I would like to voice my concern about the amount of dependencies of org.apache.tika:tika-parsers:jar:1.22.
I recently needed to detect a charset, and I found CharsetDetector from the tika-parsers jar. It seemed like a small and and easy solution to zoom in on an encoding problem.
After I started to have problems starting my application (Cannot run program "C:\opt\jdk-13\bin\java.exe" ... CreateProcess error=206), I discovered the dependencies below were added by org.apache.tika:tika-parsers:jar:1.22 (extract from mvn dependency:tree).
I do note that the classpath for my application is already quite busy, but what tika-parsers added seems a bit over the top.