feat(client): add methods to langfuse client to fetch and delete dataset runs

[marliessophie]

Important

Adds methods to Langfuse client for fetching and deleting dataset runs with error handling.

• Methods Added:
  • get_dataset_run(): Fetches a dataset run by dataset_name and run_name, returning DatasetRunWithItems.
  • get_dataset_runs(): Fetches all runs for a dataset, with pagination support, returning PaginatedDatasetRuns.
  • delete_dataset_run(): Deletes a dataset run by dataset_name and run_name, returning DeleteDatasetRunResponse.
• Error Handling:
  • All methods use handle_fern_exception() for error handling and re-raise exceptions.

^{This description was created by for 0d9bd8f. You can customize this summary. It will automatically update as commits are pushed.}

Disclaimer: Experimental PR review

Greptile Overview

Greptile Summary

Added three new methods to the Langfuse client for managing dataset runs: get_dataset_run, get_dataset_runs, and delete_dataset_run.

• get_dataset_run: fetches a single dataset run by dataset name and run name
• get_dataset_runs: retrieves paginated list of runs for a dataset
• delete_dataset_run: permanently deletes a dataset run and all its items
• Added required import statements for DatasetRunWithItems, DeleteDatasetRunResponse, and PaginatedDatasetRuns types

Issues found:

• Missing is_url_param=True flag in _url_encode() calls for path parameters across all three methods. These parameters are embedded in URL paths and passed to httpx, which handles encoding automatically in v0.28+. This inconsistency with the existing get_dataset method (line 2445) could cause double-encoding issues.

Confidence Score: 3/5

• This PR can be merged after addressing the URL encoding parameter issue.
• The implementation follows existing patterns and adds straightforward CRUD operations for dataset runs. However, the missing is_url_param=True flags could cause encoding issues with special characters in dataset/run names. The fix is simple but critical for correctness.
• Fix the _url_encode() calls in langfuse/_client/client.py to include is_url_param=True for all path parameters

Important Files Changed

File Analysis

Filename	Score	Overview
langfuse/_client/client.py	3/5	added three dataset run methods (get_dataset_run, get_dataset_runs, delete_dataset_run); missing is_url_param=True flag in URL encoding calls for path parameters

Sequence Diagram

sequenceDiagram
    participant User
    participant Langfuse
    participant API as Langfuse API
    
    Note over User,API: Get Dataset Run
    User->>Langfuse: get_dataset_run(dataset_name, run_name)
    Langfuse->>Langfuse: _url_encode(dataset_name)
    Langfuse->>Langfuse: _url_encode(run_name)
    Langfuse->>API: datasets.get_run(dataset_name, run_name)
    API-->>Langfuse: DatasetRunWithItems
    Langfuse-->>User: DatasetRunWithItems
    
    Note over User,API: Get Dataset Runs (Paginated)
    User->>Langfuse: get_dataset_runs(dataset_name, page, limit)
    Langfuse->>Langfuse: _url_encode(dataset_name)
    Langfuse->>API: datasets.get_runs(dataset_name, page, limit)
    API-->>Langfuse: PaginatedDatasetRuns
    Langfuse-->>User: PaginatedDatasetRuns
    
    Note over User,API: Delete Dataset Run
    User->>Langfuse: delete_dataset_run(dataset_name, run_name)
    Langfuse->>Langfuse: _url_encode(dataset_name)
    Langfuse->>Langfuse: _url_encode(run_name)
    Langfuse->>API: datasets.delete_run(dataset_name, run_name)
    API-->>Langfuse: DeleteDatasetRunResponse
    Langfuse-->>User: DeleteDatasetRunResponse

Loading

[greptile-apps]

_{1 file reviewed, no comments}

_{Edit Code Review Agent Settings | Greptile}

[hassiebp]

We should have all arguments here as keyword arguments

[hassiebp]

We should have all arguments here as keyword arguments

[hassiebp]

We should have all arguments here as keyword arguments

[hassiebp]

If run_name is being passed as a query param, we would have to set url_param=True in _url_encode to avoid double encoding by newer httpx versions.

In any case, we should add a few tests here to cover all works as expected with the encoding 👍🏾