github-actions[bot] commented on code in PR #63322:
URL: https://github.com/apache/doris/pull/63322#discussion_r3256322828
##########
be/src/util/json/json_parser.h:
##########
@@ -53,16 +55,31 @@ Field getValueAsField(const Element& element) {
if (element.isUInt64()) {
return
Field::create_field<TYPE_LARGEINT>(static_cast<int128_t>(element.getUInt64()));
}
+ if (element.isInt128()) {
+ if (preserve_number_as_string) {
+ return
Field::create_field<TYPE_STRING>(String(element.getRawNumber()));
+ }
+ return Field::create_field<TYPE_LARGEINT>(element.getInt128());
+ }
+ if (element.isBigInteger()) {
+ if (preserve_number_as_string) {
+ return
Field::create_field<TYPE_STRING>(String(element.getRawNumber()));
+ }
+ return {};
+ }
Review Comment:
When raw preservation is enabled for any decimal typed path in the Variant
column, the whole document is parsed through the ondemand `Node` tree. For an
unrelated path containing a JSON integer larger than UInt64, `isBigInteger()`
reaches this branch with `preserve_number_as_string == false` and returns a
null `Field`, so a value that used to be present in an untyped path is silently
stored as NULL/missing just because another path in the schema needs decimal
preservation. For example, with a decimal typed path `a`, loading `{"a": 1.23,
"b": 340282366920938463463374607431768211456}` loses `b`. Please preserve
non-decimal big integers as a string (or otherwise keep the previous behavior)
instead of returning null, and add a regression where a decimal typed path
coexists with an unrelated oversized integer path.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
