name}`;
if (token.publicId !== null) {
res += ` PUBLIC "${token.publicId}"`;
}
else if (token.systemId !== null) {
res += ' SYSTEM';
}
if (token.systemId !== null) {
res += ` "${token.systemId}"`;
}
res += '>';
this.push(res);
}
/** Emits a serialized start tag token into the output stream. */
emitStartTag(token) {
let res = `<${token.tagName}`;
for (const attr of token.attrs) {
res += ` ${attr.name}="${escapeAttribute(attr.value)}"`;
}
res += token.selfClosing ? '/>' : '>';
this.push(res);
}
/** Emits a serialized end tag token into the output stream. */
emitEndTag(token) {
this.push(`</${token.tagName}>`);
}
/** Emits a serialized text token into the output stream. */
emitText({ text }) {
this.push(!this.parserFeedbackSimulator.inForeignContent &&
html.hasUnescapedText(this.tokenizer.lastStartTagName, true)
? text
: escapeText(text));
}
/** Emits a serialized comment token into the output stream. */
emitComment(token) {
this.push(`<!--${token.text}-->`);
}
/** Emits a raw HTML string into the output stream. */
emitRaw(html) {
this.push(html);
}
}
//# sourceMappingURL=index.js.map

View file

@ -0,0 +1 @@
{"version":3,"file":"index.js","sourceRoot":"","sources":["../lib/index.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,IAAI,EAAc,MAAM,QAAQ,CAAC;AAC1C,OAAO,EACH,SAAS,GAOZ,MAAM,mBAAmB,CAAC;AAC3B,OAAO,EAAE,UAAU,EAAE,eAAe,EAAE,MAAM,wBAAwB,CAAC;AAErE;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;GAgDG;AACH,MAAM,OAAO,eAAgB,SAAQ,SAAS;IAC1C,wDAAwD;IACxD;QACI,KAAK,CAAC,EAAE,sBAAsB,EAAE,IAAI,EAAE,CAAC,CAAC;IAC5C,CAAC;IAEQ,eAAe,CAAC,KAAa;QAClC,4DAA4D;QAC5D,2DAA2D;QAC3D,KAAK,CAAC,eAAe,CAAC,KAAK,CAAC,CAAC;QAC7B,OAAO,EAAE,CAAC;IACd,CAAC;IAEO,WAAW,CAAC,QAAwB;QACxC,MAAM,EAAE,iBAAiB,EAAE,IAAI,EAAE,GAAG,IAAI,CAAC,SAAS,CAAC,YAAY,CAAC;QAChE,MAAM,KAAK,GAAG,QAAQ,CAAC,WAAW,GAAG,iBAAiB,CAAC;QACvD,MAAM,GAAG,GAAG,QAAQ,CAAC,SAAS,GAAG,iBAAiB,CAAC;QAEnD,OAAO,IAAI,CAAC,KAAK,CAAC,KAAK,EAAE,GAAG,CAAC,CAAC;IAClC,CAAC;IAED,SAAS;IACU,oBAAoB,CAAC,SAAiB,EAAE,KAAe;QACtE,IAAI,CAAC,KAAK,CAAC,oBAAoB,CAAC,SAAS,EAAE,KAAK,CAAC,EAAE;YAC/C,IAAI,CAAC,OAAO,CAAC,IAAI,CAAC,WAAW,CAAC,KAAK,CAAC,kBAAmB,CAAC,CAAC,CAAC;SAC7D;QAED,2DAA2D;QAC3D,2CAA2C;QAC3C,IAAI,CAAC,uBAAuB,CAAC,eAAe,GAAG,KAAK,CAAC;QACrD,OAAO,IAAI,CAAC;IAChB,CAAC;IAED,cAAc;IACK,UAAU,CAAC,SAAiB,EAAE,KAAe;QAC5D,IAAI,CAAC,IAAI,CAAC,SAAS,EAAE,KAAK,EAAE,IAAI,CAAC,WAAW,CAAC,KAAK,CAAC,kBAAmB,CAAC,CAAC,CAAC;IAC7E,CAAC;IAED,qEAAqE;IAC9D,WAAW,CAAC,KAAc;QAC7B,IAAI,GAAG,GAAG,aAAa,KAAK,CAAC,IAAI,EAAE,CAAC;QAEpC,IAAI,KAAK,CAAC,QAAQ,KAAK,IAAI,EAAE;YACzB,GAAG,IAAI,YAAY,KAAK,CAAC,QAAQ,GAAG,CAAC;SACxC;aAAM,IAAI,KAAK,CAAC,QAAQ,KAAK,IAAI,EAAE;YAChC,GAAG,IAAI,SAAS,CAAC;SACpB;QAED,IAAI,KAAK,CAAC,QAAQ,KAAK,IAAI,EAAE;YACzB,GAAG,IAAI,KAAK,KAAK,CAAC,QAAQ,GAAG,CAAC;SACjC;QAED,GAAG,IAAI,GAAG,CAAC;QAEX,IAAI,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;IACnB,CAAC;IAED,iEAAiE;IAC1D,YAAY,CAAC,KAAe;QAC/B,IAAI,GAAG,GAAG,IAAI,KAAK,CAAC,OAAO,EAAE,CAAC;QAE9B,KAAK,MAAM,IAAI,IAAI,KAAK,CAAC,KAAK,EAAE;YAC5B,GAAG,IAAI,IAAI,IAAI,CAAC,IAAI,KAAK,eAAe,CAAC,IAAI,CAAC,KAAK,CAAC,GAAG,CAAC;SAC3D;QAED,GAAG,IAAI,KAAK,CAAC,WAAW,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,GAAG,CAAC;QAEtC,IAAI,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;IACnB,CAAC;IAED,+DAA+D;IACxD,UAAU,CAAC,KAAa;QAC3B,IAAI,CAAC,IAAI,CAAC,KAAK,KAAK,CAAC,OAAO,GAAG,CAAC,CAAC;IACrC,CAAC;IAED,4DAA4D;IACrD,QAAQ,CAAC,EAAE,IAAI,EAAQ;QAC1B,IAAI,CAAC,IAAI,CACL,CAAC,IAAI,CAAC,uBAAuB,CAAC,gBAAgB;YAC1C,IAAI,CAAC,gBAAgB,CAAC,IAAI,CAAC,SAAS,CAAC,gBAAgB,EAAE,IAAI,CAAC;YAC5D,CAAC,CAAC,IAAI;YACN,CAAC,CAAC,UAAU,CAAC,IAAI,CAAC,CACzB,CAAC;IACN,CAAC;IAED,+DAA+D;IACxD,WAAW,CAAC,KAAc;QAC7B,IAAI,CAAC,IAAI,CAAC,OAAO,KAAK,CAAC,IAAI,KAAK,CAAC,CAAC;IACtC,CAAC;IAED,sDAAsD;IAC/C,OAAO,CAAC,IAAY;QACvB,IAAI,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;IACpB,CAAC;CACJ"}