Hi,

 

 

the easiest way is to let FME construct the basic workbench for you, then adapt it as necessary.

 

 

Open Workbench, go to File / New... / Generate workspace using wizard.

 

 

Specify the input and output formats and necessary connection parameters (for Oracle) and let FME do the rest.

 

 

As easy as it gets :-)

 

 

David

---

The three datasets you mention are conceptually very different so you really need three solutions:

 

A geodatabase (be it a personal geodatabase; *.mdb file, or a file geodatabase, a directory, *.gdb) is the same as an Oracle geodatabase (a set of data including spatial and attribute rules and relationships). So the wizard can handle this. Just keep an eye on the attribute type because attribute type casts can go wrong.

 

 

A shapefile is one dataset. The wizard can handle this as well,  you just need one wizard per shapefile. Once again, keep an eye on the casts and, in this case, the attribute names. A shapefile is a very old format, names of the attributes cannot exceed 11 (?) characters. A bit of renaming can yield more "realistic" names.

 

 

A layerfile is not really a dataset but a set of presentation rules (symbology, scale levels etc) that contains a pointer to one or more datasets (shapefiles or geodatabases). The symbols and other rules cannot be handled by FME and since layer files points to datasets you might have already converted the data they are pointing to.

 

 

The best solution for layer files is to not try and convert them but change the datasource (can only be done using ESRI software, eather manually or using Python, see ESRI docs for this) and have the layerfile point to the newly converted data instead.

 

 

Namaste,

 

George Schenk

 

---