Processors run before Schema type

Users sometimes report confusion in Huwise around configuring processors.

We just updated the documentation for a specific processor to help make it clear, but here’s the key thing to understand about processing and types:

The bottom line — processors run before types are applied

What you see in the interface may reflect the schema type, but processors still run on the original source value until typing is applied at the end.

1. Source data is ingested (raw values)

2. Processors run on those raw values

3. Schema types are applied afterward

In more detail

Huwise runs processors on source values before schema typing is applied. This means the value displayed in the interface may already look typed, while processors are still using the original underlying source value.

So if your source contains values like 0, 1, 0.0, 1.0, "No", "False", etc., processors operate on those underlying values, even if the field is typed as Boolean in the Schema.

As a result, a value may appear as false in the interface while still being processed as 0.0 internally.

An example

If you use Transform boolean columns to multivalued field and your source uses the values 0.0  and  1.0:

• In the Processing view, you might see false

• But the processor is still operating on the underlying source value 0.0

• Result: unless you explicitly list 0.0 under Values considered as “false”, the processor will not treat that value as false

What to do when it’s an issue

• Check the original source values, not just the displayed values in Huwise

• Configure the processor with those exact source values, like in the example of the Transform boolean columns to multivalued field processor

• If needed, add a processor to rewrite the values into the format expected by downstream processors