CSV.parse not removing BOM character about csv HOT 5 CLOSED

Right. So I should do:

csv = CSV.parse(File.read(t, encoding: 'bom|utf-8'), headers: true)

Thanks very much for your help! Sorry to raise an issue when the issue is with my understanding 👍

from csv.

I had to do this with a file today as well (parsing an export from 新生銀行)

CSV.foreach(file, encoding: 'bom|utf-8') - thanks for posting your solution @davich

from csv.

Right. So I should do:

csv = CSV.parse(File.read(t, encoding: 'bom|utf-8'), headers: true)

Thanks very much for your help! Sorry to raise an issue when the issue is with my understanding 👍

I have to manually remove BOM in order to test the response in my rspec.

BOM = "\xEF\xBB\xBF"
expect(CSV.parse(response.body.delete_prefix(BOM))).to eq [['abcdefg']]

It is not elegant, but maybe it is intended since it says: BOM is for opening a file not parse target string. in #23

from csv.

BOM is for file not string data.

I don't know why some users want to use BOM for string data.
Can you show your use case?

from csv.

Per @kou 's comment, it's a little curious that response.body would contain a BOM. But if it does, what you did seems to make sense.

I see a mention of using IO.read on https://bugs.ruby-lang.org/issues/15210 from @nobu but I'm not sure it makes sense just to remove a few characters from a string.

from csv.