Currently it is possible to decode Array values from columns, by specifying Array Element separator (default being ";"). But there is quite a bit of data that uses outer "wrappers", f.ex commonly we might have:
id,title,"[1.0, 0.4, 16.98]"
wherein brackets are used around the sequence of values.
We should allow supporting such notation by addition to CsvSchema. Could start by allowing use of start/end String; no need to get fancy at first (but could conceivably support regexps if need be, i.e. take java.util.regex.Pattern start/end markers).
NOTE: this issue is specifically for READING such decorated values: different issue (#495) filed for WRITING.