Character encoding issues with places-of-worship.csv

[c6andy]

Steps to reproduce

For example:

1. Go to: https://campaign-lab-constituency-explorer.c6digital.dev/constituency/E14001334
2. Click “Places of Worship”
3. The first listed place of worship in the list is called “���������� ������������������”

Screenshots

Why does this happen

I think this happens because this file was originally encoded as ISO-8859-1, but was possibly saved as UTF-8 without re-encoding? Something like that.

Proposed solution

The importer could probably sort this out, but I don’t think that would be a future-proof solution, because I think we’d end up overfitting to the current data. We’d be better off fixing this in the source data, and then whenever new source data arrives, ensuring it’s correctly encoded as UTF-8 prior to importing.

Anything else

This reminds me that we ought to note in the README that all source data is currently expected to be encoded as UTF-8, except schools-england.csv, which is expected to be ISO-8859-1. I’ve done that in #57.