Slurp functionality about jq.py HOT 17 CLOSED

No, there's no support for slurp. Mind if I ask what your use case is?

from jq.py.

I am trying to combine multiple JSON arrays (in strings) into one, and
while reading on the jq documentation, which I am already using for its
more common uses, I noticed the "slurp" option.

On Fri, Jan 15, 2016 at 6:35 PM Michael Williamson [email protected]
wrote:

No, there's no support for slurp. Mind if I ask what your use case is?

—
Reply to this email directly or view it on GitHub
#17 (comment).

from jq.py.

By looking further into the manual, I found that .[0] + .[1] does exactly what I need.

from jq.py.

Can this get re-opened? My use case is to replace Subprocess calls to jq by using this extension. This requires that the extension offer the same feature set as the command line binary. Just from a short look into the source code of jq it should being possible to use the slurp mode with the library.

from jq.py.

What's your use case? Just trying to understand why explicitly passing in an array wouldn't work.

from jq.py.

Let's say you have this input:

{"name":"Bob Dylan", "tags":"{Artist}{Singer}"}{"name": "Michael Jackson"}

which is not valid json but a stream of valid json objects. The following jq command:

jq -s . input.txt

gives you a valid json array:

[
  {
    "name": "Bob Dylan",
    "tags": "{Artist}{Singer}"
  },
  {
    "name": "Michael Jackson"
  }
]

from jq.py.

Ah, I see. Seems reasonable. I don't really have time to add new functionality, but feel free to make a pull request.

from jq.py.

@mwilliamson Ok. Let me dig into it.. I'll give my best! :)

from jq.py.

This relates to --slurpfile which allows reading in and processing multiple files. The can be referred to using variables in the jq filter:

jq -n --slurpfile old $1 --slurpfile new $2 '{old: $old[0], new: $new[0]}'

If someone was working on slurp functionality, it may be worth considering this as well.

My use case is computing a JSON that represents something like the diff between old and new.

from jq.py.

+1 - I was using this library through urlwatch and wanted to be able to slurp as I'm handling a list of objects (not wrapped in an array) from an HTTP endpoint

from jq.py.

I've added slurp functionality and released a prerelease version to PyPI. Feedback would be welcome!

from jq.py.

@mwilliamson any plans on releasing the slurp-enabled version as a full release?

from jq.py.

As above, some feedback would be useful to know if the current implementation actually solves the problem or not before making a release.

from jq.py.

Well, don't know about the case above but it certainly works for me and works as expected 🙂

from jq.py.

This has now been released as part of 1.5.0.

from jq.py.

Thanks, @mwilliamson! 🎉

From the code, I gather that referring to different inputs using variables in the jq filter, like in jq, hasn't been implemented:

jq -n --slurpfile old $1 --slurpfile new $2 '{old: $old[0], new: $new[0]}'

Is that right? -- I can open a separate feature request issue for that, just would like to make sure.

(My use case is computing a JSON that represents something like the diff between two inputs, which currently requires invocation of jq as a separate process.)

from jq.py.

Correct, the change is only to add the slurp argument to input_text().

from jq.py.