pq-json

pq is a Python command-line JSON processor


Keywords
json, json-parser, python, python-3
License
MIT
Install
pip install pq-json==0.0.3.1

Documentation

pq-json

pq is a Python command-line JSON processor - almost like (atleast inspired by 🐶️) jq, but using Python focusing on simplicity and convenience with familiar syntax.

Install

pip install pq-json

Here is a simple example for parsing JSON. Output from pq is pretty printed using Rich.

$ echo '{"text": "Text here", "header": "Header", "list": [1,2,3]}' | pq
{
  "text": "Text here",
  "header": "Header",
  "list": [
    1,
    2,
    3
  ]
}

Pretty Parse

Filters

The processing is handled with filters, like in jq. j represents the current input object in a filter.

$ echo '{"example": "data", "nothing": "interesting"}' | pq "j['example']"
"data"
$ echo '{"example": "data", "nothing": "interesting"}' | pq "j['example']"
"data"

As default, None will not be passed.

$ echo '{"example": "data", "nothing": "interesting"}' | pq "j.get('nada')"

List slicing

JSON arrays are converted to Python list.

$ echo '[{"name": "eric", "age": 22}, {"name": "daniel", "age": 44}]' | pq "j[0]"
{
  "name": "eric",
  "age": 22
}
$ echo '[{"name": "eric", "age": 22}, {"name": "daniel", "age": 44}]' | pq "j[-1]"
{
  "name": "daniel",
  "age": 44
}

An array always iterates to the next filter. Here we are using the slice symbol [:] to highlight that we are working with an array.

$ echo '[{"name": "eric", "age": 22}, {"name": "daniel", "age": 44}]' | pq "j[:]"
{
  "name": "eric",
  "age": 22
}
{
  "name": "daniel",
  "age": 44
}

Want to send the full array to the next filter? Wrap array around [] brackets like:

$ echo '[1,2,3,4,5]' | pq "[j[:]]"
[
  1,
  2,
  3,
  4,
  5
]

Pipes

Pipes let you chain multiple filters by produce output to the filter to the right of the pipe. Under the hood a pipeline is a chain of generators. An array will for example yield multiple elements to the right.

input:

["a", "b", "c", "d"]
pq "j[0:2] | j.upper()"
      |         |
   filter1   filter2

produces the following result:
"A"
"B"

In this case, the expression j.upper() will be run each time.

j[0] -> "a".upper() -> "A"
j[1] -> "b".upper() -> "B"

Another example:

$ echo '[1,2,3,4,5,6,7,8,9]' | pq 'j[:] | j**2+50'
51
54
59
66
75
86
99
114
131

Array constructs

Above example outputs a list of integers. It's possible to wrap it all into a single array by using [] around the full expression.

$ echo '[1,2,3,4,5,6,7,8,9]' | pq '[j[:] | j**2+50]'
[51, 54, 59, 66, 75, 86, 99, 114, 131]

Object constructs

$ echo '{"name":"jan", "age":4, "parents": ["lisa", "dan"]}' | pq '{"name": j["name"], "parents": j["parents"]}'
{
  "name": "jan",
  "parents": [
    "lisa",
    "dan"
  ]
}

Custom modules

It's possible to add additional modules to global scope from file or input string. We can declare a variable for example.

echo '[1,2,3]' | pq --module 'two=2' 'j[1]*two'
4

Other examples

We can easily use built-in functions in Python

$ echo "[1,2,3,4,5,6]" | pq 'sum(j) | {"total": j}'
{
  "total": 21
}