py2neo: batch create indexed node with multiple key:value pairs

Question

With

batch = neo4j.WriteBatch(graph_db)
for n in nodes:
    batch.create_indexed_node_or_fail(index, key, value, {node properties})
batch.submit()

you can easily create indexed nodes in a batch.

But how can I create a node in a batch and add multiple key:value pairs at the same time? Or add the node to multiple indices?

One idea I thought of is to first create the nodes and then add them to the indices in a second batch:

for n in nodes:
    batch.create({node properties})
results = batch.submit()

for r in results:
    batch add node to index1
    batch add node to index2
batch.submit()

But this seems to complicated. Py2neo is so amazing, there must be an easier way.

---

Based on Nigel's answer, I tried to this for a list of nodes:

batch = neo4j.WriteBatch(graph_db)
item_in_batch = 0

for n in list_of_nodes:
    batch.create(n) # this works, all nodes are created
    batch.add_indexed_node("people", "key", "value from n", item_in_batch)
    batch.add_indexed_node("people", "key2", "other value from n", item_in_batch)
    i += 1
batch.submit()

Some nodes end up in my index, but not all. When I search the index for ('key:value-from-n") it returns a node but not the correct one. How can I access the 'internal' ID of the nodes when I try to add them to an index?

---

I tried the (currently testing) version 1.6.0 of py2neo and it works!

n = batch.create(node)

now returns a request object which can be used within the batch to refer to the node.

An example is in the docs for 1.6.0

Answer 1

You shouldn't need two batches. You should be able to create and add in the same batch. Something like:

batch = neo4j.WriteBatch(graph_db)
batch.create({"name": "Alice})
batch.add_indexed_node("people", "name", "Alice", 0)
batch.add_indexed_node("employees", "id", 2345678, 0)
batch.add_indexed_node("fighter_pilots", "call_sign", "Silver Eagle", 0)
batch.submit()

For unique indexing, you could use add_indexed_node_or_fail instead.

And thanks for the compliment :-)