I want to find all attribute names that contain the source table name in a tabular (CSV)translation. Because I end up with several million features per input table I don't want to use substring on the fme_basename attribute - I want to extract the source table name from the Source parameter once per source table.
There might be a substring operation in Advanced Text Editor, who knows?
Thanks,
Ian
Why not aggregate or use the sampler per input ?
Cheers,
Itay
If you worry about performance, I would take try to pre-process and analyze your data first.
If your schema doesn't change too often, you could run all your features once through an AttributeExploder, then through a DuplicateRemover on "_attr_name" to get only the unique values, then finally through a Sorter before outputting it to something like a Logger or an Excel writer.
This should give you a nice list of all the possible attribute names to work with. If the list is manageable, you could read the relevant parts of it into an AttributeFilter, thereby avoiding all the string search operations each time you run the workspace.
David
Any functionality in the advanced editor should be exposed in the menu on the left hand side, as parameters, attributes, functions, etc. So if you don't see something there then it isn't supported.
2013-SP1 does have some additional string functionality added to it. Padding, formatting, and string length are the ones I see right now, though I believe more will be added.
In this case I would use the advanced editor, but I would also use a Sampler first. If all the features have the same schema, then you only need the first one in order to check its attribute names.
Hope this helps
Mark