Files
manipulator/README.rst
2017-01-17 15:24:30 +01:00

73 lines
2.2 KiB
ReStructuredText

manipulator
=============
Python data manipulation made braindead.
Installation
------------
::
pip install manipulator
Usage
-----
``manipulator`` mainly exposes three functions, ``get``, ``update``, and ``set``.
``get`` retrieves data, ``update`` transforms it based on its form, and ``set``
transforms it by simply resetting. Transformations can be applied in place or
on a copy. The default is in place, because copying is expensive. If you want a
copy of the data, set the keyword argument ``in_place`` to ``True``.
It uses a query "language" not unlike CSS, albeit much, much simpler. The only
two entities in this "language" are IDs—denoted by ``#``—and Classes—denoted by
``.``. IDs are unique, whereas Classes are collections of all leaf values that
conform.
A few motivating examples (a more exhaustive list can be found in the ``test`` directory):
.. code-block:: python
import manipulator
manipulator.get({"k": [1, [2], 3]}, "#k #1 #0")
# => 2 (note how list indices are coerced into integers)
manipulator.get([{"k": "v"}, {"k": "v2", "k2": "v3"}], ".k")
# => ["v", "v2"]
manipulator.get([{"k": "v"},
{"k": {
"a": [{"k": 10}, {"k": 11}]
}
}], ".k #1 #a .k")
# => [10, 11]
manipulator.set({"k": [1, [2], 3]}, "#k #1 #0", 3)
# => {"k": [1, [3], 3]} (in place)
manipulator.set({"k": [1, [2], 3]}, "#k #1 #0", 3, in_place=False)
# => {"k": [1, [3], 3]} (will create a copy of the data)
manipulator.set([{"k": "v"},
{"k": {
"a": [{"k": 10}, {"k": 11}]
}
}], ".k #1 #a .k", 100)
# => [{"k": "v"}, {"k": {"a": [{"k": 100}, {"k": 100}]}}]
manipulator.update({"k": [1, [2], 3]}, "#k #1 #0", lambda x: x-1)
# => {"k": [1, [1], 3]} (in place, use in_place=False to copy)
manipulator.set([{"k": "v"},
{"k": {
"a": [{"k": 10}, {"k": 11}]
}
}], ".k #1 #a .k", lambda x: x+1)
# => [{"k": "v"}, {"k": {"a": [{"k": 11}, {"k": 12}]}}]
That is all.
Have fun!