DOMParser - 儿童不是DOM对象
问题描述:
DOMParser有一种奇怪的行为。当我使用“text/xml”作为参数时,我得到我的对象,并且每次使用子对象(如parentNodes)时,该子对象本身就是一个DOM对象。但是,当我使用“text/html”作为参数时,孩子不是DOM对象。为什么是这样,我怎么能为所有的孩子有DOM对象?DOMParser - 儿童不是DOM对象
这里是我做的:
parser = new DOMParser();
doc = parser.parseFromString(stringContainingHTMLSource, "text/html").getElementsByTagName('p');
console.log(doc[0].childNodes[0]);
我childNode返回元素,但不是作为一个DOM对象...
编辑: 这里是我的递归函数:
var getParents = function(node, parentNodes){
if(node.nodeName == 'span'){
parentNodes.push(node.attributes[0].nodeValue);
} else if(node.nodeName == 'p' && node.attributes.length > 0) {
parentNodes.push(node.nodeName);
parentNodes.push(node.attributes[0].nodeValue);
} else {
parentNodes.push(node.nodeName);
}
if(node.parentNode.nodeName != '#document'){
getParents(node.parentNode, parentNodes);
}
return parentNodes;
};
var parse = function(node, vertical, horizontal, paragraph){
if(node.childNodes.length > 0){
for(var int = 0; int < node.childNodes.length; int++){
parse(node.childNodes[int], vertical, horizontal, paragraph);
}
} else{
var object = {};
var attributes = getParents(node, []);
for(var int = 0; int < attributes.length; int++) {
// right alignment
if(/text-align/i.test(attributes[int])){
object.alignment = attributes[int].split(": ")[1].replace(';','');
} else if (/color/i.test(attributes[int])) {
// color
object.color = attributes[int].split(":")[1];
} else if (attributes[int] == 'em') {
// italic
if (object.italics) {
delete object.bold;
object.bolditalics = true;
} else {
object.italics = true;
}
} else if (attributes[int] == 'strong') {
// bold
if (object.italics) {
delete object.italics;
object.bolditalics = true;
} else {
object.bold = true;
}
} else if (attributes[int] == 'u') {
// underline
object.decoration = 'underline';
} else if (attributes[int] == 's') {
// strike
object.decoration = 'lineThrough';
}
}
object.text = node.textContent;
pdfContent[vertical][horizontal].push(object);
}
};
for(var vertical = 0; vertical < payment.htmlContent.length; vertical++) {
for(var horizontal = 0; horizontal < payment.htmlContent[vertical].length; horizontal++) {
var parser = new DOMParser();
var paragraphs = parser.parseFromString(payment.htmlContent[vertical][horizontal], "text/xml").getElementsByTagName('p');
for (var paragraph = 0; paragraph < paragraphs.length; paragraph++) {
for (var num = 0; num < paragraphs[paragraph].childNodes.length; num++) {
parse(paragraphs[paragraph].childNodes[num], vertical, horizontal, paragraph);
}
}
}
}
答
我对这些值做了一些假设,然后我在代码中添加了一些验证码,如if(node.attributes.length>0)
,它似乎可行。
var payment={htmlContent:[['<p>some<em>text</em></p>', '<p>some<span>text<strong>here</strong></span></p>'],['<p>some<s>text</s></p>', '<p>some<span style="color:#FF00FF">text</span></p>']]};
var getParents = function(node, parentNodes){
if(node.nodeName == 'span'){
if(node.attributes.length>0)
parentNodes.push(node.attributes[0].nodeValue);
} else if(node.nodeName == 'p' && node.attributes.length > 0) {
parentNodes.push(node.nodeName);
if(node.attributes.length>0)
parentNodes.push(node.attributes[0].nodeValue);
} else {
parentNodes.push(node.nodeName);
}
if(node.parentNode.nodeName != '#document'){
getParents(node.parentNode, parentNodes);
}
return parentNodes;
};
var parse = function(node, vertical, horizontal, paragraph){
if(node.childNodes.length > 0){
for(var int = 0; int < node.childNodes.length; int++){
parse(node.childNodes[int], vertical, horizontal, paragraph);
}
} else{
var object = {};
var attributes = getParents(node, []);
console.log(attributes);
for(var int = 0; int < attributes.length; int++) {
// right alignment
if(/text-align/i.test(attributes[int])){
object.alignment = attributes[int].split(": ")[1].replace(';','');
} else if (/color/i.test(attributes[int])) {
// color
object.color = attributes[int].split(":")[1];
} else if (attributes[int] == 'em') {
// italic
if (object.italics) {
delete object.bold;
object.bolditalics = true;
} else {
object.italics = true;
}
} else if (attributes[int] == 'strong') {
// bold
if (object.italics) {
delete object.italics;
object.bolditalics = true;
} else {
object.bold = true;
}
} else if (attributes[int] == 'u') {
// underline
object.decoration = 'underline';
} else if (attributes[int] == 's') {
// strike
object.decoration = 'lineThrough';
}
}
object.text = node.textContent;
if(!pdfContent[vertical])pdfContent[vertical]=[];
if(!pdfContent[vertical][horizontal])
pdfContent[vertical][horizontal]=[];
pdfContent[vertical][horizontal].push(object);
}
};
var pdfContent = [];
for(var vertical = 0; vertical < payment.htmlContent.length; vertical++) {
for(var horizontal = 0; horizontal < payment.htmlContent[vertical].length; horizontal++) {
var parser = new DOMParser();
var paragraphs = parser.parseFromString(payment.htmlContent[vertical][horizontal], "text/xml").getElementsByTagName('p');
for (var paragraph = 0; paragraph < paragraphs.length; paragraph++) {
for (var num = 0; num < paragraphs[paragraph].childNodes.length; num++) {
parse(paragraphs[paragraph].childNodes[num], vertical, horizontal, paragraph);
}
}
}
}
for(var i=0; i<pdfContent.length; i++){
for(var j=0; j<pdfContent[i].length; j++){
document.querySelector('#log').textContent+=pdfContent[i][j].toSource();
}
}
<p id="log"></p>
你所说的 “他们不是DOM对象” 呢?你能给我们一个你传递的字符串和输出的例子吗?乍一看我会说这是因为childNodes返回节点。你可能只想用'doc [0] .children [0]'来检查元素。 – Kaiido
@Kaiido嗯,这就是我想要做的......我递归地进入所有childNodes(递归函数),当我到达我的最后一个孩子时,我通过我的父节点递归地回去(以获得孩子所有的nodeNames )。使用xml,它工作得很好,但带有html,当我到达最后一个childNode时,它不再有父节点,因为它只是一个简单的元素。 (将用递归函数更新我的问题...) – ncohen
是的,请更新,对我来说,它在控制台中工作,手动进入最后一个节点并再次爬升。但是,也许你会对[TreeWalker](https://developer.mozilla.org/en/docs/Web/API/TreeWalker)对这类任务感兴趣 – Kaiido