Character encoding issue

Shuguang Xiang

unread,

Jan 11, 2021, 4:57:46 AM1/11/21

to debezium

Hi Jiri and Gunnar,

I reported this issue before and it was fixed in 1.4.0.Alpha1.

I applied 1.4.0.Alpha2 in our production env but still saw the error. The connector is using windows-1252 charset for a column defined as latin1 charset. I have attached more info in the following Jira. Could you help take a look?

https://issues.redhat.com/browse/DBZ-2604

Best regards,

Shuguang

Shuguang Xiang

unread,

Jan 13, 2021, 9:54:04 AM1/13/21

to debezium

The issue mentioned here is different from DBZ-2604.

In DBZ-2922, when column whose collation is set but charset is not set, the server charset is chosen. Theoretically, the column charset should be derived from collation and applied.

https://issues.redhat.com/browse/DBZ-2922

Best regards,

Shuguang

Shuguang Xiang

unread,

Jan 26, 2021, 9:12:56 PM1/26/21

to debezium

Hi team,

Could you help take a look at this pull request? From the test result, it could resolve column charset from column level collation. It is taking database charset.
https://github.com/debezium/debezium/pull/2067