grid_engine
Description
grid_engine is a framework for generating and manipulating grids. It provides a number of classes and functions for generating and manipulating grids. Each grid is composed of Cell objects and is defined by a Blueprint. A grid can be generated from a blueprint, loaded from a file, or created manually. It can also be pickled for later use. It can be rendered as a 2D image, an animated GIF or an ASCII string. Grids provide a number of relevant methods for pathfinding, cell manipulation, and more.
Installation
To install grid_engine, run the following command:
```bash
pip install gridengine_framework
```
Usage
To use gridengine you can import any number of the submodules and utilize its respective features or you can import the main module.
```python
import grid_engine as ge
# Create a grid
grid = ge.grid.Grid(cell_size=10, grid_dimensions=(1000, 1000))
# Save a grid
ge.grid.save_grid(grid)
# Load a grid
loaded_grid = ge.grid.Grid.load_grid(1)
```
grid_engine also provides a command line interface. To use it, run the following command:
```bash
python -m grid_engine --help
# Output:
# usage: gridengine [-h] [-i] [-b BLUEPRINT] [--ascii] [-l LOAD] [-t] [-ns NOISE_SCALE] [-no NOISE_OCTAVES] [-nr NOISE_ROUGHNESS] [-r ROWS] [-c COLUMNS] [-s SIZE] [-S] [-T TYPE] [-v]
# Generate a visualized grid from a blueprint. For producing a blueprint, see the blueprint module.
# options:
# -h, --help show this help message and exit
# -i, --interactive Run an interactive session
# -b BLUEPRINT, --blueprint BLUEPRINT
# Load a blueprint from a file
# --ascii Print the grid as ascii
# -l LOAD, --load LOAD Load a grid from a file
# -t, --terrain Whether to generate terrain with the grid.
# -ns NOISE_SCALE, --noise-scale NOISE_SCALE
# Noise scale
# -no NOISE_OCTAVES, --noise-octaves NOISE_OCTAVES
# Noise octaves
# -nr NOISE_ROUGHNESS, --noise-roughness NOISE_ROUGHNESS
# Noise roughness
# -r ROWS, --rows ROWS Number of rows in the grid
# -c COLUMNS, --columns COLUMNS
# Number of columns in the grid
# -s SIZE, --size SIZE Size of each cell in the grid
# -S, --save Save the grid object to a file
# -T TYPE, --type TYPE Type of file to save the grid as
# -v, --verbose Verbose output
```
Examples
The following examples demonstrate the use of the grid_engine package.
CLI
The following command:
```bash
python -m grid_engine -v -S -t -ns 400 -no 75 -nr 0.55 -r 450 -c 800 -s 2
```
Will produce the following output:
```bash
Generating blueprint with cell size 2, 450 rows and 800 columns. Total_cells: 360000 ...
Success! Blueprint generated. Dimensions: (1600, 900)
Building grid from blueprint ...
Finding landmasses ...
Separating islands from landmasses ...
done
Finding start to river ...
Found largest landmass: 0 with 302951 cells
Finding lake coastal cells ...
Found 1094 cells.
Validating start/end of river ...
Start cell: pd00229(445, 228), End cell: dq00087(146, 86))
Building river ...
Getting river cells by walk ...
Generating river with end designated ...
Cells in river: 300 | Current cell: kb00082 | Next cell: kc00082
River steps: 300
Found largest landmass: 0 with 302951 cells
Finding lake coastal cells ...
Found 1094 cells.
Validating start/end of river ...
Start cell: ip00499(275, 498), End cell: dv00087(151, 86)
Building river ...
Getting river cells by walk ...
Generating river with end designated ...
Cells in river: 757 | Current cell: dz00082 | Next cell: ea000814
River steps: 757
done
Success! Grid generated.
Pickling grid ...
Success!
Pickling blueprint ...
Success!
Generating grid image ...
Importing pillow ...
Preparing raw image ...
Counting cells ...
Total cells: 360000
Shuffling cells ...
Cells drawn. 100%
Saving grid image ...
Grid ID: 8d564
```
The following image is the result of the above command:
The river generation algorithm is not perfect. I am currently working on improving it.
The above command will also produce the following files(Not included in the repository... That is, generate your own!):
-
grid.cf1b9.pickle
: A pickled Grid object.
-
blueprint.cf1b9.pickle
: A pickled TerrainGridBlueprint object.
import grid_engine
from grid_engine import Grid
_8d564 = Grid.load_grid('8d564')
print(_8d564.grid_id)
cellA = _8d564.random_cell(attr=('passable', True))
cellB = _8d564.random_cell(attr=('passable', True))
path, cost = _8d564.get_path(cellA, cellB)
print(path)