Graph.Weighted_Adjacency (and friends) don’t support `scipy.sparse.sparray`s

[flying-sheep]

Describe the bug
See title. We’re about to change everything to that API, since it’s very mature at this point. And generally also very well-supported, which is why this surprised me!

To reproduce

import igraph
import scipy.sparse

igraph.Graph.Weighted_Adjacency(scipy.sparse.random_array((50, 50)))

TypeError                                 Traceback (most recent call last)
Cell In[3], line 4
      1 import igraph
      2 import scipy.sparse
----> 4 igraph.Graph.Weighted_Adjacency(scipy.sparse.random_array((50, 50)))

File /usr/lib/python3.14/site-packages/igraph/io/adjacency.py:148, in _construct_graph_from_weighted_adjacency(cls, matrix, mode, attr, loops)
    145 if (np is not None) and isinstance(matrix, np.ndarray):
    146     matrix = matrix.tolist()
--> 148 graph, weights = super(Graph, cls)._Weighted_Adjacency(
    149     matrix,
    150     mode=mode,
    151     loops=loops,
    152 )
    153 graph.es[attr] = weights
    155 # Add vertex names if present

File /usr/lib/python3.14/site-packages/scipy/sparse/_base.py:448, in _spbase.__len__(self)
    447 def __len__(self):
--> 448     raise TypeError("sparse array length is ambiguous; use getnnz()"
    449                     " or shape[0]")

TypeError: sparse array length is ambiguous; use getnnz() or shape[0]

Version information
1.0.0