style: Fixed mypy and moved tool configs to pyproject.toml

.flake8

@@ -1,5 +1,5 @@
 [flake8]
 max-line-length = 120
 ignore = E402, E265, F403, W503, W504, E731
-exclude = .circleci, .git, venv*, docs, build
+exclude = .git, venv*, build
 per-file-ignores = **/__init__.py:F401

.github/workflows/style.yml

@@ -7,12 +7,12 @@ on:
     branches: main
 
 jobs:
-  flake8-py3:
+  flake8:
     runs-on: ${{ matrix.os }}
     strategy:
       matrix:
         os: [ubuntu-latest]
-        python: [3.7]
+        python: [3.8]
     steps:
       - uses: actions/checkout@v2
       - name: Set up Python
@@ -24,14 +24,14 @@ jobs:
         run: |
           pip install flake8
           flake8 --version
-          flake8 ./
+          flake8
 
-  isort-py3:
+  isort:
     runs-on: ${{ matrix.os }}
     strategy:
       matrix:
         os: [ubuntu-latest]
-        python: [3.7]
+        python: [3.8]
     steps:
       - uses: actions/checkout@v2
       - name: Set up Python
@@ -46,12 +46,12 @@ jobs:
           isort .
           if [ -n "$(git status --porcelain --untracked-files=no)" ]; then exit 1; else echo "All clear"; fi
 
-  mypy-py3:
+  mypy:
     runs-on: ${{ matrix.os }}
     strategy:
       matrix:
         os: [ubuntu-latest]
-        python: [3.7]
+        python: [3.8]
     steps:
       - uses: actions/checkout@v2
       - name: Set up Python
@@ -67,19 +67,19 @@ jobs:
       - name: Install dependencies
         run: |
           python -m pip install --upgrade pip
-          pip install -e .[tf] --upgrade
+          pip install -e .[all] --upgrade
           pip install mypy
       - name: Run mypy
         run: |
           mypy --version
-          mypy --config-file mypy.ini doctr/
+          mypy
 
-  pydocstyle-py3:
+  pydocstyle:
     runs-on: ${{ matrix.os }}
     strategy:
       matrix:
         os: [ubuntu-latest]
-        python: [3.7]
+        python: [3.8]
     steps:
       - uses: actions/checkout@v2
       - name: Set up Python
@@ -89,6 +89,6 @@ jobs:
           architecture: x64
       - name: Run pydocstyle
         run: |
-          pip install pydocstyle
+          pip install pydocstyle[toml]
           pydocstyle --version
-          pydocstyle doctr/
+          pydocstyle

api/app/main.py

@@ -5,11 +5,12 @@
 
 import time
 
-from app import config as cfg
-from app.routes import detection, ocr, recognition
 from fastapi import FastAPI, Request
 from fastapi.openapi.utils import get_openapi
 
+from app import config as cfg
+from app.routes import detection, ocr, recognition
+
 app = FastAPI(title=cfg.PROJECT_NAME, description=cfg.PROJECT_DESCRIPTION, debug=cfg.DEBUG, version=cfg.VERSION)
 
 

api/app/routes/detection.py

@@ -5,10 +5,10 @@
 
 from typing import List
 
-from app.schemas import DetectionOut
-from app.vision import det_predictor
 from fastapi import APIRouter, File, UploadFile, status
 
+from app.schemas import DetectionOut
+from app.vision import det_predictor
 from doctr.io import decode_img_as_tensor
 
 router = APIRouter()

api/app/routes/ocr.py

@@ -5,10 +5,10 @@
 
 from typing import List
 
-from app.schemas import OCROut
-from app.vision import predictor
 from fastapi import APIRouter, File, UploadFile, status
 
+from app.schemas import OCROut
+from app.vision import predictor
 from doctr.io import decode_img_as_tensor
 
 router = APIRouter()

api/app/routes/recognition.py

@@ -3,10 +3,10 @@
 # This program is licensed under the Apache License version 2.
 # See LICENSE or go to <https://www.apache.org/licenses/LICENSE-2.0.txt> for full license details.
 
-from app.schemas import RecognitionOut
-from app.vision import reco_predictor
 from fastapi import APIRouter, File, UploadFile, status
 
+from app.schemas import RecognitionOut
+from app.vision import reco_predictor
 from doctr.io import decode_img_as_tensor
 
 router = APIRouter()

api/tests/conftest.py

@@ -5,9 +5,10 @@
 
 import pytest
 import requests
-from app.main import app
 from httpx import AsyncClient
 
+from app.main import app
+
 
 @pytest.fixture(scope="session")
 def mock_recognition_image(tmpdir_factory):

docs/source/conf.py

@@ -35,10 +35,10 @@
 # extensions coming with Sphinx (named 'sphinx.ext.*') or your custom
 # ones.
 extensions = [
-	'sphinx.ext.autodoc',
-	'sphinx.ext.napoleon',
-	'sphinx.ext.intersphinx',
-	'sphinx.ext.viewcode',
+    'sphinx.ext.autodoc',
+    'sphinx.ext.napoleon',
+    'sphinx.ext.intersphinx',
+    'sphinx.ext.viewcode',
     'sphinx.ext.coverage',
     'sphinx.ext.mathjax',
     'sphinx.ext.autosectionlabel',
@@ -114,9 +114,10 @@
 # A list of files that should not be packed into the epub file.
 epub_exclude_files = ['search.html']
 
-# Add googleanalytics id
-# ref: https://github.com/orenhecht/googleanalytics/blob/master/sphinxcontrib/googleanalytics.py
+
 def add_ga_javascript(app, pagename, templatename, context, doctree):
+    # Add googleanalytics id
+    # ref: https://github.com/orenhecht/googleanalytics/blob/master/sphinxcontrib/googleanalytics.py
 
     metatags = context.get('metatags', '')
     metatags += """

doctr/datasets/cord.py

@@ -77,6 +77,7 @@ def __init__(
                         if len(word["text"]) > 0:
                             x = word["quad"]["x1"], word["quad"]["x2"], word["quad"]["x3"], word["quad"]["x4"]
                             y = word["quad"]["y1"], word["quad"]["y2"], word["quad"]["y3"], word["quad"]["y4"]
+                            box: Union[List[float], np.ndarray]
                             if use_polygons:
                                 # (x, y) coordinates of top left, top right, bottom right, bottom left corners
                                 box = np.array([

doctr/datasets/datasets/base.py

@@ -55,7 +55,7 @@ def __getitem__(
 
         if self.img_transforms is not None:
             # typing issue cf. https://github.com/python/mypy/issues/5485
-            img = self.img_transforms(img)  # type: ignore[call-arg]
+            img = self.img_transforms(img)
 
         if self.sample_transforms is not None:
             img, target = self.sample_transforms(img, target)

doctr/datasets/detection.py

@@ -58,7 +58,7 @@ def __init__(
             if not os.path.exists(os.path.join(self.root, img_name)):
                 raise FileNotFoundError(f"unable to locate {os.path.join(self.root, img_name)}")
 
-            polygons = np.asarray(label['polygons'], dtype=np_dtype)
+            polygons: np.ndarray = np.asarray(label['polygons'], dtype=np_dtype)
             geoms = polygons if use_polygons else np.concatenate((polygons.min(axis=1), polygons.max(axis=1)), axis=1)
 
             self.data.append((img_name, np.asarray(geoms, dtype=np_dtype)))

doctr/datasets/doc_artefacts.py

@@ -62,8 +62,8 @@ def __init__(
                 raise FileNotFoundError(f"unable to locate {os.path.join(tmp_root, img_name)}")
 
             # xmin, ymin, xmax, ymax
-            boxes = np.asarray([obj['geometry'] for obj in label], dtype=np_dtype)
-            classes = np.asarray([self.CLASSES.index(obj['label']) for obj in label], dtype=np.int64)
+            boxes: np.ndarray = np.asarray([obj['geometry'] for obj in label], dtype=np_dtype)
+            classes: np.ndarray = np.asarray([self.CLASSES.index(obj['label']) for obj in label], dtype=np.int64)
             if use_polygons:
                 # (x, y) coordinates of top left, top right, bottom right, bottom left corners
                 boxes = np.stack(

doctr/datasets/ic03.py

@@ -93,7 +93,7 @@ def __init__(
 
             # filter images without boxes
             if len(_boxes) > 0:
-                boxes = np.asarray(_boxes, dtype=np_dtype)
+                boxes: np.ndarray = np.asarray(_boxes, dtype=np_dtype)
                 # Get the labels
                 labels = [lab.text for rect in rectangles for lab in rect if lab.text]
 

doctr/datasets/ic13.py

@@ -76,7 +76,7 @@ def __init__(
                 ]
             labels = [line[-1].replace("\"", "") for line in _lines]
             # xmin, ymin, xmax, ymax
-            box_targets = np.array([list(map(int, line[:4])) for line in _lines], dtype=np_dtype)
+            box_targets: np.ndarray = np.array([list(map(int, line[:4])) for line in _lines], dtype=np_dtype)
             if use_polygons:
                 # (x, y) coordinates of top left, top right, bottom right, bottom left corners
                 box_targets = np.array(

doctr/datasets/sroie.py

@@ -76,8 +76,8 @@ def __init__(
             labels = [",".join(row[8:]) for row in _rows]
             # reorder coordinates (8 -> (4,2) ->
             # (x, y) coordinates of top left, top right, bottom right, bottom left corners) and filter empty lines
-            coords = np.stack([np.array(list(map(int, row[:8])), dtype=np_dtype).reshape((4, 2))
-                              for row in _rows], axis=0)
+            coords: np.ndarray = np.stack([np.array(list(map(int, row[:8])), dtype=np_dtype).reshape((4, 2))
+                                           for row in _rows], axis=0)
 
             if not use_polygons:
                 # xmin, ymin, xmax, ymax

doctr/datasets/svhn.py

@@ -84,7 +84,7 @@ def __init__(
                     box_dict = {k: [int(f[v[0]][()].item()) for v in vals] for k, vals in box.items()}
 
                 # Convert it to the right format
-                coords = np.array([
+                coords: np.ndarray = np.array([
                     box_dict['left'],
                     box_dict['top'],
                     box_dict['width'],
@@ -94,7 +94,7 @@ def __init__(
 
                 if use_polygons:
                     # (x, y) coordinates of top left, top right, bottom right, bottom left corners
-                    box_targets = np.stack(
+                    box_targets: np.ndarray = np.stack(
                         [
                             np.stack([coords[:, 0], coords[:, 1]], axis=-1),
                             np.stack([coords[:, 0] + coords[:, 2], coords[:, 1]], axis=-1),

doctr/datasets/svt.py

@@ -86,7 +86,7 @@ def __init__(
                     for rect in rectangles
                 ]
 
-            boxes = np.asarray(_boxes, dtype=np_dtype)
+            boxes: np.ndarray = np.asarray(_boxes, dtype=np_dtype)
             # Get the labels
             labels = [lab.text for rect in rectangles for lab in rect]
 

doctr/datasets/utils.py

@@ -73,13 +73,13 @@ def encode_string(
         A list encoding the input_string"""
 
     try:
-        return list(map(vocab.index, input_string))  # type: ignore[arg-type]
+        return list(map(vocab.index, input_string))
     except ValueError:
         raise ValueError("some characters cannot be found in 'vocab'")
 
 
 def decode_sequence(
-    input_seq: Union[np.array, SequenceType[int]],
+    input_seq: Union[np.ndarray, SequenceType[int]],
     mapping: str,
 ) -> str:
     """Given a predefined mapping, decode the sequence of numbers to a string
@@ -145,7 +145,7 @@ def encode_sequences(
         default_symbol = pad
     else:  # pad with eos symbol
         default_symbol = eos
-    encoded_data = np.full([len(sequences), target_size], default_symbol, dtype=np.int32)
+    encoded_data: np.ndarray = np.full([len(sequences), target_size], default_symbol, dtype=np.int32)
 
     # Encode the strings
     for idx, seq in enumerate(map(partial(encode_string, vocab=vocab), sequences)):
@@ -176,7 +176,7 @@ def crop_bboxes_from_image(img_path: Union[str, Path], geoms: np.ndarray) -> Lis
     Returns:
         a list of cropped images
     """
-    img = np.array(Image.open(img_path).convert('RGB'))
+    img: np.ndarray = np.array(Image.open(img_path).convert('RGB'))
     # Polygon
     if geoms.ndim == 3 and geoms.shape[1:] == (4, 2):
         return extract_rcrops(img, geoms.astype(dtype=int))

doctr/io/elements.py

@@ -141,7 +141,7 @@ def __init__(
         if geometry is None:
             # Check whether this is a rotated or straight box
             box_resolution_fn = resolve_enclosing_rbbox if len(words[0].geometry) == 4 else resolve_enclosing_bbox
-            geometry = box_resolution_fn([w.geometry for w in words])  # type: ignore[operator, misc]
+            geometry = box_resolution_fn([w.geometry for w in words])  # type: ignore[operator]
 
         super().__init__(words=words)
         self.geometry = geometry
@@ -188,7 +188,7 @@ def __init__(
             box_resolution_fn = resolve_enclosing_rbbox if isinstance(
                 lines[0].geometry, np.ndarray
             ) else resolve_enclosing_bbox
-            geometry = box_resolution_fn(line_boxes + artefact_boxes)  # type: ignore[operator, arg-type]
+            geometry = box_resolution_fn(line_boxes + artefact_boxes)  # type: ignore[operator]
 
         super().__init__(lines=lines, artefacts=artefacts)
         self.geometry = geometry
@@ -287,7 +287,11 @@ def export_as_xml(self, file_title: str = 'docTR - XML export (hOCR)') -> Tuple[
         head = SubElement(page_hocr, 'head')
         SubElement(head, 'title').text = file_title
         SubElement(head, 'meta', attrib={'http-equiv': 'Content-Type', 'content': 'text/html; charset=utf-8'})
-        SubElement(head, 'meta', attrib={'name': 'ocr-system', 'content': f"python-doctr {doctr.__version__}"})
+        SubElement(
+            head,
+            'meta',
+            attrib={'name': 'ocr-system', 'content': f"python-doctr {doctr.__version__}"},  # type: ignore[attr-defined]
+        )
         SubElement(head, 'meta', attrib={'name': 'ocr-capabilities',
                                          'content': 'ocr_page ocr_carea ocr_par ocr_line ocrx_word'})
         # Create the body
@@ -301,7 +305,7 @@ def export_as_xml(self, file_title: str = 'docTR - XML export (hOCR)') -> Tuple[
         for block in self.blocks:
             if len(block.geometry) != 2:
                 raise TypeError("XML export is only available for straight bounding boxes for now.")
-            (xmin, ymin), (xmax, ymax) = block.geometry  # type: ignore[misc]
+            (xmin, ymin), (xmax, ymax) = block.geometry
             block_div = SubElement(body, 'div', attrib={
                 'class': 'ocr_carea',
                 'id': f'block_{block_count}',
@@ -316,7 +320,7 @@ def export_as_xml(self, file_title: str = 'docTR - XML export (hOCR)') -> Tuple[
             })
             block_count += 1
             for line in block.lines:
-                (xmin, ymin), (xmax, ymax) = line.geometry  # type: ignore[misc]
+                (xmin, ymin), (xmax, ymax) = line.geometry
                 # NOTE: baseline, x_size, x_descenders, x_ascenders is currently initalized to 0
                 line_span = SubElement(paragraph, 'span', attrib={
                     'class': 'ocr_line',
@@ -327,7 +331,7 @@ def export_as_xml(self, file_title: str = 'docTR - XML export (hOCR)') -> Tuple[
                 })
                 line_count += 1
                 for word in line.words:
-                    (xmin, ymin), (xmax, ymax) = word.geometry  # type: ignore[misc]
+                    (xmin, ymin), (xmax, ymax) = word.geometry
                     conf = word.confidence
                     word_div = SubElement(line_span, 'span', attrib={
                         'class': 'ocrx_word',

doctr/io/image/__init__.py

@@ -5,4 +5,4 @@
 if is_tf_available():
     from .tensorflow import *
 elif is_torch_available():
-    from .pytorch import *  # type: ignore[misc]
+    from .pytorch import *

doctr/io/image/base.py

@@ -38,8 +38,8 @@ def read_img_as_numpy(
             raise FileNotFoundError(f"unable to access {file}")
         img = cv2.imread(str(file), cv2.IMREAD_COLOR)
     elif isinstance(file, bytes):
-        file = np.frombuffer(file, np.uint8)
-        img = cv2.imdecode(file, cv2.IMREAD_COLOR)
+        _file: np.ndarray = np.frombuffer(file, np.uint8)
+        img = cv2.imdecode(_file, cv2.IMREAD_COLOR)
     else:
         raise TypeError("unsupported object type for argument 'file'")