- 600k files in our prod system - some tiny, some gigantic
- names are related to their DB id's
- want to snapshot them all for our QA environment but that's not practical
- fake the files!
- file scanner collects: name, content-type, size into a TSV file
- very small files are embedded (base64 encoded)
- file creator reads the TSV and makes files that smell like the originals
- images all look like:
- similar for videos
- it works and is pretty fast!
it's written in Go