Are saying your data is at the daily level but you only need it at the monthly level? If so, import the data into a module dimensionalized by time and have Anaplan do the aggregation for you by using the Time mapping of DD-MM-YYYY.
If you are already importing the data at the daily level, then you can use a SYS Time module to get the parents of the day (week, month, quarter, half year, year).
If it is just a line item within your data set, do the above and do a lookup to the SYS Time - Daily module.
The text string "sep" has no meaning other than being a three character text string. If you are unable to define the text as a date within the import then you need to create a mapping that will convert the period text strings into a meaningful variable.
Create a list containing the periods as per the text in your import.
Create a mapping table dimensioned by this new list and assign Jan = 1, Feb = 2 etc
You will now be able to use YEAR, MONTH, DAY to build you date format and PERIOD to convert to into a period format using the mapping table to create the number feeding the MONTH function.
Putting all these together you should be able to convert your text into a viable period format.
You should just be able to import the data correctly using a custom time mapping,
Here is my data file, using the same naming convention you highlighted.
Then once I have initiated the import, I click Custom Mapping, to map the import to a date formatted line item.
Then a Mapping box pops up which allows you to pick the format of the data that is coming in. From there, select the top option D-M-Y (Screenshot shows the incorrect one) and on the right side, you can see a preview to show that it is working properly.
This will be cleaner, smaller, and more efficient since you don't have to use the Text line item (Date is 4 bytes, Text is 8), you can eliminate all staging lines, and can also cut out any computation required to calculate the date.
I must admit that while the solution I provided is an option I would try and ensure that your convert the text to date during the import process and use a custom date mapping as suggested by @jasonblinn .
This is the best solution for all the reasons pointed out in the very comprehensive and detailed reply.