add option for string node representation update

[prrao87]

Description

Related to the Property Graph Index retriever's as_retriever method that allows users to print the string representation of a property graph node or relationship.

I think it makes sense to revert the unintended consequence of string node representation update per #14707, where the properties dict is coerced to a string and appended to the node name prior to printing the string representation. This caused the string representation of the node to return a garbled mess of name + properties that make it hard to read when running the following code in the property graph index retriever:

retriever = index.as_retriever(
    include_text=False,  # include source text in returned nodes, default True
)

nodes = retriever.retrieve("What happened at Interleaf and Viaweb?")

for node in nodes:
    print(node.text)

The suggested revert to the change made by @logan-markewich now returns the correct results, pretty-printed as follows:

Paul Graham -> WORKED_ON -> Viaweb
Viaweb -> PART_OF -> Yahoo
Paul Graham -> WORKED_ON -> Technology companies
Paul Graham -> WORKED_ON -> World Wide Web

New Package?

Did I fill in the tool.llamahub section in the pyproject.toml and provide a detailed README.md for my new integration or package?

• Yes
• No

Version Bump?

Did I bump the version in the pyproject.toml file of the package I am updating? (Except for the llama-index-core package)

• Yes
• No

Type of Change

Please delete options that are not relevant.

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• This change requires a documentation update

How Has This Been Tested?

Tested on Kùzu, Neo4j and FalkorDB by rerunning the example notebooks end-to-end, and nothing breaks.

• Added new unit/integration tests
• Added new notebook (that tests end-to-end)
• I stared at the code and made sure it makes sense

Suggested Checklist:

• I have performed a self-review of my own code
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• I have added Google Colab support for the newly added notebooks.
• My changes generate no new warnings
• I have added tests that prove my fix is effective or that my feature works
• New and existing unit tests pass locally with my changes
• I ran make format; make lint to appease the lint gods

[logan-markewich]

@prrao87 I'm pretty sure i made this change because people wanted the node properties in the string representation lol

Instead of removing, there should be a toggle (likely on the retriever)

[prrao87]

Interesting.. the call to {triplet[n]!s} per this line is hardcoded tho, so I'm wondering what a good toggle could be - would it be in the subretriever, or at this upper level in types.py?

[logan-markewich]

@prrao87 maybe something roughly like

.as_retriever(include_properties=True)

if self.include_properties
  node_text = f"{node} ({node.properties}) -> ..."
else:
  node_text = f"{node} -> ..."

[prrao87]

@prrao87 maybe something roughly like

.as_retriever(include_properties=True)

if self.include_properties
  node_text = f"{node} ({node.properties}) -> ..."
else:
  node_text = f"{node} -> ..."

Was exactly trying this but am getting stuck in inheritance hell lol. It applies to the subretriever but I think I need to have the same kwarg in all the retrievers that are being subclassed for the PropertyGraphIndex 😅.

[logan-markewich]

@prrao87 yea, there's the base class and 5 other retrievers, not too bad :)

[prrao87]

yea, there's the base class and 5 other retrievers, not too bad :)

Now that you confirmed this, yes, not too bad at all 😅. Fixing now.

[prrao87]

Done - I left the default value as False because this way, none of the existing example notebooks that call as_retriever would need to be rerun. If users want, they can always call the retriever this way:

retriever = index.as_retriever(include_properties=True)