Source data error in reusable dataflow

Hi,

We have a REUSABLE dataflow loading a (parameterized) Synapse table from a (also parameterized) parquet file.

It failed with a source value greater than the 8,000 maximum character size Synapse accommodates. Our source is Salesforce, which does not have a substring function we could resort to address this.

Do we have an option to have the dataflow simply truncate the source value by default ?

Please advise.