That is an interesting question indeed, but unfortunately the answer is that the extra length isn't detected or flagged with your scenario using a schema file and RCP.
Depending upon how important this is to you, there are some methods that I can think of to get this information, but they do involve quite a bit of work. You can read the data in using a dummy schema defining the columns as varchar with no length, apply an aggregator to get the maximum length and then compare that with the appropriate values in the schema file to output a flag on whether a column's length has been exceeded. It is neither simple nor efficient if you have large amounts of data.
Sequential file stage with schemafile crops text
Moderators: chulett, rschirm, roy
You could make an assumption that any string that runs out the full length "may have been" cropped and raise a flag when that happens with very simple transformer logic (for example have 2 outputs from a transformer, normal path as already happens and second path that writes all the records that have length = max to a dataset for evaluation). You can read the evaluation dataset to see if the vast majority of these are cropped records and try to find a way to handle it from there.
The transformer stage is frequently the way to avoid implicit type conversion warnings, transforming larger strings into smaller strings, and other such things as might require an explicit transformation.
You could try replacing the transformer with a copy or a modify to see if the warnings you want to see will appear.
Mike
You could try replacing the transformer with a copy or a modify to see if the warnings you want to see will appear.
Mike