ISSUE-6612 FIX: parse long field in GenricJsonRecord (#6612) #6622
Conversation
|
/pulsarbot run-failure-checks |
|
/pulsarbot run-failure-checks |
6 similar comments
|
/pulsarbot run-failure-checks |
|
/pulsarbot run-failure-checks |
|
/pulsarbot run-failure-checks |
|
/pulsarbot run-failure-checks |
|
/pulsarbot run-failure-checks |
|
/pulsarbot run-failure-checks |
|
/pulsarbot run-failure-checks |
2 similar comments
|
/pulsarbot run-failure-checks |
|
/pulsarbot run-failure-checks |
| return fn.asDouble(); | ||
| } else if (fn.isDouble()) { | ||
| return fn.asDouble(); | ||
| } else if (fn.isNumber()) { |
There was a problem hiding this comment.
I think the integration tests failed due to this change.
I think we should keep the original behavior and add the following
else if (fn.isLong()) {
return fn.asLong();
} else if (fn.isNumber()) {
return fn.numberValue();
}
There was a problem hiding this comment.
if fn is long, then fn.aslong() should equals to fn.numberValue()。
The differences compared to the previous code are:
- If fn is BigDecimal, the new code will return BigDecimal rather than double
- If fn is BigInteger, the new code will return BigInteger rather than text
- If fn is float, the new code will return float rather than double
So i modify the code like:
} else if (fn.isBoolean()) {
return fn.asBoolean();
} else if (fn.isFloatingPointNumber()) {
return fn.asDouble();
} else if (fn.isBigInteger()) {
if (fn.canConvertToLong()) {
return fn.asLong();
} else {
return fn.asText();
}
} else if (fn.isNumber()) {
return fn.numberValue();
} else {
return fn.asText();
}I am trying to do integration tests locally but it takes too long to download snapshot jars from maven. I'll try again, and if fail again, can I push directly and do ci test?
There was a problem hiding this comment.
Is there anything wrong with CI workflow? I reset to commit the one before the first commit of this PR, and it still fails.
There was a problem hiding this comment.
I found this pull #6660, and it has been merged.
I should try again.
There was a problem hiding this comment.
Finally pass the CI
|
/pulsarbot run-failure-checks |
1 similar comment
|
/pulsarbot run-failure-checks |
|
/pulsarbot run-failure-checks |
4 similar comments
|
/pulsarbot run-failure-checks |
|
/pulsarbot run-failure-checks |
|
/pulsarbot run-failure-checks |
|
/pulsarbot run-failure-checks |
lynnmatrix
force-pushed
the
fix/json_decode_long
branch
from
f9f57f0
to
26cd98e
Compare
|
/pulsarbot run-failure-checks |
2 similar comments
|
/pulsarbot run-failure-checks |
|
/pulsarbot run-failure-checks |
|
/pulsarbot run-failure-checks |
lynnmatrix
force-pushed
the
fix/json_decode_long
branch
2 times, most recently
from
03b803d
to
d4ee5a6
Compare
|
/pulsarbot run-failure-checks |
1 similar comment
|
/pulsarbot run-failure-checks |
lynnmatrix
force-pushed
the
fix/json_decode_long
branch
from
d4ee5a6
to
6905380
Compare
|
/pulsarbot run-failure-checks |
2 similar comments
|
/pulsarbot run-failure-checks |
|
/pulsarbot run-failure-checks |
…pache#6622) Fixes apache#6612 ### Motivation If message sent in json schema, long field will be decoded as int if its value below Integer.MAX_VALUE, other wise decoded as string. For example, the json message below: ```json { "timestamp": 1585204833128 } ``` will be decoded as ```json { "timestamp": "1585204833128" } ``` ### Modifications Add field type check in GenericJsonRecord
Fixes #6612 ### Motivation If message sent in json schema, long field will be decoded as int if its value below Integer.MAX_VALUE, other wise decoded as string. For example, the json message below: ```json { "timestamp": 1585204833128 } ``` will be decoded as ```json { "timestamp": "1585204833128" } ``` ### Modifications Add field type check in GenericJsonRecord (cherry picked from commit 1aad3b7)
…pache#6622) Fixes apache#6612 ### Motivation If message sent in json schema, long field will be decoded as int if its value below Integer.MAX_VALUE, other wise decoded as string. For example, the json message below: ```json { "timestamp": 1585204833128 } ``` will be decoded as ```json { "timestamp": "1585204833128" } ``` ### Modifications Add field type check in GenericJsonRecord
Fixes #6612
Motivation
If message sent in json schema, long field will be decoded as int if its value below Integer.MAX_VALUE, other wise decoded as string.
For example, the json message below:
{ "timestamp": 1585204833128 }will be decoded as
{ "timestamp": "1585204833128" }Modifications
Add field type check in GenericJsonRecord
Verifying this change
Does this pull request potentially affect one of the following parts:
Documentation