dan/node-html-parser
1
0
Fork 0
mirror of https://github.com/danbulant/node-html-parser synced 2026-05-19 20:38:46 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

remove duplicated items in querySelectAll

Browse source
This commit is contained in:
taoqf 2020-01-26 21:57:27 +08:00
parent c145442179
commit 881d708e72
2 changed files with 19 additions and 8 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

15
src/index.ts
View file

@ -219,7 +219,7 @@ export class HTMLElement extends Node {
currentBlock.push(text);
}
}
}
}
dfs(this);
return blocks
.map(function (block) {
@ -353,11 +353,12 @@ export class HTMLElement extends Node {
} else {
if (selector.includes(',')) {
const selectors = selector.split(',') as string[];
let result = [] as HTMLElement[];
selectors.forEach((s) => {
result = result.concat(this.querySelectorAll(s.trim()));
});
return result;
return Array.from(selectors.reduce((pre, cur) => {
const result = this.querySelectorAll(cur.trim()) as HTMLElement[];
return result.reduce((p, c) => {
return p.add(c);
}, pre);
}, new Set<HTMLElement>()));
}
matcher = new Matcher(selector);
}
@ -801,7 +802,7 @@ export function parse(data: string, options?: {
// this is a comment
if (options.comment) {
// Only keep what is in between <!-- and -->
const text = data.substring(lastTextPos - 3 , lastTextPos - match[0].length + 4);
const text = data.substring(lastTextPos - 3, lastTextPos - match[0].length + 4);
currentParent.appendChild(new CommentNode(text));
}
continue;

12
test/html.js
View file

@ -129,7 +129,7 @@ describe('HTML Parser', function () {
it('should parse picture element', function () {
var root = parseHTML('<picture><source srcset="/images/example-1.jpg 1200w, /images/example-2.jpg 1600w" sizes="100vw"><img src="/images/example.jpg" alt="Example"/></picture>');
var picture = new HTMLElement('picture', {}, '');
var source = picture.appendChild(new HTMLElement('source', {}, 'srcset="/images/example-1.jpg 1200w, /images/example-2.jpg 1600w" sizes="100vw"'));
var img = picture.appendChild(new HTMLElement('img', {}, 'src="/images/example.jpg" alt="Example"'));
@ -341,6 +341,16 @@ describe('HTML Parser', function () {
root.querySelectorAll('#id span').should.eql(root.firstChild.firstChild.childNodes);
root.querySelectorAll('#id, #id .b').should.eql([root.firstChild, root.firstChild.firstChild.firstChild]);
});
it('should return just one element', function () {
var root = parseHTML('<time class="date">');
root.querySelectorAll('time,.date').should.eql([root.firstChild]);
});
it('should return elements in order', function () {
var root = parseHTML('<img src=""><p>hello</p>');
var results = root.querySelectorAll('p,img');
results[0].rawText.should.eql('hello');
results[1].should.eql(root.firstChild);
});
});
describe('#structuredText', function () {