jq: sed for json

I like those tools - working with json files often calls for some simple processing and it's a bit silly to write an app for each case. But I wish they would start embracing some common standard. So far we've got:

* jsonselect

* jsonpath

* jsonquery

* jaql

* unql (query part)

which all need the same query dsl, but use a slightly different syntax. I'd really like to see them unified one day, like xpath is the default for xml.

Previous submissions:

http://news.ycombinator.com/item?id=4562609

http://news.ycombinator.com/item?id=4679933

I would like something like this but for querying Python objects (dict containing lists of dicts with particular key-values that are needed, etc.) sort of like a declarative way of cleanly creating a new data structure from a response, or even yielding elements that match the definition I give. Does anything like this exist already? I am constantly writing different code to do this.

awk for json would be really nice (where patterns and records match substructures ...)

this reminds me a bit of powershell's XML/object filtering except less painful.

great work! btw, is there a versioned json toolkit?