Metadata-Version: 2.4
Name: stockfish
Version: 4.0.3
Summary: Wraps the open-source Stockfish chess engine for easy integration into python.
Home-page: https://github.com/py-stockfish/stockfish
Author: Ilya Zhelyabuzhsky
Author-email: zhelyabuzhsky@icloud.com
License: MIT
Keywords: chess stockfish
Classifier: Programming Language :: Python
Classifier: Natural Language :: English
Classifier: Operating System :: Unix
Classifier: License :: OSI Approved :: MIT License
Classifier: Development Status :: 5 - Production/Stable
Classifier: Topic :: Games/Entertainment :: Board Games
Classifier: Topic :: Software Development :: Libraries :: Python Modules
Classifier: Programming Language :: Python :: 3.8
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: Programming Language :: Python :: 3.14
Classifier: Programming Language :: Python :: Implementation :: CPython
Requires-Python: >=3.10
Description-Content-Type: text/markdown
License-File: LICENSE
Dynamic: author
Dynamic: author-email
Dynamic: classifier
Dynamic: description
Dynamic: description-content-type
Dynamic: home-page
Dynamic: keywords
Dynamic: license
Dynamic: license-file
Dynamic: requires-python
Dynamic: summary

# Stockfish

> [!NOTE]
> This section refers to the technical application. If you are looking for information regarding the status of this project and the original repo, please look [here](https://github.com/py-stockfish/stockfish/tree/master#status-of-the-project).

Implements an easy-to-use Stockfish class to integrates the Stockfish chess engine with Python.

## Install

```bash
$ pip install stockfish
```

#### Ubuntu

```bash
$ sudo apt install stockfish
```

#### Mac OS

```bash
$ brew install stockfish
```

## API Documentation

See [API Documentation](https://py-stockfish.github.io/stockfish/) for more information.

## Technical details and setup

* The codebase is compatible with Python 3.10+
* Run `pip install -r requirements.txt` to get all necessary dependencies.
* Pytest is used to run the unit tests in `/tests`
* For linting use `pre-commit` by running `pre-commit install` once and the pre-commit hooks will be executed automatically

## Features and usage examples

### Initialize Stockfish class

You should install the stockfish engine in your operating system globally or specify path to binary file in class constructor

```python
from stockfish import Stockfish

stockfish = Stockfish(path="/Users/zhelyabuzhsky/Work/stockfish/stockfish-9-64")
```

There are some default engine settings used by this wrapper. For increasing Stockfish's strength and speed, the "Threads" and "Hash" parameters can be modified.

```python
{
    "Debug Log File": "",
    "Contempt": 0,
    "Min Split Depth": 0,
    "Threads": 1, # More threads will make the engine stronger, but should be kept at less than the number of logical processors on your computer.
    "Ponder": False,
    "Hash": 16, # Default size is 16 MB. It's recommended that you increase this value, but keep it as some power of 2. E.g., if you're fine using 2 GB of RAM, set Hash to 2048 (11th power of 2).
    "MultiPV": 1,
    "Skill Level": 20,
    "Move Overhead": 10,
    "Minimum Thinking Time": 20,
    "Slow Mover": 100,
    "UCI_Chess960": False,
    "UCI_LimitStrength": False,
    "UCI_Elo": 1350
}
```

You can change them, as well as the default search depth, during your Stockfish class initialization:

```python
stockfish = Stockfish(path="/Users/zhelyabuzhsky/Work/stockfish/stockfish-9-64", depth=18, parameters={"Threads": 2, "Minimum Thinking Time": 30})
```

These parameters can also be updated at any time by calling the "update_engine_parameters" function:

```python
stockfish.update_engine_parameters({"Hash": 2048, "UCI_Chess960": True}) # Gets stockfish to use a 2GB hash table, and also to play Chess960.
```

When you're done using the Stockfish engine process, you can send the "quit" uci command to it with:

```python
stockfish.send_quit_command()
```

The `__del__()` method of the Stockfish class will call send_quit_command(), but it's technically not guaranteed python will call `__del__()` when the Stockfish object goes out of scope. So even though it'll probably not be needed, it doesn't hurt to call send_quit_command() yourself.

### Set position by a sequence of moves from the starting position

```python
stockfish.set_position(["e2e4", "e7e6"])
```

### Update position by making a sequence of moves from the current position

```python
stockfish.make_moves_from_current_position(["g4d7", "a8b8", "f1d1"])
```

### Set position by Forsyth–Edwards Notation (FEN)

If you'd like to first check if your fen is valid, call the is_fen_valid() function below.
Also, if you want to play Chess960, it's recommended you first update the "UCI_Chess960" engine parameter to be True, before calling set_fen_position.

```python
stockfish.set_fen_position("rnbqkbnr/pppp1ppp/4p3/8/4P3/8/PPPP1PPP/RNBQKBNR w KQkq - 0 2")
```

### Check whether the given FEN is valid

This function returns a bool saying whether the passed in FEN is valid (both syntax wise and whether the position represented is legal).
The function isn't perfect and won't catch all cases, but generally it should return the correct answer.
For example, one exception is positions which are legal, but have no legal moves.
I.e., for checkmates and stalemates, this function will incorrectly say the fen is invalid.

```python
stockfish.is_fen_valid("rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1")
```

```text
True
```

```python
stockfish.is_fen_valid("rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq -") # will return False, in this case because the FEN is missing two of the six required fields.
```

```text
False
```

### Get best move

```python
stockfish.get_best_move()
```

```text
d2d4
```

It's possible to specify remaining time on black and/or white clock. Time is in milliseconds.

```python
stockfish.get_best_move(wtime=1000, btime=1000)
```

### Get best move based on a time constraint

```python
stockfish.get_best_move_time(1000)
```

Time constraint is in milliseconds

```text
e2e4
```

### Check if a move is legal in the current position

```python
stockfish.is_move_correct('a2a3')
```

```text
True
```

### Get info on the top n moves

Get moves, centipawns, and mates for the top n moves. If the move is a mate, the Centipawn value will be None, and vice versa. Note that if you have stockfish on a weaker elo or skill level setting, the top moves returned by this
function will still be for full strength.

```python
stockfish.get_top_moves(3)
# [
#   {'Move': 'f5h7', 'Centipawn': None, 'Mate': 1},
#   {'Move': 'f5d7', 'Centipawn': 713, 'Mate': None},
#   {'Move': 'f5h5', 'Centipawn': -31, 'Mate': None}
# ]
```

Optional parameter `verbose` (default `False`) specifies whether to include the full info from the engine in the returned dictionary, including SelectiveDepth, Nodes, NodesPerSecond, Time, MultiPVLine, and WDL if available.

```py
stockfish.get_top_moves(1, verbose=True)
# [{
#   "Move":"d6e7",
#   "Centipawn":-408,
#   "Mate":"None",
#   "Nodes":"2767506",
#   "NodesPerSecond":"526442",
#   "Time":"5257",
#   "SelectiveDepth":"31",
#   "MultiPVLine":"1",
#   "WDL":"0 0 1000"
# }]
```

Optional parameter `num_nodes` specifies the number of nodes to search. If num_nodes is 0, then the engine will search until the configured depth is reached.

### Get perft information

The [perft](https://www.chessprogramming.org/Perft) command is used to test the move generation. It counts the total number of leaf nodes to a certain depth, and shows how this node count is divided amongst all legal moves
of the current position.

The `depth` parameter should be an integer greater than zero and specifies the search depth.

```python
stockfish.get_perft(3)
# (8902, {'a2a3': 380, 'b2b3': 420, 'c2c3': 420, 'd2d3': 539,
#         'e2e3': 599, 'f2f3': 380, 'g2g3': 420, 'h2h3': 380,
#         'a2a4': 420, 'b2b4': 421, 'c2c4': 441, 'd2d4': 560,
#         'e2e4': 600, 'f2f4': 401, 'g2g4': 421, 'h2h4': 420,
#         'b1a3': 400, 'b1c3': 440, 'g1f3': 440, 'g1h3': 400})
```

### Flip

Flip the side to move.
```python
stockfish.get_fen_position()
# rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1
stockfish.flip()
stockfish.get_fen_position()
# rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR b KQkq - 0 1
```

### Set perspective of the evaluation

You can set the perspective of the evaluation to be from the perspective of the side to move, or from the perspective of White. Currently this setting only applies to `get_top_moves()`.

```py
# Set the perspective of the evaluation to be from the point of view of the side to move
stockfish.set_turn_perspective(True)

# Set the perspective of the evaluation to be from White's perspective
stockfish.set_turn_perspective(False)

# Get the current perspective of the evaluation
is_turn_perspective = stockfish.get_turn_perspective()

```

### Get Stockfish's win/draw/loss stats for the side to move in the current position

Before calling this function, it is recommended that you first check if your version of Stockfish is recent enough to display WDL stats. To do this,
use the "does_current_engine_version_have_wdl_option()" function below.

```python
stockfish.get_wdl_stats() # include the argument get_as_tuple=True if you'd like to have a tuple returned instead of a list.
```

```text
[87, 894, 19]
```

### Find if your version of Stockfish is recent enough to display WDL stats

```python
stockfish.does_current_engine_version_have_wdl_option()
```

```text
True
```

### Set the engine's skill level (ignoring ELO rating)

```python
stockfish.set_skill_level(15)
```

### Set the engine's ELO rating (ignoring skill level)

```python
stockfish.set_elo_rating(1350)
```

### Put the engine back to full strength (if you've previously lowered the ELO or skill level)
```python
stockfish.resume_full_strength()
```

### Set the engine's depth

```python
stockfish.set_depth(15)
```

### Get the engine's current parameters

Returns a deep copy of the dictionary storing the engine's current parameters.

```python
stockfish.get_engine_parameters()
```

```text
{
    "Debug Log File": "",
    "Contempt": 0,
    "Min Split Depth": 0,
    "Threads": 1,
    "Ponder": False,
    "Hash": 16,
    "MultiPV": 1,
    "Skill Level": 20,
    "Move Overhead": 10,
    "Minimum Thinking Time": 20,
    "Slow Mover": 100,
    "UCI_Chess960": False,
    "UCI_LimitStrength": False,
    "UCI_Elo": 1350
}
```

### Reset the engine's parameters to the default

```python
stockfish.reset_engine_parameters()
```

### Get the current board position in Forsyth–Edwards notation (FEN)

```python
stockfish.get_fen_position()
```

```text
rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1
```

### Get the current board visual

```python
stockfish.get_board_visual()
```

```text
+---+---+---+---+---+---+---+---+
| r | n | b | q | k | b | n | r | 8
+---+---+---+---+---+---+---+---+
| p | p | p | p | p | p | p | p | 7
+---+---+---+---+---+---+---+---+
|   |   |   |   |   |   |   |   | 6
+---+---+---+---+---+---+---+---+
|   |   |   |   |   |   |   |   | 5
+---+---+---+---+---+---+---+---+
|   |   |   |   |   |   |   |   | 4
+---+---+---+---+---+---+---+---+
|   |   |   |   |   |   |   |   | 3
+---+---+---+---+---+---+---+---+
| P | P | P | P | P | P | P | P | 2
+---+---+---+---+---+---+---+---+
| R | N | B | Q | K | B | N | R | 1
+---+---+---+---+---+---+---+---+
  a   b   c   d   e   f   g   h
```

This function has an optional boolean (True by default) as a parameter that indicates whether the board should be seen from the view of white. So it is possible to get the board from black's point of view like this:

```python
stockfish.get_board_visual(False)
```

```text
+---+---+---+---+---+---+---+---+
| R | N | B | K | Q | B | N | R | 1
+---+---+---+---+---+---+---+---+
| P | P | P | P | P | P | P | P | 2
+---+---+---+---+---+---+---+---+
|   |   |   |   |   |   |   |   | 3
+---+---+---+---+---+---+---+---+
|   |   |   |   |   |   |   |   | 4
+---+---+---+---+---+---+---+---+
|   |   |   |   |   |   |   |   | 5
+---+---+---+---+---+---+---+---+
|   |   |   |   |   |   |   |   | 6
+---+---+---+---+---+---+---+---+
| p | p | p | p | p | p | p | p | 7
+---+---+---+---+---+---+---+---+
| r | n | b | k | q | b | n | r | 8
+---+---+---+---+---+---+---+---+
  h   g   f   e   d   c   b   a
```

### Get the current position's evaluation in centipawns or mate in x

Stockfish searches to the specified depth and evaluates the current position:
```python
stockfish.get_evaluation()
```

Instead of using the depth, you can also specify the time the engine should take to evaluate:
```python
stockfish.get_evaluation(searchtime=2000) # searchtime in milliseconds
```

A dictionary is returned representing the evaluation. Two example return values:

```text
{"type":"cp", "value":12}

{"type":"mate", "value":-3}
```

If stockfish.get_turn_perspective() is True, then the eval value is relative to the side to move.
Otherwise, positive is advantage white, negative is advantage black.

### Get the current position's 'static evaluation'

```python
stockfish.get_static_eval()
```

Sends the 'eval' command to Stockfish. This will get it to 'directly' evaluate the current position
(i.e., no search is involved), and output a float value (not a whole number centipawn).

If one side is in check or mated, recent versions of Stockfish will output 'none' for the static eval.
In this case, the function will return None.

Some example return values:

```text
-5.27

0.28

None
```

### Run benchmark

#### BenchmarkParameters

```python
params = BenchmarkParameters(**kwargs)
```

parameters required to run the benchmark function. kwargs can be used to set custom values.

```text
ttSize: range(1,128001)
threads: range(1,513)
limit: range(1,10001)
fenFile: "path/to/file.fen"
limitType: "depth", "perft", "nodes", "movetime"
evalType: "mixed", "classical", "NNUE"
```

```python
stockfish.benchmark(params)
```

This will run the bench command with BenchmarkParameters.
It is an additional custom non-UCI command, mainly for debugging.
Do not use this command during a search!

### Get the major version of the stockfish engine being used

E.g., if the engine being used is Stockfish 14.1 or Stockfish 14, then the function would return 14.
Meanwhile, if a development build of the engine is being used (not an official release), then the function returns an
int with 5 or 6 digits, representing the date the engine was compiled on.
For example, 20122 is returned for the development build compiled on January 2, 2022.

```python
stockfish.get_stockfish_major_version()
```

```text
15
```

### Find if the version of Stockfish being used is a development build

```python
stockfish.is_development_build_of_engine()
```

```text
False
```

### Find what is on a certain square

If the square is empty, the None object is returned. Otherwise, one of 12 enum members of a custom
Stockfish.Piece enum will be returned. Each of the 12 members of this enum is named in the following pattern:
_colour_ followed by _underscore_ followed by _piece name_, where the colour and piece name are in all caps.
For example, say the current position is the starting position:

```python
stockfish.get_what_is_on_square("e1") # returns Stockfish.Piece.WHITE_KING
stockfish.get_what_is_on_square("d8") # returns Stockfish.Piece.BLACK_QUEEN
stockfish.get_what_is_on_square("h2") # returns Stockfish.Piece.WHITE_PAWN
stockfish.get_what_is_on_square("b5") # returns None
```

### Find if a move will be a capture (and if so, what type of capture)

The argument must be a string that represents the move, using the notation that Stockfish uses (i.e., the coordinate of the starting square followed by the coordinate of the ending square).
The function will return one of the following enum members from a custom Stockfish.Capture enum: DIRECT_CAPTURE, EN_PASSANT, or NO_CAPTURE.
For example, say the current position is the one after 1.e4 Nf6 2.Nc3 e6 3.e5 d5.

```python
stockfish.will_move_be_a_capture("c3d5")  # returns Stockfish.Capture.DIRECT_CAPTURE
stockfish.will_move_be_a_capture("e5f6")  # returns Stockfish.Capture.DIRECT_CAPTURE
stockfish.will_move_be_a_capture("e5d6")  # returns Stockfish.Capture.EN_PASSANT
stockfish.will_move_be_a_capture("f1e2")  # returns Stockfish.Capture.NO_CAPTURE
```

### StockfishException

The `StockfishException` is a newly defined Exception type. It is thrown when the underlying Stockfish process created by the wrapper crashes. This can happen when an incorrect input like an invalid FEN (for example `8/8/8/3k4/3K4/8/8/8 w - - 0 1` with both kings next to each other) is given to Stockfish. \
Not all invalid inputs will lead to a `StockfishException`, but only those which cause the Stockfish process to crash. \
To handle a `StockfishException` when using this library, import the `StockfishException` from the library and use a `try/except`-block:

```python
from stockfish import StockfishException

try:
    # Evaluation routine

except StockfishException:
    # Error handling
```

### Debug view

You can (de-)activate the debug view option with the `set_debug_view` function. Like this you can see all communication between the engine and the library.

```python
stockfish.set_debug_view(True)
```

## Contributing

Most contributions will involve making updates to `stockfish/models.py`. To test your changes, download a version of stockfish and paste the executable in the `stockfish` folder. Then, create a file in the `stockfish` folder called `main.py`. Both the executable and `main.py` will be ignored by git.
In `main.py`, start with something like the following:

```python
from models import Stockfish

def main():
    sf = Stockfish(path = "name of your stockfish executable")
    # Use this object as you wish to test your changes.

if __name__ == "__main__":
    main()
```
Then when navigating to the `stockfish` folder in the terminal, you can run this `main.py` file simply with `python main.py`.
Once you're satisfied with your changes to `models.py`, see the section below for how to run the project's entire test suite.

## Testing

For your stockfish executable (the same one mentioned in the previous section), paste it also in the project's root directory. Then in `models.py`,
temporarily modify this line so that the `path`'s default value is changed from "stockfish" to the name of your stockfish executable:

```python
path: str = "stockfish",
```

Then in the project's root directory, you can run:

```bash
$ python setup.py test
```
To skip some of the slower tests, run:
```bash
$ python setup.py skip_slow_tests
```

## Security

If you discover any security related issues, please report it via the [Private vulnerability reporting](https://github.com/py-stockfish/stockfish/security) instead of using the issue tracker.

## Status of the project

> **Note**
> This is just a brief summary. For more information, please look [here](https://github.com/zhelyabuzhsky/stockfish/issues/130).

Due to the [unfortunate death](https://github.com/zhelyabuzhsky/stockfish/pull/112#issuecomment-1367800036) of [Ilya Zhelyabuzhsky](https://github.com/zhelyabuzhsky), the original [repo](https://github.com/zhelyabuzhsky/stockfish) is no longer maintained. For this reason, this fork was created, which continues the project and is currently maintained by [johndoknjas](https://github.com/johndoknjas) and [kieferro](https://github.com/kieferro).
The official PyPi releases for the [Stockfish package](https://pypi.org/project/stockfish/) will also be created from this repo in the future.

Please submit all bug reports and PRs to this repo instead of the old one.

## Credits

- We want to sincerely thank [Ilya Zhelyabuzhsky](https://github.com/zhelyabuzhsky), the original founder of this project for writing and maintaining the code and for his contributions to the open source community.
- We also want to thank all the [other contributors](https://github.com/py-stockfish/stockfish/graphs/contributors) for working on this project.

## License

MIT License. Please see [License File](https://github.com/py-stockfish/stockfish/blob/master/LICENSE) for more information.
