Exercism tracks inherit exercise definitions from the [problem-specifications] repository in the form of description files
(from which exercise READMEs are [generated](#generating-exercise-readmes))
## Implementing an exercise
### Exercise structure
```Bash
exercises/[EXERCISE]/
├── [EXERCISE].py
├── [EXERCISE]_test.py
├── example.py
├── .meta
│ ├── template.j2
│ ├── additional_tests.json
│ └── hints.md
└── README.md
```
Files:
| File | Description | Source |
|:--- |:--- |:--- |
| [[EXERCISE].py](exercises/two-fer/two_fer.py) | Solution stub | Manually created by the implementer |
| [[EXERCISE]_test.py](exercises/two-fer/two_fer_test.py) | Exercise test suite | Automatically generated if `.meta/template.j2` is present, otherwise manually created by the implementer |
| [example.py](exercises/two-fer/example.py) | Example solution used to automatically verify the `[EXERCISE]_test.py` suite | Manually created by the implementer |
| [.meta/template.j2](exercises/two-fer/.meta/template.j2) | Test generation template; if present used to automatically generate `[EXERCISE]_test.py` (See [generator documentation](docs/GENERATOR.md)) | Manually created by implementer |
| [.meta/additional_tests.json](exercises/word-count/.meta/additional_tests.json) | Defines additional track-specific test cases; if `.meta/template.j2` is also present these test will be incorporated into the automatically generated `[EXERCISE]_test.py` | Manually created by the implementer |
| [.meta/hints.md](exercises/high-scores/.meta/hints.md) | Contains track-specific hints that are automatically included in the generated `README.md` file | Manually created by the implementer |
If an unimplemented exercise has a `canonical-data.json` file in the [problem-specifications] repository, a generation template must be created. See the [test generator documentation](docs/GENERATOR.md) for more information.
If an unimplemented exercise does not have a `canonical-data.json` file, the test file must be written manually (use existing test files for examples).
### Example solutions
Example solution files serve two purposes:
1. Verification of the tests
2. Example implementation for mentor/student reference
### config.json
[`config.json`](config.json) is used by the website to determine which exercises to load an in what order. It also contains some exercise metadata, such as difficulty, labels, and if the exercise is a core exercise. New entries should be places just before the first exercise that is marked `"deprecated": true`:
```JSON
{
"slug": "current-exercise",
"uuid": "aaaaaaaa-aaaa-aaaa-aaaa-aaaaaaaaaaaa",
"core": false,
"unlocked_by": null,
"difficulty": 1,
"topics": [
"strings"
]
},
<<<HERE
{
"slug": "old-exercise",
"uuid": "aaaaaaaa-aaaa-aaaa-aaaa-aaaaaaaaaaaa",
"core": false,
"unlocked_by": null,
"difficulty": 2,
"topics": null,
"status": "deprecated"
},
```
Fields
<table>
<tr>
<td>slug</td>
<td>Hyphenated lowercase exercise name</td>
</tr>
<tr>
<td>uuid</td>
<td>Generate using <code>configlet uuid</code></td>
</tr>
<tr>
<td>core</td>
<td>Set to <code>false</code>; core exercises are decided by track maintainers</td>
</tr>
<tr>
<td>unlocked_by</td>
<td>Slug for the core exercise that unlocks the new one</td>
</tr>
<tr>
<td>difficulty</td>
<td><code>1</code> through <code>10</code>. Discuss with reviewer if uncertain.</td>
</tr>
<tr>
<td>topics</td>
<td>Array of relevant topics from the <ahref="https://github.com/exercism/problem-specifications/blob/master/TOPICS.txt">topics list</a></td>
</tr>
</table>
## Implementing Track-specific Exercises
Similar to implementing a canonical exercise that has no `canonical-data.json`, but the exercise README will also need to be written manually. Carefully follow the structure of generated exercise READMEs.
