| Bug #97226 | Incorrect charset comparison to 'UTF-8' when reading a longtext field as NClob | ||
|---|---|---|---|
| Submitted: | 15 Oct 2019 9:36 | Modified: | 16 Nov 2019 11:36 |
| Reporter: | Oleksandr Shkurat | Email Updates: | |
| Status: | No Feedback | Impact on me: | |
| Category: | Connector / J | Severity: | S2 (Serious) |
| Version: | 8.0.17 | OS: | Any |
| Assigned to: | CPU Architecture: | Any | |
| Tags: | charset, NClob, UTF-8 | ||
[15 Oct 2019 9:39]
Oleksandr Shkurat
Screenshot from IDEA on breakpoint
Attachment: NClob_Issue.png (image/png, text), 159.30 KiB.
[16 Oct 2019 11:36]
Alexander Soklakov
Hi Oleksandr, I can not reproduce this error. Please provide the exact table definition, c/J connection string and the code snippet how you create the ResultSet. It would be great if you can provide a reproducible test case.
[17 Nov 2019 1:00]
Bugs System
No feedback was provided for this bug for over a month, so it is being suspended automatically. If you are able to provide the information that was originally requested, please do so and change the status of the bug back to "Open".
Description: My application works with MySQL DB using Spring Data JPA 2.1.11 (Spring boot 2.19), which uses Hibernate 5.3.13 and MySQL Connector 8.0.17. I've declared an entity with a field annotated with @Type(type = "materialized_nclob") So that, there was created a table with a field of type 'longtext' type. The DB, table and this field are marked as characterset=utf8, collation=utf8_general_ci. Now I've got a problem with a reading of this field from DB. The problem occurs inside the class `com.mysql.cj.jdbc.result.ResultSetImpl` in its method `public NClob getNClob(int columnIndex)` There is an exception with message `Can not call getNClob() when field's charset isn't UTF-8` How to repeat: Save some string value to a `longtext` field and then read it Suggested fix: Looks like `UTF-8` is non-unique identifier for UFT8 codepage. That's why the code: ` if (fieldEncoding != null && fieldEncoding.equals("UTF-8")) { String asString = this.getStringForNClob(columnIndex); return asString == null ? null : new com.mysql.cj.jdbc.NClob(asString, this.getExceptionInterceptor()); } else { throw new SQLException("Can not call getNClob() when field's charset isn't UTF-8"); }` doesn't work correctly, because in my case fieldEncoding has the value 'utf8'. Possibly there can be implemented a set of valid identifiers for utf8. For example: Sets.newHashSet("UTF-8", "utf8")