将大文件切成块并使用 ajax 和 html5 FileReader 上传
我要实现的是:
我在前端使用html5文件api读取文件,然后使用ajax将文件内容上传到php后端,文件小也可以.但是,如果文件足够大,它会导致 chrome 崩溃.所以我使用file.slice将大文件分割成块,当所有块都上传到php时,将块合并成一个完整的块.
In the front end, I use the html5 file api to read the file, and then upload the file's content to the php backend using ajax, and it's ok if the filesize is small. However,if the file is big enough, it causes chrome to crash. So I split the large file into chunks using file.slice, when all chunks are uploaded to the php, merge the chunks into a single complete one.
代码如下:
前端:
<style>
#container {
min-width:300px;
min-height:200px;
border:3px dashed #000;
}
</style>
<div id='container'>
</div>
<script>
function addDNDListener(obj){
obj.addEventListener('dragover',function(e){
e.preventDefault();
e.stopPropagation();
},false);
obj.addEventListener('dragenter',function(e){
e.preventDefault();
e.stopPropagation();
},false);
obj.addEventListener('drop',function(e){
e.preventDefault();
e.stopPropagation();
var ul = document.createElement("ul");
var filelist = e.dataTransfer.files;
for(var i=0;i<filelist.length;i++){
var file = filelist[i];
var li = document.createElement('li');
li.innerHTML = '<label id="'+file.name+'">'+file.name+':</label> <progress value="0" max="100"></progress>';
ul.appendChild(li);
}
document.getElementById('container').appendChild(ul);
for(var i=0;i<filelist.length;i++){
var file = filelist[i];
uploadFile(file);
}
},false);
}
function uploadFile(file){
var loaded = 0;
var step = 1024*1024;
var total = file.size;
var start = 0;
var progress = document.getElementById(file.name).nextSibling;
var reader = new FileReader();
reader.onprogress = function(e){
loaded += e.loaded;
progress.value = (loaded/total) * 100;
};
reader.onload = function(e){
var xhr = new XMLHttpRequest();
var upload = xhr.upload;
upload.addEventListener('load',function(){
if(loaded <= total){
blob = file.slice(loaded,loaded+step+1);
reader.readAsBinaryString(blob);
}else{
loaded = total;
}
},false);
xhr.open("POST", "upload.php?fileName="+file.name+"&nocache="+new Date().getTime());
xhr.overrideMimeType("application/octet-stream");
xhr.sendAsBinary(e.target.result);
};
var blob = file.slice(start,start+step+1);
reader.readAsBinaryString(blob);
}
window.onload = function(){
addDNDListener(document.getElementById('container'));
if(!XMLHttpRequest.prototype.sendAsBinary){
XMLHttpRequest.prototype.sendAsBinary = function(datastr) {
function byteValue(x) {
return x.charCodeAt(0) & 0xff;
}
var ords = Array.prototype.map.call(datastr, byteValue);
var ui8a = new Uint8Array(ords);
try{
this.send(ui8a);
}catch(e){
this.send(ui8a.buffer);
}
};
}
};
</script>
php 代码:
<?php
$filename = "upload/".$_GET['fileName'];
//$filename = "upload/".$_GET['fileName']."_".$_GET['nocache'];
$xmlstr = $GLOBALS['HTTP_RAW_POST_DATA'];
if(empty($xmlstr)){
$xmlstr = file_get_contents('php://input');
}
$is_ok = false;
while(!$is_ok){
$file = fopen($filename,"ab");
if(flock($file,LOCK_EX)){
fwrite($file,$xmlstr);
flock($file,LOCK_UN);
fclose($file);
$is_ok = true;
}else{
fclose($file);
sleep(3);
}
}
问题是,文件的chunk全部上传到服务器并合并为一个新文件后,文件总大小比原来的要小,合并后的文件坏了.问题出在哪里以及如何解决?
The problem is, after the chunks of the file all being uploaded to the server and merged into a new one, the total file size is smaller than the original, and the merged one is broken. Where is the problem and how to fix it?
推荐答案
- 使用 readAsBinaryString fn 是不好的做法
- SendAsBinary 也被贬低了
- 读取块内容只是纯粹的傻瓜.切片就足够了.
xhr.send(blob.slice(0,10))
- 切片也是不必要的,除非服务器不接受如此大的文件(例如 Dropbox 受限的 REST API)
- 因此,如果有人试图巧妙地使用工作线程、base64 或 FileReader 上传大文件 - 不要这样做,这都是不必要的.
只有当您决定在将文件发送到服务器之前对其进行加密/解密/压缩时,才可以读取/切片文件.
但仅在有限的时间内,直到所有浏览器开始支持流.
那你应该看看 fetch 和 ReadableStream
Only time it's okey to read/slice the file is if you are deciding to encrypt/decrypt/zip the files before sending it to the server.
But only for a limited time until all browser start supporting streams.
Then you should take a look at fetch and ReadableStream
fetch(url, {method: 'post', body: new ReadableStream({...})})
<小时>
如果您只需要将 blob 转发到服务器,只需执行以下操作:xhr.send(blob_or_file)
并且浏览器将负责(正确地)读取它并且不消耗任何内存.并且文件可以是文件/blob 的大小
if you just need to forward the blob to the server, just do:
xhr.send(blob_or_file)
and the browser will take care of reading it (correctly) and not consume any memory. And the file can be however large the file/blob is
相关文章