rgeval 0.1.0__tar.gz

rgeval-0.1.0/.gitignore

@@ -0,0 +1,262 @@
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[codz]
+*$py.class
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+share/python-wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+MANIFEST
+
+# PyInstaller
+#   Usually these files are written by a python script from a template
+#   before PyInstaller builds the exe, so as to inject date/other infos into it.
+*.manifest
+*.spec
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.nox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+*.py.cover
+*.lcov
+.hypothesis/
+.pytest_cache/
+cover/
+
+# Translations
+*.mo
+*.pot
+
+# Django stuff:
+*.log
+local_settings.py
+db.sqlite3
+db.sqlite3-journal
+
+# Flask stuff:
+instance/
+.webassets-cache
+
+# Scrapy stuff:
+.scrapy
+
+# Sphinx documentation
+docs/_build/
+
+# PyBuilder
+.pybuilder/
+target/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# IPython
+profile_default/
+ipython_config.py
+
+# pyenv
+#   For a library or package, you might want to ignore these files since the code is
+#   intended to run in multiple environments; otherwise, check them in:
+# .python-version
+
+# pipenv
+#   According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
+#   However, in case of collaboration, if having platform-specific dependencies or dependencies
+#   having no cross-platform support, pipenv may install dependencies that don't work, or not
+#   install all needed dependencies.
+# Pipfile.lock
+
+# uv
+#   Similar to Pipfile.lock, it is generally recommended to include uv.lock in version control.
+#   This is especially recommended for binary packages to ensure reproducibility, and is more
+#   commonly ignored for libraries.
+# uv.lock
+
+# poetry
+#   Similar to Pipfile.lock, it is generally recommended to include poetry.lock in version control.
+#   This is especially recommended for binary packages to ensure reproducibility, and is more
+#   commonly ignored for libraries.
+#   https://python-poetry.org/docs/basic-usage/#commit-your-poetrylock-file-to-version-control
+# poetry.lock
+# poetry.toml
+
+# pdm
+#   Similar to Pipfile.lock, it is generally recommended to include pdm.lock in version control.
+#   pdm recommends including project-wide configuration in pdm.toml, but excluding .pdm-python.
+#   https://pdm-project.org/en/latest/usage/project/#working-with-version-control
+# pdm.lock
+# pdm.toml
+.pdm-python
+.pdm-build/
+
+# pixi
+#   Similar to Pipfile.lock, it is generally recommended to include pixi.lock in version control.
+# pixi.lock
+#   Pixi creates a virtual environment in the .pixi directory, just like venv module creates one
+#   in the .venv directory. It is recommended not to include this directory in version control.
+.pixi/*
+!.pixi/config.toml
+
+# PEP 582; used by e.g. github.com/David-OConnor/pyflow and github.com/pdm-project/pdm
+__pypackages__/
+
+# Celery stuff
+celerybeat-schedule*
+celerybeat.pid
+
+# Redis
+*.rdb
+*.aof
+*.pid
+
+# RabbitMQ
+mnesia/
+rabbitmq/
+rabbitmq-data/
+
+# ActiveMQ
+activemq-data/
+
+# SageMath parsed files
+*.sage.py
+
+# Environments
+.env
+.envrc
+.venv
+env/
+venv/
+ENV/
+env.bak/
+venv.bak/
+
+# Spyder project settings
+.spyderproject
+.spyproject
+
+# Rope project settings
+.ropeproject
+
+# mkdocs/Zensical documentation
+/site
+
+# mypy
+.mypy_cache/
+.dmypy.json
+dmypy.json
+
+# Pyre type checker
+.pyre/
+
+# pytype static type analyzer
+.pytype/
+
+# Cython debug symbols
+cython_debug/
+
+# PyCharm
+#   JetBrains specific template is maintained in a separate JetBrains.gitignore that can
+#   be found at https://github.com/github/gitignore/blob/main/Global/JetBrains.gitignore
+#   and can be added to the global gitignore or merged into this file.  For a more nuclear
+#   option (not recommended) you can uncomment the following to ignore the entire idea folder.
+# .idea/
+
+# Abstra
+#   Abstra is an AI-powered process automation framework.
+#   Ignore directories containing user credentials, local state, and settings.
+#   Learn more at https://abstra.io/docs
+.abstra/
+
+# Visual Studio Code
+#   Visual Studio Code specific template is maintained in a separate VisualStudioCode.gitignore that
+#   can be found at https://github.com/github/gitignore/blob/main/Global/VisualStudioCode.gitignore
+#   and can be added to the global gitignore or merged into this file. However, if you prefer, you
+#   could uncomment the following to ignore the entire vscode folder
+# .vscode/
+# Temporary file for partial code execution
+tempCodeRunnerFile.py
+
+# Ruff stuff:
+.ruff_cache/
+
+# PyPI configuration file
+.pypirc
+
+# Marimo
+marimo/_static/
+marimo/_lsp/
+__marimo__/
+
+# Streamlit
+.streamlit/secrets.toml
+
+# ---------------------------------------------------------------------------
+# Project-specific
+# ---------------------------------------------------------------------------
+# Generated at build time by hatch-vcs from the git tag. Committing it makes the
+# tree dirty after every build, and setuptools_scm then reports a .dev version
+# instead of the tag.
+src/rgeval/_version.py
+
+# No datasets or model predictions in this repo. Datasets live in
+# https://github.com/nl2graphql-bench/datasets; predictions are an input the
+# user supplies, and scoring writes its own output directories.
+*.jsonl
+*.jsonl.gz
+*.xlsx
+*.parquet
+*.csv
+2026_evals/
+2026_input/
+2026_results/
+2024_evals/
+2024_input/
+2024_results/
+emnlp_inputs/
+emnlp_outputs/
+every_prediction/
+review_evals/
+test_input_files/
+test_output_files/
+
+# The 2,100-row hidden benchmark must never be committed anywhere, public or
+# private: it is what leaderboard submissions are scored against.
+secret_data.jsonl
+secret_dataset*
+secret_*
+*hidden*
+
+# Credentials
+credentials.json
+config.local.yaml
+*.pem
+*.key

rgeval-0.1.0/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 NL2GraphQL Bench contributors
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

rgeval-0.1.0/PKG-INFO

@@ -0,0 +1,158 @@
+Metadata-Version: 2.5
+Name: rgeval
+Version: 0.1.0
+Summary: Semantic-equivalence scoring for generated GraphQL queries
+Project-URL: Homepage, https://nl2graphql-bench.github.io/
+Project-URL: Repository, https://github.com/nl2graphql-bench/evaluation
+Project-URL: Datasets, https://github.com/nl2graphql-bench/datasets
+Project-URL: Issues, https://github.com/nl2graphql-bench/evaluation/issues
+Author: Keya Battu, Parth Mehta, Harikrishnan Balagopal, Manish Kesarwani
+License-Expression: MIT
+License-File: LICENSE
+Keywords: benchmark,evaluation,graphql,nl2graphql,text-to-graphql
+Classifier: Development Status :: 4 - Beta
+Classifier: Intended Audience :: Science/Research
+Classifier: License :: OSI Approved :: MIT License
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.9
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
+Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
+Requires-Python: >=3.9
+Requires-Dist: graphql-core<3.3,>=3.2
+Requires-Dist: matplotlib>=3.5
+Requires-Dist: networkx<3,>=2.8
+Requires-Dist: pyyaml>=6.0
+Provides-Extra: datasets
+Requires-Dist: openpyxl>=3.1; extra == 'datasets'
+Requires-Dist: pandas>=2.0; extra == 'datasets'
+Provides-Extra: dev
+Requires-Dist: build>=1.0; extra == 'dev'
+Requires-Dist: pytest>=7; extra == 'dev'
+Requires-Dist: twine>=5.0; extra == 'dev'
+Provides-Extra: ui
+Requires-Dist: fastapi>=0.100; extra == 'ui'
+Requires-Dist: uvicorn>=0.23; extra == 'ui'
+Description-Content-Type: text/markdown
+
+# rgeval
+
+Semantic-equivalence scoring for generated GraphQL queries. The scorer behind
+the [NL2GraphQL Bench leaderboard](https://nl2graphql-bench.github.io/).
+
+A generated GraphQL query can be correct without matching the reference text:
+field order varies, aliases rename things, fragments can be inlined. RGEval
+parses and validates both queries against the schema, normalises them, and
+compares them as graphs, so it scores meaning rather than string equality.
+
+## Install
+
+```bash
+pip install rgeval
+```
+
+Reading the benchmark `.xlsx` files needs the extra:
+
+```bash
+pip install "rgeval[datasets]"
+```
+
+## Use as a library
+
+```python
+from rgeval import score_pair
+
+schema = """
+type Query { city(id: Int!): City }
+type City { id: Int! name: String population: Float }
+"""
+gold = 'query { city(id: 1) { name population } }'
+pred = 'query { city(id: 1) { population name } }'   # different field order
+
+score_pair(schema, gold, pred)        # True
+```
+
+`score_pair` returns `False` rather than raising for a query that cannot be
+parsed or does not validate against the schema: an invalid prediction scores 0.
+
+```python
+score_pair(schema, gold, pred, variables='{"cityId": 1}')
+```
+
+## Score the public benchmark
+
+Get the data from
+[nl2graphql-bench/datasets](https://github.com/nl2graphql-bench/datasets):
+
+```bash
+gunzip -k test_data.jsonl.gz
+```
+
+Use the `CleanSchema` field. `OriginalGraphQLSchema` and
+`RefinedGraphQLSchema` are truncated at 32,767 characters in some rows and will
+not build.
+
+Predictions are JSONL, one object per instance, with the model output in a field
+starting with `pred_`:
+
+```json
+{"id": 1, "gold": "query { ... }", "pred_mymodel": "query { ... }"}
+```
+
+```bash
+mkdir -p 2026_input/myrun
+cp predictions.jsonl 2026_input/myrun/test_main_predictions.jsonl
+rgeval-score-2026 --input_folder myrun
+```
+
+Results land in `2026_results/myrun/results.json` with an overall score and a
+per-category breakdown.
+
+Name prediction files `test_main_*` or `test_part2_*`. The scorer reads the part
+from the filename, because the two parts of the public benchmark number their
+rows from 1 each and a `DataID` alone does not say which part a row is from.
+
+## How scoring works
+
+| stage | module |
+|---|---|
+| parse and validate against the schema | `rgeval.pipeline.AST` |
+| AST normalisation | `rgeval.pipeline.normalizeAST` |
+| graph construction | `rgeval.pipeline.graphNet` |
+| cycle handling | `rgeval.pipeline.cycles` |
+| reduction | `rgeval.pipeline.newRed` |
+| comparison | `rgeval.pipeline.compare_jsons` |
+
+`rgeval.score_pair` runs all six in order. Import a stage directly only to
+inspect or replace it.
+
+## Requirements
+
+Python 3.9+. `networkx` is pinned below 3.0: the graph code uses APIs removed in
+3.x.
+
+## What is not here
+
+No datasets and no model predictions. Datasets are in
+[nl2graphql-bench/datasets](https://github.com/nl2graphql-bench/datasets);
+predictions are an input you supply.
+
+The 2,100-instance hidden benchmark is not published anywhere. It is what
+submissions are scored against, so if it were public a model could be tuned on
+it and the score would mean nothing.
+
+## Citation
+
+AutoGraphQL: A Bring-Your-Own-Schema Framework for Adapting Open-Weight LLMs to
+GraphQL APIs. EMNLP 2026, Industry Track.
+<https://openreview.net/forum?id=NcuPXBFKAO>
+
+## License
+
+MIT, see `LICENSE`. `NOTICE` records why: the source repository carried
+CC BY-NC-ND 4.0, which is not an OSI-approved open-source license and bars both
+commercial use and derivative works, neither of which suits a package people
+install and extend. The other NL2GraphQL Bench components and the upstream
+StepZen dataset are all MIT.

rgeval-0.1.0/README.md

@@ -0,0 +1,119 @@
+# rgeval
+
+Semantic-equivalence scoring for generated GraphQL queries. The scorer behind
+the [NL2GraphQL Bench leaderboard](https://nl2graphql-bench.github.io/).
+
+A generated GraphQL query can be correct without matching the reference text:
+field order varies, aliases rename things, fragments can be inlined. RGEval
+parses and validates both queries against the schema, normalises them, and
+compares them as graphs, so it scores meaning rather than string equality.
+
+## Install
+
+```bash
+pip install rgeval
+```
+
+Reading the benchmark `.xlsx` files needs the extra:
+
+```bash
+pip install "rgeval[datasets]"
+```
+
+## Use as a library
+
+```python
+from rgeval import score_pair
+
+schema = """
+type Query { city(id: Int!): City }
+type City { id: Int! name: String population: Float }
+"""
+gold = 'query { city(id: 1) { name population } }'
+pred = 'query { city(id: 1) { population name } }'   # different field order
+
+score_pair(schema, gold, pred)        # True
+```
+
+`score_pair` returns `False` rather than raising for a query that cannot be
+parsed or does not validate against the schema: an invalid prediction scores 0.
+
+```python
+score_pair(schema, gold, pred, variables='{"cityId": 1}')
+```
+
+## Score the public benchmark
+
+Get the data from
+[nl2graphql-bench/datasets](https://github.com/nl2graphql-bench/datasets):
+
+```bash
+gunzip -k test_data.jsonl.gz
+```
+
+Use the `CleanSchema` field. `OriginalGraphQLSchema` and
+`RefinedGraphQLSchema` are truncated at 32,767 characters in some rows and will
+not build.
+
+Predictions are JSONL, one object per instance, with the model output in a field
+starting with `pred_`:
+
+```json
+{"id": 1, "gold": "query { ... }", "pred_mymodel": "query { ... }"}
+```
+
+```bash
+mkdir -p 2026_input/myrun
+cp predictions.jsonl 2026_input/myrun/test_main_predictions.jsonl
+rgeval-score-2026 --input_folder myrun
+```
+
+Results land in `2026_results/myrun/results.json` with an overall score and a
+per-category breakdown.
+
+Name prediction files `test_main_*` or `test_part2_*`. The scorer reads the part
+from the filename, because the two parts of the public benchmark number their
+rows from 1 each and a `DataID` alone does not say which part a row is from.
+
+## How scoring works
+
+| stage | module |
+|---|---|
+| parse and validate against the schema | `rgeval.pipeline.AST` |
+| AST normalisation | `rgeval.pipeline.normalizeAST` |
+| graph construction | `rgeval.pipeline.graphNet` |
+| cycle handling | `rgeval.pipeline.cycles` |
+| reduction | `rgeval.pipeline.newRed` |
+| comparison | `rgeval.pipeline.compare_jsons` |
+
+`rgeval.score_pair` runs all six in order. Import a stage directly only to
+inspect or replace it.
+
+## Requirements
+
+Python 3.9+. `networkx` is pinned below 3.0: the graph code uses APIs removed in
+3.x.
+
+## What is not here
+
+No datasets and no model predictions. Datasets are in
+[nl2graphql-bench/datasets](https://github.com/nl2graphql-bench/datasets);
+predictions are an input you supply.
+
+The 2,100-instance hidden benchmark is not published anywhere. It is what
+submissions are scored against, so if it were public a model could be tuned on
+it and the score would mean nothing.
+
+## Citation
+
+AutoGraphQL: A Bring-Your-Own-Schema Framework for Adapting Open-Weight LLMs to
+GraphQL APIs. EMNLP 2026, Industry Track.
+<https://openreview.net/forum?id=NcuPXBFKAO>
+
+## License
+
+MIT, see `LICENSE`. `NOTICE` records why: the source repository carried
+CC BY-NC-ND 4.0, which is not an OSI-approved open-source license and bars both
+commercial use and derivative works, neither of which suits a package people
+install and extend. The other NL2GraphQL Bench components and the upstream
+StepZen dataset are all MIT.

rgeval-0.1.0/pyproject.toml

@@ -0,0 +1,80 @@
+[build-system]
+requires = ["hatchling>=1.21", "hatch-vcs>=0.4"]
+build-backend = "hatchling.build"
+
+[project]
+name = "rgeval"
+# Version comes from the git tag via hatch-vcs, so a release cannot publish a
+# number that disagrees with the tag it was cut from. Do not hardcode it here.
+dynamic = ["version"]
+description = "Semantic-equivalence scoring for generated GraphQL queries"
+readme = "README.md"
+requires-python = ">=3.9"
+license = "MIT"
+license-files = ["LICENSE"]
+authors = [
+  { name = "Keya Battu" },
+  { name = "Parth Mehta" },
+  { name = "Harikrishnan Balagopal" },
+  { name = "Manish Kesarwani" },
+]
+keywords = ["graphql", "nl2graphql", "evaluation", "benchmark", "text-to-graphql"]
+classifiers = [
+  "Development Status :: 4 - Beta",
+  "Intended Audience :: Science/Research",
+  "License :: OSI Approved :: MIT License",
+  "Programming Language :: Python :: 3",
+  "Programming Language :: Python :: 3.9",
+  "Programming Language :: Python :: 3.10",
+  "Programming Language :: Python :: 3.11",
+  "Programming Language :: Python :: 3.12",
+  "Programming Language :: Python :: 3.13",
+  "Topic :: Scientific/Engineering :: Artificial Intelligence",
+]
+dependencies = [
+  # networkx must stay below 3.0: the graph code uses APIs removed in 3.x.
+  "networkx>=2.8,<3",
+  # graphql-core must stay below 3.3: from 3.3 a node's `directives` is None
+  # when absent, where 3.2 gave [], and pipeline/normalizeAST.py iterates it
+  # directly (TypeError: 'NoneType' object is not iterable).
+  "graphql-core>=3.2,<3.3",
+  # pipeline/newRed.py and pipeline/preProcessor.py read YAML config.
+  "pyyaml>=6.0",
+  # pipeline/graphNet.py and pipeline/Graph.py import pyplot at module level.
+  "matplotlib>=3.5",
+]
+
+[project.optional-dependencies]
+# Reading the benchmark xlsx files and running the dataset preparation scripts.
+datasets = ["pandas>=2.0", "openpyxl>=3.1"]
+# app.py, the local UI for inspecting a scored run.
+ui = ["fastapi>=0.100", "uvicorn>=0.23"]
+dev = ["pytest>=7", "build>=1.0", "twine>=5.0"]
+
+[project.urls]
+Homepage = "https://nl2graphql-bench.github.io/"
+Repository = "https://github.com/nl2graphql-bench/evaluation"
+Datasets = "https://github.com/nl2graphql-bench/datasets"
+Issues = "https://github.com/nl2graphql-bench/evaluation/issues"
+
+[project.scripts]
+rgeval-score-2026 = "rgeval.eval_2026:main"
+rgeval-score-2024 = "rgeval.eval_2024:main"
+
+[tool.hatch.version]
+source = "vcs"
+# Tags are v-prefixed: v0.1.0 -> 0.1.0
+tag-pattern = "v(?P<version>.+)"
+
+[tool.hatch.build.hooks.vcs]
+# Generated at build time; the package reads its version from here.
+version-file = "src/rgeval/_version.py"
+
+[tool.hatch.build.targets.wheel]
+packages = ["src/rgeval"]
+
+[tool.hatch.build.targets.sdist]
+include = ["src/rgeval", "tests", "README.md", "LICENSE", "pyproject.toml"]
+
+[tool.pytest.ini_options]
+testpaths = ["tests"]

rgeval-0.1.0/src/rgeval/__init__.py

@@ -0,0 +1,15 @@
+"""RGEval: semantic-equivalence scoring for generated GraphQL queries.
+
+Compares a generated query against a reference by normalising both against the
+schema and comparing them as graphs, so a query that is correct but written
+differently is not scored wrong.
+"""
+try:
+    # Written at build time by hatch-vcs from the git tag.
+    from ._version import __version__
+except ImportError:                      # running from a source tree, not installed
+    __version__ = "0.0.0.dev0"
+
+from .scoring import score_pair
+
+__all__ = ["score_pair", "__version__"]

rgeval-0.1.0/src/rgeval/_version.py

@@ -0,0 +1,24 @@
+# file generated by vcs-versioning
+# don't change, don't track in version control
+from __future__ import annotations
+
+__all__ = [
+    "__version__",
+    "__version_tuple__",
+    "version",
+    "version_tuple",
+    "__commit_id__",
+    "commit_id",
+]
+
+version: str
+__version__: str
+__version_tuple__: tuple[int | str, ...]
+version_tuple: tuple[int | str, ...]
+commit_id: str | None
+__commit_id__: str | None
+
+__version__ = version = '0.1.0'
+__version_tuple__ = version_tuple = (0, 1, 0)
+
+__commit_id__ = commit_id = None