On 5/12/20 2:51 AM, Xiangdong Huang wrote:
Hi Giorgio, Thanks for reminding the details. When we designed the TsFile, we wanted to use UTF-8.
Ok thanks, so std::string can handle it.
I checked the source code, sometimes we clearly claim "utf-8" (e.g., device name and measurement name) but something we do not (e.g., data point value). Maybe we should claim the encoding all the time to avoid potential bugs. Besides, as my computer is utf-8 by default, the String in Java also uses utf-8 (and I think that is why there is no issue on my computer)
It is weird, docs says not the same: https://docs.oracle.com/javase/10/docs/api/java/lang/Character.html#unicode But if i execute the test: jozoppi@giorgio-XPS-15-7590:~/vcpkg$ java StringTest UTF-8 is arr1 equals to arr2 : true is arr1 equals to arr3 : false jozoppi@giorgio-XPS-15-7590:~/vcpkg$ Interesting. BR, Giorgio,
