Fix Range Header processing logic for NetworkStreamer

[t83714]

Not all HTTP servers implement the correct behaviour of processing Range header for partial download.

Some HTTP server may ignore the range-end parameter and return the file content from the start index to the end of the file.

e.g.:
Request:

Range: bytes=0-2097151
Sec-Fetch-Mode: cors
User-Agent: Mozilla/5.0 

Response:

accept-ranges: bytes
Content-Length: 8256106
Content-Range: bytes 0-8256106/8256106
server: nginx/1.13.12
status: 206, 206 Partial Content

As we increase the _start index by chunkSize config option at each round of chunk processing (instead of increasing the _start by the actual size of the response), papaparse may incorrect re-download & parse unnecessary chunk:

PapaParse/papaparse.js

Lines 673 to 688 in 94d7bf9

	this._start += this._config.chunkSize;
	};
	this._chunkLoaded = function()
	{
	if (xhr.readyState !== 4)
	return;
	if (xhr.status < 200 || xhr.status >= 400)
	{
	this._chunkError();
	return;
	}
	this._finished = !this._config.chunkSize || this._start > getFileSize(xhr);
	this.parseChunk(xhr.responseText);

This PR allows Papaparse to handle this situation and avoid downloading file content for many times.

[pokoli]

Thanks for your proposal.

Please do not modify the minified file it will be automatically modified on our release process.

I've added a comment on the PR, once fixed I will be happy to merge the fix.

[t83714]

@pokoli Could you please have another look?
I have removed the empty lines and changes to the minified file.

[pokoli]

@t83714 I've merged it. Thanks!