new: preserve_adb_keys in PyG to ArangoDB

[aMahanna]

Warrants a 1.1.0 release

Note: For notebook testing purposes:

• Use !pip install git+https://github.com/arangoml/pyg-adapter.git@feature/adbpyg-map instead of !pip install adbpyg-adapter==1.1.0
• Use !git clone -b feature/adbpyg-map --single-branch https://github.com/arangoml/pyg-adapter.git instead of !git clone -b 1.1.0 --single-branch https://github.com/arangoml/pyg-adapter.git

What's new

Introduces the preserve_adb_keys parameter in ArangoDB to PyG. When set to True, will store the ArangoDB vertex & edge _key values in the (newly created) PyG graph. When calling PyG to ArangoDB, PyG graphs that have a _key or _id property will have their vertex & edge IDs persisted. See below for an example

Additional Changes

1. Feature: New supported metagraph value type for ArangoDB to PyG & PyG to ArangoDB: Set[str]
2. Bug: In PyG to ArangoDB, when explicit_metagraph=True, omitted PyG node & edge properties were still being brought over into ArangoDB (this has been fixed)
3. encapsulate some PyG to ArangoDB & ArangoDB to PyG logic in new helper methods: __set_adb_data(),__set_pyg_data()
4. code cleanup
5. docstring updates
6. test assertion logic refactor (the existing validation logic was out of date, as we were not accounting for the various supported values in our ArangoDB & PyG Metagraphs)
7. new: pytest_exception_interact in conftest.py for post-failure ArangoDB cleanup
8. new test cases
9. new 1.1.0 notebook

Demo (preserve_adb_keys)

# With `preserve_adb_keys=True`, the adapter will preserve the ArangoDB vertex & edge _key values into the (newly created) PyG graph.
# Users can then re-import their PyG graph into ArangoDB using the same _key values 
pyg_g = adbpyg_adapter.arangodb_graph_to_pyg("imdb", preserve_adb_keys=True)

# pyg_g["Movies"]["_key"] --> ["1", "2", ..., "1682"]
# pyg_g["Users"]["_key"] --> ["1", "2", ..., "943"]
# pyg_g[("Users", "Ratings", "Movies")]["_key"] --> ["2732620466", ..., "2730643624"]

# Let's add a new PyG User Node by updating the _key property
pyg_g["Users"]["_key"].append("new-user-here-944")

# Note: Prior to the re-import, we must manually set the number of nodes in the PyG graph, since the `arangodb_graph_to_pyg` API creates featureless node data
pyg_g["Movies"].num_nodes = len(pyg_g["Movies"]["_key"]) # 1682
pyg_g["Users"].num_nodes = len(pyg_g["Users"]["_key"]) # 944 (prev. 943)

# Re-import PyG graph into ArangoDB
adbpyg_adapter.pyg_to_arangodb("imdb", pyg_g, on_duplicate="update")

Demo (Set[str] as a metagraph value)

metagraph = {
    "nodeTypes": {
        "v0": {"x", "y"}, # equivalent to {"x": "x", "y": "y"}
    },
    "edgeTypes": {
        ("v0", "e0", "v0"): {"edge_attr"},
    },
}

adbpyg_adapter.pyg_to_arangodb("FakeData", data, metagraph, explicit_metagraph=False)

metagraph = {
    "vertexCollections": {
        "v0": {"x", "y"}, # equivalent to {"x": "x", "y": "y"}
    },
    "edgeCollections": {
        "e0": {"edge_attr"},
    },
}

pyg_g = adbpyg_adapter.arangodb_to_pyg("FakeData", metagraph)

[joerg84]

I believe the entry Changelog is missing

[aMahanna]

I believe the entry Changelog is missing

@joerg84

No need, as the release.yml action will auto-generate a Changelog entry using commit messages upon triggering a new release (in this case it would be 1.1.0).

Example: #9