I am using a featurewriter to update 3 three fields. For this update, I am only exposing the fields that need to be updated, because I have found that works best.
However, I need the rest of the fields in the dataset downstream.
I ran an attributeexposer to reveal the other fields, but they all come in with null values.
These fields do exist in the dataset, and they do have values. How do I access the fields in such a way that I can see the values in them?
Thank you,
Randy McGregor
Best answer by nielsgerrits
The workspace does not load the fields before the featurewriter, but the fields are in the dataset. I may just not be understanding how featurewriter works (a strong possibility). I was hoping I could update the three fields first - having ONLY those three exposed, then expose the rest of the fields later. Apparently, it doesn't work that way?
The FeatureWriter writes, it does not read. So if you had a (Feature)Reader upstream of the FeatureWriter, reading these records, you should have these attributes. If you did not, you won't have them as the workspace never loaded the data.
But there seems to be a bug where attributes not assigned in the FeatureWriter, do not continue through the additional outputports. Which is solved in 2023.1.
+54I'm having a hard time visualizing the workflow you have, can you share a printscreen of the workspace from the reader to where you need the attributes?
The writer updates three fields, the three fields plus the match field are crossed out with black. There are 7 additional fields I need for another transformation - Red. I am trying to reveal them and their values. I picked the fields from a list. They are all from the database, but they come in blank. Thank you!
+54The writer updates three fields, the three fields plus the match field are crossed out with black. There are 7 additional fields I need for another transformation - Red. I am trying to reveal them and their values. I picked the fields from a list. They are all from the database, but they come in blank. Thank you!
Were the fields already loaded before the FeatureWriter? If you put the AttributeExposer before the FeatureWriter, are they having values then? If not, the fields were never loaded. If they are, you might be tripped over a bug mentioned in this topic?
The workspace does not load the fields before the featurewriter, but the fields are in the dataset. I may just not be understanding how featurewriter works (a strong possibility). I was hoping I could update the three fields first - having ONLY those three exposed, then expose the rest of the fields later. Apparently, it doesn't work that way?
+54The workspace does not load the fields before the featurewriter, but the fields are in the dataset. I may just not be understanding how featurewriter works (a strong possibility). I was hoping I could update the three fields first - having ONLY those three exposed, then expose the rest of the fields later. Apparently, it doesn't work that way?
The FeatureWriter writes, it does not read. So if you had a (Feature)Reader upstream of the FeatureWriter, reading these records, you should have these attributes. If you did not, you won't have them as the workspace never loaded the data.
But there seems to be a bug where attributes not assigned in the FeatureWriter, do not continue through the additional outputports. Which is solved in 2023.1.
Thank you very much. That makes sense. The structure of FME sometimes confuses me. I appreciate your time.
+54Thank you very much. That makes sense. The structure of FME sometimes confuses me. I appreciate your time.
Happy if I can help. It is originally an ETL tool, where the readers do the extract and the writers the load part. It can do a lot more nowadays.
Enter your E-mail address. We'll send you an e-mail with instructions to reset your password.
