[FEATURE] implement `__repr__` and `__str__`

[kornerc]

Both str() and repr() of a (Python) PowerGridModel return the default representation of an object in Python:

pgm: PowerGridModel = ...

str(pgm)
# '<power_grid_model._core.power_grid_model.PowerGridModel object at 0x00000209EAB2DA00>'

repr(pgm)
# '<power_grid_model._core.power_grid_model.PowerGridModel object at 0x00000209EAB2DA00>'

It would be nice to have a more meaningful string representation.

This behaviour can be changed by implementing the methods __str__ and __repr__ (if __repr__ is implemented but __str__ not, the implementation of __repr__ will be used for str() and repr()).

Usually, __repr__ can be used to directly reconstruct the object and __str__ is used for pretty printing:

from pathlib import Path

path = Path("./a/relative/path")

str(path)
# 'a\\relative\\path'

repr(path)
# "WindowsPath('a/relative/path')"

from pathlib import WindowsPath
assert path = eval(repr(path))

I guess implementing __repr__ so that an object of PowerGridModel can be directly reconstructed is not the right choice.
For example, a pandas DataFrame can't be reconstructed and the same implementations are used for __repr__ and __str__:

import pandas as pd

repr(pd.DataFrame({"a": [1, 2, 3], "b": [4, 5, 6]}))
# '   a  b\n0  1  4\n1  2  5\n2  3  6'

pandapower also uses the same implementation for both and returns the number of elements per element type

import pandapower as pp

net: pp.pandapowerNet = ...

str(net)
# This pandapower network includes the following parameter tables:\n   - bus (146 elements)\n   - load (139 elements)\n   - switch (11 elements)\n   - ext_grid (1 element)\n   - line (147 elements)\n   - trafo (2 elements)\n   - controller (2 elements)

print(net)
# This pandapower network includes the following parameter tables:
#    - bus (146 elements)
#    - load (139 elements)
#    - switch (11 elements)
#    - ext_grid (1 element)
#    - line (147 elements)
#    - trafo (2 elements)
#    - controller (2 elements)

Should we have a similar implementation like pandapower by leveraging internally PowerGridModel.all_component_count.

It could look similar to:

pgm.all_component_count
# {<ComponentType.node: 'node'>: 146, <ComponentType.line: 'line'>: 147, <ComponentType.source: 'source'>: 1, <ComponentType.sym_load: 'sym_load'>: 417, <ComponentType.transformer: 'transformer'>: 2, <ComponentType.link: 'link'>: 11}

print(pgm)
# PowerGridModel
#   - nodes: 146
#   - lines: 147
#   - sources: 1
#   - sym_loads: 417
#   - transformers: 2
#   - links: 11

[TonyXiang8787]

Hi @kornerc,

It would be a good idea to implement this as the count of all the component types. Feel free to propose such an implementation.