Description

Looking at our fly_skeleton_legs.json test data: Our final json_str has keys directed, graph, links, multigraph, nodes.

directed is a boolean. (UNCHANGED from input graph)
graph is a dict with keys name and num_edges_inserted. (UNCHANGED from input graph)
links is a list of dicts with keys edge_insert_idx, key, source, target, type, for each edge.
multigraph is a boolean = True. (UNCHANGED from input graph)
nodes is a list of dicts with keys id for each node.

input object to JSON encoder	python object location in graph dict	encoded object as JSON string
`Node(name='neck', weight=1.0)`	`"links":[{... source: { Node(name='neck', weight=1.0) }`	`{"py/object": "sleap.skeleton.Node", "py/state": {"py/tuple": ["neck", 1.0]}}`
`Node(name='head', weight=1.0)`	`"links":[{... target: { Node(name='head', weight=1.0) }`	`{"py/object": "sleap.skeleton.Node", "py/state": {"py/tuple": ["head", 1.0]}}`
`<EdgeType.BODY: 1>`	`"links":[{... type: { <EdgeType.BODY: 1> }`	`{"py/reduce": [{"py/type": "sleap.skeleton.EdgeType"}, {"py/tuple": [1]}, null, null, null]}`
`Node(name='head', weight=1.0)`	`"nodes": [{'id': Node(name='head', weight=1.0)}, ..., ]`	`{"id": {"py/id": 2}}`

This is because Skeleton._graph is passed through json_graph.node_link_data, which returns a dictionary with node-link formatted data (see documentation here.

def node_link_data():
 multigraph = G.is_multigraph()

# Allow 'key' to be omitted from attrs if the graph is not a multigraph.
key = None if not multigraph else key
if len({source, target, key}) < 3:
    raise nx.NetworkXError("Attribute names are not unique.")
data = {
    "directed": G.is_directed(),
    "multigraph": multigraph,
    "graph": G.graph,
    "nodes": [{**G.nodes[n], name: n} for n in G],
}
if multigraph:
    data[link] = [
        {**d, source: u, target: v, key: k}
        for u, v, k, d in G.edges(keys=True, data=True)
    ]
else:
    data[link] = [{**d, source: u, target: v} for u, v, d in G.edges(data=True)]
return data

Note that our node-linked data has key links and not link.
- We do not have networkx pinned. I am not sure why that hasn't been a problem.

Then the node-link formatted data is passed to jsonpickle.encode(data).
- Serializes the SLEAP class Node. Each Node object, which has attributes name and weight, is encoded as a Python object sleap.skeleton.Node. The Node's internal state is represented as a Python tuple containing the name (as a string) and weight (as a numerical value): ["name", weight].
  - In the case where the node is an int, such as when the skeleton has a node_to_idx mapping (https://github.com/talmolab/sleap/blob/3c7f5afa8dd952fdee8ef3c8ac94b0689dac9cdb/sleap/skeleton.py#L1002-L1006), i.e. when the skeleton is from Labels, the node is returned as an integer without the name and weight attributes.
- Serializes the SLEAP class EdgeType. EdgeType inherits from Python's Enum class. The possible edge types are BODY = 1 and SYMMETRY = 2. When serialized, EdgeType is represented using Python's reduce function, storing the type as sleap.skeleton.EdgeType and the value as a Python tuple containing the Enum value.

If the object has been "seen" before, it will not be encoded as the full JSON string but referenced by its py/id, which starts at 1 and indexes the objects in the order they are seen so that the second time the first object is used, it will be referenced as {"py/id": 1}.

This is tricky because looking at fly_skeleton_legs.json we can see that the first time a node is used in a source or target in the links list, it is encoded as a py/object with a py/state. Then it is referenced later on using the py/id. nodes then uses only the py/id to reference the Node.
However this is not what you would expect from looking at the order in of the graph passed from jsongraph.node_link_data() to jsonpickle.encode() (shown below) which has nodes first and links after.
The same is true for the order of source, target and type: in the graph below type comes before source and target so unless explicitly ordered after source and target the py/id will not be consistent with the legacy data.
For this reason, a method is added to the SkeletonEncoder called _encode_links that is used within the classmethod encode to encode the links first, and within the links list, encode the source, target and then the type.

If the exact order wasn't followed, and the py/ids were not the same as the legacy data, the EdgeType would be decoded as a Node and the edges would not be formed correctly in the decoded Skeleton using Skeleton.from_json.

{'directed': True, 'multigraph': True, 'graph': {'name': 'skeleton_legs.mat', 'num_edges_inserted': 23}, 'nodes': [{'id': Node(name='head', weight=1.0)}, {'id': Node(name='neck', weight=1.0)}, {'id': Node(name='thorax', weight=1.0)}, {'id': Node(name='abdomen', weight=1.0)}, {'id': Node(name='wingL', weight=1.0)}, {'id': Node(name='wingR', weight=1.0)}, {'id': Node(name='forelegL1', weight=1.0)}, {'id': Node(name='forelegL2', weight=1.0)}, {'id': Node(name='forelegL3', weight=1.0)}, {'id': Node(name='forelegR1', weight=1.0)}, {'id': Node(name='forelegR2', weight=1.0)}, {'id': Node(name='forelegR3', weight=1.0)}, {'id': Node(name='midlegL1', weight=1.0)}, {'id': Node(name='midlegL2', weight=1.0)}, {'id': Node(name='midlegL3', weight=1.0)}, {'id': Node(name='midlegR1', weight=1.0)}, {'id': Node(name='midlegR2', weight=1.0)}, {'id': Node(name='midlegR3', weight=1.0)}, {'id': Node(name='hindlegL1', weight=1.0)}, {'id': Node(name='hindlegL2', weight=1.0)}, {'id': Node(name='hindlegL3', weight=1.0)}, {'id': Node(name='hindlegR1', weight=1.0)}, {'id': Node(name='hindlegR2', weight=1.0)}, {'id': Node(name='hindlegR3', weight=1.0)}], 'links': [{'edge_insert_idx': 1, 'type': <EdgeType.BODY: 1>, 'source': Node(name='neck', weight=1.0), 'target': Node(name='head', weight=1.0), 'key': 0}, {'edge_insert_idx': 0, 'type': <EdgeType.BODY: 1>, 'source': Node(name='thorax', weight=1.0), 'target': Node(name='neck', weight=1.0), 'key': 0}, {'edge_insert_idx': 2, 'type': <EdgeType.BODY: 1>, 'source': Node(name='thorax', weight=1.0), 'target': Node(name='abdomen', weight=1.0), 'key': 0}, {'edge_insert_idx': 3, 'type': <EdgeType.BODY: 1>, 'source': Node(name='thorax', weight=1.0), 'target': Node(name='wingL', weight=1.0), 'key': 0}, {'edge_insert_idx': 4, 'type': <EdgeType.BODY: 1>, 'source': Node(name='thorax', weight=1.0), 'target': Node(name='wingR', weight=1.0), 'key': 0}, {'edge_insert_idx': 5, 'type': <EdgeType.BODY: 1>, 'source': Node(name='thorax', weight=1.0), 'target': Node(name='forelegL1', weight=1.0), 'key': 0}, {'edge_insert_idx': 8, 'type': <EdgeType.BODY: 1>, 'source': Node(name='thorax', weight=1.0), 'target': Node(name='forelegR1', weight=1.0), 'key': 0}, {'edge_insert_idx': 11, 'type': <EdgeType.BODY: 1>, 'source': Node(name='thorax', weight=1.0), 'target': Node(name='midlegL1', weight=1.0), 'key': 0}, {'edge_insert_idx': 14, 'type': <EdgeType.BODY: 1>, 'source': Node(name='thorax', weight=1.0), 'target': Node(name='midlegR1', weight=1.0), 'key': 0}, {'edge_insert_idx': 17, 'type': <EdgeType.BODY: 1>, 'source': Node(name='thorax', weight=1.0), 'target': Node(name='hindlegL1', weight=1.0), 'key': 0}, {'edge_insert_idx': 20, 'type': <EdgeType.BODY: 1>, 'source': Node(name='thorax', weight=1.0), 'target': Node(name='hindlegR1', weight=1.0), 'key': 0}, {'edge_insert_idx': 6, 'type': <EdgeType.BODY: 1>, 'source': Node(name='forelegL1', weight=1.0), 'target': Node(name='forelegL2', weight=1.0), 'key': 0}, {'edge_insert_idx': 7, 'type': <EdgeType.BODY: 1>, 'source': Node(name='forelegL2', weight=1.0), 'target': Node(name='forelegL3', weight=1.0), 'key': 0}, {'edge_insert_idx': 9, 'type': <EdgeType.BODY: 1>, 'source': Node(name='forelegR1', weight=1.0), 'target': Node(name='forelegR2', weight=1.0), 'key': 0}, {'edge_insert_idx': 10, 'type': <EdgeType.BODY: 1>, 'source': Node(name='forelegR2', weight=1.0), 'target': Node(name='forelegR3', weight=1.0), 'key': 0}, {'edge_insert_idx': 12, 'type': <EdgeType.BODY: 1>, 'source': Node(name='midlegL1', weight=1.0), 'target': Node(name='midlegL2', weight=1.0), 'key': 0}, {'edge_insert_idx': 13, 'type': <EdgeType.BODY: 1>, 'source': Node(name='midlegL2', weight=1.0), 'target': Node(name='midlegL3', weight=1.0), 'key': 0}, {'edge_insert_idx': 15, 'type': <EdgeType.BODY: 1>, 'source': Node(name='midlegR1', weight=1.0), 'target': Node(name='midlegR2', weight=1.0), 'key': 0}, {'edge_insert_idx': 16, 'type': <EdgeType.BODY: 1>, 'source': Node(name='midlegR2', weight=1.0), 'target': Node(name='midlegR3', weight=1.0), 'key': 0}, {'edge_insert_idx': 18, 'type': <EdgeType.BODY: 1>, 'source': Node(name='hindlegL1', weight=1.0), 'target': Node(name='hindlegL2', weight=1.0), 'key': 0}, {'edge_insert_idx': 19, 'type': <EdgeType.BODY: 1>, 'source': Node(name='hindlegL2', weight=1.0), 'target': Node(name='hindlegL3', weight=1.0), 'key': 0}, {'edge_insert_idx': 21, 'type': <EdgeType.BODY: 1>, 'source': Node(name='hindlegR1', weight=1.0), 'target': Node(name='hindlegR2', weight=1.0), 'key': 0}, {'edge_insert_idx': 22, 'type': <EdgeType.BODY: 1>, 'source': Node(name='hindlegR2', weight=1.0), 'target': Node(name='hindlegR3', weight=1.0), 'key': 0}]}

A test has been added which loads a skeleton from a JSON file, get the graph with json_graph.node_link_data from networkx, encodes the graph with the new SkeletonEncoder.encode() method, then uses Skeleton.from_json to deserialize the JSON string (with jsonpickle.decode, then json_graph.node_link_graph). The Skeleton.matches() method is used to test the equivalence of the deserialized Skeletons. [X] This test could be parametrized to do with various skeleton fixtures. [X] We should make sure to do this with a template fixture at least. [X] An additional test could be added to see if the serialized JSON strings are equivalent.
- Important: This test currently fails when attrs is unconstrained (see github run) because with the version of attrs pulled in, the nodes are json_pickle fails to decode the Nodes which have been encoded as tuples by our SkeletonEncoder.encode() (to maintain backwards compatibility).
  - """The change in attrs version used by jsonpickle to write/read the skeleton.json is causing problems, namely, attrs >=22.2.0 returns the object state as a dictionary instead of a tuple. This results in incompatibilities when attempting to deserialize skeleton.json files (w/ an older attrs <22.2.0) that were serialized with attrs >=22.2.0. Based on the discussion in https://github.com/python-attrs/attrs/pull/1085 it sounds like attrs ==22.2.0 was unable to deserialize pickled data from attrs <22.2.0, and attrs ==23.1.0 (>=23.1.0?) at least fixes backwards compatibility. The long-term plan for attrs (discussed in https://github.com/python-attrs/attrs/issues/1091) is to remove support for deserializing tuples after giving everyone a few versions of attrs to convert their data over to a dictionary serialization.""" from https://github.com/talmolab/sleap/issues/1918 .

Types of changes

[ ] Bugfix
[ ] New feature
[X] Refactor / Code style update (no logical changes)
[ ] Build / CI changes
[ ] Documentation Update
[ ] Other (explain)

Does this address any currently open issues?

1470 #1918

This pull request accompanies #1961 in handling JSON encoding and decoding internally instead of relying on jsonpickle.

Outside contributors checklist

[ ] Review the guidelines for contributing to this repository
[ ] Read and sign the CLA and add yourself to the authors list
[ ] Make sure you are making a pull request against the develop branch (not main). Also you should start your branch off develop
[ ] Add tests that prove your fix is effective or that your feature works
[ ] Add necessary documentation (if appropriate)

Thank you for contributing to SLEAP!

:heart:

Summary by CodeRabbit

New Features
- Introduced a new SkeletonEncoder class for enhanced encoding of skeleton representations.
- Added methods for encoding Node and EdgeType objects into JSON formats.
Chores
- Updated version constraints for the attrs package in environment configuration files.
Tests
- Added new tests for validating the encoding and decoding of Skeleton objects.

Walkthrough

A new class named SkeletonEncoder has been introduced in the sleap/skeleton.py file, replacing the existing jsonpickle.encode functionality with a custom encoder for converting Python objects into JSON strings. The SkeletonEncoder includes methods for encoding Node and EdgeType objects, while the to_json method in the Skeleton class has been updated to utilize this new encoder. Additionally, tests for encoding and decoding Skeleton objects have been added, and the attrs package version constraints have been specified more clearly in the environment files.

Changes

File	Change Summary
sleap/skeleton.py	Added new class `SkeletonEncoder` for custom encoding; modified `to_json` method to use `SkeletonEncoder`.
environment.yml	Updated `attrs` package version constraint to `>=21.2.0`.
environment_no_cuda.yml	Updated `attrs` package version constraint to `>=21.2.0`.
tests/test_skeleton.py	Added tests for encoding and decoding `Skeleton` objects using `SkeletonEncoder`.

Possibly related PRs

1961: The introduction of the SkeletonDecoder class in this PR is directly related to the SkeletonEncoder class in the main PR, as both classes are involved in the (de)serialization of Node and EdgeType objects within the sleap/skeleton.py file.

Suggested reviewers

talmo

Poem

In the code where rabbits play,
A new encoder hops today.
Skeletons now dance with glee,
Encoding joyfully, you see!
With every node, a match so bright,
In the world of bytes, all feels right! 🐇✨

:scroll: Recent review details

**Configuration used: CodeRabbit UI** **Review profile: CHILL**

:inbox_tray: Commits

Files that changed from the base of the PR and between 743e4068ae49fdc377425854cb4157b3726beeee and 83a2704ae715a15346b22771a985d625ee297fac.

:ledger: Files selected for processing (1)

* sleap/skeleton.py (4 hunks)

:construction: Files skipped from review as they are similar to previous changes (1)

* sleap/skeleton.py

--- Thank you for using CodeRabbit. We offer it for free to the OSS community and would appreciate your support in helping us grow. If you find it useful, would you consider giving us a shout-out on your favorite social media?

:heart: Share

- [X](https://twitter.com/intent/tweet?text=I%20just%20used%20%40coderabbitai%20for%20my%20code%20review%2C%20and%20it%27s%20fantastic%21%20It%27s%20free%20for%20OSS%20and%20offers%20a%20free%20trial%20for%20the%20proprietary%20code.%20Check%20it%20out%3A&url=https%3A//coderabbit.ai) - [Mastodon](https://mastodon.social/share?text=I%20just%20used%20%40coderabbitai%20for%20my%20code%20review%2C%20and%20it%27s%20fantastic%21%20It%27s%20free%20for%20OSS%20and%20offers%20a%20free%20trial%20for%20the%20proprietary%20code.%20Check%20it%20out%3A%20https%3A%2F%2Fcoderabbit.ai) - [Reddit](https://www.reddit.com/submit?title=Great%20tool%20for%20code%20review%20-%20CodeRabbit&text=I%20just%20used%20CodeRabbit%20for%20my%20code%20review%2C%20and%20it%27s%20fantastic%21%20It%27s%20free%20for%20OSS%20and%20offers%20a%20free%20trial%20for%20proprietary%20code.%20Check%20it%20out%3A%20https%3A//coderabbit.ai) - [LinkedIn](https://www.linkedin.com/sharing/share-offsite/?url=https%3A%2F%2Fcoderabbit.ai&mini=true&title=Great%20tool%20for%20code%20review%20-%20CodeRabbit&summary=I%20just%20used%20CodeRabbit%20for%20my%20code%20review%2C%20and%20it%27s%20fantastic%21%20It%27s%20free%20for%20OSS%20and%20offers%20a%20free%20trial%20for%20proprietary%20code)

:placard: Tips

### Chat There are 3 ways to chat with [CodeRabbit](https://coderabbit.ai): - Review comments: Directly reply to a review comment made by CodeRabbit. Example: - `I pushed a fix in commit , please review it.` - `Generate unit testing code for this file.` - `Open a follow-up GitHub issue for this discussion.` - Files and specific lines of code (under the "Files changed" tab): Tag `@coderabbitai` in a new review comment at the desired location with your query. Examples: - `@coderabbitai generate unit testing code for this file.` - `@coderabbitai modularize this function.` - PR comments: Tag `@coderabbitai` in a new PR comment to ask questions about the PR branch. For the best results, please provide a very specific query, as very limited context is provided in this mode. Examples: - `@coderabbitai gather interesting stats about this repository and render them as a table. Additionally, render a pie chart showing the language distribution in the codebase.` - `@coderabbitai read src/utils.ts and generate unit testing code.` - `@coderabbitai read the files in the src/scheduler package and generate a class diagram using mermaid and a README in the markdown format.` - `@coderabbitai help me debug CodeRabbit configuration file.` Note: Be mindful of the bot's finite context window. It's strongly recommended to break down tasks such as reading entire modules into smaller chunks. For a focused discussion, use review comments to chat about specific files and their changes, instead of using the PR comments. ### CodeRabbit Commands (Invoked using PR comments) - `@coderabbitai pause` to pause the reviews on a PR. - `@coderabbitai resume` to resume the paused reviews. - `@coderabbitai review` to trigger an incremental review. This is useful when automatic reviews are disabled for the repository. - `@coderabbitai full review` to do a full review from scratch and review all the files again. - `@coderabbitai summary` to regenerate the summary of the PR. - `@coderabbitai resolve` resolve all the CodeRabbit review comments. - `@coderabbitai configuration` to show the current CodeRabbit configuration for the repository. - `@coderabbitai help` to get help. ### Other keywords and placeholders - Add `@coderabbitai ignore` anywhere in the PR description to prevent this PR from being reviewed. - Add `@coderabbitai summary` to generate the high-level summary at a specific location in the PR description. - Add `@coderabbitai` anywhere in the PR title to generate the title automatically. ### CodeRabbit Configuration File (`.coderabbit.yaml`) - You can programmatically configure CodeRabbit by adding a `.coderabbit.yaml` file to the root of your repository. - Please see the [configuration documentation](https://docs.coderabbit.ai/guides/configure-coderabbit) for more information. - If your editor has YAML language server enabled, you can add the path at the top of this file to enable auto-completion and validation: `# yaml-language-server: $schema=https://coderabbit.ai/integrations/schema.v2.json` ### Documentation and Community - Visit our [Documentation](https://coderabbit.ai/docs) for detailed information on how to use CodeRabbit. - Join our [Discord Community](https://discord.com/invite/GsXnASn26c) to get help, request features, and share feedback. - Follow us on [X/Twitter](https://twitter.com/coderabbitai) for updates and announcements.

talmolab / sleap

Handle skeleton encoding internally #1970

Description

Types of changes

Does this address any currently open issues?

1470 #1918

Outside contributors checklist

Thank you for contributing to SLEAP!

Summary by CodeRabbit

Summary by CodeRabbit

Walkthrough

Changes

Possibly related PRs

1961: The introduction of the `SkeletonDecoder` class in this PR is directly related to the `SkeletonEncoder` class in the main PR, as both classes are involved in the (de)serialization of `Node` and `EdgeType` objects within the `sleap/skeleton.py` file.

Suggested reviewers

Poem

Codecov Report

1841

talmolab / sleap

Handle skeleton encoding internally #1970

Description

Types of changes

Does this address any currently open issues?

1470 #1918

Outside contributors checklist

Thank you for contributing to SLEAP!

Summary by CodeRabbit

Summary by CodeRabbit

Walkthrough

Changes

Possibly related PRs

1961: The introduction of the SkeletonDecoder class in this PR is directly related to the SkeletonEncoder class in the main PR, as both classes are involved in the (de)serialization of Node and EdgeType objects within the sleap/skeleton.py file.

Suggested reviewers

Poem

Codecov Report

1841

1961: The introduction of the `SkeletonDecoder` class in this PR is directly related to the `SkeletonEncoder` class in the main PR, as both classes are involved in the (de)serialization of `Node` and `EdgeType` objects within the `sleap/skeleton.py` file.