Skip to main content
  • EN
  • Home
  • Forums
  • FME Form
  • Data
  • Write CSV file where filename includes date/timestamp of translation - currently getting multiple CSVs

I'm removing duplicates from a File Geodatabase (FGDB) feature class using Sorter then DuplicateFilter, then writing the Unique features back to the same FGDB feature class.

I want to write the Duplicate features to a CSV log file with the date/time of the translation in the file name so I have a CSV listing the duplicates removed each time the Workspace is run.

First I tried including @Timestamp(^Y^m^d^H^M^S) in the CSV File Name in the Writer but I ended up with multiple CSVs, presumably because features were being written as they were received by the Writer. I haven't noticed this problem with FGDBs but is that because of difference in the way the FGDB Writer works?

I then tried a Creator followed by a TimeStamper running parallel to the actual data processing with the output from the TimeStamper going to the CSV Writer along with the data. But what I get is two CSV files - one with no date/time stamp in the name and all the data, and another with the date/time stamp in the name but no data.

Would "Append to file" in the CSV Writer properties solve this? If so, how can I ensure that the CSV file with the date/time stamp in the name is created first?

I know I could record the date/time against each feature and append to a single log file but I'd prefer the clarity of separate files.

 

You need to merge the timestamper with the rest of your data, e.g. with a FeatureMerger

I agree with @egomm, but I would highly recommend setting "Suppliers first" to yes to avoid it blocking the data flow and consuming huge amounts of memory if you have a lot of data.

Another way to do it is to use a private Python scripted parameter to give you the timestamp at the start of the translation, e.g.

from datetime import datetime
return datetime.now().strftime('%Y%m%d%H%M%S')

Useful if you need the same timestamp many places in your workspace and you want to avoid all those FeatureMergers.

You could also use a variable setter and a variable retriever

Yes, there is more than one way. Another approach: once create the destination CSV file with temporary name (e.g. "temp.csv") by a FeatureWriter, and then rename it to the timestamp. i.e. move the file finally.

You need to merge the timestamper with the rest of your data, e.g. with a FeatureMerger

I am having the same issue and tried this solution but the I'm stumped on the join. I can choose _timestamp in supplier but what do I choose to join on Requester? If I choose a unique identifier nothing outputs from the Merged.

I am having the same issue and tried this solution but the I'm stumped on the join. I can choose _timestamp in supplier but what do I choose to join on Requester? If I choose a unique identifier nothing outputs from the Merged.

You need an unconditional join (1 = 1) to get the timestamp on all features.

Reply


Community Stats

30,108
Posts
114,295
Replies
37,745
Members
Terms & ConditionsCookie settings
A product of
Safe Software respectfully acknowledges that we live, learn and work on the traditional and unceded territories of the Kwantlen, Katzie, and Semiahmoo First Nations.