Many web application require a seed file -- a file, written in the same language as a web application, where I have data to pre-populate a database. This makes sense.
Nonetheless, I sometimes see that some parts of that data is extracted into separate, text files to be read from the main seed.{rb, py, php} file and inserted into a database.
It can be a list of countries with languages and locales. Or a list of products for an internet-shop with descriptions, prices, etc. Or a list of articles for a blog. Whatever.
I wonder, why are kept in separate, text files which then read from the main seed.{rb, py, php} file and inserted into a database instead of keeping that in a seed file itself? What are the benefits of this approach? And when is this approach necessary?