Update to FastDoubleParser v1.0.1 to fix BigDecimal decoding problem - #1331
Conversation
|
As per my note on FasterXML/jackson-databind#4694 I think I'd like to see if there's any chance of getting FDP fix and new version before adding work-arounds. In the meantime what would be valuable would be (for now) failing test case first, to check fix, guard against regression. |
cowtowncoder
left a comment
cowtowncoder
left a comment
There was a problem hiding this comment.
So: makes sense, but will wait for a bit before merging to see if there's any chance to get fix via FDP before 2.18.0 release. Or not.
ede3461 to
acbece6
Compare
|
Thank you for clean up @pjfanning ! Will make sure we'll either add this work-around, or upgrade if FDP fix available, before 2.18.0 final. |
pjfanning
changed the title
| if (valueStr.length() < 500) { | ||
| return new BigDecimal(valueStr); | ||
| } | ||
| return JavaBigDecimalParser.parseBigDecimal(valueStr); |
There was a problem hiding this comment.
effectively, the same code as the parse(char[]) method but slight optimisation is that we avoid converting to a char array when we call the JavaBigDecimalParser method
cowtowncoder
changed the title
BigDecimal decoding problem
|
@cowtowncoder The bug is an edge case, is this change worth backporting to 2.17? I'm on the fence. |
|
Merged, backporting in 2.17(.3) -- just the library upgrade tho, not other changes. Not certain 2.17.3 will be released (not many fixes, this was 3rd) but just in case. |
2 participants
see FasterXML/jackson-databind#4694 (comment)
Pick up fix for wrandelshofer/FastDoubleParser#80.
Thanks to @wrandelshofer for his continued support and quick response times.
https://github.com/wrandelshofer/FastDoubleParser/releases/tag/v1.0.1