dict-picker

Dict picker (dictionary picker)

Retrieve data from Python dictionaries and lists.

Installation

With pip:

pip install -U dict-picker

Or with conda:

conda install "dict-picker" -c conda-forge

Usage

pick_by_path

The pick_by_path function is used to extract values from nested dictionaries using a path-like string to identify the desired value. The function takes three arguments: a dictionary object, a string representing the path to the desired value and optionally a delimiter.

The path string consists of a sequence of keys, separated by the delimiter character ('/' by default), where each key represents a level in the nested dictionary. If the key has the value '*', it matches any key on that level, or sets it to check all items in the list.

The function returns the value at the end of the path or a list of values if the path contains a wildcard. If the path is not found, the function returns None.

from dict_picker import pick_by_path

example_dict = {
    "foo": "bar",
    "baz": {
        "qux": "quux",
        "quuux": "corge"
    },
    "fred": {
        "wilma": "betty",
        "barney": "pebbles"
    },
    "arr": [
        { 'id': 123 },
        { 'id': 456 },
        { 'id': 789 },
        { 'name': 'bubbles' },
    ],
}

assert pick_by_path(example_dict, "foo") == "bar"
assert pick_by_path(example_dict, "baz/qux") == "quux"
assert pick_by_path(example_dict, "fred/wilma") == "betty"
assert pick_by_path(example_dict, "*/qux") == "quux"
assert pick_by_path(example_dict, "fred/*") == {"wilma": "betty", "barney": "pebbles"}
assert pick_by_path(example_dict, "*/quux") is None
assert pick_by_path(example_dict, "arr/*/id") == [123, 456, 789]

# slice syntax
assert pick_by_path(example_dict, "arr/0") == {'id': 123}
assert pick_by_path(example_dict, "arr/1/id") == 456
assert pick_by_path(example_dict, "arr/-1") == {'name': 'bubbles'}
assert pick_by_path(example_dict, "arr/1:/id") == [456, 789, None]
assert pick_by_path(example_dict, "arr/::-2/id") == [789, 123]

Parameters:

• obj -- A dictionary object to search for the desired value.
• path: str -- A string representing the path to the desired value.
• delimiter: str -- A string used to separate keys in the path string. Default is "/".

Return value:

• Returns the value found at the end of the path string, or a list of values if the path contains a wildcard. If the path is not found, None is returned.

pick_by_paths

The pick_by_paths function is similar to pick_by_path, but can extract values from multiple paths at once. It takes a dictionary object and a list of path strings as arguments, and returns a list of values found at the end of each path string or None if a path is not found. This works faster than running pick_by_path in a loop.

from dict_picker import pick_by_path, pick_by_paths

example_dict = {
    "foo": "bar",
    "baz": {
        "qux": "quux",
        "quuux": "corge"
    },
    "fred": {
        "wilma": "betty",
        "barney": "pebbles"
    },
    "arr": [
        { 'id': 123 },
        { 'id': 456 },
        { 'id': 789 },
        { 'name': 'bubbles' },
    ],
}

assert pick_by_paths(example_dict, ["fred/*","*/quux", "arr/*/id",]) == [{"wilma": "betty", "barney": "pebbles"}, None, [123, 456, 789]]

Build from sources

Cargo

curl --proto '=https' --tlsv1.2 -sSf <https://sh.rustup.rs> | sh

For other installation options, see the official website.

Maturin

To bind python and rust, pyo3 is used. The Maturin library is used to make it easy to work with.

python -m pip install maturin

Build

python -m maturin build --release

Wheel will be released under the system and python in which it will be built. Read more about the compilation.

The finished wheel can be found in the target/wheels folder.

Build with docker

docker run --rm -v $(pwd):/io ghcr.io/pyo3/maturin build --release  # or other maturin arguments

On macOS, it is better to use docker instead of podman.

Roadmap

• Arbitrary levels skip operator **;

• Search inside an array of dictionaries. For example:

  {
      arr:  [
          { id: 123 },
          { id: 456 },
          { id: 789 },
      ]
  }
  'arr/*/id' -> [123, 456, 789]
  

• Search by tuple of patterns. Returns the result as a list of found values.

• add option to access lists via python slice syntax (issue #2):

  {
      arr:  [
          { id: 123 },
          { id: 456 },
          { id: 789 },
      ]
  }
  'arr/1/id' -> 456
  'arr/1' -> { id: 123 }
  'arr/:2/id' -> [123, 456]
  'arr/1:' -> [{ id: 456 },{ id: 789 }]
  'arr/::2/id' -> [123, 789]