If you intend to index the file instead of having Splunk monitoring/tailing it, then I think the CLI "oneshot" command is indeed the best solution.
# splunk add oneshot <path_to_your_file>
You could also create a new monitor input for that file from the CLI, this will index your file but it won't be at all the same than "uploading" it to Splunk as a one-time thing, as Splunk will be keeping track of it.
# splunk add monitor <path_to_your_file>
Another possibility is to toss the file in $SPLUNK_HOME/var/spool/splunk which is set up by default as a batch input (see $SPLUNK_HOME/etc/system/default/inputs.conf). Note that the file will be indexed destructively, so you may want to copy it there, not move it.
You can also use the built-in sinkhole "$SPLUNK_HOME/var/spool/splunk" to index a file in an ad-hoc fashion. Place any applicable file in there and it will be indexed by Splunk. Historically you could not control how those files were indexed, but now you can.
Just add an extra line to the top of your file that looks something like this:
***SPLUNK*** sourcetype=YourSourcetype index=YourIndex host=foo1 ...
And Splunk will index your file with those parameters.
Read more about it in the Splunk docs: http://www.splunk.com/base/Documentation/4.1.5/Admin/Assignmetadatatoeventsdynamically
I use the built-in sinkhole when I am doing a quick test of a new log-type to evaluate it before setting up a perm monitor or when doing an investigation and I need to get a bunch of evidence into Splunk from a system that wasn't already monitored by Splunk.