Read files with different encodings

[Jolanrensen]

As was brought to our attention in Kotlin/dataframe#1555, it appears that Deephaven CSV reads files as whatever the default encoding is in the JVM.

This can produce some unexpected results if you are reading a file that has a different encoding. For instance, let's say your JVM is running in US-ASCII (this can be reproduced by adding JAVA_TOOL_OPTIONS=-Dfile.encoding=US-ASCII to the environment variables) and you're trying to parse a UTF-8 csv like this: scratch.csv.

whereas if we set the default encoding of the JVM to UTF-8:

I wonder if it's possible if you could add a setting to CsvSpecs that can make it possible to read from any given encoding without having to change the environment variable.

[devinrsmith]

Thanks for the report. We'll triage this and report back.