Error : Type mismatch: cannot convert from int to BigDecimal

[_AnonymousUser]

Hi,
I am working with an Oracle database and number column are translated to BigDecimal in java. But I always have a type mismatch.
Is there a workaround ? Besides, the number column is the PK and is loaded with an oracle sequence.
Thanks,
Jean Marie

[Anonymous]

Hello
Convert int to BigDecimal, try(for example, int id)

new BigDecimal(Integer.toString(row1.id))

Best regards
shong

[Anonymous]

I might put this as bug because the bigdecimal type is used by default when i retrieve a schema from oracle in the metadata repository.
So my workaround id to change the type from bigDecimal to long for each table in the repository.
Jean Marie

[Anonymous]

Hello

I might put this as bug because the bigdecimal type is used by default when i retrieve a schema from oracle in the metadata repository

What's the data type in your table? number?numeric?decimal?
Best regards

shong

[Anonymous]

I think the systematic use of bigDecimal to translate Oracle number is not good.
There should be a test and if the precision is 0, then use int|Integer or bigInteger.

[Anonymous]

Hi Evrybody,
Shong, i think that is a serious problem, cause it gnenrates a java exception , and we can't do any thing sure at our level.
The solution i tried is to get the number type as long type (modifying the BigDecimal type in the Tmap), But some unexpected behaviour can occurs.
Best regards

[Anonymous]

Hi
J've the same problem :
- the datatype of Oracle column is INTEGER
- the datatype (DB TYpe) in Talend schema is NUMBER
Why not INTEGER ?
- the java type in Talend is BigDecimal
Note:
In TOS 2.3 it was: INTEGER->NUMBER-> Long
Expected, for me, is: INTEGER->INTEGER-> Int
Any solution ?
Best Regards
msosmo

[Anonymous]

Hi
Anybody has seen this problem ?
msomso

[Anonymous]

Same issue for me, automatic mapping of NUMBER to BigDecimal is not such a good idea.
Especially since NUMBER without precision is heavily used in a good number of oracle data model design as a datatype for the "id" primary key column of tables.
So I will try to modify the Oracle mapping file . It can be accessed via Windows > preferences >Talend > metadata of TalendType.
The file is pretty self explanatory and specifies the allowed and default mappings for both ways : dbToTalend and talendToDb and both languages : java and perl.
But Alexandre is right, Talend shoud have a way to detect if precision and length are provided in the oracle NUMBER datatype and then map accordingly.
Hope it helps.
David.

[Anonymous]

Hi,
J think J wasn't clear: the problem occurs before the "mapping file" translation.
I work with the INTEGER database columns, and they appear as NUMBER in Talend when the schema based on the table is generated.
Next it, the translation NUMBER -> BigDecimal seems be ok for me.
J think it is a bug, but how can it be rounded ?
Regards
msomso