Naming format of update source data files

When Forge processes update source data files, it is important to keep two issues in mind concerning the names of the data files.

The update files should be processed by Forge in order of their creation. The reason is that if a specific record appears in more than one update file, you want the latest update to be processed last, so that it will override earlier versions when the Dgraph loads the update record files.
Forge reads the files in strict lexicographic order of their filenames. Therefore, you should use a naming scheme that ensures the processing of the update files in chronological order of their creation (i.e., last created, last processed).

For these reasons, it is strongly recommended that you use a timestamp format as the naming scheme for the filenames. If necessary, use leading zeros to force the desired numeric order. For example, if you have two files named 9.xml and 10.xml, Forge will process 10.xml before 9.xml; therefore, you must rename 9.xml to 09.xml so that it is processed before 10.xml.