Implementation details

fq uses a gojq fork that can be found at https://github.com/wader/gojq/tree/fq (the "fq" branch)
cli readline uses raw mode so blocks ctrl-c to become a SIGINT
TODO: scope and scopedump functions used to implement REPL completion
TODO: Custom object interface used to traverse fq's field tree and to allowing a terse syntax for comparing and working with fields, accessing child fields and special properties like _range.

Decoder implementation help

Main goal in the end is to produce a tree that is user-friendly and easy to work with. So there are always excepts to these rules and sometimes it might be better to let the decoder code be a bit ugly over producing a tree that is hard to understand.
Try use same names, symbols, constant number base etc as in specification
TODO: Decode only what you know. If possible let "parent" decide what to do with unknown bits by using *Decode*Len/Range/Limit funcitions
Use sub decoders if possible for frames, metadata etc. You can pass data between them. Also makes it possible to call them separately
Try to no decode to much as one field or value. A length encoded int could be two fields, a flags byte can be struct with bit fields.
Try to have add symbols for all named constants.

Send log package output and stderr to a file that can be tail -f:ed:

LOGFILE=/tmp/log go run main.go ... 2>/tmp/log

gojq execution debug:

GOJQ_DEBUG=1 go run -tags debug main.go ...