admin/n8n-enterprise-unlocked
1
0
Fork 0
mirror of https://github.com/Abdulazizzn/n8n-enterprise-unlocked.git synced 2025-12-17 10:02:05 +00:00
Code Issues Packages Projects Releases Wiki Activity

feat(HTML Extract Node): Better text extraction, option to specify selectors to skip, option to clean up text data (#8586)

Browse Source
This commit is contained in:
Michael Kret
2024-02-12 12:52:51 +02:00
committed by GitHub
parent 510bf8905d
commit 32281d12d7
8 changed files with 555 additions and 12 deletions
Download Patch File Download Diff File Expand all files Collapse all files

47
packages/nodes-base/nodes/Html/utils.ts
View File

@@ -1,26 +1,59 @@
import type { IDataObject } from 'n8n-workflow';
import type { IValueData, Cheerio } from './types';
import { convert } from 'html-to-text';
// The extraction functions
const extractFunctions: {
[key: string]: ($: Cheerio, valueData: IValueData) => string | undefined;
[key: string]: ($: Cheerio, valueData: IValueData, nodeVersion: number) => string | undefined;
} = {
attribute: ($: Cheerio, valueData: IValueData): string | undefined =>
$.attr(valueData.attribute!),
html: ($: Cheerio, _valueData: IValueData): string | undefined => $.html() || undefined,
text: ($: Cheerio, _valueData: IValueData): string | undefined => $.text(),
text: ($: Cheerio, _valueData: IValueData, nodeVersion: number): string | undefined => {
if (nodeVersion <= 1.1) return $.text() || undefined;
const html = $.html() || '';
let options;
if (_valueData.skipSelectors) {
options = {
selectors: _valueData.skipSelectors.split(',').map((s) => ({
selector: s.trim(),
format: 'skip',
})),
};
}
return convert(html, options);
},
value: ($: Cheerio, _valueData: IValueData): string | undefined => $.val(),
};
/**
* Simple helper function which applies options
*/
export function getValue($: Cheerio, valueData: IValueData, options: IDataObject) {
const value = extractFunctions[valueData.returnValue]($, valueData);
if (options.trimValues === false || value === undefined) {
export function getValue(
$: Cheerio,
valueData: IValueData,
options: IDataObject,
nodeVersion: number,
) {
let value = extractFunctions[valueData.returnValue]($, valueData, nodeVersion);
if (value === undefined) {
return value;
}
return value.trim();
if (options.trimValues) {
value = value.trim();
}
if (options.cleanUpText) {
value = value
.replace(/^\s+|\s+$/g, '')
.replace(/(\r\n|\n|\r)/gm, '')
.replace(/\s+/g, ' ');
}
return value;
}