add some structure to the project

Author: janek-cosmose

.gitignore

@@ -1,2 +1,3 @@
 __pycache__
 .envrc
+parse_qwantz.egg-info

add_test_case.sh

@@ -3,5 +3,5 @@
 number="$1"
 
 stdout_file=test/expected_outputs/$number.txt
-python parse_qwantz.py $COMIC_DIR/"${number}"* > "$stdout_file"
+parse-qwantz $COMIC_DIR/"${number}"* > "$stdout_file"
 cp $COMIC_DIR/"${number}"* test/comics/$number.png

dev_requirements.txt

@@ -0,0 +1,2 @@
+from parse_qwantz.cli import app
+from parse_qwantz.main import main

parse_qwantz/__init__.py

@@ -1,6 +1,6 @@
 from typing import NamedTuple
 
-from pixels import Pixel
+from parse_qwantz.pixels import Pixel
 
 
 class Box(NamedTuple):

box.py parse_qwantz/box.py

@@ -2,13 +2,16 @@
 
 import typer
 
-from parse_qwantz import main
+from parse_qwantz.main import main
 
+app = typer.Typer()
 
+
+@app.command()
 def cli(image_path: Path, debug: bool = typer.Option(False, help="Enable debug features.")):
     """Generate transcripts for Ryan North's Dinosaur Comics from https://qwantz.com"""
     main(image_path, debug=debug)
 
 
 if __name__ == '__main__':
-    typer.run(cli)
+    app()

cli.py parse_qwantz/cli.py

@@ -1,10 +1,10 @@
 from collections import deque
 from itertools import product
 
-from box import Box
-from pixels import Pixel
-from simple_image import SimpleImage
-from shape import get_shape, get_adjacent_pixels, get_box
+from parse_qwantz.box import Box
+from parse_qwantz.pixels import Pixel
+from parse_qwantz.simple_image import SimpleImage
+from parse_qwantz.shape import get_shape, get_adjacent_pixels, get_box
 
 
 def get_thought(pixel: Pixel, image: SimpleImage) -> tuple[Box, list[Pixel]] | None:

colors.py parse_qwantz/colors.py

@@ -1,13 +1,13 @@
 from logging import getLogger
 
-from box import Box
-from fonts import ALL_FONTS
-from lines import Line, get_line
-from text_lines import TextLine, try_text_line
-from detect_thought import get_thought
-from pixels import Pixel
-from shape import get_shape
-from simple_image import SimpleImage
+from parse_qwantz.box import Box
+from parse_qwantz.fonts import ALL_FONTS
+from parse_qwantz.lines import Line, get_line
+from parse_qwantz.text_lines import TextLine, try_text_line
+from parse_qwantz.detect_thought import get_thought
+from parse_qwantz.pixels import Pixel
+from parse_qwantz.shape import get_shape
+from parse_qwantz.simple_image import SimpleImage
 
 logger = getLogger()
 

detect_thought.py parse_qwantz/detect_thought.py

@@ -1,22 +1,23 @@
 import string
 from dataclasses import dataclass
+from importlib.resources import as_file
 from itertools import product
-from typing import NamedTuple
+from pathlib import Path
+from typing import NamedTuple, ContextManager
 
 from PIL import Image
 
-from box import Box
-from pixels import Pixel
-from simple_image import SimpleImage
+from parse_qwantz.box import Box
+from parse_qwantz.pixels import Pixel
+from parse_qwantz.simple_image import SimpleImage
 
 PRINTABLE = string.printable.strip()
 FORBIDDEN_CHARS = '\\_`|~'
 
-REGULAR13_SHAPE_FILE = 'img/regular13.png'
-REGULAR12_SHAPE_FILE = 'img/regular12.png'
-REGULAR11_SHAPE_FILE = 'img/regular11.png'
-REGULAR9_SHAPE_FILE = 'img/regular9.png'
-REGULAR8_SHAPE_FILE = 'img/regular8.png'
+FONT_SIZES = [(13, 'Regular'), (12, 'Condensed'), (11, 'Small'), (9, 'Mini'), (8, 'Tiny')]
+SHIFTED_VARIANTS = {
+    13: {',': 1, ':': 1, '.': -1},
+}
 
 
 class CharBox(NamedTuple):
@@ -90,9 +91,10 @@ def __repr__(self):
 
     @classmethod
     def from_file(
-        cls, file_path: str, name: str, shifted_variants: dict[str, int] | None = None
+        cls, file_path_context_manager: ContextManager[Path], name: str, shifted_variants: dict[str, int] | None = None
     ) -> "Font":
-        image = SimpleImage.from_image(Image.open(file_path))
+        with file_path_context_manager as file_path:
+            image = SimpleImage.from_image(Image.open(file_path))
         width = image.width // len(PRINTABLE)
         height = image.height
         shapes = {}
@@ -150,10 +152,11 @@ def get_bitmask(pixel: Pixel, image: SimpleImage, width: int, height: int) -> in
     return bitmask
 
 
-REGULAR_FONT = Font.from_file(REGULAR13_SHAPE_FILE, 'Regular', shifted_variants={',': 1, ':': 1, '.': -1})
-CONDENSED_FONT = Font.from_file(REGULAR12_SHAPE_FILE, 'Condensed')
-SMALL_FONT = Font.from_file(REGULAR11_SHAPE_FILE, 'Small')
-MINI_FONT = Font.from_file(REGULAR9_SHAPE_FILE, 'Mini')
-TINY_FONT = Font.from_file(REGULAR8_SHAPE_FILE, 'Tiny')
-
-ALL_FONTS = [REGULAR_FONT, SMALL_FONT, TINY_FONT, CONDENSED_FONT, MINI_FONT]
+ALL_FONTS = [
+    Font.from_file(
+        file_path_context_manager=as_file(Path('parse_qwantz', 'img', f'regular{size}.png')),
+        name=name,
+        shifted_variants=SHIFTED_VARIANTS.get(size, {}),
+    )
+    for size, name in FONT_SIZES
+]

elements.py parse_qwantz/elements.py

@@ -1,8 +1,8 @@
 from itertools import groupby, product
 
-from pixels import Pixel
-from simple_image import SimpleImage
-from shape import get_shape, get_box
+from parse_qwantz.pixels import Pixel
+from parse_qwantz.simple_image import SimpleImage
+from parse_qwantz.shape import get_shape, get_box
 
 Line = tuple[Pixel, Pixel]
 

fonts.py parse_qwantz/fonts.py

@@ -3,18 +3,18 @@
 
 from PIL import Image, ImageDraw
 
-from box import Box, get_interval_distance
-from colors import Color
-from text_blocks import get_text_blocks, TextBlock
-from elements import get_elements
-from match_blocks import match_blocks, Character_s
-from match_lines import match_lines, Character, OFF_PANEL, UnmatchedLine
-from match_thought import match_thought
-from pixels import Pixel
-from shape import get_box
-from simple_image import SimpleImage
-from prepare_image import apply_mask
-from logger import get_logger
+from parse_qwantz.box import Box, get_interval_distance
+from parse_qwantz.colors import Color
+from parse_qwantz.text_blocks import get_text_blocks, TextBlock
+from parse_qwantz.elements import get_elements
+from parse_qwantz.match_blocks import match_blocks, Character_s
+from parse_qwantz.match_lines import match_lines, Character, OFF_PANEL, UnmatchedLine
+from parse_qwantz.match_thought import match_thought
+from parse_qwantz.pixels import Pixel
+from parse_qwantz.shape import get_box
+from parse_qwantz.simple_image import SimpleImage
+from parse_qwantz.prepare_image import apply_mask
+from parse_qwantz.logger import get_logger
 
 logger = get_logger()
 

img/mask.png parse_qwantz/img/mask.png

@@ -1,8 +1,8 @@
 from typing import Iterable
 
-from text_blocks import TextBlock
-from match_lines import Target, Character
-from text_lines import TextLine
+from parse_qwantz.text_blocks import TextBlock
+from parse_qwantz.match_lines import Target, Character
+from parse_qwantz.text_lines import TextLine
 
 Character_s = Character | tuple[Character, Character]
 

img/regular11.png parse_qwantz/img/regular11.png

@@ -2,12 +2,12 @@
 from dataclasses import dataclass
 from typing import Iterable
 
-from box import Box
-from lines import Line
-from text_blocks import TextBlock
-from pixels import Pixel
-from simple_image import SimpleImage
-from text_lines import TextLine
+from parse_qwantz.box import Box
+from parse_qwantz.lines import Line
+from parse_qwantz.text_blocks import TextBlock
+from parse_qwantz.pixels import Pixel
+from parse_qwantz.simple_image import SimpleImage
+from parse_qwantz.text_lines import TextLine
 
 logger = logging.getLogger()
 

img/regular12.png parse_qwantz/img/regular12.png

@@ -1,8 +1,8 @@
 from typing import Iterable
 
-from box import Box
-from text_blocks import TextBlock
-from match_lines import Character
+from parse_qwantz.box import Box
+from parse_qwantz.text_blocks import TextBlock
+from parse_qwantz.match_lines import Character
 
 
 def match_thought(

img/regular13.png parse_qwantz/img/regular13.png

@@ -3,7 +3,7 @@
 
 from PIL import Image
 
-from colors import Color
+from parse_qwantz.colors import Color
 
 
 class Pixel(NamedTuple):

img/regular8.png parse_qwantz/img/regular8.png

@@ -0,0 +1,20 @@
+from functools import cache
+from importlib.resources import as_file
+from pathlib import Path
+
+from PIL import Image
+
+DIM = (735, 500)
+
+MASK_FILE_PATH = Path('parse_qwantz', 'img', 'mask.png')
+
+
+@cache
+def get_mask_image():
+    with as_file(MASK_FILE_PATH) as image_path:
+        return Image.open(image_path)
+
+
+def apply_mask(image: Image):
+    all_white = Image.new(mode='RGB', size=DIM, color=(255, 255, 255))
+    return Image.composite(image, all_white, get_mask_image())

img/regular9.png parse_qwantz/img/regular9.png

@@ -1,9 +1,9 @@
 from typing import Collection
 
-from box import Box
-from colors import Color
-from pixels import Pixel
-from simple_image import SimpleImage
+from parse_qwantz.box import Box
+from parse_qwantz.colors import Color
+from parse_qwantz.pixels import Pixel
+from parse_qwantz.simple_image import SimpleImage
 
 
 def get_box(shape: Collection[Pixel], padding=0) -> Box:

lines.py parse_qwantz/lines.py

@@ -2,9 +2,9 @@
 
 from PIL import Image
 
-from box import Box
-from colors import Color
-from pixels import Pixel, get_pixels
+from parse_qwantz.box import Box
+from parse_qwantz.colors import Color
+from parse_qwantz.pixels import Pixel, get_pixels
 
 
 @dataclass

logger.py parse_qwantz/logger.py

@@ -1,12 +1,12 @@
 from itertools import chain, groupby
 from typing import Iterable, NamedTuple
 
-from box import Box, get_interval_distance
-from fonts import Font
-from colors import Color
-from text_lines import TextLine, try_text_line
-from pixels import Pixel
-from simple_image import SimpleImage
+from parse_qwantz.box import Box, get_interval_distance
+from parse_qwantz.fonts import Font
+from parse_qwantz.colors import Color
+from parse_qwantz.text_lines import TextLine, try_text_line
+from parse_qwantz.pixels import Pixel
+from parse_qwantz.simple_image import SimpleImage
 
 
 class TextBlock(NamedTuple):

parse_qwantz.py parse_qwantz/main.py

@@ -2,10 +2,10 @@
 
 from dataclasses import dataclass
 
-from box import Box
-from fonts import Font, CharBox
-from pixels import Pixel
-from simple_image import SimpleImage
+from parse_qwantz.box import Box
+from parse_qwantz.fonts import Font, CharBox
+from parse_qwantz.pixels import Pixel
+from parse_qwantz.simple_image import SimpleImage
 
 
 @dataclass

match_blocks.py parse_qwantz/match_blocks.py

@@ -0,0 +1,29 @@
+[project]
+name = "parse_qwantz"
+authors = [{name = "Jan Szejko", email = "[email protected]"}]
+version = "2023.1.26"
+description = "Transcript generator for Dinosaur Comics"
+requires-python = ">=3.10"
+classifiers = [
+    "Development Status :: 3 - Alpha",
+    "Programming Language :: Python :: 3.10",
+    "Environment :: Console",
+    "Intended Audience :: End Users/Desktop",
+    # "License :: OSI Approved :: BSD License",
+    "Natural Language :: English",
+    "Operating System :: OS Independent",
+    "Topic :: Multimedia :: Graphics :: Graphics Conversion",
+    "Scientific/Engineering :: Image Recognition",
+    "Typing :: Typed",
+]
+dependencies = [
+  "Pillow",
+  "typer",
+]
+# TODO: readme, license
+
+[project.optional-dependencies]
+dev = ["pytest"]
+
+[project.scripts]
+parse-qwantz = "parse_qwantz:app"