text2sql-eval-toolkit 1.3.0__tar.gz → 1.4.0__tar.gz

{text2sql_eval_toolkit-1.3.0/src/text2sql_eval_toolkit.egg-info → text2sql_eval_toolkit-1.4.0}/PKG-INFO

@@ -1,10 +1,14 @@
 Metadata-Version: 2.4
 Name: text2sql-eval-toolkit
-Version: 1.3.0
+Version: 1.4.0
 Summary: A toolkit for evaluating text-to-SQL pipelines
 Author-email: Oktie Hassanzadeh <lastname@us.ibm.com>
 License: MIT
 Project-URL: Homepage, https://github.com/IBM/text2sql-eval-toolkit
+Project-URL: Documentation, https://text2sql-eval-toolkit.readthedocs.io/
+Project-URL: Repository, https://github.com/IBM/text2sql-eval-toolkit
+Project-URL: Changelog, https://github.com/IBM/text2sql-eval-toolkit/blob/main/CHANGELOG.md
+Project-URL: Issues, https://github.com/IBM/text2sql-eval-toolkit/issues
 Classifier: Programming Language :: Python :: 3
 Classifier: License :: OSI Approved :: MIT License
 Classifier: Operating System :: OS Independent
@@ -30,6 +34,12 @@ Requires-Dist: langchain-core>=0.3.0
 Requires-Dist: openai>=1.0.0
 Requires-Dist: google-genai>=1.0.0
 Requires-Dist: huggingface_hub<1.0,>=0.30
+Provides-Extra: litellm
+Requires-Dist: litellm>=1.50; extra == "litellm"
+Provides-Extra: docs
+Requires-Dist: mkdocs>=1.6; extra == "docs"
+Requires-Dist: mkdocs-material>=9.5; extra == "docs"
+Requires-Dist: mkdocstrings[python]>=0.26; extra == "docs"
 Provides-Extra: db2
 Requires-Dist: ibm-db>=3.2.6; extra == "db2"
 Requires-Dist: ibm-db-sa>=0.4.0; extra == "db2"
@@ -71,7 +81,7 @@ Dynamic: license-file
 > Supports execution-based metrics, multiple ground truths, LLM-as-judge, and rich error analysis.
 
 <p align="center">
-  <img src="dashboard-screenshot.png" alt="Text2SQL Evaluation Dashboard: benchmark overview with tiles for each dataset" width="560"/>
+  <img src="https://raw.githubusercontent.com/IBM/text2sql-eval-toolkit/main/dashboard-screenshot.png" alt="Text2SQL Evaluation Dashboard: benchmark overview with tiles for each dataset" width="560"/>
 </p>
 
 Text-to-SQL systems aim to translate natural language questions into executable SQL queries, enabling intuitive access to structured data. While recent advances in large language models have significantly improved generation quality, evaluating these systems remains a complex and critical challenge.
@@ -101,26 +111,31 @@ Whether you're building new models, comparing existing ones, or diagnosing perfo
 
 ## Documentation
 
-- **[docs/dashboard/](docs/dashboard/)** — the evaluation dashboard: features,
+- **[Documentation site](https://text2sql-eval-toolkit.readthedocs.io/)** — the
+  guide (installation, the five stages, the data model, benchmarks, models,
+  LLM-as-judge, the CLI and configuration) and the
+  [API reference](https://text2sql-eval-toolkit.readthedocs.io/en/latest/reference/) for
+  every exported function and class.
+- **[docs/dashboard/](https://github.com/IBM/text2sql-eval-toolkit/tree/main/docs/dashboard)** — the evaluation dashboard: features,
   shareable links, the query index, capability tiers, and deployment.
-- **[data/benchmarks/README.md](data/benchmarks/README.md)** — benchmark
+- **[data/benchmarks/README.md](https://github.com/IBM/text2sql-eval-toolkit/blob/main/data/benchmarks/README.md)** — benchmark
   definitions and configuration.
-- **[data/benchmarks/dbs/README.md](data/benchmarks/dbs/README.md)** — database
+- **[data/benchmarks/dbs/README.md](https://github.com/IBM/text2sql-eval-toolkit/blob/main/data/benchmarks/dbs/README.md)** — database
   setup for each benchmark.
-- **[CONTRIBUTING.md](CONTRIBUTING.md)** — toolchain, tests, and coding standards.
+- **[CONTRIBUTING.md](https://github.com/IBM/text2sql-eval-toolkit/blob/main/CONTRIBUTING.md)** — toolchain, tests, and coding standards.
 
 ## Components
 
 <p align="center">
-  <img src="toolkit.png" alt="Text-to-SQL Toolkit Components" width="500"/>
+  <img src="https://raw.githubusercontent.com/IBM/text2sql-eval-toolkit/main/toolkit.png" alt="Text-to-SQL Toolkit Components" width="500"/>
 </p>
 
-- **Evaluation** ([`scripts/evaluation`](scripts/evaluation)/`src/evaluation.py`): Includes a script and library for evaluating text-to-SQL model outputs using various metrics. See [`scripts/evaluation/README.md`](scripts/evaluation/README.md) for details.
-- **Text-to-SQL Inference** ([`scripts/inference`](scripts/inference)): Provides a simple LLM inference pipeline for generating SQL queries from natural language inputs. Run `python scripts/inference/run_inference.py -h` for more information.
-- **SQL Execution** ([`scripts/execution`](scripts/execution)): Runs the ground truth and predicted SQLs for a given benchmark and saves the dataframes for evaluation and error analysis. Run `python scripts/execution/run_execution.py -h` for more information.
-- **Results and Error Analysis** ([`scripts/analysis`](scripts/analysis)): Contains scripts and utilities for analyzing evaluation results, identifying common error patterns, and generating summary statistics and visualizations. Useful for debugging and improving model performance.
-- **SQL Profiling** ([`scripts/profiling`](scripts/profiling)) tools to profile SQL queries to gather query characteristics to facilitate better analysis of results and errors.
-- **Evaluation Dashboard** ([`dashboard`](dashboard)): Optional FastAPI + React web UI for browsing benchmarks and pipeline metrics, error analysis (search, filters, cross-pipeline disagreement), side-by-side comparison of result summaries, editing LLM-as-judge YAML, and launching evaluations with job status. See [docs/dashboard/](docs/dashboard/) for features, deployment, and development.
+- **Evaluation** ([`scripts/evaluation`](https://github.com/IBM/text2sql-eval-toolkit/tree/main/scripts/evaluation)/`src/evaluation.py`): Includes a script and library for evaluating text-to-SQL model outputs using various metrics. See [`scripts/evaluation/README.md`](https://github.com/IBM/text2sql-eval-toolkit/blob/main/scripts/evaluation/README.md) for details.
+- **Text-to-SQL Inference** ([`scripts/inference`](https://github.com/IBM/text2sql-eval-toolkit/tree/main/scripts/inference)): Provides a simple LLM inference pipeline for generating SQL queries from natural language inputs. Run `python scripts/inference/run_inference.py -h` for more information.
+- **SQL Execution** ([`scripts/execution`](https://github.com/IBM/text2sql-eval-toolkit/tree/main/scripts/execution)): Runs the ground truth and predicted SQLs for a given benchmark and saves the dataframes for evaluation and error analysis. Run `python scripts/execution/run_execution.py -h` for more information.
+- **Results and Error Analysis** ([`scripts/analysis`](https://github.com/IBM/text2sql-eval-toolkit/tree/main/scripts/analysis)): Contains scripts and utilities for analyzing evaluation results, identifying common error patterns, and generating summary statistics and visualizations. Useful for debugging and improving model performance.
+- **SQL Profiling** ([`scripts/profiling`](https://github.com/IBM/text2sql-eval-toolkit/tree/main/scripts/profiling)) tools to profile SQL queries to gather query characteristics to facilitate better analysis of results and errors.
+- **Evaluation Dashboard** ([`dashboard`](https://github.com/IBM/text2sql-eval-toolkit/tree/main/dashboard)): Optional FastAPI + React web UI for browsing benchmarks and pipeline metrics, error analysis (search, filters, cross-pipeline disagreement), side-by-side comparison of result summaries, editing LLM-as-judge YAML, and launching evaluations with job status. See [docs/dashboard/](https://github.com/IBM/text2sql-eval-toolkit/tree/main/docs/dashboard) for features, deployment, and development.
 
 ## Setup
 
@@ -155,7 +170,7 @@ text2sql-eval-toolkit results fetch
 ```
 
 After this completes, the dashboard and analysis scripts will work
-against the downloaded artefacts. See [docs/dashboard/](docs/dashboard/) to launch
+against the downloaded artefacts. See [docs/dashboard/](https://github.com/IBM/text2sql-eval-toolkit/tree/main/docs/dashboard) to launch
 the UI.
 
 To fetch only a specific benchmark:
@@ -216,13 +231,13 @@ pip install -e .
 
 ## Additional Requirements
 
-To use features that rely on LLM inference, you need to set the required environment variables for the relevant services. For watsonx.ai, this includes setting the following environment variables: `WATSONX_APIKEY`, `WATSONX_API_BASE`, and `WATSONX_PROJECTID`. See [env.example](env.example) for all available configuration options.
+To use features that rely on LLM inference, you need to set the required environment variables for the relevant services. For watsonx.ai, this includes setting the following environment variables: `WATSONX_APIKEY`, `WATSONX_API_BASE`, and `WATSONX_PROJECTID`. See [env.example](https://github.com/IBM/text2sql-eval-toolkit/blob/main/env.example) for all available configuration options.
 
 ### Benchmarks
 
 The toolkit comes with pre-defined public benchmarks including BIRD-SQL, Spider, Beaver, and Archer. These benchmarks require initial setup such as downloading databases or configuring connection strings. The toolkit also includes smaller test benchmarks (subsets of the full benchmarks) for quick validation and CI/CD pipelines.
 
-**📚 See [data/benchmarks/README.md](data/benchmarks/README.md) for complete benchmark details, setup instructions, and configuration.**
+**📚 See [data/benchmarks/README.md](https://github.com/IBM/text2sql-eval-toolkit/blob/main/data/benchmarks/README.md) for complete benchmark details, setup instructions, and configuration.**
 
 ## Usage
 
@@ -301,7 +316,7 @@ agentic.run_pipeline(
 )
 ```
 
-See the docstrings of the exported functions/classes in `text2sql_eval_toolkit.__init__` for the full list of public APIs.
+Every exported function and class is documented in the **[API reference](https://text2sql-eval-toolkit.readthedocs.io/en/latest/reference/)**, generated from the docstrings.
 
 ### Running Experiments
 
@@ -310,7 +325,7 @@ See the docstrings of the exported functions/classes in `text2sql_eval_toolkit._
 python scripts/run_experiment.py [benchmark_name]
 ```
 
-This will initialize a pipeline, perform inference to generate SQLs and output dataframes, and run the evaluation script to gather accuracy metrics. Benchmark names and configurations are defined in [`data/benchmarks.json`](data/benchmarks.json) (for full benchmarks) or [`data/test-benchmarks.json`](data/test-benchmarks.json) (for test benchmarks). See [data/benchmarks/README.md](data/benchmarks/README.md) for more details. Run `python scripts/run_experiment.py -h` for details on how to set different input arguments for the script.
+This will initialize a pipeline, perform inference to generate SQLs and output dataframes, and run the evaluation script to gather accuracy metrics. Benchmark names and configurations are defined in [`data/benchmarks.json`](https://github.com/IBM/text2sql-eval-toolkit/blob/main/data/benchmarks.json) (for full benchmarks) or [`data/test-benchmarks.json`](https://github.com/IBM/text2sql-eval-toolkit/blob/main/data/test-benchmarks.json) (for test benchmarks). See [data/benchmarks/README.md](https://github.com/IBM/text2sql-eval-toolkit/blob/main/data/benchmarks/README.md) for more details. Run `python scripts/run_experiment.py -h` for details on how to set different input arguments for the script.
 
 **Multiple Benchmarks:**
 ```bash
@@ -324,7 +339,7 @@ python scripts/run_all_benchmarks.py
 python scripts/run_all_benchmarks.py --test --run_all_baselines
 ```
 
-The `run_all_benchmarks.py` script supports separate model configurations for standard and agentic baselines to optimize runtime. See [data/benchmarks/README.md](data/benchmarks/README.md#model-configuration) for detailed configuration options.
+The `run_all_benchmarks.py` script supports separate model configurations for standard and agentic baselines to optimize runtime. See [data/benchmarks/README.md](https://github.com/IBM/text2sql-eval-toolkit/blob/main/data/benchmarks/README.md#model-configuration) for detailed configuration options.
 
 ### Running Individual Components
 
@@ -334,7 +349,7 @@ To run evaluation only:
 python scripts/evaluation/run_evaluation.py [-h] [--output_file OUTPUT_FILE] [--summary_file SUMMARY_FILE] [--csv_summary_file CSV_SUMMARY_FILE] [--use_llm_judge] input_file
 ```
 
-See [`scripts/evaluation/README.md`](scripts/evaluation/README.md) for details.
+See [`scripts/evaluation/README.md`](https://github.com/IBM/text2sql-eval-toolkit/blob/main/scripts/evaluation/README.md) for details.
 
 To run inference only:
 
@@ -371,7 +386,7 @@ To produce the results dashboard along with results summary and error analysis M
 python scripts/analysis/make_summary_report.py
 ```
 
-The output will be written to [data/results/README.md](data/results/README.md).
+The output will be written to [data/results/README.md](https://github.com/IBM/text2sql-eval-toolkit/blob/main/data/results/README.md).
 
 ### Evaluation dashboard
 
@@ -387,7 +402,7 @@ individual record can be linked to directly.
 
 The dashboard runs as a local tool with every capability enabled, or as a
 shared read-only site with optional sign-in for a small allowlist. See
-**[docs/dashboard/](docs/dashboard/)** for features, the URL scheme, the query
+**[docs/dashboard/](https://github.com/IBM/text2sql-eval-toolkit/tree/main/docs/dashboard)** for features, the URL scheme, the query
 index, capability tiers, and deployment.
 
 ## Project Structure
@@ -441,7 +456,7 @@ A good way to familiarize yourself with the codebase and contribution process is
 
 **Note: We appreciate your effort, and want to avoid a situation where a contribution requires extensive rework (by you or by us), sits in backlog for a long time, or cannot be accepted at all!**
 
-For detailed contribution guidelines, please see [CONTRIBUTING.md](CONTRIBUTING.md).
+For detailed contribution guidelines, please see [CONTRIBUTING.md](https://github.com/IBM/text2sql-eval-toolkit/blob/main/CONTRIBUTING.md).
 
 ### Pull Request Process
 
@@ -453,7 +468,7 @@ For detailed contribution guidelines, please see [CONTRIBUTING.md](CONTRIBUTING.
 
 ## License
 
-This project is licensed under the Apache License 2.0. See the [LICENSE](LICENSE) file for more details.
+This project is licensed under the Apache License 2.0. See the [LICENSE](https://github.com/IBM/text2sql-eval-toolkit/tree/main/LICENSE) file for more details.
 
 ## Citation
 
@@ -475,9 +490,9 @@ If you use this toolkit in research, please cite:
 
 ## Additional Resources
 
-- [CODE_OF_CONDUCT.md](CODE_OF_CONDUCT.md) - Community guidelines
-- [SECURITY.md](SECURITY.md) - Security policy and vulnerability reporting
-- [MAINTAINERS.md](MAINTAINERS.md) - Project maintainers
-- [CHANGELOG.md](CHANGELOG.md) - Project changelog
+- [CODE_OF_CONDUCT.md](https://github.com/IBM/text2sql-eval-toolkit/blob/main/CODE_OF_CONDUCT.md) - Community guidelines
+- [SECURITY.md](https://github.com/IBM/text2sql-eval-toolkit/blob/main/SECURITY.md) - Security policy and vulnerability reporting
+- [MAINTAINERS.md](https://github.com/IBM/text2sql-eval-toolkit/blob/main/MAINTAINERS.md) - Project maintainers
+- [CHANGELOG.md](https://github.com/IBM/text2sql-eval-toolkit/blob/main/CHANGELOG.md) - Project changelog
 
 [issues]: https://github.com/IBM/text2sql-eval-toolkit/issues/new

{text2sql_eval_toolkit-1.3.0 → text2sql_eval_toolkit-1.4.0}/README.md

@@ -4,7 +4,7 @@
 > Supports execution-based metrics, multiple ground truths, LLM-as-judge, and rich error analysis.
 
 <p align="center">
-  <img src="dashboard-screenshot.png" alt="Text2SQL Evaluation Dashboard: benchmark overview with tiles for each dataset" width="560"/>
+  <img src="https://raw.githubusercontent.com/IBM/text2sql-eval-toolkit/main/dashboard-screenshot.png" alt="Text2SQL Evaluation Dashboard: benchmark overview with tiles for each dataset" width="560"/>
 </p>
 
 Text-to-SQL systems aim to translate natural language questions into executable SQL queries, enabling intuitive access to structured data. While recent advances in large language models have significantly improved generation quality, evaluating these systems remains a complex and critical challenge.
@@ -34,26 +34,31 @@ Whether you're building new models, comparing existing ones, or diagnosing perfo
 
 ## Documentation
 
-- **[docs/dashboard/](docs/dashboard/)** — the evaluation dashboard: features,
+- **[Documentation site](https://text2sql-eval-toolkit.readthedocs.io/)** — the
+  guide (installation, the five stages, the data model, benchmarks, models,
+  LLM-as-judge, the CLI and configuration) and the
+  [API reference](https://text2sql-eval-toolkit.readthedocs.io/en/latest/reference/) for
+  every exported function and class.
+- **[docs/dashboard/](https://github.com/IBM/text2sql-eval-toolkit/tree/main/docs/dashboard)** — the evaluation dashboard: features,
   shareable links, the query index, capability tiers, and deployment.
-- **[data/benchmarks/README.md](data/benchmarks/README.md)** — benchmark
+- **[data/benchmarks/README.md](https://github.com/IBM/text2sql-eval-toolkit/blob/main/data/benchmarks/README.md)** — benchmark
   definitions and configuration.
-- **[data/benchmarks/dbs/README.md](data/benchmarks/dbs/README.md)** — database
+- **[data/benchmarks/dbs/README.md](https://github.com/IBM/text2sql-eval-toolkit/blob/main/data/benchmarks/dbs/README.md)** — database
   setup for each benchmark.
-- **[CONTRIBUTING.md](CONTRIBUTING.md)** — toolchain, tests, and coding standards.
+- **[CONTRIBUTING.md](https://github.com/IBM/text2sql-eval-toolkit/blob/main/CONTRIBUTING.md)** — toolchain, tests, and coding standards.
 
 ## Components
 
 <p align="center">
-  <img src="toolkit.png" alt="Text-to-SQL Toolkit Components" width="500"/>
+  <img src="https://raw.githubusercontent.com/IBM/text2sql-eval-toolkit/main/toolkit.png" alt="Text-to-SQL Toolkit Components" width="500"/>
 </p>
 
-- **Evaluation** ([`scripts/evaluation`](scripts/evaluation)/`src/evaluation.py`): Includes a script and library for evaluating text-to-SQL model outputs using various metrics. See [`scripts/evaluation/README.md`](scripts/evaluation/README.md) for details.
-- **Text-to-SQL Inference** ([`scripts/inference`](scripts/inference)): Provides a simple LLM inference pipeline for generating SQL queries from natural language inputs. Run `python scripts/inference/run_inference.py -h` for more information.
-- **SQL Execution** ([`scripts/execution`](scripts/execution)): Runs the ground truth and predicted SQLs for a given benchmark and saves the dataframes for evaluation and error analysis. Run `python scripts/execution/run_execution.py -h` for more information.
-- **Results and Error Analysis** ([`scripts/analysis`](scripts/analysis)): Contains scripts and utilities for analyzing evaluation results, identifying common error patterns, and generating summary statistics and visualizations. Useful for debugging and improving model performance.
-- **SQL Profiling** ([`scripts/profiling`](scripts/profiling)) tools to profile SQL queries to gather query characteristics to facilitate better analysis of results and errors.
-- **Evaluation Dashboard** ([`dashboard`](dashboard)): Optional FastAPI + React web UI for browsing benchmarks and pipeline metrics, error analysis (search, filters, cross-pipeline disagreement), side-by-side comparison of result summaries, editing LLM-as-judge YAML, and launching evaluations with job status. See [docs/dashboard/](docs/dashboard/) for features, deployment, and development.
+- **Evaluation** ([`scripts/evaluation`](https://github.com/IBM/text2sql-eval-toolkit/tree/main/scripts/evaluation)/`src/evaluation.py`): Includes a script and library for evaluating text-to-SQL model outputs using various metrics. See [`scripts/evaluation/README.md`](https://github.com/IBM/text2sql-eval-toolkit/blob/main/scripts/evaluation/README.md) for details.
+- **Text-to-SQL Inference** ([`scripts/inference`](https://github.com/IBM/text2sql-eval-toolkit/tree/main/scripts/inference)): Provides a simple LLM inference pipeline for generating SQL queries from natural language inputs. Run `python scripts/inference/run_inference.py -h` for more information.
+- **SQL Execution** ([`scripts/execution`](https://github.com/IBM/text2sql-eval-toolkit/tree/main/scripts/execution)): Runs the ground truth and predicted SQLs for a given benchmark and saves the dataframes for evaluation and error analysis. Run `python scripts/execution/run_execution.py -h` for more information.
+- **Results and Error Analysis** ([`scripts/analysis`](https://github.com/IBM/text2sql-eval-toolkit/tree/main/scripts/analysis)): Contains scripts and utilities for analyzing evaluation results, identifying common error patterns, and generating summary statistics and visualizations. Useful for debugging and improving model performance.
+- **SQL Profiling** ([`scripts/profiling`](https://github.com/IBM/text2sql-eval-toolkit/tree/main/scripts/profiling)) tools to profile SQL queries to gather query characteristics to facilitate better analysis of results and errors.
+- **Evaluation Dashboard** ([`dashboard`](https://github.com/IBM/text2sql-eval-toolkit/tree/main/dashboard)): Optional FastAPI + React web UI for browsing benchmarks and pipeline metrics, error analysis (search, filters, cross-pipeline disagreement), side-by-side comparison of result summaries, editing LLM-as-judge YAML, and launching evaluations with job status. See [docs/dashboard/](https://github.com/IBM/text2sql-eval-toolkit/tree/main/docs/dashboard) for features, deployment, and development.
 
 ## Setup
 
@@ -88,7 +93,7 @@ text2sql-eval-toolkit results fetch
 ```
 
 After this completes, the dashboard and analysis scripts will work
-against the downloaded artefacts. See [docs/dashboard/](docs/dashboard/) to launch
+against the downloaded artefacts. See [docs/dashboard/](https://github.com/IBM/text2sql-eval-toolkit/tree/main/docs/dashboard) to launch
 the UI.
 
 To fetch only a specific benchmark:
@@ -149,13 +154,13 @@ pip install -e .
 
 ## Additional Requirements
 
-To use features that rely on LLM inference, you need to set the required environment variables for the relevant services. For watsonx.ai, this includes setting the following environment variables: `WATSONX_APIKEY`, `WATSONX_API_BASE`, and `WATSONX_PROJECTID`. See [env.example](env.example) for all available configuration options.
+To use features that rely on LLM inference, you need to set the required environment variables for the relevant services. For watsonx.ai, this includes setting the following environment variables: `WATSONX_APIKEY`, `WATSONX_API_BASE`, and `WATSONX_PROJECTID`. See [env.example](https://github.com/IBM/text2sql-eval-toolkit/blob/main/env.example) for all available configuration options.
 
 ### Benchmarks
 
 The toolkit comes with pre-defined public benchmarks including BIRD-SQL, Spider, Beaver, and Archer. These benchmarks require initial setup such as downloading databases or configuring connection strings. The toolkit also includes smaller test benchmarks (subsets of the full benchmarks) for quick validation and CI/CD pipelines.
 
-**📚 See [data/benchmarks/README.md](data/benchmarks/README.md) for complete benchmark details, setup instructions, and configuration.**
+**📚 See [data/benchmarks/README.md](https://github.com/IBM/text2sql-eval-toolkit/blob/main/data/benchmarks/README.md) for complete benchmark details, setup instructions, and configuration.**
 
 ## Usage
 
@@ -234,7 +239,7 @@ agentic.run_pipeline(
 )
 ```
 
-See the docstrings of the exported functions/classes in `text2sql_eval_toolkit.__init__` for the full list of public APIs.
+Every exported function and class is documented in the **[API reference](https://text2sql-eval-toolkit.readthedocs.io/en/latest/reference/)**, generated from the docstrings.
 
 ### Running Experiments
 
@@ -243,7 +248,7 @@ See the docstrings of the exported functions/classes in `text2sql_eval_toolkit._
 python scripts/run_experiment.py [benchmark_name]
 ```
 
-This will initialize a pipeline, perform inference to generate SQLs and output dataframes, and run the evaluation script to gather accuracy metrics. Benchmark names and configurations are defined in [`data/benchmarks.json`](data/benchmarks.json) (for full benchmarks) or [`data/test-benchmarks.json`](data/test-benchmarks.json) (for test benchmarks). See [data/benchmarks/README.md](data/benchmarks/README.md) for more details. Run `python scripts/run_experiment.py -h` for details on how to set different input arguments for the script.
+This will initialize a pipeline, perform inference to generate SQLs and output dataframes, and run the evaluation script to gather accuracy metrics. Benchmark names and configurations are defined in [`data/benchmarks.json`](https://github.com/IBM/text2sql-eval-toolkit/blob/main/data/benchmarks.json) (for full benchmarks) or [`data/test-benchmarks.json`](https://github.com/IBM/text2sql-eval-toolkit/blob/main/data/test-benchmarks.json) (for test benchmarks). See [data/benchmarks/README.md](https://github.com/IBM/text2sql-eval-toolkit/blob/main/data/benchmarks/README.md) for more details. Run `python scripts/run_experiment.py -h` for details on how to set different input arguments for the script.
 
 **Multiple Benchmarks:**
 ```bash
@@ -257,7 +262,7 @@ python scripts/run_all_benchmarks.py
 python scripts/run_all_benchmarks.py --test --run_all_baselines
 ```
 
-The `run_all_benchmarks.py` script supports separate model configurations for standard and agentic baselines to optimize runtime. See [data/benchmarks/README.md](data/benchmarks/README.md#model-configuration) for detailed configuration options.
+The `run_all_benchmarks.py` script supports separate model configurations for standard and agentic baselines to optimize runtime. See [data/benchmarks/README.md](https://github.com/IBM/text2sql-eval-toolkit/blob/main/data/benchmarks/README.md#model-configuration) for detailed configuration options.
 
 ### Running Individual Components
 
@@ -267,7 +272,7 @@ To run evaluation only:
 python scripts/evaluation/run_evaluation.py [-h] [--output_file OUTPUT_FILE] [--summary_file SUMMARY_FILE] [--csv_summary_file CSV_SUMMARY_FILE] [--use_llm_judge] input_file
 ```
 
-See [`scripts/evaluation/README.md`](scripts/evaluation/README.md) for details.
+See [`scripts/evaluation/README.md`](https://github.com/IBM/text2sql-eval-toolkit/blob/main/scripts/evaluation/README.md) for details.
 
 To run inference only:
 
@@ -304,7 +309,7 @@ To produce the results dashboard along with results summary and error analysis M
 python scripts/analysis/make_summary_report.py
 ```
 
-The output will be written to [data/results/README.md](data/results/README.md).
+The output will be written to [data/results/README.md](https://github.com/IBM/text2sql-eval-toolkit/blob/main/data/results/README.md).
 
 ### Evaluation dashboard
 
@@ -320,7 +325,7 @@ individual record can be linked to directly.
 
 The dashboard runs as a local tool with every capability enabled, or as a
 shared read-only site with optional sign-in for a small allowlist. See
-**[docs/dashboard/](docs/dashboard/)** for features, the URL scheme, the query
+**[docs/dashboard/](https://github.com/IBM/text2sql-eval-toolkit/tree/main/docs/dashboard)** for features, the URL scheme, the query
 index, capability tiers, and deployment.
 
 ## Project Structure
@@ -374,7 +379,7 @@ A good way to familiarize yourself with the codebase and contribution process is
 
 **Note: We appreciate your effort, and want to avoid a situation where a contribution requires extensive rework (by you or by us), sits in backlog for a long time, or cannot be accepted at all!**
 
-For detailed contribution guidelines, please see [CONTRIBUTING.md](CONTRIBUTING.md).
+For detailed contribution guidelines, please see [CONTRIBUTING.md](https://github.com/IBM/text2sql-eval-toolkit/blob/main/CONTRIBUTING.md).
 
 ### Pull Request Process
 
@@ -386,7 +391,7 @@ For detailed contribution guidelines, please see [CONTRIBUTING.md](CONTRIBUTING.
 
 ## License
 
-This project is licensed under the Apache License 2.0. See the [LICENSE](LICENSE) file for more details.
+This project is licensed under the Apache License 2.0. See the [LICENSE](https://github.com/IBM/text2sql-eval-toolkit/tree/main/LICENSE) file for more details.
 
 ## Citation
 
@@ -408,9 +413,9 @@ If you use this toolkit in research, please cite:
 
 ## Additional Resources
 
-- [CODE_OF_CONDUCT.md](CODE_OF_CONDUCT.md) - Community guidelines
-- [SECURITY.md](SECURITY.md) - Security policy and vulnerability reporting
-- [MAINTAINERS.md](MAINTAINERS.md) - Project maintainers
-- [CHANGELOG.md](CHANGELOG.md) - Project changelog
+- [CODE_OF_CONDUCT.md](https://github.com/IBM/text2sql-eval-toolkit/blob/main/CODE_OF_CONDUCT.md) - Community guidelines
+- [SECURITY.md](https://github.com/IBM/text2sql-eval-toolkit/blob/main/SECURITY.md) - Security policy and vulnerability reporting
+- [MAINTAINERS.md](https://github.com/IBM/text2sql-eval-toolkit/blob/main/MAINTAINERS.md) - Project maintainers
+- [CHANGELOG.md](https://github.com/IBM/text2sql-eval-toolkit/blob/main/CHANGELOG.md) - Project changelog
 
 [issues]: https://github.com/IBM/text2sql-eval-toolkit/issues/new

text2sql_eval_toolkit-1.3.0/dashboard/dist/assets/BenchmarkDetail-DQVQUBUE.js → text2sql_eval_toolkit-1.4.0/dashboard/dist/assets/BenchmarkDetail-gEcRuzmW.js

@@ -1,4 +1,4 @@
-import{R as p,u as Ne,a as Be,r as l,b as Oe,F as Ke,c as We,d as ee,m as He,i as qe,_ as te,W as Ue,e as Ve,C as Qe,f as Xe,s as Ge,g as Je,h as Ye,A as Ze,P as i,j as ue,k as s,I as ve,l as et,D as tt,B as se,n as st,T as at,o as it,p as nt,q as je,t as rt,v as lt,w as ot,x as dt,y as ct,z as ut}from"./index-C9VP9iwL.js";import{d as _e,u as $e,L as ae,a as mt,b as pt}from"./defaultItemToString-BtCahKYf.js";const{ItemMouseMove:ht,MenuMouseLeave:ft,ToggleButtonBlur:gt,FunctionCloseMenu:bt}=$e.stateChangeTypes;function Ie(o,E){const{changes:h,type:S}=E;switch(S){case ht:return o;case ft:return h.highlightedIndex===o.highlightedIndex?o:h;case gt:case bt:return{...h,selectedItem:o.selectedItem};default:return h}}const ie=p.forwardRef(({autoAlign:o=!1,className:E,decorator:h,disabled:S=!1,direction:z="bottom",items:R,label:A,["aria-label"]:I,ariaLabel:Q,itemToString:b=_e,itemToElement:L=null,renderSelectedItem:N,type:X="default",size:D,onChange:y,id:F,titleText:B="",hideLabel:m,helperText:O="",translateWithId:M,light:K,invalid:G,invalidText:$,warn:ne,warnText:P,initialSelectedItem:W,selectedItem:H,downshiftProps:J,readOnly:w,slug:v,...re},Y)=>{const C=Ne("enable-v12-dynamic-floating-styles"),{refs:j,floatingStyles:q,middlewareData:le}=Be(C||o?{placement:z,strategy:"fixed",middleware:[Ge({apply({rects:a,elements:c}){Object.assign(c.floating.style,{width:`${a.reference.width}px`})}}),o&&Je(),o&&Ye()],whileElementsMounted:Xe}:{});l.useEffect(()=>{var a;if(C||o){const c={...q,visibility:(a=le.hide)!=null&&a.referenceHidden?"hidden":"visible"};Object.keys(c).forEach(_=>{j.floating.current&&(j.floating.current.style[_]=c[_])})}},[q,o,j.floating]);const e=Oe(),{isFluid:t}=l.useContext(Ke),n=l.useCallback(({selectedItem:a})=>{y&&y({selectedItem:a??null})},[y]),u=l.useCallback((a,c)=>a!==null&&typeof a=="object"&&"disabled"in a&&a.disabled===!0,[]),d=l.useCallback(a=>{const{highlightedIndex:c}=a;if(c!==void 0&&c>-1){const Z=document.querySelectorAll(`li.${e}--list-box__menu-item[role="option"]`)[c];Z&&Z.scrollIntoView({behavior:"smooth",block:"nearest"})}},[e]),f=l.useMemo(()=>R,[R]),k=l.useMemo(()=>({items:f,itemToString:b,initialSelectedItem:W,onSelectedItemChange:n,stateReducer:Ie,isItemDisabled:u,onHighlightedIndexChange:d,...J}),[f,b,W,n,Ie,u,d,J]);H!==void 0&&(k.selectedItem=H);const{isOpen:g,getToggleButtonProps:Ce,getLabelProps:ke,getMenuProps:me,getItemProps:Te,selectedItem:x,highlightedIndex:Ee}=$e(k),T=X==="inline",r=We({id:F,readOnly:w,disabled:S??!1,invalid:G??!1,invalidText:$,warn:ne??!1,warnText:P}),[Se,oe]=l.useState(!1),De=ee(`${e}--dropdown`,{[`${e}--dropdown--invalid`]:r.invalid,[`${e}--dropdown--warning`]:r.warn,[`${e}--dropdown--open`]:g,[`${e}--dropdown--focus`]:Se,[`${e}--dropdown--inline`]:T,[`${e}--dropdown--disabled`]:r.disabled,[`${e}--dropdown--light`]:K,[`${e}--dropdown--readonly`]:w,[`${e}--dropdown--${D}`]:D,[`${e}--list-box--up`]:z==="top",[`${e}--autoalign`]:o}),Fe=ee(`${e}--label`,{[`${e}--label--disabled`]:r.disabled,[`${e}--visually-hidden`]:m}),Me=ee(`${e}--form__helper-text`,{[`${e}--form__helper-text--disabled`]:r.disabled}),Pe=ee(`${e}--dropdown__wrapper`,`${e}--list-box__wrapper`,E,{[`${e}--dropdown__wrapper--inline`]:T,[`${e}--list-box__wrapper--inline`]:T,[`${e}--dropdown__wrapper--inline--invalid`]:T&&r.invalid,[`${e}--list-box__wrapper--inline--invalid`]:T&&r.invalid,[`${e}--list-box__wrapper--fluid--invalid`]:t&&r.invalid,[`${e}--list-box__wrapper--slug`]:v,[`${e}--list-box__wrapper--decorator`]:h}),U=Ce({"aria-label":I||Q}),pe=O&&!t?p.createElement("div",{id:r.helperId,className:Me},O):null,he=a=>{oe(a.type==="focus"&&!x)},fe=l.useRef(null),ze=He(U.ref,Y,fe),[V,ge]=l.useState(),[be,de]=l.useState(!1),xe=l.useCallback(a=>{["ArrowDown","ArrowUp"," ","Enter"].includes(a.key)?be&&a.key===" "&&(V&&clearTimeout(V),ge(setTimeout(()=>{de(!1)},3e3))):(de(!0),V&&clearTimeout(V),ge(setTimeout(()=>{de(!1)},3e3))),["ArrowDown"].includes(a.key)&&oe(!1),["Enter"].includes(a.key)&&!x&&!g&&oe(!0),U.onKeyDown&&(a.key!=="ArrowUp"||g&&a.key==="ArrowUp")&&U.onKeyDown(a)},[be,V,U]),Re=l.useMemo(()=>w?{onClick:a=>{var c;a.preventDefault(),(c=fe.current)==null||c.focus()},onKeyDown:a=>{["ArrowDown","ArrowUp"," ","Enter"].includes(a.key)&&a.preventDefault()}}:{onKeyDown:xe},[w,xe]),Ae=l.useMemo(()=>me({ref:C||o?j.setFloating:null}),[o,me,j.setFloating,C]),ce=v??h,ye=qe(ce,Ze)?l.cloneElement(ce,{size:"mini"}):ce,we=ke(),Le=l.isValidElement(B)?{id:we.id}:we;return p.createElement("div",te({className:Pe},re),B&&p.createElement("label",te({className:Fe},Le),B),p.createElement(ae,{onFocus:he,onBlur:he,size:D,className:De,invalid:r.invalid,invalidText:$,invalidTextId:r.invalidId,warn:r.warn,warnText:P,warnTextId:r.warnId,light:K,isOpen:g,ref:C||o?j.setReference:null,id:F},r.invalid&&p.createElement(Ue,{className:`${e}--list-box__invalid-icon`}),r.warn&&p.createElement(Ve,{className:`${e}--list-box__invalid-icon ${e}--list-box__invalid-icon--warning`}),p.createElement("button",te({type:"button",className:`${e}--list-box__field`,disabled:r.disabled,"aria-disabled":w?!0:void 0,"aria-describedby":!T&&!r.invalid&&!r.warn&&pe?r.helperId:r.invalid?r.invalidId:r.warn?r.warnId:void 0,title:x&&b!==void 0?b(x):_e(A)},U,Re,{ref:ze}),p.createElement("span",{className:`${e}--list-box__label`},x?N?N(x):b(x):A),p.createElement(ae.MenuIcon,{isOpen:g,translateWithId:M})),v?ye:h?p.createElement("div",{className:`${e}--list-box__inner-wrapper--decorator`},ye):"",p.createElement(ae.Menu,Ae,g&&f.map((a,c)=>{const _=Te({item:a,index:c}),Z=b(a);return p.createElement(ae.MenuItem,te({key:_.id,isActive:x===a,isHighlighted:Ee===c,title:Z,disabled:_["aria-disabled"]},_),L?L(a):b(a),x===a&&p.createElement(Qe,{className:`${e}--list-box__menu-item__selected-icon`}))}))),!T&&!t&&!r.validation&&pe)});ie.displayName="Dropdown";ie.propTypes={"aria-label":i.string,ariaLabel:ue(i.string),autoAlign:i.bool,className:i.string,decorator:i.node,direction:i.oneOf(["top","bottom"]),disabled:i.bool,downshiftProps:i.object,helperText:i.node,hideLabel:i.bool,id:i.string.isRequired,initialSelectedItem:i.oneOfType([i.object,i.string,i.number]),invalid:i.bool,invalidText:i.node,itemToElement:i.func,itemToString:i.func,items:i.array.isRequired,label:i.node.isRequired,light:ue(i.bool),onChange:i.func,readOnly:i.bool,renderSelectedItem:i.func,selectedItem:i.oneOfType([i.object,i.string,i.number]),size:pt,slug:ue(i.node),titleText:i.node.isRequired,translateWithId:i.func,type:mt,warn:i.bool,warnText:i.node};const vt=({benchmarkId:o,style:E,onSelectPipeline:h,onOpenToolkitInsights:S,onOpenPipelineCompare:z,onOpenProfileCompare:R,onOpenErrorAnalysis:A})=>{const[I,Q]=l.useState(null),[b,L]=l.useState(!0),[N,X]=l.useState(null),[D,y]=l.useState(1),[F,B]=l.useState(10),[m,O]=l.useState(null),[M,K]=l.useState("overall");if(l.useEffect(()=>{(async()=>{try{L(!0),X(null),Q(null);const n=await(await ct(ut(`/api/benchmarks/${o}/summary/by-category`))).json();Q(n),O(n.default_sort_metric),K("overall"),y(1)}catch(t){X(t.message||"Failed to load benchmark summary")}finally{L(!1)}})()},[o]),N)return s.jsx(ve,{kind:"error",title:`Error loading summary for ${o}`,subtitle:N,lowContrast:!0});if(b)return s.jsxs("div",{style:{...E,display:"flex",flexDirection:"column",gap:"1rem"},children:[s.jsxs("h3",{style:{margin:0},children:[o," – Summary"]}),s.jsx(et,{description:`Loading benchmark summary for ${o}…`,status:"active"}),s.jsx(tt,{role:"progressbar",columnCount:5,rowCount:8})]});if(!I)return null;const G=M==="overall"?I.overall:I.categories[M]||[],$=new Set;G.forEach(e=>{Object.keys(e.metrics).forEach(t=>{t!=="num_records"&&t!=="num_evaluated"&&$.add(t)})});const ne=[{key:"name",header:"Pipeline"},...Array.from($).map(e=>({key:e,header:e}))],P=[...G].sort((e,t)=>{var d,f,k,g;if(!m)return 0;const n=((f=(d=e.metrics)==null?void 0:d[m])==null?void 0:f.average)??0;return(((g=(k=t.metrics)==null?void 0:k[m])==null?void 0:g.average)??0)-n}),W=e=>{var n;if(!m)return 0;const t=(n=e.metrics)==null?void 0:n[m];return typeof t=="number"?t:t&&typeof t=="object"&&typeof t.average=="number"?t.average:0},H=e=>{var n;if(!m)return 0;const t=(n=e.metrics)==null?void 0:n[m];return t&&typeof t=="object"&&typeof t.stddev=="number"?Math.max(0,t.stddev):0},J=e=>{var d;if(!m)return{low:0,high:0,hasCI:!1};const t=(d=e.metrics)==null?void 0:d[m];if(t&&typeof t=="object"&&typeof t.ci95_low=="number"&&typeof t.ci95_high=="number")return{low:t.ci95_low,high:t.ci95_high,hasCI:!0};const n=W(e),u=H(e);return{low:Math.max(0,n-u),high:n+u,hasCI:!1}},w=P.map(e=>({name:e.name,value:W(e),stddev:H(e),ci:J(e)})),v=w.reduce((e,t)=>Math.max(e,t.value,t.ci.high),0)||1,re=P.length,Y=(D-1)*F,C=Y+F,j=P.slice(Y,C),q=Array.from($),le=["overall",...Object.keys(I.categories)];return s.jsxs("div",{style:{...E,display:"flex",flexDirection:"column",gap:"0.5rem"},children:[I.has_full_results===!1&&s.jsx(ve,{kind:"info",title:"Summary-only mode",subtitle:`The full evaluation results file (${o}-predictions_eval.json) was not found. Overall pipeline metrics are shown from the summary file, but category breakdown and error analysis are unavailable. Download pre-computed results with: text2sql-eval-toolkit results fetch --benchmarks ${o} or run the evaluation pipeline locally (or fix TEXT2SQL_DATA_ROOT to a directory that already contains this file).`,lowContrast:!0}),s.jsxs("div",{style:{display:"flex",gap:"1rem",alignItems:"center"},children:[s.jsxs("h3",{style:{margin:0},children:[o," – Summary"]}),s.jsxs("div",{style:{marginLeft:"auto",display:"flex",gap:"0.75rem",alignItems:"center"},children:[S&&s.jsx(se,{kind:"ghost",size:"sm",onClick:()=>S(),children:"Metric Insights"}),z&&s.jsx(se,{kind:"ghost",size:"sm",onClick:()=>z(),children:"Pipeline Compare"}),R&&s.jsx(se,{kind:"ghost",size:"sm",onClick:()=>R(),children:"Profile Compare"}),A&&s.jsx(se,{kind:"ghost",size:"sm",onClick:()=>A(),children:"Error Analysis"}),s.jsx("div",{style:{minWidth:"220px"},children:s.jsx(ie,{id:"category-dropdown",titleText:"Query Category",label:M,items:le,selectedItem:M,onChange:e=>{K(e.selectedItem),y(1)}})}),q.length>0&&s.jsx("div",{style:{minWidth:"260px"},children:s.jsx(ie,{id:"sort-metric-dropdown",titleText:"Sort by metric",label:m||"Select metric",items:q,selectedItem:m,onChange:e=>{O(e.selectedItem),y(1)}})})]})]}),s.jsxs("div",{style:{border:"1px solid rgba(255,255,255,0.12)",borderRadius:"6px",padding:"0.75rem",background:"rgba(255,255,255,0.02)"},children:[s.jsxs("div",{style:{marginBottom:"0.5rem",fontWeight:600},children:["Pipeline performance by ",m||"selected metric"]}),s.jsx("div",{style:{maxHeight:"260px",overflow:"auto",display:"flex",flexDirection:"column",gap:"0.45rem"},children:w.map(e=>{const t=Math.max(2,e.value/v*100),n=Math.max(0,e.ci.low),u=Math.min(v,e.ci.high),d=n/v*100,f=u/v*100,k=Math.max(0,f-d),g=f>=99.9;return s.jsxs("div",{title:e.ci.hasCI?`${e.name}
+import{R as p,u as Ne,a as Be,r as l,b as Oe,F as Ke,c as We,d as ee,m as He,i as qe,_ as te,W as Ue,e as Ve,C as Qe,f as Xe,s as Ge,g as Je,h as Ye,A as Ze,P as i,j as ue,k as s,I as ve,l as et,D as tt,B as se,n as st,T as at,o as it,p as nt,q as je,t as rt,v as lt,w as ot,x as dt,y as ct,z as ut}from"./index-DLyduwv0.js";import{d as _e,u as $e,L as ae,a as mt,b as pt}from"./defaultItemToString-ClUBp4QC.js";const{ItemMouseMove:ht,MenuMouseLeave:ft,ToggleButtonBlur:gt,FunctionCloseMenu:bt}=$e.stateChangeTypes;function Ie(o,E){const{changes:h,type:S}=E;switch(S){case ht:return o;case ft:return h.highlightedIndex===o.highlightedIndex?o:h;case gt:case bt:return{...h,selectedItem:o.selectedItem};default:return h}}const ie=p.forwardRef(({autoAlign:o=!1,className:E,decorator:h,disabled:S=!1,direction:z="bottom",items:R,label:A,["aria-label"]:I,ariaLabel:Q,itemToString:b=_e,itemToElement:L=null,renderSelectedItem:N,type:X="default",size:D,onChange:y,id:F,titleText:B="",hideLabel:m,helperText:O="",translateWithId:M,light:K,invalid:G,invalidText:$,warn:ne,warnText:P,initialSelectedItem:W,selectedItem:H,downshiftProps:J,readOnly:w,slug:v,...re},Y)=>{const C=Ne("enable-v12-dynamic-floating-styles"),{refs:j,floatingStyles:q,middlewareData:le}=Be(C||o?{placement:z,strategy:"fixed",middleware:[Ge({apply({rects:a,elements:c}){Object.assign(c.floating.style,{width:`${a.reference.width}px`})}}),o&&Je(),o&&Ye()],whileElementsMounted:Xe}:{});l.useEffect(()=>{var a;if(C||o){const c={...q,visibility:(a=le.hide)!=null&&a.referenceHidden?"hidden":"visible"};Object.keys(c).forEach(_=>{j.floating.current&&(j.floating.current.style[_]=c[_])})}},[q,o,j.floating]);const e=Oe(),{isFluid:t}=l.useContext(Ke),n=l.useCallback(({selectedItem:a})=>{y&&y({selectedItem:a??null})},[y]),u=l.useCallback((a,c)=>a!==null&&typeof a=="object"&&"disabled"in a&&a.disabled===!0,[]),d=l.useCallback(a=>{const{highlightedIndex:c}=a;if(c!==void 0&&c>-1){const Z=document.querySelectorAll(`li.${e}--list-box__menu-item[role="option"]`)[c];Z&&Z.scrollIntoView({behavior:"smooth",block:"nearest"})}},[e]),f=l.useMemo(()=>R,[R]),k=l.useMemo(()=>({items:f,itemToString:b,initialSelectedItem:W,onSelectedItemChange:n,stateReducer:Ie,isItemDisabled:u,onHighlightedIndexChange:d,...J}),[f,b,W,n,Ie,u,d,J]);H!==void 0&&(k.selectedItem=H);const{isOpen:g,getToggleButtonProps:Ce,getLabelProps:ke,getMenuProps:me,getItemProps:Te,selectedItem:x,highlightedIndex:Ee}=$e(k),T=X==="inline",r=We({id:F,readOnly:w,disabled:S??!1,invalid:G??!1,invalidText:$,warn:ne??!1,warnText:P}),[Se,oe]=l.useState(!1),De=ee(`${e}--dropdown`,{[`${e}--dropdown--invalid`]:r.invalid,[`${e}--dropdown--warning`]:r.warn,[`${e}--dropdown--open`]:g,[`${e}--dropdown--focus`]:Se,[`${e}--dropdown--inline`]:T,[`${e}--dropdown--disabled`]:r.disabled,[`${e}--dropdown--light`]:K,[`${e}--dropdown--readonly`]:w,[`${e}--dropdown--${D}`]:D,[`${e}--list-box--up`]:z==="top",[`${e}--autoalign`]:o}),Fe=ee(`${e}--label`,{[`${e}--label--disabled`]:r.disabled,[`${e}--visually-hidden`]:m}),Me=ee(`${e}--form__helper-text`,{[`${e}--form__helper-text--disabled`]:r.disabled}),Pe=ee(`${e}--dropdown__wrapper`,`${e}--list-box__wrapper`,E,{[`${e}--dropdown__wrapper--inline`]:T,[`${e}--list-box__wrapper--inline`]:T,[`${e}--dropdown__wrapper--inline--invalid`]:T&&r.invalid,[`${e}--list-box__wrapper--inline--invalid`]:T&&r.invalid,[`${e}--list-box__wrapper--fluid--invalid`]:t&&r.invalid,[`${e}--list-box__wrapper--slug`]:v,[`${e}--list-box__wrapper--decorator`]:h}),U=Ce({"aria-label":I||Q}),pe=O&&!t?p.createElement("div",{id:r.helperId,className:Me},O):null,he=a=>{oe(a.type==="focus"&&!x)},fe=l.useRef(null),ze=He(U.ref,Y,fe),[V,ge]=l.useState(),[be,de]=l.useState(!1),xe=l.useCallback(a=>{["ArrowDown","ArrowUp"," ","Enter"].includes(a.key)?be&&a.key===" "&&(V&&clearTimeout(V),ge(setTimeout(()=>{de(!1)},3e3))):(de(!0),V&&clearTimeout(V),ge(setTimeout(()=>{de(!1)},3e3))),["ArrowDown"].includes(a.key)&&oe(!1),["Enter"].includes(a.key)&&!x&&!g&&oe(!0),U.onKeyDown&&(a.key!=="ArrowUp"||g&&a.key==="ArrowUp")&&U.onKeyDown(a)},[be,V,U]),Re=l.useMemo(()=>w?{onClick:a=>{var c;a.preventDefault(),(c=fe.current)==null||c.focus()},onKeyDown:a=>{["ArrowDown","ArrowUp"," ","Enter"].includes(a.key)&&a.preventDefault()}}:{onKeyDown:xe},[w,xe]),Ae=l.useMemo(()=>me({ref:C||o?j.setFloating:null}),[o,me,j.setFloating,C]),ce=v??h,ye=qe(ce,Ze)?l.cloneElement(ce,{size:"mini"}):ce,we=ke(),Le=l.isValidElement(B)?{id:we.id}:we;return p.createElement("div",te({className:Pe},re),B&&p.createElement("label",te({className:Fe},Le),B),p.createElement(ae,{onFocus:he,onBlur:he,size:D,className:De,invalid:r.invalid,invalidText:$,invalidTextId:r.invalidId,warn:r.warn,warnText:P,warnTextId:r.warnId,light:K,isOpen:g,ref:C||o?j.setReference:null,id:F},r.invalid&&p.createElement(Ue,{className:`${e}--list-box__invalid-icon`}),r.warn&&p.createElement(Ve,{className:`${e}--list-box__invalid-icon ${e}--list-box__invalid-icon--warning`}),p.createElement("button",te({type:"button",className:`${e}--list-box__field`,disabled:r.disabled,"aria-disabled":w?!0:void 0,"aria-describedby":!T&&!r.invalid&&!r.warn&&pe?r.helperId:r.invalid?r.invalidId:r.warn?r.warnId:void 0,title:x&&b!==void 0?b(x):_e(A)},U,Re,{ref:ze}),p.createElement("span",{className:`${e}--list-box__label`},x?N?N(x):b(x):A),p.createElement(ae.MenuIcon,{isOpen:g,translateWithId:M})),v?ye:h?p.createElement("div",{className:`${e}--list-box__inner-wrapper--decorator`},ye):"",p.createElement(ae.Menu,Ae,g&&f.map((a,c)=>{const _=Te({item:a,index:c}),Z=b(a);return p.createElement(ae.MenuItem,te({key:_.id,isActive:x===a,isHighlighted:Ee===c,title:Z,disabled:_["aria-disabled"]},_),L?L(a):b(a),x===a&&p.createElement(Qe,{className:`${e}--list-box__menu-item__selected-icon`}))}))),!T&&!t&&!r.validation&&pe)});ie.displayName="Dropdown";ie.propTypes={"aria-label":i.string,ariaLabel:ue(i.string),autoAlign:i.bool,className:i.string,decorator:i.node,direction:i.oneOf(["top","bottom"]),disabled:i.bool,downshiftProps:i.object,helperText:i.node,hideLabel:i.bool,id:i.string.isRequired,initialSelectedItem:i.oneOfType([i.object,i.string,i.number]),invalid:i.bool,invalidText:i.node,itemToElement:i.func,itemToString:i.func,items:i.array.isRequired,label:i.node.isRequired,light:ue(i.bool),onChange:i.func,readOnly:i.bool,renderSelectedItem:i.func,selectedItem:i.oneOfType([i.object,i.string,i.number]),size:pt,slug:ue(i.node),titleText:i.node.isRequired,translateWithId:i.func,type:mt,warn:i.bool,warnText:i.node};const vt=({benchmarkId:o,style:E,onSelectPipeline:h,onOpenToolkitInsights:S,onOpenPipelineCompare:z,onOpenProfileCompare:R,onOpenErrorAnalysis:A})=>{const[I,Q]=l.useState(null),[b,L]=l.useState(!0),[N,X]=l.useState(null),[D,y]=l.useState(1),[F,B]=l.useState(10),[m,O]=l.useState(null),[M,K]=l.useState("overall");if(l.useEffect(()=>{(async()=>{try{L(!0),X(null),Q(null);const n=await(await ct(ut(`/api/benchmarks/${o}/summary/by-category`))).json();Q(n),O(n.default_sort_metric),K("overall"),y(1)}catch(t){X(t.message||"Failed to load benchmark summary")}finally{L(!1)}})()},[o]),N)return s.jsx(ve,{kind:"error",title:`Error loading summary for ${o}`,subtitle:N,lowContrast:!0});if(b)return s.jsxs("div",{style:{...E,display:"flex",flexDirection:"column",gap:"1rem"},children:[s.jsxs("h3",{style:{margin:0},children:[o," – Summary"]}),s.jsx(et,{description:`Loading benchmark summary for ${o}…`,status:"active"}),s.jsx(tt,{role:"progressbar",columnCount:5,rowCount:8})]});if(!I)return null;const G=M==="overall"?I.overall:I.categories[M]||[],$=new Set;G.forEach(e=>{Object.keys(e.metrics).forEach(t=>{t!=="num_records"&&t!=="num_evaluated"&&$.add(t)})});const ne=[{key:"name",header:"Pipeline"},...Array.from($).map(e=>({key:e,header:e}))],P=[...G].sort((e,t)=>{var d,f,k,g;if(!m)return 0;const n=((f=(d=e.metrics)==null?void 0:d[m])==null?void 0:f.average)??0;return(((g=(k=t.metrics)==null?void 0:k[m])==null?void 0:g.average)??0)-n}),W=e=>{var n;if(!m)return 0;const t=(n=e.metrics)==null?void 0:n[m];return typeof t=="number"?t:t&&typeof t=="object"&&typeof t.average=="number"?t.average:0},H=e=>{var n;if(!m)return 0;const t=(n=e.metrics)==null?void 0:n[m];return t&&typeof t=="object"&&typeof t.stddev=="number"?Math.max(0,t.stddev):0},J=e=>{var d;if(!m)return{low:0,high:0,hasCI:!1};const t=(d=e.metrics)==null?void 0:d[m];if(t&&typeof t=="object"&&typeof t.ci95_low=="number"&&typeof t.ci95_high=="number")return{low:t.ci95_low,high:t.ci95_high,hasCI:!0};const n=W(e),u=H(e);return{low:Math.max(0,n-u),high:n+u,hasCI:!1}},w=P.map(e=>({name:e.name,value:W(e),stddev:H(e),ci:J(e)})),v=w.reduce((e,t)=>Math.max(e,t.value,t.ci.high),0)||1,re=P.length,Y=(D-1)*F,C=Y+F,j=P.slice(Y,C),q=Array.from($),le=["overall",...Object.keys(I.categories)];return s.jsxs("div",{style:{...E,display:"flex",flexDirection:"column",gap:"0.5rem"},children:[I.has_full_results===!1&&s.jsx(ve,{kind:"info",title:"Summary-only mode",subtitle:`The full evaluation results file (${o}-predictions_eval.json) was not found. Overall pipeline metrics are shown from the summary file, but category breakdown and error analysis are unavailable. Download pre-computed results with: text2sql-eval-toolkit results fetch --benchmarks ${o} or run the evaluation pipeline locally (or fix TEXT2SQL_DATA_ROOT to a directory that already contains this file).`,lowContrast:!0}),s.jsxs("div",{style:{display:"flex",gap:"1rem",alignItems:"center"},children:[s.jsxs("h3",{style:{margin:0},children:[o," – Summary"]}),s.jsxs("div",{style:{marginLeft:"auto",display:"flex",gap:"0.75rem",alignItems:"center"},children:[S&&s.jsx(se,{kind:"ghost",size:"sm",onClick:()=>S(),children:"Metric Insights"}),z&&s.jsx(se,{kind:"ghost",size:"sm",onClick:()=>z(),children:"Pipeline Compare"}),R&&s.jsx(se,{kind:"ghost",size:"sm",onClick:()=>R(),children:"Profile Compare"}),A&&s.jsx(se,{kind:"ghost",size:"sm",onClick:()=>A(),children:"Error Analysis"}),s.jsx("div",{style:{minWidth:"220px"},children:s.jsx(ie,{id:"category-dropdown",titleText:"Query Category",label:M,items:le,selectedItem:M,onChange:e=>{K(e.selectedItem),y(1)}})}),q.length>0&&s.jsx("div",{style:{minWidth:"260px"},children:s.jsx(ie,{id:"sort-metric-dropdown",titleText:"Sort by metric",label:m||"Select metric",items:q,selectedItem:m,onChange:e=>{O(e.selectedItem),y(1)}})})]})]}),s.jsxs("div",{style:{border:"1px solid rgba(255,255,255,0.12)",borderRadius:"6px",padding:"0.75rem",background:"rgba(255,255,255,0.02)"},children:[s.jsxs("div",{style:{marginBottom:"0.5rem",fontWeight:600},children:["Pipeline performance by ",m||"selected metric"]}),s.jsx("div",{style:{maxHeight:"260px",overflow:"auto",display:"flex",flexDirection:"column",gap:"0.45rem"},children:w.map(e=>{const t=Math.max(2,e.value/v*100),n=Math.max(0,e.ci.low),u=Math.min(v,e.ci.high),d=n/v*100,f=u/v*100,k=Math.max(0,f-d),g=f>=99.9;return s.jsxs("div",{title:e.ci.hasCI?`${e.name}
 Mean: ${e.value.toFixed(3)}
 95% CI: [${e.ci.low.toFixed(3)}, ${e.ci.high.toFixed(3)}]`:`${e.name}
 Mean: ${e.value.toFixed(3)}`,style:{display:"grid",gridTemplateColumns:"260px 1fr 56px",gap:"0.6rem",alignItems:"center"},children:[s.jsx("div",{title:e.name,style:{whiteSpace:"nowrap",overflow:"hidden",textOverflow:"ellipsis",fontSize:"0.82rem",opacity:.9},children:e.name}),s.jsxs("div",{style:{height:"12px",background:"rgba(255,255,255,0.1)",borderRadius:"999px",overflow:"hidden",position:"relative"},children:[s.jsx("div",{style:{width:`${t}%`,height:"100%",background:"linear-gradient(90deg, #0f62fe, #78a9ff)",borderRadius:"999px"}}),u>n&&s.jsxs(s.Fragment,{children:[s.jsx("div",{style:{position:"absolute",left:`${d}%`,width:`${k}%`,top:"5.5px",height:"1px",background:"rgba(255,255,255,0.7)"}}),s.jsx("div",{style:{position:"absolute",left:`${d}%`,top:"3px",width:"1px",height:"6px",background:"rgba(255,255,255,0.7)"}}),s.jsx("div",{style:{position:"absolute",...g?{right:0}:{left:`${f}%`},top:"3px",width:"1px",height:"6px",background:"rgba(255,255,255,0.7)"}})]})]}),s.jsx("div",{style:{textAlign:"right",fontSize:"0.82rem"},children:e.value.toFixed(3)})]},`chart-${e.name}`)})})]}),s.jsxs("div",{style:{maxHeight:"360px",overflow:"auto"},children:[s.jsx("div",{style:{fontSize:"0.82rem",opacity:.85,marginBottom:"0.45rem"},children:"Tip: click a pipeline row in the table below for detailed pipeline analysis."}),s.jsx(st,{rows:j.map(e=>({id:e.name,name:e.name,...Object.fromEntries(Array.from($).map(t=>{const n=e.metrics[t];return n&&typeof n=="object"&&"average"in n?[t,n.average.toFixed(3)]:[t,n??""]}))})),headers:ne,size:"sm",children:({rows:e,headers:t,getHeaderProps:n})=>s.jsx(at,{children:s.jsxs(it,{"aria-label":"Pipeline summary",children:[s.jsx(nt,{children:s.jsx(je,{children:t.map(u=>{const{key:d,...f}=n({header:u});return s.jsx(rt,{...f,children:u.header},d)})})}),s.jsx(lt,{children:e.map(u=>s.jsx(je,{style:{cursor:h?"pointer":"default"},onClick:()=>h==null?void 0:h(u.id),children:u.cells.map(d=>s.jsx(ot,{children:d.value},d.id))},u.id))})]})})})]}),s.jsx(dt,{page:D,pageSize:F,pageSizes:[10,25,50],totalItems:re,onChange:({page:e,pageSize:t})=>{y(e),B(t)}})]})};export{vt as BenchmarkDetail};

text2sql_eval_toolkit-1.4.0/dashboard/dist/assets/ComboBox-BActzty7.js

@@ -0,0 +1 @@
+import{b as Ce,d as M,R as r,ag as qe,_ as z,P as n,a6 as Vt,r as o,u as kt,a as Lt,F as Mt,Q as Ft,ah as j,c as Pt,i as At,X as Ue,ai as X,m as Wt,W as Ot,e as Bt,C as jt,f as zt,g as Ht,h as Kt,A as qt,a9 as Ut,a8 as Qt,aj as Xt,ak as Gt,al as Jt,j as xe}from"./index-DLyduwv0.js";import{d as tt,c as nt,L as we,b as Yt}from"./defaultItemToString-ClUBp4QC.js";var Qe,Xe;const Ge={"clear.all":"clear.all","clear.selection":"clear.selection"},Zt={[Ge["clear.all"]]:"Clear all selected items",[Ge["clear.selection"]]:"Clear selected item"},vt=d=>Zt[d];function lt({clearSelection:d,selectionCount:c,translateWithId:h=vt,disabled:f,readOnly:g,onClearSelection:_,...F}){const I=Ce(),H=M(`${I}--list-box__selection`,{[`${I}--tag--filter`]:c,[`${I}--list-box__selection--multi`]:c}),P=h(c?"clear.all":"clear.selection"),$=M(`${I}--tag`,`${I}--tag--filter`,`${I}--tag--high-contrast`,{[`${I}--tag--disabled`]:f});function K(C){C.stopPropagation(),!(f||g)&&(d(C),_&&_(C))}return c?r.createElement("div",{className:$},r.createElement("span",{className:`${I}--tag__label`,title:c==null?void 0:c.toString()},c),r.createElement("button",{"aria-label":P,className:`${I}--tag__close-icon`,disabled:f||g,onClick:K,tabIndex:-1,title:P,type:"button","aria-disabled":g?!0:void 0},Qe||(Qe=r.createElement(qe,null)))):r.createElement("button",z({},F,{"aria-label":P,className:H,disabled:f||g,onClick:K,tabIndex:-1,title:P,type:"button","aria-disabled":g?!0:void 0}),Xe||(Xe=r.createElement(qe,null)))}lt.propTypes={clearSelection:n.func.isRequired,disabled:n.bool,readOnly:n.bool,onClearSelection:n.func,onClick:n.func,onMouseUp:n.func,onKeyDown:n.func,selectionCount:n.number,translateWithId:n.func};var Je;const Ye={"close.menu":"close.menu","open.menu":"open.menu"},en={[Ye["close.menu"]]:"Close",[Ye["open.menu"]]:"Open"},tn=d=>en[d],at=r.forwardRef(({isOpen:d,translateWithId:c=tn,...h},f)=>{const g=Ce(),_=M({[`${g}--list-box__menu-icon`]:!0,[`${g}--list-box__menu-icon--open`]:d}),F=c(d?"close.menu":"open.menu");return r.createElement("button",z({},h,{"aria-label":F,title:F,className:_,type:"button",tabIndex:-1,ref:f}),Je||(Je=r.createElement(Vt,null)))});at.propTypes={isOpen:n.bool.isRequired,translateWithId:n.func};const{InputBlur:nn,InputKeyDownEnter:Ze,FunctionToggleMenu:ln,ToggleButtonClick:an,ItemMouseMove:sn,InputKeyDownArrowUp:rn,InputKeyDownArrowDown:on,MenuMouseLeave:cn,ItemClick:un,FunctionSelectItem:dn}=nt.stateChangeTypes,ve=()=>!0,ye=({item:d,inputValue:c})=>{if(c===null||c==="")return!0;const h=d.toLowerCase(),f=c.toLowerCase();return h.startsWith(f)},et=({initialSelectedItem:d,itemToString:c,selectedItem:h,prevSelectedItem:f})=>h!==null&&typeof h<"u"?c(h):typeof f>"u"&&d!==null&&typeof d<"u"?c(d):"",fn=({items:d,itemToString:c=tt},h)=>{if(!h)return-1;const f=h.toLowerCase();for(let g=0;g<d.length;g++){const _=c(d[g]).toLowerCase();if(!d[g].disabled&&_.indexOf(f)!==-1)return g}return-1},st=o.forwardRef((d,c)=>{var He,Ke;const h=o.useRef(0),f=o.useRef(null),{["aria-label"]:g="Choose an item",ariaLabel:_,autoAlign:F=!1,className:I,decorator:H,direction:P="bottom",disabled:$=!1,downshiftActions:K,downshiftProps:C,helperText:ee,id:Ee,initialSelectedItem:te,invalid:ne,invalidText:le,items:b,itemToElement:ae=null,itemToString:s=tt,light:it,onChange:R,onInputChange:se,onToggleClick:Te,placeholder:rt,readOnly:A,selectedItem:E,shouldFilterItem:_e=ve,size:ot,titleText:ie,translateWithId:$e,typeahead:G=!1,warn:ct,warnText:re,allowCustomValue:N=!1,slug:oe,inputProps:ut,...dt}=d,W=kt("enable-v12-dynamic-floating-styles")||F,{refs:x,floatingStyles:Se,middlewareData:ft}=Lt(W?{placement:P,strategy:"fixed",middleware:F?[Ht(),Kt()]:void 0,whileElementsMounted:zt}:{}),De=(He=x==null?void 0:x.reference)==null?void 0:He.current,ce=De instanceof HTMLElement?De.clientWidth:void 0;o.useEffect(()=>{var e;if(W){const t={...Se,visibility:(e=ft.hide)!=null&&e.referenceHidden?"hidden":"visible"};Object.keys(t).forEach(u=>{x.floating.current&&(x.floating.current.style[u]=t[u])}),ce&&x.floating.current&&(x.floating.current.style.width=ce+"px")}},[W,Se,x.floating,ce]);const[a,J]=o.useState(et({initialSelectedItem:te,itemToString:s,selectedItem:E})),[Y,Z]=o.useState("");o.useEffect(()=>{if(G){if(a.length>=h.current)if(a){const e=b.filter(t=>ye({item:s(t),inputValue:a}));if(e.length>0){const t=s(e[0]);Z(t.slice(a.length))}else Z("")}else Z("");else Z("");h.current=a.length}},[G,a,b,s,ye]);const Re=o.useRef(!1),O=o.useRef(""),[ue,de]=o.useState(!1),i=Ce(),{isFluid:fe}=o.useContext(Mt),w=o.useRef(null),me=Ft(),[Ne,mt]=o.useState(!1),Ve=o.useRef(a),pe=o.useRef(E);o.useEffect(()=>{Re.current=ue,ue&&de(!1)},[ue]),o.useEffect(()=>{if(pe.current!==E){const e=et({initialSelectedItem:te,itemToString:s,selectedItem:E,prevSelectedItem:pe.current});a!==e&&(J(e),R({selectedItem:E,inputValue:e})),pe.current=E}},[E]);const q=(e,t,u)=>e.filter(l=>G?ye({item:t(l),inputValue:u}):_e?_e({item:l,itemToString:t,inputValue:u}):ve());o.useEffect(()=>{Ve.current!==a&&(Ve.current=a,se==null||se(a))},[a]);const ke=()=>{w!=null&&w.current&&w.current.focus()},pt=e=>q(b,s,e||null),Le=e=>fn({...d,items:pt(e)},e),gt=o.useCallback((e,t)=>{const{type:u,changes:l}=t,{highlightedIndex:y}=l;switch(u){case nn:{if(N&&y===-1){const m=e.inputValue??"",p=typeof l.selectedItem>"u"?e.selectedItem:l.selectedItem;if(p!==null&&typeof p<"u"&&s(p)===m&&b.some(L=>j(L,p)))return l;const k=m===""?null:b.find(L=>s(L)===m)??m,Q=typeof k=="string"&&k!==""&&!b.some(L=>j(L,k));return!j(p,k)&&R&&(R({selectedItem:k,inputValue:m}),O.current=Q?m:""),{...l,selectedItem:k}}if(e.inputValue&&y===-1&&l.selectedItem)return{...l,inputValue:s(l.selectedItem)};if(!N){const m=e.inputValue??"";if(!(!!m&&b.some(D=>s(D)===m))){const D=e.selectedItem!==null?s(e.selectedItem):"";return{...l,inputValue:D}}}return l}case Ze:if(!N)if(e.highlightedIndex!==-1){const p=q(b,s,a)[e.highlightedIndex];if(p&&!p.disabled)return{...l,selectedItem:p,inputValue:s(p)}}else{const m=Le(a);if(m!==-1){const p=b[m];if(p&&!p.disabled)return{...l,selectedItem:p,inputValue:s(p)}}if(e.selectedItem!==null)return{...l,selectedItem:null,inputValue:a}}return{...l,isOpen:!0};case ln:case an:if(e.isOpen&&!l.isOpen&&!N){const m=e.inputValue??"";if(!(!!m&&b.some(D=>s(D)===m))){const D=e.selectedItem!==null?s(e.selectedItem):"";return{...l,inputValue:D}}}return l;case cn:return{...l,highlightedIndex:e.highlightedIndex};case rn:case on:return y===-1?{...l,highlightedIndex:0}:l;case sn:return{...l,highlightedIndex:e.highlightedIndex};default:return l}},[N,a,s,b,R]),ht=e=>t=>{var u,l;if(Te&&Te(t),A){t.preventDownshiftDefault=!0,(u=t==null?void 0:t.persist)==null||u.call(t);return}t.target===w.current&&e&&(t.preventDownshiftDefault=!0,(l=t==null?void 0:t.persist)==null||l.call(t))},S=Pt({id:Ee,readOnly:A,disabled:$||!1,invalid:ne||!1,invalidText:le,warn:ct||!1,warnText:re}),bt=M(`${i}--combo-box`,{[`${i}--combo-box--invalid--focused`]:ne&&Ne,[`${i}--list-box--up`]:P==="top",[`${i}--combo-box--warning`]:S.warn,[`${i}--combo-box--readonly`]:A,[`${i}--autoalign`]:W}),It=M(`${i}--label`,{[`${i}--label--disabled`]:$}),Me=`combobox-helper-text-${me}`,Fe=`combobox-warn-text-${me}`,Pe=`combobox-invalid-text-${me}`,xt=M(`${i}--form__helper-text`,{[`${i}--form__helper-text--disabled`]:$}),wt=M(`${i}--list-box__wrapper`,[I,{[`${i}--list-box__wrapper--fluid--invalid`]:fe&&ne,[`${i}--list-box__wrapper--slug`]:oe,[`${i}--list-box__wrapper--decorator`]:H}]),yt=M(`${i}--text-input`,{[`${i}--text-input--empty`]:!a,[`${i}--combo-box--input--focus`]:Ne}),ge=oe??H,Ae=At(ge,qt),he=Ae?o.cloneElement(ge,{size:"mini"}):ge,{getInputProps:Ct,getItemProps:Et,getLabelProps:Tt,getMenuProps:We,getToggleButtonProps:_t,isOpen:T,highlightedIndex:U,selectedItem:$t,closeMenu:Oe,openMenu:be,reset:Be,selectItem:V,setHighlightedIndex:Ie,setInputValue:v,toggleMenu:B}=nt({items:q(b,s,a),inputValue:a,itemToString:e=>s(e),onInputValueChange({inputValue:e}){const t=e||"";J(t),Ie(Le(t))},onHighlightedIndexChange:({highlightedIndex:e})=>{if(e>-1){const u=document.querySelectorAll(`li.${i}--list-box__menu-item[role="option"]`)[e];u&&u.scrollIntoView({behavior:"smooth",block:"nearest"})}},initialSelectedItem:te,inputId:Ee,stateReducer:gt,isItemDisabled(e,t){return e==null?void 0:e.disabled},...C,onStateChange:({type:e,selectedItem:t})=>{var u;(u=C==null?void 0:C.onStateChange)==null||u.call(C,{type:e,selectedItem:t}),!Re.current&&(e===un||e===dn||e===Ze)&&typeof t<"u"&&!j(E,t)&&(b.some(l=>j(l,t))&&(O.current=""),R({selectedItem:t}))}}),St=typeof E<"u"?E:$t;o.useEffect(()=>{K&&(K.current={closeMenu:Oe,openMenu:be,reset:Be,selectItem:V,setHighlightedIndex:Ie,setInputValue:v,toggleMenu:B})},[Oe,be,Be,V,Ie,v,B]);const Dt=_t({disabled:$||A,onClick:ht(T),onMouseUp(e){T&&e.stopPropagation()}}),je=e=>{var t;mt(e.type==="focus"),!((t=f.current)!=null&&t.value)&&e.type==="blur"&&V(null)},Rt=A?{onKeyDown:e=>{e.key!=="Tab"&&e.preventDefault()},onClick:e=>{e.preventDefault(),e.currentTarget.focus()}}:{},Nt=S.invalid&&le&&Pe||S.warn&&re&&Fe||ee&&!fe&&Me||void 0,ze=o.useMemo(()=>We({ref:W?x.setFloating:null}),[W,_,g,We,x.setFloating]);return o.useEffect(()=>{if(w.current&&f.current&&Y){const e=a.length,t=e+Y.length;f.current.value=a+Y,f.current.setSelectionRange(e,t)}},[a,Y]),r.createElement("div",{className:wt},ie&&r.createElement(Ue,z({as:"label",className:It},Tt()),ie),r.createElement(we,{onFocus:je,onBlur:je,className:bt,disabled:$,invalid:S.invalid,invalidText:le,invalidTextId:Pe,isOpen:T,light:it,size:ot,warn:S.warn,ref:W?x.setReference:null,warnText:re,warnTextId:Fe},r.createElement("div",{className:`${i}--list-box__field`},r.createElement("input",z({disabled:$,className:yt,type:"text",tabIndex:0,"aria-haspopup":"listbox",title:(Ke=w==null?void 0:w.current)==null?void 0:Ke.value},Ct({"aria-label":ie?void 0:_||g,"aria-controls":ze.id,placeholder:rt,value:a,...ut,onChange:e=>{const t=e.target.value,u=N&&O.current&&a===O.current&&t==="";J(t),v(t),u&&(de(!0),R({selectedItem:null,inputValue:""}),V(null),O.current="")},ref:Wt(w,c,f),onKeyDown:e=>{var t,u;if(X(e,Ut)&&e.stopPropagation(),X(e,Qt)&&(!a||N)&&(B(),U!==-1&&V(q(b,s,a)[U]),N&&T&&a&&U===-1&&(O.current=a,R({selectedItem:null,inputValue:a})),e.preventDownshiftDefault=!0,(t=e==null?void 0:e.persist)==null||t.call(e)),X(e,Xt)&&a&&e.target===w.current&&T&&(B(),e.preventDownshiftDefault=!0,(u=e==null?void 0:e.persist)==null||u.call(e)),X(e,Gt)&&e.code!=="Numpad7"&&e.target.setSelectionRange(0,0),X(e,Jt)&&e.code!=="Numpad1"&&e.target.setSelectionRange(e.target.value.length,e.target.value.length),e.altKey&&e.key=="ArrowDown"&&(e.preventDownshiftDefault=!0,T||B()),e.altKey&&e.key=="ArrowUp"&&(e.preventDownshiftDefault=!0,T&&B()),!a&&U==-1&&e.key=="Enter"&&(T||B(),V(null),e.preventDownshiftDefault=!0,e.currentTarget.ariaExpanded==="false"&&be()),G&&e.key==="Tab"){const l=b.find(y=>s(y).toLowerCase().startsWith(a.toLowerCase()));if(l){const y=s(l);v(y),V(l)}}}}),dt,Rt,{readOnly:A,"aria-describedby":Nt})),S.invalid&&r.createElement(Ot,{className:`${i}--list-box__invalid-icon`}),S.warn&&r.createElement(Bt,{className:`${i}--list-box__invalid-icon ${i}--list-box__invalid-icon--warning`}),a&&r.createElement(lt,{clearSelection:()=>{de(!0),J(""),R({selectedItem:null}),V(null),O.current="",ke()},translateWithId:$e,disabled:$||A,onClearSelection:ke,selectionCount:0}),r.createElement(at,z({},Dt,{isOpen:T,translateWithId:$e}))),oe?he:H?r.createElement("div",{className:`${i}--list-box__inner-wrapper--decorator`},Ae?he:r.createElement("span",null,he)):"",r.createElement(we.Menu,ze,T?q(b,s,a).map((e,t)=>{var L;const l=e!==null&&typeof e=="object"&&"text"in e&&ae?(L=e.text)==null?void 0:L.toString():s(e),y=Et({item:e,index:t}),m=y["aria-disabled"],{"aria-disabled":p,"aria-selected":D,...k}=y,Q=j(St,e);return r.createElement(we.MenuItem,z({key:y.id,isActive:Q,isHighlighted:U===t,title:l,disabled:m},k,{"aria-selected":Q}),ae?ae(e):s(e),Q&&r.createElement(jt,{className:`${i}--list-box__menu-item__selected-icon`}))}):null)),ee&&!S.invalid&&!S.warn&&!fe&&r.createElement(Ue,{as:"div",id:Me,className:xt},ee))});st.displayName="ComboBox";st.propTypes={allowCustomValue:n.bool,"aria-label":n.string,ariaLabel:xe(n.string),autoAlign:n.bool,className:n.string,decorator:n.node,direction:n.oneOf(["top","bottom"]),disabled:n.bool,downshiftProps:n.object,downshiftActions:n.exact({current:n.any}),helperText:n.node,id:n.string.isRequired,initialSelectedItem:n.oneOfType([n.object,n.string,n.number]),invalid:n.bool,invalidText:n.node,itemToElement:n.func,itemToString:n.func,items:n.array.isRequired,light:xe(n.bool),onChange:n.func.isRequired,onInputChange:n.func,onToggleClick:n.func,placeholder:n.string,readOnly:n.bool,selectedItem:n.oneOfType([n.object,n.string,n.number]),shouldFilterItem:n.func,size:Yt,slug:xe(n.node),titleText:n.node,translateWithId:n.func,typeahead:n.bool,warn:n.bool,warnText:n.node,inputProps:n.object};export{st as C};

text2sql_eval_toolkit-1.3.0/dashboard/dist/assets/ErrorAnalysis-ho3kmDX5.js → text2sql_eval_toolkit-1.4.0/dashboard/dist/assets/ErrorAnalysis-BpplWvLW.js

@@ -1 +1 @@
-import{r,k as t,B as x,I as h,E as Ae,S as ne,G as y,l as ut,D as pt,n as ft,T as mt,o as xt,p as gt,q as Ge,t as ht,v as yt,w as jt,x as bt,H as St,y as le,z as R}from"./index-C9VP9iwL.js";import{C as Ve}from"./ComboBox-Cz5Wuuu_.js";import{b as Be,f as wt,S as He}from"./metricInsightsSelect-B46SO2Wy.js";import{R as be,h as We}from"./ResultTableView-Dq7hZwq5.js";import"./defaultItemToString-BtCahKYf.js";function _t(a,M,n){const $=M.find(ae=>ae.record_id===a),U=Object.keys(($==null?void 0:$.predictions)??{});return n&&U.includes(n)?n:U[0]??null}function Ue(a){if(a==null)return"N/A";if(typeof a=="number")return Number.isFinite(a)?String(a):"N/A";if(typeof a=="boolean")return a?"true":"false";if(typeof a=="string")return a;try{return JSON.stringify(a)}catch{return String(a)}}function Xe(a,M){const n=a.trim();return n?n.replaceAll("_"," "):M}const Rt=({benchmarkId:a,onBack:M,initialFilters:n,initialPage:$,initialPageSize:U,initialRecordId:ae,onStateChange:w})=>{const[L,Se]=r.useState([]),[Ye,Ke]=r.useState(0),_=$??1,k=U??25,[we,Ze]=r.useState(""),[u,D]=r.useState(()=>(n==null?void 0:n.pipeline)??""),[j,X]=r.useState(()=>(n==null?void 0:n.metric)??"execution_accuracy"),[P,ie]=r.useState(()=>(n==null?void 0:n.value)??"0"),[q,oe]=r.useState(()=>(n==null?void 0:n.op)??"eq"),[m,I]=r.useState(()=>(n==null?void 0:n.pipeline2)??""),[b,Y]=r.useState(()=>(n==null?void 0:n.metric2)??"subset_non_empty_execution_accuracy"),[C,ce]=r.useState(()=>(n==null?void 0:n.disagree)??!1),[_e,ke]=r.useState(!0),[Ce,Te]=r.useState(null),o=ae??null,[T,de]=r.useState(null),ue=r.useCallback((e,s)=>_t(e,s,u),[u]);r.useEffect(()=>{if(!o||T||L.length===0)return;const e=ue(o,L);e&&de(e)},[o,T,L,ue]);const ve=r.useMemo(()=>({pipeline:u,metric:j,value:P,op:q,pipeline2:m,metric2:b,disagree:C}),[u,j,P,q,m,b,C]),J=r.useCallback(e=>{w==null||w({filters:ve,page:e.page??_,pageSize:e.pageSize??k,record:e.record!==void 0?e.record:o})},[ve,w,_,k,o]),pe=r.useCallback(e=>J({page:e}),[J]),Ee=r.useCallback(e=>J({record:e}),[J]);r.useEffect(()=>{w==null||w({filters:{pipeline:u,metric:j,value:P,op:q,pipeline2:m,metric2:b,disagree:C},page:_,pageSize:k,record:o})},[w,u,j,P,q,m,b,C,_,k,o]);const[c,Re]=r.useState(null),[Fe,fe]=r.useState(!1),[Me,me]=r.useState(null),[O,Q]=r.useState("detail"),[xe,ge]=r.useState(null),[et,K]=r.useState(!1),[Le,Z]=r.useState(null),[tt,he]=r.useState(""),[v,F]=r.useState(""),[Pe,A]=r.useState(!1),[qe,G]=r.useState(null),[E,V]=r.useState(null),[st,B]=r.useState(!1),[ze,z]=r.useState(null),[Ne,N]=r.useState(null),[$e,rt]=r.useState([]),[S,De]=r.useState(null),[Ie,Je]=r.useState(null),ee=r.useMemo(()=>Be((S==null?void 0:S.metrics)??[]),[S]);r.useEffect(()=>{let e=!1;return(async()=>{try{Je(null);const d=await(await le(R("/api/evaluation-metric-definitions"))).json();e||De(d)}catch(l){e||(De(null),Je((l==null?void 0:l.message)||"Failed to load metric definitions"))}})(),()=>{e=!0}},[]),r.useEffect(()=>{var d;if(!((d=S==null?void 0:S.metrics)!=null&&d.length))return;const e=Be(S.metrics),s=wt(e);if(s.length===0)return;const l=new Set(s);X(i=>l.has(i)?i:s[0]),Y(i=>l.has(i)?i:s[1]??s[0])},[S]);const H=async e=>{const s=(e==null?void 0:e.page)??_,l=(e==null?void 0:e.pageSize)??k,d=(e==null?void 0:e.search)??we,i=(e==null?void 0:e.pipeline)??u,f=(e==null?void 0:e.metric)??j,g=(e==null?void 0:e.value)??P,te=(e==null?void 0:e.op)??q,se=(e==null?void 0:e.pipeline2)??m,ye=(e==null?void 0:e.metric2)??b,re=(e==null?void 0:e.disagree)??C;try{ke(!0),Te(null),Se([]);const p=new URLSearchParams;p.set("page",String(s)),p.set("page_size",String(l)),d&&p.set("q",d),i&&(p.set("pipeline",i),f&&p.set("metric",f),g&&p.set("value",g),te&&p.set("op",te)),i&&se&&re&&(p.set("pipeline2",se),ye&&p.set("metric2",ye),p.set("disagree","true"));const W=await(await le(R(`/api/benchmarks/${a}/errors?${p.toString()}`))).json();Se(W.items),Ke(W.total)}catch(p){Te(p.message||"Failed to load error records")}finally{ke(!1)}},[nt]=r.useState(()=>(n==null?void 0:n.pipeline)??"");r.useEffect(()=>{(async()=>{var s;try{const l=await fetch(R(`/api/benchmarks/${a}/summary/by-category`));if(!l.ok)return;const i=[...(await l.json()).overall??[]].sort((g,te)=>{var re,p,je,W;const se=Number(((p=(re=g.metrics)==null?void 0:re.subset_non_empty_execution_accuracy)==null?void 0:p.average)??-1);return Number(((W=(je=te.metrics)==null?void 0:je.subset_non_empty_execution_accuracy)==null?void 0:W.average)??-1)-se});rt(i.map(g=>g.name));const f=((s=i[0])==null?void 0:s.name)??"";if(!f)return;D(g=>g||f),I(g=>g||f),nt||H({pipeline:f})}catch{}})()},[a]);const lt=()=>{pe(1),H({page:1})};r.useEffect(()=>{H()},[a,_,k]);const at=[{key:"record_id",header:"Record ID"},{key:"question",header:"Question"},{key:"metric1_pipeline",header:"Pipeline 1"},{key:"metric1_score",header:Xe(j,"Metric 1")},{key:"metric2_pipeline",header:"Pipeline 2"},{key:"metric2_score",header:Xe(b,"Metric 2")}],it=r.useMemo(()=>L.map(e=>{var d,i;const s=u?(d=e.predictions)==null?void 0:d[u]:void 0,l=m?(i=e.predictions)==null?void 0:i[m]:void 0;return{id:e.record_id,record_id:e.record_id,question:e.question,metric1_pipeline:u||"—",metric1_score:u?Ue(s==null?void 0:s[j]):"—",metric2_pipeline:m||"—",metric2_score:m?Ue(l==null?void 0:l[b]):"—"}}),[L,j,b,u,m]);r.useEffect(()=>{if(!o||!T)return;(async()=>{try{fe(!0),me(null);const s=new URLSearchParams;s.set("pipeline",T);const l=await le(R(`/api/benchmarks/${a}/errors/${o}/detail?${s.toString()}`));Re(await l.json())}catch(s){me(s.message||"Failed to load record details")}finally{fe(!1)}})()},[a,o,T]);const Oe=()=>{Ee(null),de(null),Re(null),me(null),fe(!1),Q("detail"),ge(null),Z(null),K(!1),he(""),F(""),A(!1),G(null),V(null),B(!1),z(null),N(null)};r.useEffect(()=>{Q("detail"),ge(null),Z(null),K(!1),he(""),F(""),A(!1),G(null),V(null),B(!1),z(null),N(null)},[o]);const ot=async()=>{if(o&&(Q("raw"),!xe))try{K(!0),Z(null);const e=await le(R(`/api/benchmarks/${a}/errors/${o}`));ge(await e.json())}catch(e){Z(e.message||"Failed to load raw JSON")}finally{K(!1)}},Qe=(e,s)=>{he(s),F(e),A(!1),G(null),V(null),B(!1),z(null),N(null),Q("modify")},ct=async()=>{if(!(!o||!v.trim()))try{A(!0),G(null);const e=await fetch(R(`/api/benchmarks/${a}/execute`),{method:"POST",headers:{"Content-Type":"application/json"},body:JSON.stringify({sql:v,record_id:o,db_id:c==null?void 0:c.db_id})}),s=await e.json();if(!e.ok)throw new Error((s==null?void 0:s.detail)||`HTTP ${e.status}`);V(s),z(null),N(null)}catch(e){G(e.message||"Failed to execute SQL"),V(null)}finally{A(!1)}},dt=async()=>{if(!(!o||!v.trim()||!window.confirm("Are you confident this query should be added to benchmark ground truth SQLs?")))try{B(!0),z(null),N(null);const s=await fetch(R(`/api/benchmarks/${a}/ground-truth-sql`),{method:"POST",headers:{"Content-Type":"application/json"},body:JSON.stringify({record_id:o,sql:v})}),l=await s.json();if(!s.ok)throw new Error((l==null?void 0:l.detail)||`HTTP ${s.status}`);N(l.message)}catch(s){z(s.message||"Failed to add query to benchmark ground truth")}finally{B(!1)}};return t.jsxs("div",{style:{display:"flex",flexDirection:"column",gap:"0.5rem"},children:[t.jsxs("div",{style:{display:"flex",alignItems:"center",justifyContent:"space-between",gap:"0.75rem"},children:[t.jsxs("h3",{style:{margin:0},children:["Error analysis – ",a]}),M&&t.jsx(x,{kind:"ghost",size:"sm",onClick:M,children:"Back"})]}),Ie&&t.jsx(h,{kind:"warning",title:"Metric list unavailable",subtitle:Ie,lowContrast:!0}),t.jsxs("div",{style:{display:"flex",flexDirection:"column",gap:"0.5rem"},children:[t.jsxs("div",{style:{display:"grid",gridTemplateColumns:"repeat(auto-fit, minmax(180px, 1fr))",gap:"0.5rem",alignItems:"end"},children:[t.jsx(Ae,{id:"error-search",labelText:"Search",placeholder:"Question text or record id",value:we,onChange:e=>Ze(e.target.value)}),t.jsx(Ve,{id:"pipeline-1",titleText:"Pipeline 1 (optional)",placeholder:"e.g. wxai:openai/gpt-oss-120b-greedy-zero-shot-chatapi",items:$e,itemToString:e=>e??"",selectedItem:u||null,onChange:({selectedItem:e})=>D(e??""),onInputChange:e=>D(e),allowCustomValue:!0}),t.jsx(ne,{id:"metric-select",labelText:"Metric",value:j,onChange:e=>X(e.target.value),disabled:ee.length===0,children:ee.map(e=>t.jsx(He,{label:e.label,children:e.metrics.map(s=>t.jsx(y,{value:s.name,text:s.name,title:s.description},s.name))},e.label))}),t.jsx(ne,{id:"metric2-select",labelText:"Metric 2 (for disagreement)",value:b,onChange:e=>Y(e.target.value),disabled:!C||ee.length===0,children:ee.map(e=>t.jsx(He,{label:e.label,children:e.metrics.map(s=>t.jsx(y,{value:s.name,text:s.name,title:s.description},`${s.name}-m2`))},`${e.label}-m2`))}),t.jsxs(ne,{id:"op-select",labelText:"Operator",value:q,onChange:e=>oe(e.target.value),children:[t.jsx(y,{value:"eq",text:"="}),t.jsx(y,{value:"ne",text:"≠"}),t.jsx(y,{value:"lt",text:"<"}),t.jsx(y,{value:"gt",text:">"}),t.jsx(y,{value:"le",text:"≤"}),t.jsx(y,{value:"ge",text:"≥"})]})]}),t.jsxs("div",{style:{display:"flex",flexWrap:"wrap",alignItems:"flex-end",gap:"0.5rem",width:"100%"},children:[t.jsxs("div",{style:{flex:"1 1 400px",display:"grid",gridTemplateColumns:"repeat(auto-fit, minmax(180px, 1fr))",gap:"0.5rem",alignItems:"end",minWidth:0},children:[t.jsx(Ae,{id:"metric-value",labelText:"Metric value",value:P,onChange:e=>ie(e.target.value)}),t.jsx(Ve,{id:"pipeline-2",titleText:"Pipeline 2 (for disagreement)",placeholder:"Second pipeline id",items:$e,itemToString:e=>e??"",selectedItem:m||null,onChange:({selectedItem:e})=>I(e??""),onInputChange:e=>I(e),allowCustomValue:!0}),t.jsxs(ne,{id:"disagree-select",labelText:"P1 vs P2 disagree?",value:C?"true":"false",onChange:e=>ce(e.target.value==="true"),children:[t.jsx(y,{value:"false",text:"No"}),t.jsx(y,{value:"true",text:"Yes"})]})]}),t.jsx(x,{kind:"primary",size:"sm",onClick:lt,disabled:_e,style:{flex:"0 0 auto",marginLeft:"auto"},children:"Apply filters"})]}),t.jsxs("div",{style:{display:"flex",flexWrap:"wrap",alignItems:"center",gap:"0.5rem",padding:"0.65rem 0.75rem",borderRadius:"6px",border:"1px solid rgba(15, 98, 254, 0.15)",background:"rgba(15, 98, 254, 0.03)"},children:[t.jsx("span",{style:{fontSize:"0.75rem",fontWeight:600,letterSpacing:"0.02em",color:"var(--cds-text-secondary, #525252)",marginRight:"0.25rem"},children:"Quick presets"}),t.jsx(x,{kind:"secondary",size:"sm",disabled:!u,onClick:()=>{const e=u;e&&(D(e),I(e),X("execution_accuracy"),Y("subset_non_empty_execution_accuracy"),ie("0"),oe("eq"),ce(!0),pe(1),H({page:1,pipeline:e,pipeline2:e,metric:"execution_accuracy",metric2:"subset_non_empty_execution_accuracy",value:"0",op:"eq",disagree:!0}))},children:"Exec=0 & subset=1"}),t.jsx(x,{kind:"secondary",size:"sm",disabled:!u,onClick:()=>{const e=u;e&&(D(e),I(e),X("subset_non_empty_execution_accuracy"),Y("llm_score"),ie("0"),oe("eq"),ce(!0),pe(1),H({page:1,pipeline:e,pipeline2:e,metric:"subset_non_empty_execution_accuracy",metric2:"llm_score",value:"0",op:"eq",disagree:!0}))},children:"Subset=0 & llm=1"})]})]}),Ce&&t.jsx(h,{kind:"error",title:"Error loading error records",subtitle:Ce,lowContrast:!0}),_e?t.jsxs("div",{style:{display:"flex",flexDirection:"column",gap:"0.75rem"},children:[t.jsx(ut,{description:`Loading error records for ${a}…`,status:"active"}),t.jsx(pt,{role:"progressbar",columnCount:6,rowCount:10})]}):t.jsxs(t.Fragment,{children:[t.jsx("div",{style:{maxHeight:"420px",overflow:"auto"},children:t.jsx(ft,{rows:it,headers:at,size:"sm",children:({rows:e,headers:s,getHeaderProps:l})=>t.jsx(mt,{children:t.jsxs(xt,{"aria-label":"Error records",children:[t.jsx(gt,{children:t.jsx(Ge,{children:s.map(d=>{const{key:i,...f}=l({header:d});return t.jsx(ht,{...f,children:d.header},i)})})}),t.jsx(yt,{children:e.map(d=>t.jsx(Ge,{style:{cursor:"pointer"},onClick:()=>{const i=String(d.id),f=ue(i,L);f&&(Ee(i),de(f))},children:d.cells.map(i=>t.jsx(jt,{children:i.value},i.id))},d.id))})]})})})}),t.jsx(bt,{page:_,pageSize:k,pageSizes:[10,25,50,100],totalItems:Ye,onChange:({page:e,pageSize:s})=>J({page:e,pageSize:s})})]}),o&&t.jsxs(t.Fragment,{children:[t.jsx("div",{onClick:Oe,style:{position:"fixed",inset:0,background:"rgba(0,0,0,0.35)",zIndex:7400}}),t.jsxs("div",{style:{position:"fixed",top:"3rem",right:0,bottom:0,width:"min(900px, 92vw)",zIndex:7500,background:"#ffffff",color:"#161616",borderLeft:"1px solid rgba(0,0,0,0.12)",padding:"0.85rem",overflow:"auto",display:"flex",flexDirection:"column",gap:"0.75rem"},children:[t.jsxs("div",{style:{display:"flex",justifyContent:"space-between",alignItems:"center"},children:[t.jsx("h3",{style:{margin:0},children:O==="raw"?`Raw JSON – ${o}`:O==="modify"?`Modify Query – ${o}`:`Record detail – ${o}${T?` (${T})`:""}`}),t.jsxs("div",{style:{display:"flex",gap:"0.35rem"},children:[O!=="detail"&&t.jsx(x,{kind:"ghost",size:"sm",onClick:()=>Q("detail"),children:"Back to detail"}),t.jsx(x,{kind:"ghost",size:"sm",onClick:Oe,children:"X"})]})]}),O==="raw"?t.jsxs(t.Fragment,{children:[et&&t.jsx(h,{kind:"info",title:"Loading raw JSON...",subtitle:"Fetching full record payload from predictions_eval",lowContrast:!0}),Le&&t.jsx(h,{kind:"error",title:"Failed to load raw JSON",subtitle:Le,lowContrast:!0}),xe&&t.jsx("section",{children:t.jsx("pre",{style:{margin:"0.3rem 0",padding:"0.6rem",background:"#f4f4f4",borderRadius:"4px",whiteSpace:"pre-wrap",border:"1px solid rgba(15,98,254,0.2)",color:"#161616"},children:JSON.stringify(xe,null,2)})})]}):O==="modify"?t.jsxs(t.Fragment,{children:[t.jsxs("section",{children:[t.jsx("h4",{style:{margin:"0.25rem 0",color:"#0f62fe"},children:"Source SQL"}),t.jsx("div",{style:{marginBottom:"0.35rem"},children:tt||"Custom query"})]}),t.jsx(St,{id:"error-analysis-modify-query-sql",labelText:"Editable SQL",rows:14,value:v,onChange:e=>F(e.target.value)}),t.jsx("div",{style:{display:"flex",justifyContent:"flex-end"},children:t.jsx(x,{kind:"secondary",size:"sm",disabled:Pe||!v.trim(),onClick:()=>void ct(),children:"Execute"})}),Pe&&t.jsx(h,{kind:"info",title:"Executing SQL...",subtitle:"Running query against the benchmark backend",lowContrast:!0}),qe&&t.jsx(h,{kind:"error",title:"SQL execution failed",subtitle:qe,lowContrast:!0}),ze&&t.jsx(h,{kind:"error",title:"Failed to update ground truth",subtitle:ze,lowContrast:!0}),Ne&&t.jsx(h,{kind:"success",title:"Ground truth updated",subtitle:Ne,lowContrast:!0}),E&&t.jsxs(t.Fragment,{children:[t.jsxs("section",{children:[t.jsx("h4",{style:{margin:"0.25rem 0",color:"#0f62fe"},children:"Execution summary"}),t.jsx("pre",{style:{margin:"0.3rem 0",padding:"0.6rem",background:"#f4f4f4",borderRadius:"4px",whiteSpace:"pre-wrap",color:"#161616"},children:JSON.stringify({db_type:E.db_type,db_id:E.db_id,execution_time_ms:E.execution_time_ms,row_count:E.row_count,column_count:E.column_count},null,2)})]}),t.jsx(be,{title:"Execution result",rawData:E.result}),t.jsx("div",{style:{display:"flex",justifyContent:"flex-end"},children:t.jsx(x,{kind:"primary",size:"sm",disabled:st||!v.trim(),onClick:()=>void dt(),children:"Add to benchmark ground truth"})})]})]}):t.jsxs(t.Fragment,{children:[Fe&&t.jsx(h,{kind:"info",title:"Loading details...",subtitle:"Fetching full record detail",lowContrast:!0}),Me&&t.jsx(h,{kind:"error",title:"Failed to load details",subtitle:Me,lowContrast:!0}),c&&t.jsxs(t.Fragment,{children:[t.jsxs("section",{children:[t.jsx("h4",{style:{margin:"0.25rem 0",color:"#0f62fe"},children:"Question"}),t.jsx("div",{style:{whiteSpace:"pre-wrap"},children:c.question||"N/A"})]}),t.jsxs("section",{children:[t.jsx("h4",{style:{margin:"0.25rem 0",color:"#0f62fe"},children:"Ground truth SQL"}),(c.ground_truth_sql||[]).map((e,s)=>t.jsxs("div",{style:{marginBottom:"0.55rem"},children:[t.jsx("pre",{style:{margin:"0.3rem 0",padding:"0.6rem",background:"#f4f4f4",borderRadius:"4px",whiteSpace:"pre-wrap",border:"1px solid rgba(15,98,254,0.2)",color:"#161616"},children:t.jsx("code",{dangerouslySetInnerHTML:{__html:We(e)}})}),t.jsx("div",{style:{display:"flex",justifyContent:"flex-end"},children:t.jsx(x,{kind:"ghost",size:"sm",onClick:()=>Qe(e,`Ground truth SQL ${s+1}`),children:"Modify Query"})})]},`gt-sql-${s}`))]}),t.jsxs("section",{children:[t.jsx("h4",{style:{margin:"0.25rem 0",color:"#0f62fe"},children:"Predicted SQL"}),t.jsx("pre",{style:{margin:"0.3rem 0",padding:"0.6rem",background:"#f4f4f4",borderRadius:"4px",whiteSpace:"pre-wrap",border:"1px solid rgba(15,98,254,0.2)",color:"#161616"},children:t.jsx("code",{dangerouslySetInnerHTML:{__html:We(c.predicted_sql||"N/A")}})}),t.jsx("div",{style:{display:"flex",justifyContent:"flex-end"},children:t.jsx(x,{kind:"ghost",size:"sm",disabled:!c.predicted_sql,onClick:()=>Qe(c.predicted_sql||"","Predicted SQL"),children:"Modify Query"})})]}),t.jsxs("section",{children:[t.jsx("h4",{style:{margin:"0.25rem 0",color:"#0f62fe"},children:"Evaluation metrics"}),t.jsx("pre",{style:{margin:"0.3rem 0",padding:"0.6rem",background:"#f4f4f4",borderRadius:"4px",whiteSpace:"pre-wrap",color:"#161616"},children:JSON.stringify(c.evaluation_metrics??{},null,2)})]}),(c.ground_truth_results||[]).map((e,s)=>{var l;return t.jsx(be,{title:`Ground truth result ${s+1}`,rawData:e,totalRows:(l=c.ground_truth_result_row_counts)==null?void 0:l[s]},`gt-result-table-${s}`)}),t.jsx(be,{title:"Predicted result",rawData:c.predicted_result,totalRows:c.predicted_result_row_count}),t.jsxs("section",{children:[t.jsx("h4",{style:{margin:"0.25rem 0",color:"#0f62fe"},children:"Prompt"}),t.jsx("pre",{style:{margin:"0.3rem 0",padding:"0.6rem",background:"#f4f4f4",borderRadius:"4px",whiteSpace:"pre-wrap",color:"#161616"},children:c.prompt||"N/A"})]}),t.jsxs("section",{children:[t.jsx("h4",{style:{margin:"0.25rem 0",color:"#0f62fe"},children:"LLM judge"}),t.jsxs("div",{style:{marginBottom:"0.25rem"},children:["Score: ",c.llm_judge_score??"N/A"]}),t.jsx("pre",{style:{margin:"0.3rem 0",padding:"0.6rem",background:"#f4f4f4",borderRadius:"4px",whiteSpace:"pre-wrap",color:"#161616"},children:c.llm_judge_explanation||"N/A"})]}),(c.sql_execution_error||c.inference_error)&&t.jsxs("section",{children:[t.jsx("h4",{style:{margin:"0.25rem 0",color:"#0f62fe"},children:"Errors"}),t.jsx("pre",{style:{margin:"0.3rem 0",padding:"0.6rem",background:"#f4f4f4",borderRadius:"4px",whiteSpace:"pre-wrap",color:"#161616"},children:JSON.stringify({sql_execution_error:c.sql_execution_error,inference_error:c.inference_error},null,2)})]}),t.jsx("div",{style:{marginTop:"0.25rem",display:"flex",justifyContent:"flex-end"},children:t.jsx(x,{kind:"secondary",size:"sm",onClick:()=>void ot(),children:"View Raw JSON"})})]})]})]})]})]})};export{Rt as ErrorAnalysis};
+import{r,k as t,B as x,I as h,E as Ae,S as ne,G as y,l as ut,D as pt,n as ft,T as mt,o as xt,p as gt,q as Ge,t as ht,v as yt,w as jt,x as bt,H as St,y as le,z as R}from"./index-DLyduwv0.js";import{C as Ve}from"./ComboBox-BActzty7.js";import{b as Be,f as wt,S as He}from"./metricInsightsSelect-unZgHJkx.js";import{R as be,h as We}from"./ResultTableView-C-syf-qW.js";import"./defaultItemToString-ClUBp4QC.js";function _t(a,M,n){const $=M.find(ae=>ae.record_id===a),U=Object.keys(($==null?void 0:$.predictions)??{});return n&&U.includes(n)?n:U[0]??null}function Ue(a){if(a==null)return"N/A";if(typeof a=="number")return Number.isFinite(a)?String(a):"N/A";if(typeof a=="boolean")return a?"true":"false";if(typeof a=="string")return a;try{return JSON.stringify(a)}catch{return String(a)}}function Xe(a,M){const n=a.trim();return n?n.replaceAll("_"," "):M}const Rt=({benchmarkId:a,onBack:M,initialFilters:n,initialPage:$,initialPageSize:U,initialRecordId:ae,onStateChange:w})=>{const[L,Se]=r.useState([]),[Ye,Ke]=r.useState(0),_=$??1,k=U??25,[we,Ze]=r.useState(""),[u,D]=r.useState(()=>(n==null?void 0:n.pipeline)??""),[j,X]=r.useState(()=>(n==null?void 0:n.metric)??"execution_accuracy"),[P,ie]=r.useState(()=>(n==null?void 0:n.value)??"0"),[q,oe]=r.useState(()=>(n==null?void 0:n.op)??"eq"),[m,I]=r.useState(()=>(n==null?void 0:n.pipeline2)??""),[b,Y]=r.useState(()=>(n==null?void 0:n.metric2)??"subset_non_empty_execution_accuracy"),[C,ce]=r.useState(()=>(n==null?void 0:n.disagree)??!1),[_e,ke]=r.useState(!0),[Ce,Te]=r.useState(null),o=ae??null,[T,de]=r.useState(null),ue=r.useCallback((e,s)=>_t(e,s,u),[u]);r.useEffect(()=>{if(!o||T||L.length===0)return;const e=ue(o,L);e&&de(e)},[o,T,L,ue]);const ve=r.useMemo(()=>({pipeline:u,metric:j,value:P,op:q,pipeline2:m,metric2:b,disagree:C}),[u,j,P,q,m,b,C]),J=r.useCallback(e=>{w==null||w({filters:ve,page:e.page??_,pageSize:e.pageSize??k,record:e.record!==void 0?e.record:o})},[ve,w,_,k,o]),pe=r.useCallback(e=>J({page:e}),[J]),Ee=r.useCallback(e=>J({record:e}),[J]);r.useEffect(()=>{w==null||w({filters:{pipeline:u,metric:j,value:P,op:q,pipeline2:m,metric2:b,disagree:C},page:_,pageSize:k,record:o})},[w,u,j,P,q,m,b,C,_,k,o]);const[c,Re]=r.useState(null),[Fe,fe]=r.useState(!1),[Me,me]=r.useState(null),[O,Q]=r.useState("detail"),[xe,ge]=r.useState(null),[et,K]=r.useState(!1),[Le,Z]=r.useState(null),[tt,he]=r.useState(""),[v,F]=r.useState(""),[Pe,A]=r.useState(!1),[qe,G]=r.useState(null),[E,V]=r.useState(null),[st,B]=r.useState(!1),[ze,z]=r.useState(null),[Ne,N]=r.useState(null),[$e,rt]=r.useState([]),[S,De]=r.useState(null),[Ie,Je]=r.useState(null),ee=r.useMemo(()=>Be((S==null?void 0:S.metrics)??[]),[S]);r.useEffect(()=>{let e=!1;return(async()=>{try{Je(null);const d=await(await le(R("/api/evaluation-metric-definitions"))).json();e||De(d)}catch(l){e||(De(null),Je((l==null?void 0:l.message)||"Failed to load metric definitions"))}})(),()=>{e=!0}},[]),r.useEffect(()=>{var d;if(!((d=S==null?void 0:S.metrics)!=null&&d.length))return;const e=Be(S.metrics),s=wt(e);if(s.length===0)return;const l=new Set(s);X(i=>l.has(i)?i:s[0]),Y(i=>l.has(i)?i:s[1]??s[0])},[S]);const H=async e=>{const s=(e==null?void 0:e.page)??_,l=(e==null?void 0:e.pageSize)??k,d=(e==null?void 0:e.search)??we,i=(e==null?void 0:e.pipeline)??u,f=(e==null?void 0:e.metric)??j,g=(e==null?void 0:e.value)??P,te=(e==null?void 0:e.op)??q,se=(e==null?void 0:e.pipeline2)??m,ye=(e==null?void 0:e.metric2)??b,re=(e==null?void 0:e.disagree)??C;try{ke(!0),Te(null),Se([]);const p=new URLSearchParams;p.set("page",String(s)),p.set("page_size",String(l)),d&&p.set("q",d),i&&(p.set("pipeline",i),f&&p.set("metric",f),g&&p.set("value",g),te&&p.set("op",te)),i&&se&&re&&(p.set("pipeline2",se),ye&&p.set("metric2",ye),p.set("disagree","true"));const W=await(await le(R(`/api/benchmarks/${a}/errors?${p.toString()}`))).json();Se(W.items),Ke(W.total)}catch(p){Te(p.message||"Failed to load error records")}finally{ke(!1)}},[nt]=r.useState(()=>(n==null?void 0:n.pipeline)??"");r.useEffect(()=>{(async()=>{var s;try{const l=await fetch(R(`/api/benchmarks/${a}/summary/by-category`));if(!l.ok)return;const i=[...(await l.json()).overall??[]].sort((g,te)=>{var re,p,je,W;const se=Number(((p=(re=g.metrics)==null?void 0:re.subset_non_empty_execution_accuracy)==null?void 0:p.average)??-1);return Number(((W=(je=te.metrics)==null?void 0:je.subset_non_empty_execution_accuracy)==null?void 0:W.average)??-1)-se});rt(i.map(g=>g.name));const f=((s=i[0])==null?void 0:s.name)??"";if(!f)return;D(g=>g||f),I(g=>g||f),nt||H({pipeline:f})}catch{}})()},[a]);const lt=()=>{pe(1),H({page:1})};r.useEffect(()=>{H()},[a,_,k]);const at=[{key:"record_id",header:"Record ID"},{key:"question",header:"Question"},{key:"metric1_pipeline",header:"Pipeline 1"},{key:"metric1_score",header:Xe(j,"Metric 1")},{key:"metric2_pipeline",header:"Pipeline 2"},{key:"metric2_score",header:Xe(b,"Metric 2")}],it=r.useMemo(()=>L.map(e=>{var d,i;const s=u?(d=e.predictions)==null?void 0:d[u]:void 0,l=m?(i=e.predictions)==null?void 0:i[m]:void 0;return{id:e.record_id,record_id:e.record_id,question:e.question,metric1_pipeline:u||"—",metric1_score:u?Ue(s==null?void 0:s[j]):"—",metric2_pipeline:m||"—",metric2_score:m?Ue(l==null?void 0:l[b]):"—"}}),[L,j,b,u,m]);r.useEffect(()=>{if(!o||!T)return;(async()=>{try{fe(!0),me(null);const s=new URLSearchParams;s.set("pipeline",T);const l=await le(R(`/api/benchmarks/${a}/errors/${o}/detail?${s.toString()}`));Re(await l.json())}catch(s){me(s.message||"Failed to load record details")}finally{fe(!1)}})()},[a,o,T]);const Oe=()=>{Ee(null),de(null),Re(null),me(null),fe(!1),Q("detail"),ge(null),Z(null),K(!1),he(""),F(""),A(!1),G(null),V(null),B(!1),z(null),N(null)};r.useEffect(()=>{Q("detail"),ge(null),Z(null),K(!1),he(""),F(""),A(!1),G(null),V(null),B(!1),z(null),N(null)},[o]);const ot=async()=>{if(o&&(Q("raw"),!xe))try{K(!0),Z(null);const e=await le(R(`/api/benchmarks/${a}/errors/${o}`));ge(await e.json())}catch(e){Z(e.message||"Failed to load raw JSON")}finally{K(!1)}},Qe=(e,s)=>{he(s),F(e),A(!1),G(null),V(null),B(!1),z(null),N(null),Q("modify")},ct=async()=>{if(!(!o||!v.trim()))try{A(!0),G(null);const e=await fetch(R(`/api/benchmarks/${a}/execute`),{method:"POST",headers:{"Content-Type":"application/json"},body:JSON.stringify({sql:v,record_id:o,db_id:c==null?void 0:c.db_id})}),s=await e.json();if(!e.ok)throw new Error((s==null?void 0:s.detail)||`HTTP ${e.status}`);V(s),z(null),N(null)}catch(e){G(e.message||"Failed to execute SQL"),V(null)}finally{A(!1)}},dt=async()=>{if(!(!o||!v.trim()||!window.confirm("Are you confident this query should be added to benchmark ground truth SQLs?")))try{B(!0),z(null),N(null);const s=await fetch(R(`/api/benchmarks/${a}/ground-truth-sql`),{method:"POST",headers:{"Content-Type":"application/json"},body:JSON.stringify({record_id:o,sql:v})}),l=await s.json();if(!s.ok)throw new Error((l==null?void 0:l.detail)||`HTTP ${s.status}`);N(l.message)}catch(s){z(s.message||"Failed to add query to benchmark ground truth")}finally{B(!1)}};return t.jsxs("div",{style:{display:"flex",flexDirection:"column",gap:"0.5rem"},children:[t.jsxs("div",{style:{display:"flex",alignItems:"center",justifyContent:"space-between",gap:"0.75rem"},children:[t.jsxs("h3",{style:{margin:0},children:["Error analysis – ",a]}),M&&t.jsx(x,{kind:"ghost",size:"sm",onClick:M,children:"Back"})]}),Ie&&t.jsx(h,{kind:"warning",title:"Metric list unavailable",subtitle:Ie,lowContrast:!0}),t.jsxs("div",{style:{display:"flex",flexDirection:"column",gap:"0.5rem"},children:[t.jsxs("div",{style:{display:"grid",gridTemplateColumns:"repeat(auto-fit, minmax(180px, 1fr))",gap:"0.5rem",alignItems:"end"},children:[t.jsx(Ae,{id:"error-search",labelText:"Search",placeholder:"Question text or record id",value:we,onChange:e=>Ze(e.target.value)}),t.jsx(Ve,{id:"pipeline-1",titleText:"Pipeline 1 (optional)",placeholder:"e.g. wxai:openai/gpt-oss-120b-greedy-zero-shot-chatapi",items:$e,itemToString:e=>e??"",selectedItem:u||null,onChange:({selectedItem:e})=>D(e??""),onInputChange:e=>D(e),allowCustomValue:!0}),t.jsx(ne,{id:"metric-select",labelText:"Metric",value:j,onChange:e=>X(e.target.value),disabled:ee.length===0,children:ee.map(e=>t.jsx(He,{label:e.label,children:e.metrics.map(s=>t.jsx(y,{value:s.name,text:s.name,title:s.description},s.name))},e.label))}),t.jsx(ne,{id:"metric2-select",labelText:"Metric 2 (for disagreement)",value:b,onChange:e=>Y(e.target.value),disabled:!C||ee.length===0,children:ee.map(e=>t.jsx(He,{label:e.label,children:e.metrics.map(s=>t.jsx(y,{value:s.name,text:s.name,title:s.description},`${s.name}-m2`))},`${e.label}-m2`))}),t.jsxs(ne,{id:"op-select",labelText:"Operator",value:q,onChange:e=>oe(e.target.value),children:[t.jsx(y,{value:"eq",text:"="}),t.jsx(y,{value:"ne",text:"≠"}),t.jsx(y,{value:"lt",text:"<"}),t.jsx(y,{value:"gt",text:">"}),t.jsx(y,{value:"le",text:"≤"}),t.jsx(y,{value:"ge",text:"≥"})]})]}),t.jsxs("div",{style:{display:"flex",flexWrap:"wrap",alignItems:"flex-end",gap:"0.5rem",width:"100%"},children:[t.jsxs("div",{style:{flex:"1 1 400px",display:"grid",gridTemplateColumns:"repeat(auto-fit, minmax(180px, 1fr))",gap:"0.5rem",alignItems:"end",minWidth:0},children:[t.jsx(Ae,{id:"metric-value",labelText:"Metric value",value:P,onChange:e=>ie(e.target.value)}),t.jsx(Ve,{id:"pipeline-2",titleText:"Pipeline 2 (for disagreement)",placeholder:"Second pipeline id",items:$e,itemToString:e=>e??"",selectedItem:m||null,onChange:({selectedItem:e})=>I(e??""),onInputChange:e=>I(e),allowCustomValue:!0}),t.jsxs(ne,{id:"disagree-select",labelText:"P1 vs P2 disagree?",value:C?"true":"false",onChange:e=>ce(e.target.value==="true"),children:[t.jsx(y,{value:"false",text:"No"}),t.jsx(y,{value:"true",text:"Yes"})]})]}),t.jsx(x,{kind:"primary",size:"sm",onClick:lt,disabled:_e,style:{flex:"0 0 auto",marginLeft:"auto"},children:"Apply filters"})]}),t.jsxs("div",{style:{display:"flex",flexWrap:"wrap",alignItems:"center",gap:"0.5rem",padding:"0.65rem 0.75rem",borderRadius:"6px",border:"1px solid rgba(15, 98, 254, 0.15)",background:"rgba(15, 98, 254, 0.03)"},children:[t.jsx("span",{style:{fontSize:"0.75rem",fontWeight:600,letterSpacing:"0.02em",color:"var(--cds-text-secondary, #525252)",marginRight:"0.25rem"},children:"Quick presets"}),t.jsx(x,{kind:"secondary",size:"sm",disabled:!u,onClick:()=>{const e=u;e&&(D(e),I(e),X("execution_accuracy"),Y("subset_non_empty_execution_accuracy"),ie("0"),oe("eq"),ce(!0),pe(1),H({page:1,pipeline:e,pipeline2:e,metric:"execution_accuracy",metric2:"subset_non_empty_execution_accuracy",value:"0",op:"eq",disagree:!0}))},children:"Exec=0 & subset=1"}),t.jsx(x,{kind:"secondary",size:"sm",disabled:!u,onClick:()=>{const e=u;e&&(D(e),I(e),X("subset_non_empty_execution_accuracy"),Y("llm_score"),ie("0"),oe("eq"),ce(!0),pe(1),H({page:1,pipeline:e,pipeline2:e,metric:"subset_non_empty_execution_accuracy",metric2:"llm_score",value:"0",op:"eq",disagree:!0}))},children:"Subset=0 & llm=1"})]})]}),Ce&&t.jsx(h,{kind:"error",title:"Error loading error records",subtitle:Ce,lowContrast:!0}),_e?t.jsxs("div",{style:{display:"flex",flexDirection:"column",gap:"0.75rem"},children:[t.jsx(ut,{description:`Loading error records for ${a}…`,status:"active"}),t.jsx(pt,{role:"progressbar",columnCount:6,rowCount:10})]}):t.jsxs(t.Fragment,{children:[t.jsx("div",{style:{maxHeight:"420px",overflow:"auto"},children:t.jsx(ft,{rows:it,headers:at,size:"sm",children:({rows:e,headers:s,getHeaderProps:l})=>t.jsx(mt,{children:t.jsxs(xt,{"aria-label":"Error records",children:[t.jsx(gt,{children:t.jsx(Ge,{children:s.map(d=>{const{key:i,...f}=l({header:d});return t.jsx(ht,{...f,children:d.header},i)})})}),t.jsx(yt,{children:e.map(d=>t.jsx(Ge,{style:{cursor:"pointer"},onClick:()=>{const i=String(d.id),f=ue(i,L);f&&(Ee(i),de(f))},children:d.cells.map(i=>t.jsx(jt,{children:i.value},i.id))},d.id))})]})})})}),t.jsx(bt,{page:_,pageSize:k,pageSizes:[10,25,50,100],totalItems:Ye,onChange:({page:e,pageSize:s})=>J({page:e,pageSize:s})})]}),o&&t.jsxs(t.Fragment,{children:[t.jsx("div",{onClick:Oe,style:{position:"fixed",inset:0,background:"rgba(0,0,0,0.35)",zIndex:7400}}),t.jsxs("div",{style:{position:"fixed",top:"3rem",right:0,bottom:0,width:"min(900px, 92vw)",zIndex:7500,background:"#ffffff",color:"#161616",borderLeft:"1px solid rgba(0,0,0,0.12)",padding:"0.85rem",overflow:"auto",display:"flex",flexDirection:"column",gap:"0.75rem"},children:[t.jsxs("div",{style:{display:"flex",justifyContent:"space-between",alignItems:"center"},children:[t.jsx("h3",{style:{margin:0},children:O==="raw"?`Raw JSON – ${o}`:O==="modify"?`Modify Query – ${o}`:`Record detail – ${o}${T?` (${T})`:""}`}),t.jsxs("div",{style:{display:"flex",gap:"0.35rem"},children:[O!=="detail"&&t.jsx(x,{kind:"ghost",size:"sm",onClick:()=>Q("detail"),children:"Back to detail"}),t.jsx(x,{kind:"ghost",size:"sm",onClick:Oe,children:"X"})]})]}),O==="raw"?t.jsxs(t.Fragment,{children:[et&&t.jsx(h,{kind:"info",title:"Loading raw JSON...",subtitle:"Fetching full record payload from predictions_eval",lowContrast:!0}),Le&&t.jsx(h,{kind:"error",title:"Failed to load raw JSON",subtitle:Le,lowContrast:!0}),xe&&t.jsx("section",{children:t.jsx("pre",{style:{margin:"0.3rem 0",padding:"0.6rem",background:"#f4f4f4",borderRadius:"4px",whiteSpace:"pre-wrap",border:"1px solid rgba(15,98,254,0.2)",color:"#161616"},children:JSON.stringify(xe,null,2)})})]}):O==="modify"?t.jsxs(t.Fragment,{children:[t.jsxs("section",{children:[t.jsx("h4",{style:{margin:"0.25rem 0",color:"#0f62fe"},children:"Source SQL"}),t.jsx("div",{style:{marginBottom:"0.35rem"},children:tt||"Custom query"})]}),t.jsx(St,{id:"error-analysis-modify-query-sql",labelText:"Editable SQL",rows:14,value:v,onChange:e=>F(e.target.value)}),t.jsx("div",{style:{display:"flex",justifyContent:"flex-end"},children:t.jsx(x,{kind:"secondary",size:"sm",disabled:Pe||!v.trim(),onClick:()=>void ct(),children:"Execute"})}),Pe&&t.jsx(h,{kind:"info",title:"Executing SQL...",subtitle:"Running query against the benchmark backend",lowContrast:!0}),qe&&t.jsx(h,{kind:"error",title:"SQL execution failed",subtitle:qe,lowContrast:!0}),ze&&t.jsx(h,{kind:"error",title:"Failed to update ground truth",subtitle:ze,lowContrast:!0}),Ne&&t.jsx(h,{kind:"success",title:"Ground truth updated",subtitle:Ne,lowContrast:!0}),E&&t.jsxs(t.Fragment,{children:[t.jsxs("section",{children:[t.jsx("h4",{style:{margin:"0.25rem 0",color:"#0f62fe"},children:"Execution summary"}),t.jsx("pre",{style:{margin:"0.3rem 0",padding:"0.6rem",background:"#f4f4f4",borderRadius:"4px",whiteSpace:"pre-wrap",color:"#161616"},children:JSON.stringify({db_type:E.db_type,db_id:E.db_id,execution_time_ms:E.execution_time_ms,row_count:E.row_count,column_count:E.column_count},null,2)})]}),t.jsx(be,{title:"Execution result",rawData:E.result}),t.jsx("div",{style:{display:"flex",justifyContent:"flex-end"},children:t.jsx(x,{kind:"primary",size:"sm",disabled:st||!v.trim(),onClick:()=>void dt(),children:"Add to benchmark ground truth"})})]})]}):t.jsxs(t.Fragment,{children:[Fe&&t.jsx(h,{kind:"info",title:"Loading details...",subtitle:"Fetching full record detail",lowContrast:!0}),Me&&t.jsx(h,{kind:"error",title:"Failed to load details",subtitle:Me,lowContrast:!0}),c&&t.jsxs(t.Fragment,{children:[t.jsxs("section",{children:[t.jsx("h4",{style:{margin:"0.25rem 0",color:"#0f62fe"},children:"Question"}),t.jsx("div",{style:{whiteSpace:"pre-wrap"},children:c.question||"N/A"})]}),t.jsxs("section",{children:[t.jsx("h4",{style:{margin:"0.25rem 0",color:"#0f62fe"},children:"Ground truth SQL"}),(c.ground_truth_sql||[]).map((e,s)=>t.jsxs("div",{style:{marginBottom:"0.55rem"},children:[t.jsx("pre",{style:{margin:"0.3rem 0",padding:"0.6rem",background:"#f4f4f4",borderRadius:"4px",whiteSpace:"pre-wrap",border:"1px solid rgba(15,98,254,0.2)",color:"#161616"},children:t.jsx("code",{dangerouslySetInnerHTML:{__html:We(e)}})}),t.jsx("div",{style:{display:"flex",justifyContent:"flex-end"},children:t.jsx(x,{kind:"ghost",size:"sm",onClick:()=>Qe(e,`Ground truth SQL ${s+1}`),children:"Modify Query"})})]},`gt-sql-${s}`))]}),t.jsxs("section",{children:[t.jsx("h4",{style:{margin:"0.25rem 0",color:"#0f62fe"},children:"Predicted SQL"}),t.jsx("pre",{style:{margin:"0.3rem 0",padding:"0.6rem",background:"#f4f4f4",borderRadius:"4px",whiteSpace:"pre-wrap",border:"1px solid rgba(15,98,254,0.2)",color:"#161616"},children:t.jsx("code",{dangerouslySetInnerHTML:{__html:We(c.predicted_sql||"N/A")}})}),t.jsx("div",{style:{display:"flex",justifyContent:"flex-end"},children:t.jsx(x,{kind:"ghost",size:"sm",disabled:!c.predicted_sql,onClick:()=>Qe(c.predicted_sql||"","Predicted SQL"),children:"Modify Query"})})]}),t.jsxs("section",{children:[t.jsx("h4",{style:{margin:"0.25rem 0",color:"#0f62fe"},children:"Evaluation metrics"}),t.jsx("pre",{style:{margin:"0.3rem 0",padding:"0.6rem",background:"#f4f4f4",borderRadius:"4px",whiteSpace:"pre-wrap",color:"#161616"},children:JSON.stringify(c.evaluation_metrics??{},null,2)})]}),(c.ground_truth_results||[]).map((e,s)=>{var l;return t.jsx(be,{title:`Ground truth result ${s+1}`,rawData:e,totalRows:(l=c.ground_truth_result_row_counts)==null?void 0:l[s]},`gt-result-table-${s}`)}),t.jsx(be,{title:"Predicted result",rawData:c.predicted_result,totalRows:c.predicted_result_row_count}),t.jsxs("section",{children:[t.jsx("h4",{style:{margin:"0.25rem 0",color:"#0f62fe"},children:"Prompt"}),t.jsx("pre",{style:{margin:"0.3rem 0",padding:"0.6rem",background:"#f4f4f4",borderRadius:"4px",whiteSpace:"pre-wrap",color:"#161616"},children:c.prompt||"N/A"})]}),t.jsxs("section",{children:[t.jsx("h4",{style:{margin:"0.25rem 0",color:"#0f62fe"},children:"LLM judge"}),t.jsxs("div",{style:{marginBottom:"0.25rem"},children:["Score: ",c.llm_judge_score??"N/A"]}),t.jsx("pre",{style:{margin:"0.3rem 0",padding:"0.6rem",background:"#f4f4f4",borderRadius:"4px",whiteSpace:"pre-wrap",color:"#161616"},children:c.llm_judge_explanation||"N/A"})]}),(c.sql_execution_error||c.inference_error)&&t.jsxs("section",{children:[t.jsx("h4",{style:{margin:"0.25rem 0",color:"#0f62fe"},children:"Errors"}),t.jsx("pre",{style:{margin:"0.3rem 0",padding:"0.6rem",background:"#f4f4f4",borderRadius:"4px",whiteSpace:"pre-wrap",color:"#161616"},children:JSON.stringify({sql_execution_error:c.sql_execution_error,inference_error:c.inference_error},null,2)})]}),t.jsx("div",{style:{marginTop:"0.25rem",display:"flex",justifyContent:"flex-end"},children:t.jsx(x,{kind:"secondary",size:"sm",onClick:()=>void ot(),children:"View Raw JSON"})})]})]})]})]})]})};export{Rt as ErrorAnalysis};

text2sql_eval_toolkit-1.4.0/dashboard/dist/assets/LLMJudgeConfigView-DXZQKioY.js

@@ -0,0 +1,10 @@
+import{R as f,J as F,r as i,y as C,z as j,k as s,B as _,K as O,I as L,E as z,L as D,H as q}from"./index-DLyduwv0.js";import{C as B}from"./ComboBox-BActzty7.js";import"./defaultItemToString-ClUBp4QC.js";var b,A,I,M;const P=f.forwardRef(function({children:S,size:n=16,...c},h){return f.createElement(F,{width:n,height:n,ref:h,xmlns:"http://www.w3.org/2000/svg",viewBox:"0 0 32 32",fill:"currentColor",...c},b||(b=f.createElement("path",{d:"M12 12H14V24H12z"})),A||(A=f.createElement("path",{d:"M18 12H20V24H18z"})),I||(I=f.createElement("path",{d:"M4,6V8H6V28a2,2,0,0,0,2,2H24a2,2,0,0,0,2-2V8h2V6ZM8,28V8H24V28Z"})),M||(M=f.createElement("path",{d:"M12 2H20V4H12z"})),S)}),R="llm_judge_default_config",G=/^[A-Za-z0-9][A-Za-z0-9_.-]{0,127}$/,Q={model:{id:"anthropic:claude-sonnet-4-5",max_tokens:1e3,temperature:0},prompt_template:`You are evaluating whether a predicted SQL query correctly answers a question.
+
+Question: {question}
+Ground truth SQL: {ground_truth_sql}
+Predicted SQL: {predicted_sql}
+Ground truth result: {ground_truth_df}
+Predicted result: {predicted_df}
+
+Answer Yes if the prediction correctly answers the question, No otherwise.
+Start your reply with Yes or No, then explain.`},K=()=>{const[v,S]=i.useState([]),[n,c]=i.useState(null),[h,u]=i.useState(""),[N,a]=i.useState(null),[k,r]=i.useState(null),[p,g]=i.useState(!1),[d,w]=i.useState(null),y=i.useCallback(async()=>{const e=await(await C(j("/api/llm-judge/configs"))).json();return S(e.items),e.items},[]),x=i.useCallback(async t=>{try{g(!0),a(null),r(null);const l=await(await C(j(`/api/llm-judge/configs/${t.name}`))).json();u(JSON.stringify(l,null,2))}catch(e){a(e.message||"Failed to load config"),u("")}finally{g(!1)}},[]);i.useEffect(()=>{(async()=>{try{a(null);const l=(await y()).find(o=>o.name===R);l&&(c(l),await x(l))}catch(e){a(e.message||"Failed to load config list")}})()},[x,y]);const H=()=>{a(null),r(null),c(null),w(""),u(JSON.stringify(Q,null,2))},V=()=>{w(null),u(""),a(null)},$=async()=>{const t=d!==null,e=t?(d||"").trim():n==null?void 0:n.name;if(!e)return;if(t&&!G.test(e)){a("A config name must start with a letter or digit and contain only letters, digits, dots, dashes and underscores.");return}if(t&&v.some(o=>o.name===e)){a(`A config named "${e}" already exists. Pick another name.`);return}let l;try{l=JSON.parse(h)}catch(o){a(`The config is not valid JSON: ${o.message}`);return}try{g(!0),a(null),r(null),await C(j(`/api/llm-judge/configs/${e}`),{method:"PUT",headers:{"Content-Type":"application/json"},body:JSON.stringify(l)});const m=(await y()).find(E=>E.name===e)||null;c(m),w(null),r(t?`Created "${e}". It is now selectable in the judge playground.`:`Saved "${e}".`)}catch(o){a(o.message||"Failed to save config")}finally{g(!1)}},J=async()=>{if(!(n!=null&&n.user_defined))return;const t=n.name;try{g(!0),a(null),r(null);const l=await(await C(j(`/api/llm-judge/configs/${t}`),{method:"DELETE"})).json(),m=(await y()).find(E=>E.name===t)||null;c(m),m?await x(m):u(""),r(l!=null&&l.reverted_to_packaged?`Deleted the edit to "${t}"; the packaged config is back in use.`:`Deleted "${t}".`)}catch(e){a(e.message||"Failed to delete config")}finally{g(!1)}},T=d!==null||n!==null;return s.jsxs("div",{style:{display:"flex",flexDirection:"column",gap:"0.5rem"},children:[s.jsxs("div",{style:{display:"flex",alignItems:"center",justifyContent:"space-between",gap:"1rem"},children:[s.jsx("h3",{style:{margin:0},children:"LLM-as-judge configuration"}),s.jsx(_,{kind:"ghost",size:"sm",renderIcon:O,onClick:H,disabled:p,children:"New config"})]}),N&&s.jsx(L,{kind:"error",title:"Error",subtitle:N,lowContrast:!0,onCloseButtonClick:()=>a(null)}),k&&s.jsx(L,{kind:"success",title:"Saved",subtitle:k,lowContrast:!0,onCloseButtonClick:()=>r(null)}),d!==null?s.jsx(z,{id:"llm-config-new-name",labelText:"New config name",helperText:"Letters, digits, dots, dashes and underscores.",value:d,onChange:t=>w(t.target.value),placeholder:"llm_judge_claude_config"}):s.jsxs("div",{style:{display:"flex",alignItems:"flex-end",gap:"0.5rem"},children:[s.jsx("div",{style:{flex:1},children:s.jsx(B,{id:"llm-config-select",titleText:"Select config",items:v,itemToString:t=>t?t.name:"",selectedItem:n,onChange:t=>{const e=t.selectedItem;c(e),r(null),e?x(e):u("")},placeholder:"Choose a YAML config"})}),(n==null?void 0:n.user_defined)&&s.jsx(D,{type:"blue",size:"md",children:"edited"})]}),s.jsx(q,{id:"llm-config-editor",labelText:"Config JSON (edit and save)",rows:20,value:h,onChange:t=>u(t.target.value),disabled:!T}),s.jsxs("div",{style:{display:"flex",gap:"0.5rem"},children:[s.jsx(_,{kind:"primary",onClick:()=>void $(),disabled:!T||p,children:d!==null?"Create config":"Save config"}),d!==null&&s.jsx(_,{kind:"secondary",onClick:V,disabled:p,children:"Cancel"}),d===null&&(n==null?void 0:n.user_defined)&&s.jsx(_,{kind:"danger--tertiary",renderIcon:P,onClick:()=>void J(),disabled:p,children:"Delete"})]})]})};export{K as LLMJudgeConfigView};