ci: new build & publish workflow

- ok i didnt expect github actions does not support regex wtf

.github/workflows/build-and-draft-release.yml

@@ -1,48 +0,0 @@
-name: "Build and draft a release"
-
-on:
-  workflow_dispatch:
-  push:
-    tags:
-      - "v[0-9]+.[0-9]+.[0-9]+"
-
-permissions:
-  contents: write
-  discussions: write
-
-jobs:
-  build-and-draft-release:
-    runs-on: ubuntu-latest
-    steps:
-      - name: Checkout
-        uses: actions/checkout@v4
-        with:
-          fetch-depth: 0
-
-      - name: Set up Python environment
-        uses: actions/setup-python@v5
-        with:
-          python-version: "3.x"
-
-      - name: Build package
-        run: |
-          pip install build
-          python -m build
-
-      - name: Remove `v` in tag name
-        uses: mad9000/actions-find-and-replace-string@5
-        id: tagNameReplaced
-        with:
-          source: ${{ github.ref_name }}
-          find: "v"
-          replace: ""
-
-      - name: Draft a release
-        uses: softprops/action-gh-release@v2
-        with:
-          discussion_category_name: New releases
-          draft: true
-          generate_release_notes: true
-          files: |
-            dist/arcaea_offline_ocr-${{ steps.tagNameReplaced.outputs.value }}*.whl
-            dist/arcaea-offline-ocr-${{ steps.tagNameReplaced.outputs.value }}.tar.gz

.github/workflows/build-and-publish.yml

@@ -0,0 +1,103 @@
+name: Build, Release, Publish
+
+on:
+  workflow_dispatch:
+  push:
+    tags:
+      - "*.*.*"
+
+jobs:
+  build:
+    name: Build package
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+
+      - name: Set up Python environment
+        uses: actions/setup-python@v5
+        with:
+          python-version: "3.x"
+
+      - name: Build package
+        run: |
+          pip install build
+          python -m build
+
+      - name: Store the distribution files
+        uses: actions/upload-artifact@v4
+        with:
+          name: python-package-distributions
+          path: dist/
+
+  draft-release:
+    name: Draft a release
+    runs-on: ubuntu-latest
+    needs:
+      - build
+    permissions:
+      contents: write
+      discussions: write
+
+    steps:
+      - name: Download the distribution files
+        uses: actions/download-artifact@v4
+        with:
+          name: python-package-distributions
+          path: dist/
+
+      - name: Draft a release
+        uses: softprops/action-gh-release@v2
+        with:
+          discussion_category_name: New releases
+          draft: true
+          generate_release_notes: true
+          files: |
+            dist/*
+
+  publish-to-pypi:
+    name: Publish distribution to PyPI
+    runs-on: ubuntu-latest
+    if: startsWith(github.ref, 'refs/tags/') # only publish to PyPI on tag pushes
+    needs:
+      - build
+    environment:
+      name: pypi
+      url: https://pypi.org/p/arcaea-offline-ocr
+    permissions:
+      id-token: write
+
+    steps:
+      - name: Download the distribution files
+        uses: actions/download-artifact@v4
+        with:
+          name: python-package-distributions
+          path: dist/
+
+      - name: Publish distribution to PyPI
+        uses: pypa/gh-action-pypi-publish@release/v1
+
+  publish-to-testpypi:
+    name: Publish distribution to TestPyPI
+    runs-on: ubuntu-latest
+    needs:
+      - build
+    environment:
+      name: testpypi
+      url: https://test.pypi.org/p/arcaea-offline-ocr
+    permissions:
+      id-token: write
+
+    steps:
+      - name: Download the distribution files
+        uses: actions/download-artifact@v4
+        with:
+          name: python-package-distributions
+          path: dist/
+
+      - name: Publish distribution to TestPyPI
+        uses: pypa/gh-action-pypi-publish@release/v1
+        with:
+          repository-url: https://test.pypi.org/legacy/

pyproject.toml

@@ -1,10 +1,11 @@
 [build-system]
-requires = ["setuptools>=61.0"]
+requires = ["setuptools>=64", "setuptools-scm>=8"]
 build-backend = "setuptools.build_meta"
 
 [project]
+dynamic = ["version"]
+
 name = "arcaea-offline-ocr"
-version = "0.0.99"
 authors = [{ name = "283375", email = "log_283375@163.com" }]
 description = "Extract your Arcaea play result from screenshot."
 readme = "README.md"
@@ -19,6 +20,8 @@ classifiers = [
 "Homepage" = "https://github.com/ArcaeaOffline/core-ocr"
 "Bug Tracker" = "https://github.com/ArcaeaOffline/core-ocr/issues"
 
+[tool.setuptools_scm]
+
 [tool.isort]
 profile = "black"
 src_paths = ["src/arcaea_offline_ocr"]
@@ -34,5 +37,5 @@ generated-members = ["cv2.*"]
 disable = [
   "missing-module-docstring",
   "missing-class-docstring",
-  "missing-function-docstring"
+  "missing-function-docstring",
 ]

src/arcaea_offline_ocr/__init__.py

@@ -1,3 +0,0 @@
-from .crop import *
-from .device import *
-from .utils import *

src/arcaea_offline_ocr/b30/shared.py

@@ -1,15 +0,0 @@
-from dataclasses import dataclass
-from datetime import datetime
-from typing import Optional
-
-
-@dataclass
-class B30OcrResultItem:
-    rating_class: int
-    score: int
-    pure: Optional[int] = None
-    far: Optional[int] = None
-    lost: Optional[int] = None
-    date: Optional[datetime] = None
-    title: Optional[str] = None
-    song_id: Optional[str] = None

src/arcaea_offline_ocr/device/__init__.py

@@ -1,2 +0,0 @@
-from .common import DeviceOcrResult
-from .ocr import DeviceOcr

src/arcaea_offline_ocr/device/common.py

@@ -1,17 +0,0 @@
-from dataclasses import dataclass
-from typing import Optional
-
-
-@dataclass
-class DeviceOcrResult:
-    rating_class: int
-    score: int
-    pure: Optional[int] = None
-    far: Optional[int] = None
-    lost: Optional[int] = None
-    max_recall: Optional[int] = None
-    song_id: Optional[str] = None
-    song_id_possibility: Optional[float] = None
-    clear_status: Optional[int] = None
-    partner_id: Optional[str] = None
-    partner_id_possibility: Optional[float] = None

src/arcaea_offline_ocr/device/rois/__init__.py

@@ -1,3 +0,0 @@
-from .definition import *
-from .extractor import *
-from .masker import *

src/arcaea_offline_ocr/device/rois/definition/__init__.py

@@ -1,2 +0,0 @@
-from .auto import *
-from .common import DeviceRois

src/arcaea_offline_ocr/device/rois/definition/common.py

@@ -1,15 +0,0 @@
-from typing import Tuple
-
-Rect = Tuple[int, int, int, int]
-
-
-class DeviceRois:
-    pure: Rect
-    far: Rect
-    lost: Rect
-    score: Rect
-    rating_class: Rect
-    max_recall: Rect
-    jacket: Rect
-    clear_status: Rect
-    partner_icon: Rect

src/arcaea_offline_ocr/device/rois/extractor/__init__.py

@@ -1 +0,0 @@
-from .common import DeviceRoisExtractor

src/arcaea_offline_ocr/device/rois/extractor/common.py

@@ -1,48 +0,0 @@
-from ....crop import crop_xywh
-from ....types import Mat
-from ..definition.common import DeviceRois
-
-
-class DeviceRoisExtractor:
-    def __init__(self, img: Mat, rois: DeviceRois):
-        self.img = img
-        self.sizes = rois
-
-    def __construct_int_rect(self, rect):
-        return tuple(round(r) for r in rect)
-
-    @property
-    def pure(self):
-        return crop_xywh(self.img, self.__construct_int_rect(self.sizes.pure))
-
-    @property
-    def far(self):
-        return crop_xywh(self.img, self.__construct_int_rect(self.sizes.far))
-
-    @property
-    def lost(self):
-        return crop_xywh(self.img, self.__construct_int_rect(self.sizes.lost))
-
-    @property
-    def score(self):
-        return crop_xywh(self.img, self.__construct_int_rect(self.sizes.score))
-
-    @property
-    def jacket(self):
-        return crop_xywh(self.img, self.__construct_int_rect(self.sizes.jacket))
-
-    @property
-    def rating_class(self):
-        return crop_xywh(self.img, self.__construct_int_rect(self.sizes.rating_class))
-
-    @property
-    def max_recall(self):
-        return crop_xywh(self.img, self.__construct_int_rect(self.sizes.max_recall))
-
-    @property
-    def clear_status(self):
-        return crop_xywh(self.img, self.__construct_int_rect(self.sizes.clear_status))
-
-    @property
-    def partner_icon(self):
-        return crop_xywh(self.img, self.__construct_int_rect(self.sizes.partner_icon))

src/arcaea_offline_ocr/device/rois/masker/__init__.py

@@ -1,2 +0,0 @@
-from .auto import *
-from .common import DeviceRoisMasker

src/arcaea_offline_ocr/device/rois/masker/common.py

@@ -1,59 +0,0 @@
-from ....types import Mat
-
-
-class DeviceRoisMasker:
-    @classmethod
-    def pure(cls, roi_bgr: Mat) -> Mat:
-        raise NotImplementedError()
-
-    @classmethod
-    def far(cls, roi_bgr: Mat) -> Mat:
-        raise NotImplementedError()
-
-    @classmethod
-    def lost(cls, roi_bgr: Mat) -> Mat:
-        raise NotImplementedError()
-
-    @classmethod
-    def score(cls, roi_bgr: Mat) -> Mat:
-        raise NotImplementedError()
-
-    @classmethod
-    def rating_class_pst(cls, roi_bgr: Mat) -> Mat:
-        raise NotImplementedError()
-
-    @classmethod
-    def rating_class_prs(cls, roi_bgr: Mat) -> Mat:
-        raise NotImplementedError()
-
-    @classmethod
-    def rating_class_ftr(cls, roi_bgr: Mat) -> Mat:
-        raise NotImplementedError()
-
-    @classmethod
-    def rating_class_byd(cls, roi_bgr: Mat) -> Mat:
-        raise NotImplementedError()
-
-    @classmethod
-    def rating_class_etr(cls, roi_bgr: Mat) -> Mat:
-        raise NotImplementedError()
-
-    @classmethod
-    def max_recall(cls, roi_bgr: Mat) -> Mat:
-        raise NotImplementedError()
-
-    @classmethod
-    def clear_status_track_lost(cls, roi_bgr: Mat) -> Mat:
-        raise NotImplementedError()
-
-    @classmethod
-    def clear_status_track_complete(cls, roi_bgr: Mat) -> Mat:
-        raise NotImplementedError()
-
-    @classmethod
-    def clear_status_full_recall(cls, roi_bgr: Mat) -> Mat:
-        raise NotImplementedError()
-
-    @classmethod
-    def clear_status_pure_memory(cls, roi_bgr: Mat) -> Mat:
-        raise NotImplementedError()

src/arcaea_offline_ocr/phash_db.py

@@ -1,119 +0,0 @@
-import sqlite3
-from typing import List, Union
-
-import cv2
-import numpy as np
-
-from .types import Mat
-
-
-def phash_opencv(img_gray, hash_size=8, highfreq_factor=4):
-    # type: (Union[Mat, np.ndarray], int, int) -> np.ndarray
-    """
-    Perceptual Hash computation.
-
-    Implementation follows
-    http://www.hackerfactor.com/blog/index.php?/archives/432-Looks-Like-It.html
-
-    Adapted from `imagehash.phash`, pure opencv implementation
-
-    The result is slightly different from `imagehash.phash`.
-    """
-    if hash_size < 2:
-        raise ValueError("Hash size must be greater than or equal to 2")
-
-    img_size = hash_size * highfreq_factor
-    image = cv2.resize(img_gray, (img_size, img_size), interpolation=cv2.INTER_LANCZOS4)
-    image = np.float32(image)
-    dct = cv2.dct(image)
-    dctlowfreq = dct[:hash_size, :hash_size]
-    med = np.median(dctlowfreq)
-    diff = dctlowfreq > med
-    return diff
-
-
-def hamming_distance_sql_function(user_input, db_entry) -> int:
-    return np.count_nonzero(
-        np.frombuffer(user_input, bool) ^ np.frombuffer(db_entry, bool)
-    )
-
-
-class ImagePhashDatabase:
-    def __init__(self, db_path: str):
-        with sqlite3.connect(db_path) as conn:
-            self.hash_size = int(
-                conn.execute(
-                    "SELECT value FROM properties WHERE key = 'hash_size'"
-                ).fetchone()[0]
-            )
-            self.highfreq_factor = int(
-                conn.execute(
-                    "SELECT value FROM properties WHERE key = 'highfreq_factor'"
-                ).fetchone()[0]
-            )
-            self.built_timestamp = int(
-                conn.execute(
-                    "SELECT value FROM properties WHERE key = 'built_timestamp'"
-                ).fetchone()[0]
-            )
-
-            self.ids: List[str] = [
-                i[0] for i in conn.execute("SELECT id FROM hashes").fetchall()
-            ]
-            self.hashes_byte = [
-                i[0] for i in conn.execute("SELECT hash FROM hashes").fetchall()
-            ]
-            self.hashes = [np.frombuffer(hb, bool) for hb in self.hashes_byte]
-
-            self.jacket_ids: List[str] = []
-            self.jacket_hashes = []
-            self.partner_icon_ids: List[str] = []
-            self.partner_icon_hashes = []
-
-            for _id, _hash in zip(self.ids, self.hashes):
-                id_splitted = _id.split("||")
-                if len(id_splitted) > 1 and id_splitted[0] == "partner_icon":
-                    self.partner_icon_ids.append(id_splitted[1])
-                    self.partner_icon_hashes.append(_hash)
-                else:
-                    self.jacket_ids.append(_id)
-                    self.jacket_hashes.append(_hash)
-
-    def calculate_phash(self, img_gray: Mat):
-        return phash_opencv(
-            img_gray, hash_size=self.hash_size, highfreq_factor=self.highfreq_factor
-        )
-
-    def lookup_hash(self, image_hash: np.ndarray, *, limit: int = 5):
-        image_hash = image_hash.flatten()
-        xor_results = [
-            (id, np.count_nonzero(image_hash ^ h))
-            for id, h in zip(self.ids, self.hashes)
-        ]
-        return sorted(xor_results, key=lambda r: r[1])[:limit]
-
-    def lookup_image(self, img_gray: Mat):
-        image_hash = self.calculate_phash(img_gray)
-        return self.lookup_hash(image_hash)[0]
-
-    def lookup_jackets(self, img_gray: Mat, *, limit: int = 5):
-        image_hash = self.calculate_phash(img_gray).flatten()
-        xor_results = [
-            (id, np.count_nonzero(image_hash ^ h))
-            for id, h in zip(self.jacket_ids, self.jacket_hashes)
-        ]
-        return sorted(xor_results, key=lambda r: r[1])[:limit]
-
-    def lookup_jacket(self, img_gray: Mat):
-        return self.lookup_jackets(img_gray)[0]
-
-    def lookup_partner_icons(self, img_gray: Mat, *, limit: int = 5):
-        image_hash = self.calculate_phash(img_gray).flatten()
-        xor_results = [
-            (id, np.count_nonzero(image_hash ^ h))
-            for id, h in zip(self.partner_icon_ids, self.partner_icon_hashes)
-        ]
-        return sorted(xor_results, key=lambda r: r[1])[:limit]
-
-    def lookup_partner_icon(self, img_gray: Mat):
-        return self.lookup_partner_icons(img_gray)[0]

src/arcaea_offline_ocr/scenarios/b30/__init__.py

@@ -0,0 +1,3 @@
+from .chieri import ChieriBotV4Best30Scenario
+
+__all__ = ["ChieriBotV4Best30Scenario"]

src/arcaea_offline_ocr/scenarios/b30/base.py

@@ -0,0 +1,22 @@
+from abc import abstractmethod
+from typing import TYPE_CHECKING, List
+
+from arcaea_offline_ocr.scenarios.base import OcrScenario, OcrScenarioResult
+
+if TYPE_CHECKING:
+    from arcaea_offline_ocr.types import Mat
+
+
+class Best30Scenario(OcrScenario):
+    @abstractmethod
+    def components(self, img: "Mat", /) -> List["Mat"]: ...
+
+    @abstractmethod
+    def result(self, component_img: "Mat", /, *args, **kwargs) -> OcrScenarioResult: ...
+
+    @abstractmethod
+    def results(self, img: "Mat", /, *args, **kwargs) -> List[OcrScenarioResult]:
+        """
+        Commonly a shorthand for `[self.result(comp) for comp in self.components(img)]`
+        """
+        ...

src/arcaea_offline_ocr/scenarios/b30/chieri/__init__.py

@@ -0,0 +1,3 @@
+from .v4 import ChieriBotV4Best30Scenario
+
+__all__ = ["ChieriBotV4Best30Scenario"]

src/arcaea_offline_ocr/scenarios/b30/chieri/v4/__init__.py

@@ -0,0 +1,3 @@
+from .impl import ChieriBotV4Best30Scenario
+
+__all__ = ["ChieriBotV4Best30Scenario"]

src/arcaea_offline_ocr/scenarios/b30/chieri/v4/colors.py

@@ -27,11 +27,11 @@ FAR_BG_MAX_HSV = np.array([20, 255, 255], np.uint8)
 LOST_BG_MIN_HSV = np.array([115, 60, 150], np.uint8)
 LOST_BG_MAX_HSV = np.array([140, 255, 255], np.uint8)
 
-BYD_MIN_HSV = (158, 120, 0)
-BYD_MAX_HSV = (172, 255, 255)
+BYD_MIN_HSV = np.array([158, 120, 0], np.uint8)
+BYD_MAX_HSV = np.array([172, 255, 255], np.uint8)
 
-FTR_MIN_HSV = (145, 70, 0)
-FTR_MAX_HSV = (160, 255, 255)
+FTR_MIN_HSV = np.array([145, 70, 0], np.uint8)
+FTR_MAX_HSV = np.array([160, 255, 255], np.uint8)
 
-PRS_MIN_HSV = (45, 60, 0)
-PRS_MAX_HSV = (70, 255, 255)
+PRS_MIN_HSV = np.array([45, 60, 0], np.uint8)
+PRS_MAX_HSV = np.array([70, 255, 255], np.uint8)

src/arcaea_offline_ocr/scenarios/b30/chieri/v4/impl.py

@@ -3,10 +3,16 @@ from typing import List, Optional, Tuple
 import cv2
 import numpy as np
 
-from ....crop import crop_xywh
-from ....phash_db import ImagePhashDatabase
-from ....types import Mat
-from ...shared import B30OcrResultItem
+from arcaea_offline_ocr.crop import crop_xywh
+from arcaea_offline_ocr.providers import (
+    ImageCategory,
+    ImageIdProvider,
+    OcrKNearestTextProvider,
+)
+from arcaea_offline_ocr.scenarios.b30.base import Best30Scenario
+from arcaea_offline_ocr.scenarios.base import OcrScenarioResult
+from arcaea_offline_ocr.types import Mat
+
 from .colors import (
     BYD_MAX_HSV,
     BYD_MIN_HSV,
@@ -22,29 +28,20 @@ from .colors import (
     PURE_BG_MIN_HSV,
 )
 from .rois import ChieriBotV4Rois
-from ....providers.knn import OcrKNearestTextProvider
 
 
-class ChieriBotV4Ocr:
+class ChieriBotV4Best30Scenario(Best30Scenario):
     def __init__(
         self,
         score_knn_provider: OcrKNearestTextProvider,
         pfl_knn_provider: OcrKNearestTextProvider,
-        phash_db: ImagePhashDatabase,
+        image_id_provider: ImageIdProvider,
         factor: float = 1.0,
     ):
-        self.__phash_db = phash_db
         self.__rois = ChieriBotV4Rois(factor)
         self.pfl_knn_provider = pfl_knn_provider
         self.score_knn_provider = score_knn_provider
-
-    @property
-    def phash_db(self):
-        return self.__phash_db
-
-    @phash_db.setter
-    def phash_db(self, phash_db: ImagePhashDatabase):
-        self.__phash_db = phash_db
+        self.image_id_provider = image_id_provider
 
     @property
     def rois(self):
@@ -77,12 +74,12 @@ class ChieriBotV4Ocr:
         else:
             return max(enumerate(rating_class_results), key=lambda i: i[1])[0] + 1
 
-    def ocr_component_song_id(self, component_bgr: Mat):
+    def ocr_component_song_id_results(self, component_bgr: Mat):
         jacket_rect = self.rois.component_rois.jacket_rect.floored()
         jacket_roi = cv2.cvtColor(
             crop_xywh(component_bgr, jacket_rect), cv2.COLOR_BGR2GRAY
         )
-        return self.phash_db.lookup_jacket(jacket_roi)[0]
+        return self.image_id_provider.results(jacket_roi, ImageCategory.JACKET)
 
     def ocr_component_score_knn(self, component_bgr: Mat) -> int:
         # sourcery skip: inline-immediately-returned-variable
@@ -191,28 +188,36 @@ class ChieriBotV4Ocr:
         except Exception:
             return (None, None, None)
 
-    def ocr_component(self, component_bgr: Mat) -> B30OcrResultItem:
+    def ocr_component(self, component_bgr: Mat) -> OcrScenarioResult:
         component_blur = cv2.GaussianBlur(component_bgr, (5, 5), 0)
         rating_class = self.ocr_component_rating_class(component_blur)
-        song_id = self.ocr_component_song_id(component_bgr)
-        # title = self.ocr_component_title(component_blur)
+        song_id_results = self.ocr_component_song_id_results(component_bgr)
         # score = self.ocr_component_score(component_blur)
         score = self.ocr_component_score_knn(component_bgr)
         pure, far, lost = self.ocr_component_pfl(component_bgr)
-        return B30OcrResultItem(
-            song_id=song_id,
+        return OcrScenarioResult(
+            song_id=song_id_results[0].image_id,
+            song_id_results=song_id_results,
             rating_class=rating_class,
-            # title=title,
             score=score,
             pure=pure,
             far=far,
             lost=lost,
-            date=None,
+            played_at=None,
         )
 
-    def ocr(self, img_bgr: Mat) -> List[B30OcrResultItem]:
-        self.set_factor(img_bgr)
-        return [
-            self.ocr_component(component_bgr)
-            for component_bgr in self.rois.components(img_bgr)
-        ]
+    def components(self, img: Mat, /):
+        """
+        :param img: BGR format image
+        """
+        self.set_factor(img)
+        return self.rois.components(img)
+
+    def result(self, component_img: Mat, /):
+        return self.ocr_component(component_img)
+
+    def results(self, img: Mat, /) -> List[OcrScenarioResult]:
+        """
+        :param img: BGR format image
+        """
+        return [self.ocr_component(component) for component in self.components(img)]

src/arcaea_offline_ocr/scenarios/b30/chieri/v4/rois.py

@@ -1,8 +1,7 @@
 from typing import List
 
-from ....crop import crop_xywh
-from ....types import Mat, XYWHRect
-from ....utils import apply_factor
+from arcaea_offline_ocr.crop import crop_xywh
+from arcaea_offline_ocr.types import Mat, XYWHRect
 
 
 class ChieriBotV4ComponentRois:
@@ -19,39 +18,39 @@ class ChieriBotV4ComponentRois:
 
     @property
     def top_font_color_detect(self):
-        return apply_factor(XYWHRect(35, 10, 120, 100), self.factor)
+        return XYWHRect(35, 10, 120, 100), self.factor
 
     @property
     def bottom_font_color_detect(self):
-        return apply_factor(XYWHRect(30, 125, 175, 110), self.factor)
+        return XYWHRect(30, 125, 175, 110) * self.factor
 
     @property
     def bg_point(self):
-        return apply_factor((75, 10), self.factor)
+        return (75 * self.factor, 10 * self.factor)
 
     @property
     def rating_class_rect(self):
-        return apply_factor(XYWHRect(21, 40, 7, 20), self.factor)
+        return XYWHRect(21, 40, 7, 20) * self.factor
 
     @property
     def title_rect(self):
-        return apply_factor(XYWHRect(35, 10, 430, 50), self.factor)
+        return XYWHRect(35, 10, 430, 50) * self.factor
 
     @property
     def jacket_rect(self):
-        return apply_factor(XYWHRect(263, 0, 239, 239), self.factor)
+        return XYWHRect(263, 0, 239, 239) * self.factor
 
     @property
     def score_rect(self):
-        return apply_factor(XYWHRect(30, 60, 270, 55), self.factor)
+        return XYWHRect(30, 60, 270, 55) * self.factor
 
     @property
     def pfl_rect(self):
-        return apply_factor(XYWHRect(50, 125, 80, 100), self.factor)
+        return XYWHRect(50, 125, 80, 100) * self.factor
 
     @property
     def date_rect(self):
-        return apply_factor(XYWHRect(205, 200, 225, 25), self.factor)
+        return XYWHRect(205, 200, 225, 25) * self.factor
 
 
 class ChieriBotV4Rois:
@@ -74,27 +73,27 @@ class ChieriBotV4Rois:
 
     @property
     def top(self):
-        return apply_factor(823, self.factor)
+        return 823 * self.factor
 
     @property
     def left(self):
-        return apply_factor(107, self.factor)
+        return 107 * self.factor
 
     @property
     def width(self):
-        return apply_factor(502, self.factor)
+        return 502 * self.factor
 
     @property
     def height(self):
-        return apply_factor(240, self.factor)
+        return 240 * self.factor
 
     @property
     def vertical_gap(self):
-        return apply_factor(74, self.factor)
+        return 74 * self.factor
 
     @property
     def horizontal_gap(self):
-        return apply_factor(40, self.factor)
+        return 40 * self.factor
 
     @property
     def horizontal_items(self):
@@ -104,7 +103,7 @@ class ChieriBotV4Rois:
 
     @property
     def b33_vertical_gap(self):
-        return apply_factor(121, self.factor)
+        return 121 * self.factor
 
     def components(self, img_bgr: Mat) -> List[Mat]:
         first_rect = XYWHRect(x=self.left, y=self.top, w=self.width, h=self.height)

src/arcaea_offline_ocr/scenarios/base.py

@@ -0,0 +1,38 @@
+from abc import ABC
+from dataclasses import dataclass, field
+from datetime import datetime
+from typing import Sequence, Optional
+
+from arcaea_offline_ocr.providers import ImageIdProviderResult
+
+
+@dataclass(kw_only=True)
+class OcrScenarioResult:
+    song_id: str
+    rating_class: int
+    score: int
+
+    song_id_results: Sequence[ImageIdProviderResult] = field(default_factory=lambda: [])
+    partner_id_results: Sequence[ImageIdProviderResult] = field(
+        default_factory=lambda: []
+    )
+
+    pure: Optional[int] = None
+    pure_inaccurate: Optional[int] = None
+    pure_early: Optional[int] = None
+    pure_late: Optional[int] = None
+    far: Optional[int] = None
+    far_inaccurate: Optional[int] = None
+    far_early: Optional[int] = None
+    far_late: Optional[int] = None
+    lost: Optional[int] = None
+
+    played_at: Optional[datetime] = None
+    max_recall: Optional[int] = None
+    clear_status: Optional[int] = None
+    clear_type: Optional[int] = None
+    modifier: Optional[int] = None
+
+
+class OcrScenario(ABC):
+    pass

src/arcaea_offline_ocr/scenarios/device/__init__.py

@@ -0,0 +1,13 @@
+from .extractor import DeviceRoisExtractor
+from .impl import DeviceScenario
+from .masker import DeviceRoisMaskerAutoT1, DeviceRoisMaskerAutoT2
+from .rois import DeviceRoisAutoT1, DeviceRoisAutoT2
+
+__all__ = [
+    "DeviceRoisMaskerAutoT1",
+    "DeviceRoisMaskerAutoT2",
+    "DeviceRoisAutoT1",
+    "DeviceRoisAutoT2",
+    "DeviceRoisExtractor",
+    "DeviceScenario",
+]

src/arcaea_offline_ocr/scenarios/device/base.py

@@ -0,0 +1,8 @@
+from abc import abstractmethod
+
+from arcaea_offline_ocr.scenarios.base import OcrScenario, OcrScenarioResult
+
+
+class DeviceScenarioBase(OcrScenario):
+    @abstractmethod
+    def result(self) -> OcrScenarioResult: ...

src/arcaea_offline_ocr/scenarios/device/extractor/__init__.py

@@ -0,0 +1,3 @@
+from .base import DeviceRoisExtractor
+
+__all__ = ["DeviceRoisExtractor"]

src/arcaea_offline_ocr/scenarios/device/extractor/base.py

@@ -0,0 +1,46 @@
+from arcaea_offline_ocr.crop import crop_xywh
+from arcaea_offline_ocr.types import Mat
+
+from ..rois.base import DeviceRois
+
+
+class DeviceRoisExtractor:
+    def __init__(self, img: Mat, rois: DeviceRois):
+        self.img = img
+        self.sizes = rois
+
+    @property
+    def pure(self):
+        return crop_xywh(self.img, self.sizes.pure.rounded())
+
+    @property
+    def far(self):
+        return crop_xywh(self.img, self.sizes.far.rounded())
+
+    @property
+    def lost(self):
+        return crop_xywh(self.img, self.sizes.lost.rounded())
+
+    @property
+    def score(self):
+        return crop_xywh(self.img, self.sizes.score.rounded())
+
+    @property
+    def jacket(self):
+        return crop_xywh(self.img, self.sizes.jacket.rounded())
+
+    @property
+    def rating_class(self):
+        return crop_xywh(self.img, self.sizes.rating_class.rounded())
+
+    @property
+    def max_recall(self):
+        return crop_xywh(self.img, self.sizes.max_recall.rounded())
+
+    @property
+    def clear_status(self):
+        return crop_xywh(self.img, self.sizes.clear_status.rounded())
+
+    @property
+    def partner_icon(self):
+        return crop_xywh(self.img, self.sizes.partner_icon.rounded())

src/arcaea_offline_ocr/scenarios/device/impl.py

@@ -1,26 +1,31 @@
 import cv2
 import numpy as np
 
-from ..phash_db import ImagePhashDatabase
-from ..providers.knn import OcrKNearestTextProvider
-from ..types import Mat
-from .common import DeviceOcrResult
-from .rois.extractor import DeviceRoisExtractor
-from .rois.masker import DeviceRoisMasker
+from arcaea_offline_ocr.providers import (
+    ImageCategory,
+    ImageIdProvider,
+    OcrKNearestTextProvider,
+)
+from arcaea_offline_ocr.scenarios.base import OcrScenarioResult
+from arcaea_offline_ocr.types import Mat
+
+from .base import DeviceScenarioBase
+from .extractor import DeviceRoisExtractor
+from .masker import DeviceRoisMasker
 
 
-class DeviceOcr:
+class DeviceScenario(DeviceScenarioBase):
     def __init__(
         self,
         extractor: DeviceRoisExtractor,
         masker: DeviceRoisMasker,
         knn_provider: OcrKNearestTextProvider,
-        phash_db: ImagePhashDatabase,
+        image_id_provider: ImageIdProvider,
     ):
         self.extractor = extractor
         self.masker = masker
         self.knn_provider = knn_provider
-        self.phash_db = phash_db
+        self.image_id_provider = image_id_provider
 
     def pfl(self, roi_gray: Mat, factor: float = 1.25):
         def contour_filter(cnt):
@@ -93,14 +98,12 @@ class DeviceOcr:
         ]
         return max(enumerate(results), key=lambda i: np.count_nonzero(i[1]))[0]
 
-    def lookup_song_id(self):
-        return self.phash_db.lookup_jacket(
-            cv2.cvtColor(self.extractor.jacket, cv2.COLOR_BGR2GRAY)
+    def song_id_results(self):
+        return self.image_id_provider.results(
+            cv2.cvtColor(self.extractor.jacket, cv2.COLOR_BGR2GRAY),
+            ImageCategory.JACKET,
         )
 
-    def song_id(self):
-        return self.lookup_song_id()[0]
-
     @staticmethod
     def preprocess_char_icon(img_gray: Mat):
         h, w = img_gray.shape[:2]
@@ -114,21 +117,19 @@ class DeviceOcr:
                 np.array([[0, h], [round(w / 2), h], [0, round(h / 2)]], np.int32),
                 np.array([[w, h], [round(w / 2), h], [w, round(h / 2)]], np.int32),
             ],
-            (128),
+            (128,),
         )
         return img
 
-    def lookup_partner_id(self):
-        return self.phash_db.lookup_partner_icon(
+    def partner_id_results(self):
+        return self.image_id_provider.results(
             self.preprocess_char_icon(
                 cv2.cvtColor(self.extractor.partner_icon, cv2.COLOR_BGR2GRAY)
-            )
+            ),
+            ImageCategory.PARTNER_ICON,
         )
 
-    def partner_id(self):
-        return self.lookup_partner_id()[0]
-
-    def ocr(self) -> DeviceOcrResult:
+    def result(self):
         rating_class = self.rating_class()
         pure = self.pure()
         far = self.far()
@@ -137,20 +138,18 @@ class DeviceOcr:
         max_recall = self.max_recall()
         clear_status = self.clear_status()
 
-        hash_len = self.phash_db.hash_size**2
-        song_id, song_id_distance = self.lookup_song_id()
-        partner_id, partner_id_distance = self.lookup_partner_id()
+        song_id_results = self.song_id_results()
+        partner_id_results = self.partner_id_results()
 
-        return DeviceOcrResult(
+        return OcrScenarioResult(
+            song_id=song_id_results[0].image_id,
+            song_id_results=song_id_results,
             rating_class=rating_class,
             pure=pure,
             far=far,
             lost=lost,
             score=score,
             max_recall=max_recall,
-            song_id=song_id,
-            song_id_possibility=1 - song_id_distance / hash_len,
+            partner_id_results=partner_id_results,
             clear_status=clear_status,
-            partner_id=partner_id,
-            partner_id_possibility=1 - partner_id_distance / hash_len,
         )

src/arcaea_offline_ocr/scenarios/device/masker/__init__.py

@@ -0,0 +1,9 @@
+from .auto import DeviceRoisMaskerAuto, DeviceRoisMaskerAutoT1, DeviceRoisMaskerAutoT2
+from .base import DeviceRoisMasker
+
+__all__ = [
+    "DeviceRoisMaskerAuto",
+    "DeviceRoisMaskerAutoT1",
+    "DeviceRoisMaskerAutoT2",
+    "DeviceRoisMasker",
+]

src/arcaea_offline_ocr/scenarios/device/masker/auto.py

@@ -1,13 +1,12 @@
 import cv2
 import numpy as np
 
-from ....types import Mat
-from .common import DeviceRoisMasker
+from arcaea_offline_ocr.types import Mat
+
+from .base import DeviceRoisMasker
 
 
 class DeviceRoisMaskerAuto(DeviceRoisMasker):
-    # pylint: disable=abstract-method
-
     @staticmethod
     def mask_bgr_in_hsv(roi_bgr: Mat, hsv_lower: Mat, hsv_upper: Mat):
         return cv2.inRange(

src/arcaea_offline_ocr/scenarios/device/masker/base.py

@@ -0,0 +1,61 @@
+from abc import ABC, abstractmethod
+
+from arcaea_offline_ocr.types import Mat
+
+
+class DeviceRoisMasker(ABC):
+    @classmethod
+    @abstractmethod
+    def pure(cls, roi_bgr: Mat) -> Mat: ...
+
+    @classmethod
+    @abstractmethod
+    def far(cls, roi_bgr: Mat) -> Mat: ...
+
+    @classmethod
+    @abstractmethod
+    def lost(cls, roi_bgr: Mat) -> Mat: ...
+
+    @classmethod
+    @abstractmethod
+    def score(cls, roi_bgr: Mat) -> Mat: ...
+
+    @classmethod
+    @abstractmethod
+    def rating_class_pst(cls, roi_bgr: Mat) -> Mat: ...
+
+    @classmethod
+    @abstractmethod
+    def rating_class_prs(cls, roi_bgr: Mat) -> Mat: ...
+
+    @classmethod
+    @abstractmethod
+    def rating_class_ftr(cls, roi_bgr: Mat) -> Mat: ...
+
+    @classmethod
+    @abstractmethod
+    def rating_class_byd(cls, roi_bgr: Mat) -> Mat: ...
+
+    @classmethod
+    @abstractmethod
+    def rating_class_etr(cls, roi_bgr: Mat) -> Mat: ...
+
+    @classmethod
+    @abstractmethod
+    def max_recall(cls, roi_bgr: Mat) -> Mat: ...
+
+    @classmethod
+    @abstractmethod
+    def clear_status_track_lost(cls, roi_bgr: Mat) -> Mat: ...
+
+    @classmethod
+    @abstractmethod
+    def clear_status_track_complete(cls, roi_bgr: Mat) -> Mat: ...
+
+    @classmethod
+    @abstractmethod
+    def clear_status_full_recall(cls, roi_bgr: Mat) -> Mat: ...
+
+    @classmethod
+    @abstractmethod
+    def clear_status_pure_memory(cls, roi_bgr: Mat) -> Mat: ...

src/arcaea_offline_ocr/scenarios/device/rois/__init__.py

@@ -0,0 +1,9 @@
+from .auto import DeviceRoisAuto, DeviceRoisAutoT1, DeviceRoisAutoT2
+from .base import DeviceRois
+
+__all__ = [
+    "DeviceRois",
+    "DeviceRoisAuto",
+    "DeviceRoisAutoT1",
+    "DeviceRoisAutoT2",
+]

src/arcaea_offline_ocr/scenarios/device/rois/auto.py

@@ -1,6 +1,6 @@
-from .common import DeviceRois
+from arcaea_offline_ocr.types import XYWHRect
 
-__all__ = ["DeviceRoisAuto", "DeviceRoisAutoT1", "DeviceRoisAutoT2"]
+from .base import DeviceRois
 
 
 class DeviceRoisAuto(DeviceRois):
@@ -50,7 +50,7 @@ class DeviceRoisAutoT1(DeviceRoisAuto):
 
     @property
     def pure(self):
-        return (
+        return XYWHRect(
             self.pfl_x,
             self.layout_area_h_mid + 110 * self.factor,
             self.pfl_w,
@@ -59,7 +59,7 @@ class DeviceRoisAutoT1(DeviceRoisAuto):
 
     @property
     def far(self):
-        return (
+        return XYWHRect(
             self.pfl_x,
             self.pure[1] + self.pure[3] + 12 * self.factor,
             self.pfl_w,
@@ -68,7 +68,7 @@ class DeviceRoisAutoT1(DeviceRoisAuto):
 
     @property
     def lost(self):
-        return (
+        return XYWHRect(
             self.pfl_x,
             self.far[1] + self.far[3] + 10 * self.factor,
             self.pfl_w,
@@ -79,7 +79,7 @@ class DeviceRoisAutoT1(DeviceRoisAuto):
     def score(self):
         w = 280 * self.factor
         h = 45 * self.factor
-        return (
+        return XYWHRect(
             self.w_mid - w / 2,
             self.layout_area_h_mid - 75 * self.factor - h,
             w,
@@ -88,7 +88,7 @@ class DeviceRoisAutoT1(DeviceRoisAuto):
 
     @property
     def rating_class(self):
-        return (
+        return XYWHRect(
             self.w_mid - 610 * self.factor,
             self.layout_area_h_mid - 180 * self.factor,
             265 * self.factor,
@@ -97,7 +97,7 @@ class DeviceRoisAutoT1(DeviceRoisAuto):
 
     @property
     def max_recall(self):
-        return (
+        return XYWHRect(
             self.w_mid - 465 * self.factor,
             self.layout_area_h_mid - 215 * self.factor,
             150 * self.factor,
@@ -106,7 +106,7 @@ class DeviceRoisAutoT1(DeviceRoisAuto):
 
     @property
     def jacket(self):
-        return (
+        return XYWHRect(
             self.w_mid - 610 * self.factor,
             self.layout_area_h_mid - 143 * self.factor,
             375 * self.factor,
@@ -117,7 +117,7 @@ class DeviceRoisAutoT1(DeviceRoisAuto):
     def clear_status(self):
         w = 550 * self.factor
         h = 60 * self.factor
-        return (
+        return XYWHRect(
             self.w_mid - w / 2,
             self.layout_area_h_mid - 155 * self.factor - h,
             w * 0.4,
@@ -128,7 +128,7 @@ class DeviceRoisAutoT1(DeviceRoisAuto):
     def partner_icon(self):
         w = 90 * self.factor
         h = 75 * self.factor
-        return (self.w_mid - w / 2, 0, w, h)
+        return XYWHRect(self.w_mid - w / 2, 0, w, h)
 
 
 class DeviceRoisAutoT2(DeviceRoisAuto):
@@ -174,7 +174,7 @@ class DeviceRoisAutoT2(DeviceRoisAuto):
 
     @property
     def pure(self):
-        return (
+        return XYWHRect(
             self.pfl_x,
             self.layout_area_h_mid + 175 * self.factor,
             self.pfl_w,
@@ -183,7 +183,7 @@ class DeviceRoisAutoT2(DeviceRoisAuto):
 
     @property
     def far(self):
-        return (
+        return XYWHRect(
             self.pfl_x,
             self.pure[1] + self.pure[3] + 30 * self.factor,
             self.pfl_w,
@@ -192,7 +192,7 @@ class DeviceRoisAutoT2(DeviceRoisAuto):
 
     @property
     def lost(self):
-        return (
+        return XYWHRect(
             self.pfl_x,
             self.far[1] + self.far[3] + 35 * self.factor,
             self.pfl_w,
@@ -203,7 +203,7 @@ class DeviceRoisAutoT2(DeviceRoisAuto):
     def score(self):
         w = 420 * self.factor
         h = 70 * self.factor
-        return (
+        return XYWHRect(
             self.w_mid - w / 2,
             self.layout_area_h_mid - 110 * self.factor - h,
             w,
@@ -212,7 +212,7 @@ class DeviceRoisAutoT2(DeviceRoisAuto):
 
     @property
     def rating_class(self):
-        return (
+        return XYWHRect(
             max(0, self.w_mid - 965 * self.factor),
             self.layout_area_h_mid - 330 * self.factor,
             350 * self.factor,
@@ -221,7 +221,7 @@ class DeviceRoisAutoT2(DeviceRoisAuto):
 
     @property
     def max_recall(self):
-        return (
+        return XYWHRect(
             self.w_mid - 625 * self.factor,
             self.layout_area_h_mid - 275 * self.factor,
             150 * self.factor,
@@ -230,7 +230,7 @@ class DeviceRoisAutoT2(DeviceRoisAuto):
 
     @property
     def jacket(self):
-        return (
+        return XYWHRect(
             self.w_mid - 915 * self.factor,
             self.layout_area_h_mid - 215 * self.factor,
             565 * self.factor,
@@ -241,7 +241,7 @@ class DeviceRoisAutoT2(DeviceRoisAuto):
     def clear_status(self):
         w = 825 * self.factor
         h = 90 * self.factor
-        return (
+        return XYWHRect(
             self.w_mid - w / 2,
             self.layout_area_h_mid - 235 * self.factor - h,
             w * 0.4,
@@ -252,4 +252,4 @@ class DeviceRoisAutoT2(DeviceRoisAuto):
     def partner_icon(self):
         w = 135 * self.factor
         h = 110 * self.factor
-        return (self.w_mid - w / 2, 0, w, h)
+        return XYWHRect(self.w_mid - w / 2, 0, w, h)

src/arcaea_offline_ocr/scenarios/device/rois/base.py

@@ -0,0 +1,33 @@
+from abc import ABC, abstractmethod
+
+from arcaea_offline_ocr.types import XYWHRect
+
+
+class DeviceRois(ABC):
+    @property
+    @abstractmethod
+    def pure(self) -> XYWHRect: ...
+    @property
+    @abstractmethod
+    def far(self) -> XYWHRect: ...
+    @property
+    @abstractmethod
+    def lost(self) -> XYWHRect: ...
+    @property
+    @abstractmethod
+    def score(self) -> XYWHRect: ...
+    @property
+    @abstractmethod
+    def rating_class(self) -> XYWHRect: ...
+    @property
+    @abstractmethod
+    def max_recall(self) -> XYWHRect: ...
+    @property
+    @abstractmethod
+    def jacket(self) -> XYWHRect: ...
+    @property
+    @abstractmethod
+    def clear_status(self) -> XYWHRect: ...
+    @property
+    @abstractmethod
+    def partner_icon(self) -> XYWHRect: ...

src/arcaea_offline_ocr/types.py

@@ -34,3 +34,9 @@ class XYWHRect(NamedTuple):
             raise TypeError()
 
         return self.__class__(*[a - b for a, b in zip(self, other)])
+
+    def __mul__(self, other):
+        if not isinstance(other, (int, float)):
+            raise TypeError()
+
+        return self.__class__(*[v * other for v in self])

src/arcaea_offline_ocr/utils.py

@@ -1,11 +1,6 @@
-from collections.abc import Iterable
-from typing import TypeVar, overload
-
 import cv2
 import numpy as np
 
-from .types import XYWHRect
-
 __all__ = ["imread_unicode"]
 
 
@@ -13,27 +8,3 @@ def imread_unicode(filepath: str, flags: int = cv2.IMREAD_UNCHANGED):
     # https://stackoverflow.com/a/57872297/16484891
     # CC BY-SA 4.0
     return cv2.imdecode(np.fromfile(filepath, dtype=np.uint8), flags)
-
-
-@overload
-def apply_factor(item: int, factor: float) -> float: ...
-
-
-@overload
-def apply_factor(item: float, factor: float) -> float: ...
-
-
-T = TypeVar("T", bound=Iterable)
-
-
-@overload
-def apply_factor(item: T, factor: float) -> T: ...
-
-
-def apply_factor(item, factor: float):
-    if isinstance(item, (int, float)):
-        return item * factor
-    if isinstance(item, XYWHRect):
-        return item.__class__(*[i * factor for i in item])
-    if isinstance(item, Iterable):
-        return item.__class__([i * factor for i in item])