I was using a CSV reader that would do it based on header names and that worked fine until I thought about the fact that it might say EUR because of my own portfolio.
I then switched to an index-based reader, but that was even more problematic because depending on your date range it might not include all the fields. As discussed here: Common format for history export file (CSV)
So I feel like I need to make something custom which is a mix of both, or do pre-processing where I remove all the currencies within parenthesis.
After the scouring the internet for people who are doing the same, I found that a lot of people are not actually dealing with this problem by just ignoring it. Also discovered that some people have Finra Fee as a field.
After unsuccessfully trying to use index-based, I went back to my header-based library and found out they have a normalizer function for headers. Here I use a regex to replace any currency mentions from both my model and the CSV files; \([A-Z]{3}\)