36 KiB
竞品分析与效果对比
1 思码逸devchat
官网 https://www.devchat.ai/zh 源码 workflow https://github.com/devchat-ai/workflows services https://github.com/devchat-ai/devchat vscode extension https://github.com/devchat-ai/devchat-vscode
devchat有两种处理逻辑
- 一次用prompt + code 的方式询问大模型LLM
- 多次询问LLM，第一次LLM会返回缺少的信息，在下一次询问时补充告诉LLM（见/unit_test命令）
1.1 生成代码块解释 /explain
用的prompt是
"""
Your task is:
Explain the code.
Following the task requirements, explain the selected portion of the code. \
Note that you only explain the logic of the selected code. \
The visible code only serves as the context for you to understand the code. \
Here is the relevant context \
information for your reference:
1. selected code info: {selected_text}
2. current visible code info: {visible_text}
"""
selected_text 是编辑器中被选中的代码 visible_text 是当前编辑器中可见代码
1.2 生成单元测试代码 /unit_test
这个是最难的命令，要多次与LLM和用户交互。
第一步，AST解析代码，询问LLM想知道什么symbols，析这些symbols放到contexts里，第三步会用到
"""
You're an advanced AI test generator.
You're about to write test cases for the function `{function_name}` in the file `{file_path}`.
Before you start, you need to check if you have enough context information to write the test cases.
Here is the source code of the function:
\```
{function_content}
\```
And here are some context information that might help you write the test cases:
{context_content}
Do you think the context information is enough?
If the information is insufficient, recommend which symbols or types you need to know more about.
Return a JSON object with a single key "key_symbols" whose value is a list of strings.
- If the context information is enough, return an empty list.
- Each string is the name of a symbol or type appearing in the function that lacks context information for writing test.
- The list should contain the most important symbols and should not exceed 10 items.
JSON Format Example:
{{
"key_symbols": ["<symbol 1>", "<symbol 2>", "<symbol 3>",...]
}}
"""
第二步，询问LLM想知道什么文件，第四步会用到
"""
As an advanced AI coding assistant,
you're given the task to identify suitable reference test files that can be used as a guide
for writing test cases for a specific function in the codebase.
You're provided with a list of test files in the repository.
Infer the purpose of each test file and identify the top 3 key files
that may be relevant to the target function and can serve as a reference for writing test cases.
The reference could provide a clear example of best practices
in testing functions of a similar nature.
The target function is next_turn, located in the file .\example\game.py.
The list of test files in the repository is as follows:
- unit_tests\propose_test.py
- unit_tests\assistants\recommend_test_context.py
Answer in JSON format with a list of the top 3 key file paths under the key `files`.
Make sure each file path is from the list of test files provided above.
Example:
{
"files": ["<file path 1>", "<file path 2>", "<file path 3>"]
}
"""
第三步，带上contexts向LLM询问测试方案
"""
Help me write unit tests for the `{func_name}` function
"""
第四步，app与用户交互，用户提测试要求，根据要求向LLM询问生成单元测试代码。
1.2 生成代码文档 /docstring
类似/explain
1.3 结论
devchat的vscode插件在算法上没有任何创新
- devchat没有模型层面的优化
- 静态的prompt模板，除了
/unit_test其他命令比较简单 - 没有充分利用整个项目的信息
2 自研 ourchat
ourchat完全可以实现devchat，并且效果不会比devchat差。
2.1 测试环境
运行环境 CPU: 2.10GHz, 72核, GPU: tesla V100 16G
选同样的模型ourchat和devchat的输出几乎一样，所以以下全是用ourchat测试 测试代码, 一个贪吃蛇游戏https://github.com/autruonggiang/Simple-Snake-Game/blob/main/game.py
# Class to design the snake
class Snake:
...
# Class to design the food
class Food:
...
# Function to check the next move of snake
def next_turn(snake, food):
...
2.2 测试生成代码块解释 /explain
使用snake game 的 next_turn 函数, 返回带上think部分以便分析，在devchat是不显示的。
think部分：
Okay, so I need to explain the selected portion of this Python code. Let me look at what's given.
The visible code starts from line 51 and goes up to line 116 in game.py. It includes a function next_turn and another part that defines change_direction and ions. I should focus on the if-else blocks within next_turn since that's where the logic for moving the snake, creating food, deleting objects, etc., happens
Let me break it down step by step.
First, in next_turn, they set x and y to the head of the snake's coordinates. Then there are conditionals checking the direction. Depending on whether it's ut, or right, x or y is adjusted by SPACE_SIZE. That makes sense because moving the head would mean updating its position.
Next, they insert the new head into snake.coordinates at position 0. So if the snake was [head, tail], it becomes [new_head, head, tail]. Then a rectangle (tody) is created and inserted at index 0. I suppose this is to make sure the drawing is in order.
Then there's an if statement checking if the current x and y match the food's coordinates. If they do, score increments by 1. The label updates with the new "food" is deleted from the canvas, and a new Food() object is created. This probably reappears when eaten.
Else, it deletes the last element of snake.coordinates (the tail) and also removes the last square from the snake's drawing. That makes sense to keep the gamthout memory issues.
After that, check_collisions(snake) is called. If true, game over; else, after some milliseconds, next_turn is scheduled again using window.after with SPEED .
Wait, but why do they call next_turn inside a conditional? Oh, because the window.after schedules it in the background so that the main loop can process evend creation). So when check_collisions returns true, the game ends without scheduling further turns.
tes direction based on new_direction input, handling reversals correctly. Check_collisions probably checks for wall or self-collision, which would trigger game over.
So putting it all together, this part of the code handles how the snake moves (based on direction), eats food to grow and increase score, clears old parts of the snake, and manages collisions to end the game.
-------------------------------------------------------------------------结论部分：
To explain the selected portion of the code:
The provided Python code is an essential part of a Snake game implementation. It defines two functions (`next_turn` and `change_direction`) as well as two conditionals that handle key aspects of gameplay.
### 1. **Function: next_turn** - Creates the core logic for moving and updating the snake's position. - Takes parameters `snake` (the snake object) and `food` (a food object). - It updates the snake's head based on its current direction: - Changes x or y by `SPACE_SIZE` depending on direction ("up", "down", etc.). - Appends the new head to the front of the snake's coordinates.
### 2. **Drawing Elements** - Creates a rectangle (the snake's body segment) at the new head position. - Adds this rectangle to the list of snake parts.
### 3. **Food Interaction** - Checks if the snake has eaten the food by comparing positions. - If eaten, score increases by 1, the score label updates, and a new food is generated. - If not eaten, removes the last part of the snake (the tail) to maintain length.
### 4. **Collision Detection** - Uses `check_collisions` function to detect: - Collisions with walls or self-collision at the head. - If collision detected, calls `game_over()` which stops the game by hiding windows.
### 5. **Event Handling (Change Direction Function)** - Allows handling keyboard input for changing direction. - Ensures smooth direction change by preventing double-reversals (e.g., not allowing turning left immediately after already going right).
This logic ensures that: - The snake moves correctly in the grid based on direction. - Food is consumed to grow and increase the score. - Invalid moves like running into walls or self cause the game to end. - Proper management of the snake's body segments as it grows.
The code is part of a loop where each call to `next_turn` can also schedule itself in a background thread (using `window.after`), allowing the main event loop time to handle food generation and other tasks.
think部分：
Alright, I need to explain the given code, specifically the selected portion around lines 48 to 93 of game.py. The visible code context includes a function called next_turn which takes snake and food as arguments.
First, looking at the code structure, I see that next_turn calculates new x and y coordinates based on the current direction. It then updates the snake's coordinate list by inserting the new position at the front. This suggests movement in the direction of heading.
Next, it creates a rectangle on the canvas with these new coordinates and adds this square to the snake's squares list. The use of insert(0, ...) implies that the first element is updated each turn, which aligns with how movement should happen—adding a new segment at the front.
The code then checks if the snake's head (x, y) matches the food's position. If it does, the score increments, updates the label, deletes the old food, and creates a new Food object. This handles when the snake eats food, growing itself by not popping the tail in this case.
If the snake doesn't eat food, the tail end of the coordinates list is removed. The canvas deletes the last square of the snake's body as well, giving the illusion of movement instead of just growing indefinitely.
After each move, check_collisions is called with the current snake state. If a collision is detected (either self-collision or hitting the window boundaries), game_over() is invoked to end the game.
If no collision occurs, the function schedules the next turn using window.after(SPEED, next_turn, snake, food). This sets up a recursive loop, allowing continuous movement of the snake at the defined speed.
Putting it all together, this function manages each step of the snake moving towards the next position based on the current direction. It updates both the internal state
(coordinates) and the visual canvas to reflect this movement.
-------------------------------------------------------------------------结论部分：
The code you've highlighted is responsible for controlling how the snake moves in each turn of a Snake game. Let's break down its functionality and logic:
1. **Movement Calculation**: - The function begins by taking the current head position coordinates `(x, y)` from `snake.coordinates[0]`. - Depending on the direction (`direction`), it adjusts these coordinates to move up, down, left, or right. This is done using conditional checks for each possible direction and adjusting either `x` or `y` by subtracting or adding `SPACE_SIZE`, which likely represents one segment of the grid or the snake's body.
2. **Updating Snake Position**: - The new `(x, y)` coordinates are inserted at the beginning of `snake.coordinates` using `insert(0, (x, y))`. This adds a new head position in front of the snake’s current position list. - A new square (representing the snake's body part) is created on the canvas at the new `(x, y)` location using `canvas.create_rectangle()`, and this square is added to the front of `snake.squares` with `insert(0, square)`. This visually updates where the snake is moving to.
3. **Eating Food**: - The code checks if the new head position matches the food's coordinates (`x == food.coordinates[0] and y == food.coordinates[1]`). If so: - The score is incremented using `global score` and updated in the UI. - The existing food image on the canvas is deleted with `canvas.delete("food")`, and a new `Food()` object is created to place the food anew somewhere else on the board. This effectively grows the snake by adding without removing the tail segment, so the snake's length increases.
4. **No Food Eaten**: - If no food was eaten (`else` clause): - The last x, y coordinates are removed from `snake.coordinates` using `del snake.coordinates[-1]`. - The corresponding square (tail) is deleted from both the canvas and `snake.squares` by calling `canvas.delete(snake.squares[-1])` followed by `del snake.squares[-1]`. - This action maintains the snake's length, making it appear as though the snake has moved forward; only its position is updated, not its size.
5. **Collision Check**: - After each movement (whether or not food was eaten), it checks for collisions using `check_collisions(snake)`. If a collision is detected: - The game ends by invoking `game_over()`. - If no collision is detected, the function uses `window.after(SPEED, next_turn, snake, food)` to schedule the next turn after `SPEED` milliseconds. This creates a loop that continuously moves the snake at intervals specified in `SPEED`.
In summary, this code manages how the snake moves each turn: - Moving the head based on current direction. - Updating both the internal coordinate tracking and visual representation on the canvas
think部分：
无
-------------------------------------------------------------------------结论部分：
The selected code is the implementation of a game loop function named `next_turn` in Python. This function manages the progression of the snake's movement and interaction with food, as well as handling collisions to determine if the game should end. Here's a breakdown of its logic:
1. **Extract Head Coordinates**: The function starts by extracting the coordinates of the snake's head. These are obtained from `snake.coordinates[0]`.
2. **Update Head Position Based on Direction**: Depending on the current direction, it updates the x and y coordinates accordingly: - If the direction is "up", it decreases the y-coordinate by `SPACE_SIZE`. - If the direction is "down", it increases the y-coordinate by `SPACE_SIZE`. - If the direction is "left", it decreases the x-coordinate by `SPACE_SIZE`. - If the direction is "right", it increases the x-coordinate by `SPACE_SIZE`.
3. **Insert New Head Position**: It inserts the new coordinates (x, y) at the beginning of `snake.coordinates` to represent the snake's head moving in the specified direction.
4. **Create a Rectangle for the New Head**: Using the updated coordinates, it creates a rectangle on the canvas where the snake's head will be located with the size defined by `SPACE_SIZE`. This is done using `canvas.create_rectangle` and adds this square to both `snake.coordinates` and `snake.squares`.
5. **Check for Food Consumption**: It checks if the new head position coincides with the food's coordinates: - If yes, it increments the score by 1 and updates the label displaying the score. - It also removes the old food representation from the canvas (`canvas.delete("food")`) and generates a new food object at a random location. - If no, it removes the last segment of the snake from both `snake.coordinates` and `snake.squares` by deleting them using `del`, effectively allowing the snake to move without its tail following it.
6. **Check for Collisions**: After potentially updating or removing parts of the snake based on eating food, it checks if there are any collisions with walls or itself: - If a collision is detected (indicated by the result of `check_collisions(snake)`), it calls the game-over function (`game_over()`). - Otherwise, it schedules another call to `next_turn` after a delay specified by `SPEED`, passing along the same snake and food objects. This creates a loop that continuously updates the snake's movement every `SPEED` milliseconds until a termination condition is met (e.g., collision detected).
This function plays a critical role in maintaining game dynamics, such as moving the snake, checking for food consumption, updating scores, and detecting end conditions.
| 模型 | CPU 运行时间 | GPU 运行时间 |
|---|---|---|
| deepseek-r1:7b | 56.2s | 12.6s |
| deepseek-r1:32b | 399.7s | 72.3s |
| deepseek-coder-v2:16b | 47.4s | 12.6s |
结论：
- 7B模型有单词拼错和语法不顺。
- 输出受visible_text影响，visible_text不是越多越好。
- deepseek-coder 输出比较稳定，但总结能力偏弱。
2.3 测试生成单元测试代码 /init_test
使用snake game 的 next_turn 函数。
测试提议：
1. happy path: Test that the snake moves correctly when it eats food, increasing its length and score.
2. happy path: Verify that the snake turns left without collision, updating its position and direction.
3. happy path: Ensure the game continues smoothly when the snake moves right into empty space.
4. edge case: Check if the game correctly handles a collision with the wall, triggering game over.
5. edge case: Test that the snake dies when it collides with its own body after turning around.
6. edge case: Verify that the score does not increase and the food remains when the snake moves without eating.
-------------------------------------------------------------------------
测试代码：
```python
import unittest
from unittest.mock import patch, MagicMock
class TestNextTurn(unittest.TestCase):
def setUp(self):
# Initialize mocks for global variables
self.canvas = MagicMock()
self.snake = {'coordinates': [], 'direction': None}
self.food = {'x': 0, 'y': 0}
self.score = 0
self.game_over = MagicMock()
@patch('builtins.print') # Mock print function if used in next_turn
def test_eat_food_increases_length_and_score(self, mock_print):
"""Test that the snake moves correctly when it eats food, increasing its length and score."""
# Setup initial state
self.snake['coordinates'] = [(100, 100)]
self.snake['direction'] = "right"
self.food['x'], self.food['y'] = 110, 100
next_turn()
# Assert the snake's length increased
self.assertEqual(len(self.snake['coordinates']), 2)
# Assert score increased
self.assertEqual(self.score, 1)
# Assert new food was created
self.food.assert_called_once()
@patch('builtins.print')
def test_turn_left_without_collision(self, mock_print):
"""Verify that the snake turns left without collision, updating its position and direction."""
# Setup initial state
self.snake['coordinates'] = [(100, 100)]
self.snake['direction'] = "up"
next_turn()
# Assert the snake's direction changed
self.assertEqual(self.snake['direction'], "left")
# Assert position updated correctly (example coordinates)
self.assertEqual(self.snake['coordinates'][0], (90, 100))
@patch('builtins.print')
def test_move_right_into_empty_space(self, mock_print):
"""Ensure the game continues smoothly when the snake moves right into empty space."""
# Setup initial state
self.snake['coordinates'] = [(100, 100)]
self.snake['direction'] = "right"
global direction
direction = "right"
next_turn()
# Assert position updated correctly without increasing length
self.assertEqual(len(self.snake['coordinates']), 1)
self.assertEqual(self.snake['coordinates'][0], (110, 100))
@patch('builtins.print')
def test_wall_collision_triggers_game_over(self, mock_print):
"""Test that wall collision triggers game over."""
# Setup initial state at edge of canvas
self.snake['coordinates'] = [(0, 100)]
self.snake['direction'] = "left"
next_turn()
# Assert game_over function was called
self.game_over.assert_called_once()
@patch('builtins.print')
def test_self_collision_triggers_game_over(self, mock_print):
"""Test that self collision triggers game over."""
# Setup initial state with overlapping coordinates
self.snake['coordinates'] = [(100, 100), (90, 100)]
self.snake['direction'] = "right"
next_turn()
# Assert game_over function was called
self.game_over.assert_called_once()
@patch('builtins.print')
def test_move_without_eating_food(self, mock_print):
"""Test that moving without eating food does not change the score."""
# Setup initial state
self.snake['coordinates'] = [(100, 100)]
self.snake['direction'] = "right"
global direction
direction = "right"
next_turn()
# Assert score remains unchanged
self.assertEqual(self.score, 0)
# Assert food was not recreated
self.food.assert_not_called()
if __name__ == '__main__':
unittest.main()
```
### Explanation
- **Mocking**: The solution uses `unittest.mock` to replace global variables with mock objects, isolating the function under test.
- **Setup and Teardown**: Each test method sets up specific initial states using `setUp()`, ensuring tests are independent.
- **Assertions**: After invoking `next_turn`, each test asserts expected outcomes such as state changes, function calls, and score updates.
This approach ensures that all critical scenarios are tested efficiently, providing confidence in the function's correctness.
测试提议：
1. happy path: Test the next_turn function with a simple move up scenario where the snake eats food and grows longer.
2. happy path: Test the next_turn function when the snake moves right but does not eat any food, causing it to just move without growing.
3. edge case: Test the next_turn function with a scenario where the snake moves down and hits the boundary of the game area, triggering game over.
4. edge case: Test the next_turn function when there is no food present at the new head position of the snake, causing the tail to be removed without growing.
5. edge case: Test the next_turn function with a scenario where the snake moves left and hits another part of its body, triggering game over.
-------------------------------------------------------------------------
测试代码：
```python
import unittest
from unittest.mock import patch, MagicMock
import random
# Assuming SPACE_SIZE, WIDTH, HEIGHT, SNAKE, FOOD are defined somewhere in the game.py file
SPACE_SIZE = 20
WIDTH = 400
HEIGHT = 400
SNAKE = "green"
FOOD = "red"
SPEED = 100
class TestNextTurn(unittest.TestCase):
def setUp(self):
# Mocking the necessary components for testing
self.snake_mock = MagicMock()
self.food_mock = MagicMock()
self.canvas_mock = MagicMock()
self.label_mock = MagicMock()
self.score = 0
self.direction = "up"
def test_happy_path_snake_eats_food(self):
# TODO: Implement the happy path scenario where the snake moves up and eats food, causing it to grow longer.
pass
def test_happy_path_snake_moves_right_no_eat(self):
# TODO: Implement the happy path scenario where the snake moves right but does not eat any food, causing it to just move without growing.
pass
def test_edge_case_snake_hits_boundary(self):
# TODO: Implement the edge case scenario where the snake moves down and hits the boundary of the game area, triggering game over.
pass
def test_edge_case_no_food_at_new_head(self):
# TODO: Implement the edge case scenario when there is no food present at the new head position of the snake, causing the tail to be removed without growing.
pass
def test_edge_case_snake_hits_itself(self):
# TODO: Implement the edge case scenario where the snake moves left and hits another part of its body, triggering game over.
pass
if __name__ == "__main__":
unittest.main()
```
This code sets up a basic test suite for the `next_turn` function in the `game.py` file. Each test case is marked with TODO comments to indicate where specific implementation details need to be added. The setup method initializes mock objects for snake, food, canvas, and label, which are necessary for testing the function.
测试提议：
1. happy path: Test that the snake moves correctly when it eats food, increasing its length and score.
2. happy path: Verify that the game continues after a normal move without eating food or colliding.
3. happy path: Ensure that the snake's head moves in the correct direction based on the current direction setting.
4. edge case: Check if the game correctly ends when the snake hits a wall boundary.
5. edge case: Test that the game terminates when the snake collides with its own body.
6. edge case: Verify that the function handles unexpected direction values gracefully without crashing.
-------------------------------------------------------------------------
测试代码：
```python
import pytest
from unittest.mock import Mock, patch
from example.game import next_turn, check_collisions, Food
class TestNextTurn:
def setup_method(self):
# Initialize mock objects shared across tests
self.snake = Mock()
self.food = Mock()
self.window = Mock()
self.canvas = Mock()
# Setup initial conditions for snake and food
self.snake.coordinates = [(0, 0)]
self.food.coordinates = [Mock(), Mock()] # x and y coordinates
# Mock global variables
self.mock_score = 0
self.mock_direction = "right"
self.SPEED = 100
self.SPACE_SIZE = 20
self.WIDTH = 800
self.HEIGHT = 600
@patch('example.game.score', 0)
def test_happy_path_eat_food(self, mock_score):
# Setup: Snake eats food
self.snake.coordinates[0] = (self.food.coordinates[0],
self.food.coordinates[1])
next_turn(self.snake, self.food, window=self.window, canvas=self.canvas)
# Assert score increases and new segment is added
assert mock_score == 1
assert len(self.snake.coordinates) > 1
@patch('example.game.score', 0)
def test_happy_path_grow_segment(self, mock_score):
# Setup: Snake eats food
self.snake.coordinates[0] = (self.food.coordinates[0],
self.food.coordinates[1])
next_turn(self.snake, self.food, window=self.window, canvas=self.canvas)
# Assert snake grows by one segment
assert len(self.snake.coordinates) == 2
@patch('example.game.window')
def test_happy_path_schedule_next_turn(self, mock_window):
# Setup: No collision
next_turn(self.snake, self.food, window=mock_window, canvas=self.canvas)
# Assert window.after is called with SPEED
mock_window.after.assert_called_once_with(self.SPEED,
next_turn,
self.ref_args(),
self.ref_kwargs())
@patch('example.game.game_over')
def test_edge_case_self_collision(self, mock_game_over):
# Setup: Snake collides with itself
self.snake.coordinates = [(0, 0), (0, 0)]
next_turn(self.snake, self.food, window=self.window, canvas=self.canvas)
# Assert game over is called
mock_game_over.assert_called_once()
@patch('example.game.direction')
def test_edge_case_all_directions(self, mock_direction):
directions = ["up", "down", "left", "right"]
for direction in directions:
mock_direction.return_value = direction
next_turn(self.snake, self.food, window=self.window, canvas=self.canvas)
# Verify movement in each direction
if direction == "up":
assert self.snake.coordinates[0][1] < 0
elif direction == "down":
assert self.snake.coordinates[0][1] > 0
elif direction == "left":
assert self.snake.coordinates[0][0] < 0
else: # right
assert self.snake.coordinates[0][0] > 0
@patch('example.game.game_over')
def test_edge_case_out_of_bounds(self, mock_game_over):
# Setup: Snake moves out of bounds
self.snake.coordinates = [(self.WIDTH + 1, 0)]
next_turn(self.snake, self.food, window=self.window, canvas=self.canvas)
# Assert game over is called
mock_game_over.assert_called_once()
def main():
pytest.main([__file__])
if __name__ == "__main__":
main()
```
This test suite:
1. Uses pytest and unittest.mock for mocking dependencies
2. Tests all six requested scenarios
3. Includes proper setup and teardown with fixture methods
4. Mocks external dependencies like window, canvas, and game state
5. Verifies both successful paths and edge cases
6. Follows best practices for test organization and structure
To run the tests:
```bash
python -m pytest test_game.py -v
```
Note: This implementation assumes some knowledge of the internal implementation details of the game. You may need to adjust mocks based on actual implementation specifics.
测试提议：
1. happy path: Test that the snake moves up correctly when direction is 'up'.
2. happy path: Verify that the score increments and food regenerates when the snake eats the food.
3. happy path: Check that the game continues with updated positions if no collision occurs after a move.
4. edge case: Ensure game over is triggered when the snake collides with the boundary of the game area.
5. edge case: Test that the game correctly handles a collision where the snake's head hits its own body part.
6. edge case: Verify that the game over screen is displayed with correct text and formatting when a game over condition is met.
-------------------------------------------------------------------------
测试代码：
```python
import unittest
from unittest.mock import MagicMock, patch
# Assuming these imports are from the actual game module
from example.game import next_turn, check_collisions, Food, Snake, canvas, label, window, SNAKE, SPACE_SIZE, WIDTH, HEIGHT, SPEED, score
class TestGame(unittest.TestCase):
def setUp(self):
self.snake = MagicMock()
self.food = MagicMock()
self.canvas = MagicMock()
self.label = MagicMock()
self.window = MagicMock()
# happy path: Test that the snake moves up correctly when direction is 'up'.
@patch('example.game.direction', 'up')
def test_snake_moves_up(self):
initial_coordinates = [(10, 10)]
self.snake.coordinates = initial_coordinates.copy()
next_turn(self.snake, self.food)
expected_coordinates = [(10, 10 - SPACE_SIZE)]
self.assertEqual(self.snake.coordinates[0], expected_coordinates[0])
# happy path: Verify that the score increments and food regenerates when the snake eats the food.
@patch('example.game.direction', 'up')
def test_score_increments_and_food_regenerates(self):
initial_coordinates = [(10, 10)]
self.snake.coordinates = initial_coordinates.copy()
self.food.coordinates = [10, 10 - SPACE_SIZE]
next_turn(self.snake, self.food)
self.assertEqual(score, 1)
self.food.assert_called()
# happy path: Check that the game continues with updated positions if no collision occurs after a move.
@patch('example.game.direction', 'up')
def test_game_continues_with_updated_positions(self):
initial_coordinates = [(10, 10)]
self.snake.coordinates = initial_coordinates.copy()
next_turn(self.snake, self.food)
expected_coordinates = [(10, 10 - SPACE_SIZE)]
self.assertEqual(self.snake.coordinates[0], expected_coordinates[0])
self.window.after.assert_called_with(SPEED, next_turn, self.snake, self.food)
# edge case: Ensure game over is triggered when the snake collides with the boundary of the game area.
@patch('example.game.direction', 'up')
def test_game_over_on_boundary_collision(self):
initial_coordinates = [(10, 0)]
self.snake.coordinates = initial_coordinates.copy
| 模型 | CPU 运行时间 | GPU 运行时间 |
|---|---|---|
| deepseek-coder-v2:16b | - | 18.3s |
| deepseek-r1:32b | - | 213.4s |
| deepseek-v2:236b | 644.5s | - |
| deepseek-coder-v2:236b | 527.5s | - |
结论：
- deepseek-r1:7b和deepseek-r1:70b没有输出测试方案。
- deepseek-coder的推理能力太弱了，但是编程语法正确。
- r1和v2的整体逻辑较强，但是编程语法错误。