Harvester parses non-`.mpr` file values as strings

[mjaquiery]

Describe the bug
When parsing values from non-.mpr files (e.g. .txt or .idf), all values appear as strings rather than as their actual data type (e.g. float, int).

To Reproduce
Steps to reproduce the behavior:

1. Harvest a .txt or .idf file.
2. Observe the data_type of all columns is str

Expected behavior
data_type should be identified correctly.

Additional context
The issue is inside the parser itself. harvest.py's calls to enumerate(generator) return dictionaries where all the values are strings, rather than correctly being identified as the data type they should be.