---
license: cc-by-4.0
task_categories:
  - text-generation
language:
  - en
tags:
  - code
  - java
  - test-generation
  - software-engineering
  - benchmark
size_categories:
  - n<1K
configs:
  - config_name: default
    data_files: data/train.jsonl
---

# TestEvo-Bench / teb-generation

A benchmark for **test-case generation** in evolving Java codebases. Each row captures a single (project, revision-pair) where new tests were authored to exercise newly-added or changed production code, and surfaces both the test code and the surrounding production context needed to predict it.

## Links

- 🌐 Project website: https://www.testevo-bench.com/
- 📦 Code repository: https://anonymous.4open.science/r/testevo-bench-1150
- 🤝 Sibling dataset: [`TestEvo-Bench/teb-update`](https://huggingface.co/datasets/TestEvo-Bench/teb-update) — for **test-update** (modifying existing tests instead of generating new ones)

## Dataset summary

`teb-generation` is a benchmark for generating new test methods that exercise newly-introduced or changed production code. Every example was mined from a real commit (`rev1` → `rev2`) on a public Java project, and the new test is one that depends on the production-code change to behave correctly: it cannot be written or satisfied by reading only the pre-change codebase.

Each row corresponds to a single (project, `rev1` → `rev2`) pair. Inside the row, `test_changes` is a list of one or more individual test methods authored in that commit, each paired with its focal production method, the production-side before/after, candidate dependency methods, and the dependencies that changed in the same commit.

| | |
|---|---|
| Tasks (rows) | 746 |
| Test methods | 1,961 |
| Source projects | 84 distinct Java repositories on GitHub |
| Java versions | 8, 11, 17, 21 |
| Build system | Maven |
| Total size | ~8 MB |

A **task** is one (project, rev1 → rev2) pair — the unit a model is asked to operate on. A **test method** is one entry inside that task's `test_changes` list — a single `@Test` method that was added or modified at `rev2`. One task may contain several related test methods written together; benchmarks can be reported at either granularity.

## Schema

Each row has the following top-level fields:

| Field | Type | Description |
|---|---|---|
| `task_id` | string | Stable identifier: `<project>__<rev1_short>_<rev2_short>__<test_class>_<hash>` |
| `project_name` | string | e.g. `Javen205_IJPay-v2.8.4` |
| `git_clone_url` | string | Upstream GitHub clone URL |
| `rev1` | string | Git SHA before the change (full 40-char) |
| `rev2` | string | Git SHA after the change |
| `rev1_date` | string | ISO 8601 (e.g. `2019-12-08T06:45:32Z`) |
| `rev2_date` | string | ISO 8601 |
| `git_diff_url` | string | GitHub `compare` URL between rev1 and rev2 |
| `test_file` | string | Path of the test file within the repo |
| `test_changes` | list&lt;struct&gt; | One element per individual test method that was added/changed; see below |
| `version` | string | Upstream tag or branch (e.g. `v2.8.4`, `master`) |
| `java_version` | int | Java major version (8, 11, 17, or 21) |

Each `test_changes[]` element contains:

| Field | Type | Description |
|---|---|---|
| `change_id` | string | Unique id within the dataset |
| `test_sign` | string | JVM-style test method signature |
| `class`, `method`, `module` | string | Test class, method name, Maven module |
| `junit_selector` | string | JUnit selector usable by `mvn test -Dtest=…` |
| `old_test_code` | string \| null | The test method's source at `rev1` (null for newly-added tests) |
| `new_test_code` | string | The test method's source at `rev2` |
| `old_test_lnum`, `new_test_lnum` | list&lt;string&gt; | Line ranges of the test method at each rev |
| `old_test_file_path` | list&lt;string&gt; | Where the test file lived at `rev1` (may be empty for new files) |
| `old_production_code`, `new_production_code` | string | Source of the focal production method at each rev |
| `focal_file_path` | string | Path of the focal production file |
| `focal_method_sign` | string | JVM-style signature of the focal production method |
| `focal_all_deps_scored` | list&lt;struct&gt; | All candidate dependency methods, each with a relevance score (`nc`, `lcs_u`, `ed`, `class`, `tfidf` signal sub-scores combined into `score`) |
| `deps_changes` | list&lt;struct&gt; | Production-code dependencies that changed between `rev1` and `rev2` |

## Example row

```json
{
  "task_id": "Javen205_IJPay-v2.8.4__ceafd24_996582e__WxPayKitTest_8649b344",
  "project_name": "Javen205_IJPay-v2.8.4",
  "git_clone_url": "https://github.com/Javen205/IJPay.git",
  "rev1": "ceafd245bef465d703c7530111626de905893657",
  "rev2": "996582e5028204b7ba2938ccaf009a3715ed204b",
  "rev1_date": "2019-12-08T06:45:32Z",
  "rev2_date": "2019-12-08T07:49:34Z",
  "git_diff_url": "https://github.com/Javen205/IJPay/compare/ceafd24…...996582e…",
  "test_file": "IJPay-WxPay/src/test/java/com/ijpay/wxpay/WxPayKitTest.java",
  "test_changes": [ /* 1 or more entries — see Schema */ ],
  "version": "v2.8.4",
  "java_version": 11
}
```

## How to load

```python
from datasets import load_dataset

ds = load_dataset("TestEvo-Bench/teb-generation", split="train")
print(len(ds), ds.column_names)

# Iterate over individual test changes (the unit of evaluation)
for row in ds:
    for tc in row["test_changes"]:
        focal = tc["focal_method_sign"]
        new_test = tc["new_test_code"]
        ...
```

## Source projects + attribution

The code snippets in `old_test_code`, `new_test_code`, `old_production_code`, and `new_production_code` are derived from public Java repositories on GitHub. Each row's `git_clone_url` and `git_diff_url` link back to the upstream project; the upstream repository's own license governs reuse of those code snippets. The dataset structure and metadata are released under CC-BY-4.0 — please cite the project (see below) when using this benchmark.

## Intended uses

Evaluating LLMs and program-synthesis systems on the task of generating a new test method given (a) the production-code change, (b) the test file's prior context, and (c) candidate focal-method dependencies. The included `focal_all_deps_scored` signals support both retrieval-augmented and end-to-end setups.

## License

- **Dataset structure and metadata:** CC-BY-4.0
- **Embedded code snippets:** governed by each upstream repository's own license. See `git_clone_url` per row.