mgmeyers/linereader

Reads files line by line in browser using the HTML5 FileReader API.

linereader's Issues

Line breaks are ignored when at the end of result

Whenever this.result ends in a line break (it contains a complete line), it will be concatenated with the next line, ignoring the line break.

request to write back into the scanned document

it's a great lib
it would be great and usefull to be able to write into the soucre document
do you have anyhting like this in a smaple by any chance ?
cheers

Problem with CSV files with UTF-8

Hello, your library is very good, but there are some problems related to character encoding and chunk size.

For example: when trying to read the string "Alimentação", it reads "Alimentaç��o".

I investigated and this problem is caused by the slice method (from FileReader).

The problem happens when the file is chunked exactly over an UTF-8 character that has more than one byte. UTF-8 characters are stored with 1 up to 4 bytes. The problem will not appear if every UTF-8 character of a text file has 1 byte (for example, English language).

But when you try to use other languages, like Latin languages or Asian languages like Japanese or Chinese, you will notice that slice() method may chunk in the middle of a UTF-8 character.

Later, when the strings are glued, they end up forming one or more invalid characters. One or more questions marks appears like in the example I gave above.

I don't think there is a solution for this, because slice() method works at byte level.

newlines are ignored in Linux

I read your code and run the demo with the test.txt attached in your repo. The result I got is
1: 12345678910, but the expected results should be:
1: 1
2: 2
3: 3
....
right? It seems the newlines are ignored. I don't know whether it is system dependent thing, I am using Linux.

Publish on npmjs

It would be great to have this on npmjs.