Thanks for the report and patch, Jon! I'm far from an expert on the fst library, and the comment in the code pointing people back to this JIRA doesn't give a lot of details as to why the code in question is necessary. How did you discover class 83 was LinkedHashMap? It's also a bit weird given LinkedHashMap is a java built-in that we need to register it but not other types. Could you elaborate a bit more, preferably in the code comment?
Otherwise patch looks good. I'm assuming it's a bit difficult to write a test for this short of providing some sample data generated on fst-2.24 that the code tries to load.