I am taking a feature reader (ESRI geodatabase) and putting a feature class through an attribute manager.
The feature reader shows the guid fields correctly as guid.
However, the attribute manager shows them guid fields as buffer.
This is happening for all guid field. I've tried "reset table" and import from dataset in the attribute manager. I've also tried recreating the attribute manager. How can I fix the attribute types to be accuate?
I also tried removing the attribute manager. The next transformer (change detector) shows the guid fields as varchar(200). So strange.
@rtreecare Internally, FME doesn't have a GUID data type. It's mapped to fme_buffer. The data (GUID value) will be unchanged.
I also tried removing the attribute manager. The next transformer (change detector) shows the guid fields as varchar(200). So strange.
Ultimately, FME doesn't care what type the field is. Generally*, the only time it matters is making sure the correct data type is set in the writer
*there are some edge cases!
Ultimately, FME doesn't care what type the field is. Generally*, the only time it matters is making sure the correct data type is set in the writer
*there are some edge cases!
Thanks. The writer is the ultimate issue and I'm trying to automate the attribute types from the reader. Everything I would try to import the attributes from the reader, guid would be of type string and the feature service writer would fail to update/insert. Seems like manual work is still the best option.
@rtreecare If you're using the FeatureReader, then you can get the schema feature. The schema feature will also have the source types. This will give you the guid type:
attribute{60}.fme_data_type fme_buffer
attribute{60}.name SourceGUID
attribute{60}.native_data_type guid
On the dynamic writer, you can choose to use the native types by setting the attribute fme_format_short_name as described here