# Japanese characters transcoding: stop using Shift-JIS for the source

So, last day I got an encoding error. The data is what I though was in Shift-JIS but it could not be converted to UTF-8 because in Shift-JIS `‡` U+0087 does not exists.

![](https://cdn.hashnode.com/res/hashnode/image/upload/v1713333976175/9ff3e163-0418-444f-a1d9-5e3dbc06c526.png align="center")

It's is multiple used as title like so:

![](https://cdn.hashnode.com/res/hashnode/image/upload/v1713334233487/584977da-86eb-404b-91b9-55d392b40f37.png align="center")

[https://www.htmlsymbols.xyz/unicode/U+0087](https://www.htmlsymbols.xyz/unicode/U+0087)

# Solution:

In Windows, apparently, instead of the strict Shift-JIS, they are using CP932 which is superset of Shift-JIS (meaning it's Shift-JIS with additional characters).

### Please use `CP932` encoding when dealing with Japanese characters.

The Rubyist out there, it will be:

```ruby
CSV.read("csvfile.csv", encoding: "CP932:UTF-8")
```