DEV: Unify mypy options and warn redundant workarounds (#2223)

.github/workflows/github-ci.yaml

@@ -22,14 +22,12 @@ jobs:
     runs-on: ubuntu-20.04
     strategy:
       matrix:
-        python-version: ["3.7", "3.8", "3.9", "3.10", "3.11", "3.12"]
+        python-version: ["3.8", "3.9", "3.10", "3.11", "3.12"]
         use-crypto-lib: ["cryptography"]
         include:
           - python-version: "3.6"
             use-crypto-lib: "pycryptodome"
-          - python-version: "3.9"
-            use-crypto-lib: "pycryptodome"
-          - python-version: "3.10"
+          - python-version: "3.7"
             use-crypto-lib: ""
     steps:
     - name: Update APT packages
@@ -94,9 +92,6 @@ jobs:
     - name: Test with pytest
       run: |
         python -m coverage run --parallel-mode -m pytest tests -vv
-    - name: Test with mypy
-      run : |
-        mypy pypdf --show-error-codes --disallow-untyped-defs --disallow-incomplete-defs
     - name: Upload coverage data
       uses: actions/upload-artifact@v3
       with:
@@ -112,12 +107,6 @@ jobs:
       uses: actions/checkout@v4
       with:
         submodules: 'recursive'
-    - name: Cache Downloaded Files
-      id: cache-downloaded-files
-      uses: actions/cache@v3
-      with:
-        path: '**/tests/pdf_cache/*'
-        key: cache-downloaded-files
     - name: Setup Python 3.11
       uses: actions/setup-python@v4
       with:
@@ -137,6 +126,9 @@ jobs:
       run: |
         echo `ruff --version`
         ruff .
+    - name: Test with mypy
+      run : |
+        mypy pypdf
 
   package:
     name: Build & verify package

.pre-commit-config.yaml

@@ -48,5 +48,5 @@ repos:
     rev: 'v1.5.1'
     hooks:
       - id: mypy
+        additional_dependencies: [types-Pillow==10.0.0.2]
         files: ^pypdf/.*
-        args: [--ignore-missing-imports]

pypdf/_crypt_providers/_cryptography.py

@@ -27,11 +27,11 @@
 
 import secrets
 
-from cryptography import __version__  # type: ignore[import]
-from cryptography.hazmat.primitives import padding  # type: ignore[import]
-from cryptography.hazmat.primitives.ciphers.algorithms import AES, ARC4  # type: ignore[import]
-from cryptography.hazmat.primitives.ciphers.base import Cipher  # type: ignore[import]
-from cryptography.hazmat.primitives.ciphers.modes import CBC, ECB  # type: ignore[import]
+from cryptography import __version__
+from cryptography.hazmat.primitives import padding
+from cryptography.hazmat.primitives.ciphers.algorithms import AES, ARC4
+from cryptography.hazmat.primitives.ciphers.base import Cipher
+from cryptography.hazmat.primitives.ciphers.modes import CBC, ECB
 
 from pypdf._crypt_providers._base import CryptBase
 

pypdf/_crypt_providers/_fallback.py

@@ -34,7 +34,7 @@
 crypt_provider = ("local_crypt_fallback", "0.0.0")
 
 
-class CryptRC4(CryptBase):  # type: ignore
+class CryptRC4(CryptBase):
     def __init__(self, key: bytes) -> None:
         self.s = bytearray(range(256))
         j = 0
@@ -58,7 +58,7 @@ def decrypt(self, data: bytes) -> bytes:
         return self.encrypt(data)
 
 
-class CryptAES(CryptBase):  # type: ignore
+class CryptAES(CryptBase):
     def __init__(self, key: bytes) -> None:
         pass
 

pypdf/_crypt_providers/_pycryptodome.py

@@ -27,9 +27,9 @@
 
 import secrets
 
-from Crypto import __version__  # type: ignore[import]
-from Crypto.Cipher import AES, ARC4  # type: ignore[import]
-from Crypto.Util.Padding import pad  # type: ignore[import]
+from Crypto import __version__
+from Crypto.Cipher import AES, ARC4
+from Crypto.Util.Padding import pad
 
 from pypdf._crypt_providers._base import CryptBase
 

pypdf/_encryption.py

@@ -86,7 +86,7 @@ def encrypt_object(self, obj: PdfObject) -> PdfObject:
                 obj2[key] = self.encrypt_object(value)
             obj = obj2
         elif isinstance(obj, ArrayObject):
-            obj = ArrayObject(self.encrypt_object(x) for x in obj)  # type: ignore
+            obj = ArrayObject(self.encrypt_object(x) for x in obj)
         return obj
 
     def decrypt_object(self, obj: PdfObject) -> PdfObject:

pypdf/_merger.py

@@ -642,7 +642,7 @@ def find_outline_item(
             if isinstance(oi_enum, list):
                 # oi_enum is still an inner node
                 # (OutlineType, if recursive types were supported by mypy)
-                res = self.find_outline_item(outline_item, oi_enum)  # type: ignore
+                res = self.find_outline_item(outline_item, oi_enum)
                 if res:  # deprecated
                     return [i] + res
             elif (

pypdf/_page.py

@@ -1080,8 +1080,7 @@ def _merge_page(
             if PG.ANNOTS in page:
                 annots = page[PG.ANNOTS]
                 if isinstance(annots, ArrayObject):
-                    for ref in annots:
-                        new_annots.append(ref)
+                    new_annots.extend(annots)
 
         for res in (
             RES.EXT_G_STATE,
@@ -2323,7 +2322,7 @@ def _get_fonts(self) -> Tuple[Set[str], Set[str]]:
         assert isinstance(obj, DictionaryObject)
         fonts: Set[str] = set()
         embedded: Set[str] = set()
-        fonts, embedded = _get_fonts_walk(cast(DictionaryObject, obj), fonts, embedded)
+        fonts, embedded = _get_fonts_walk(obj, fonts, embedded)
         unembedded = fonts - embedded
         return embedded, unembedded
 

pypdf/_protocols.py

@@ -5,9 +5,9 @@
 
 try:
     # Python 3.8+: https://peps.python.org/pep-0586
-    from typing import Protocol  # type: ignore[attr-defined]
+    from typing import Protocol
 except ImportError:
-    from typing_extensions import Protocol  # type: ignore[assignment,misc]
+    from typing_extensions import Protocol  # type: ignore[assignment]
 
 from ._utils import StrByteType, StreamType
 

pypdf/_reader.py

@@ -320,7 +320,7 @@ def __init__(
         self._page_id2num: Optional[
             Dict[Any, Any]
         ] = None  # map page indirect_reference number to Page Number
-        if hasattr(stream, "mode") and "b" not in stream.mode:  # type: ignore
+        if hasattr(stream, "mode") and "b" not in stream.mode:
             logger_warning(
                 "PdfReader stream/file object is not in binary mode. "
                 "It may not be read correctly.",
@@ -1035,7 +1035,7 @@ def _build_outline_item(self, node: DictionaryObject) -> Optional[Destination]:
         except KeyError:
             if self.strict:
                 raise PdfReadError(f"Outline Entry Missing /Title attribute: {node!r}")
-            title = ""  # type: ignore
+            title = ""
 
         if "/A" in node:
             # Action, PDFv1.7 Section 12.6 (only type GoTo supported)
@@ -1074,7 +1074,7 @@ def _build_outline_item(self, node: DictionaryObject) -> Optional[Destination]:
                     f"Removed unexpected destination {dest!r} from destination",
                     __name__,
                 )
-            outline_item = self._build_destination(title, None)  # type: ignore
+            outline_item = self._build_destination(title, None)
 
         # if outline item created, add color, format, and child count if present
         if outline_item:
@@ -1229,7 +1229,7 @@ def _flatten(
             self.flattened_pages = []
 
         if PA.TYPE in pages:
-            t = pages[PA.TYPE]  # type: ignore
+            t = pages[PA.TYPE]
         # if pdf has no type, considered as a page if /Kids is missing
         elif PA.KIDS not in pages:
             t = "/Page"
@@ -1358,7 +1358,7 @@ def get_object(
                 idnum, generation = self.read_object_header(self.stream)
             except Exception:
                 if hasattr(self.stream, "getbuffer"):
-                    buf = bytes(self.stream.getbuffer())  # type: ignore
+                    buf = bytes(self.stream.getbuffer())
                 else:
                     p = self.stream.tell()
                     self.stream.seek(0, 0)
@@ -1412,7 +1412,7 @@ def get_object(
                 )
         else:
             if hasattr(self.stream, "getbuffer"):
-                buf = bytes(self.stream.getbuffer())  # type: ignore
+                buf = bytes(self.stream.getbuffer())
             else:
                 p = self.stream.tell()
                 self.stream.seek(0, 0)
@@ -1705,7 +1705,7 @@ def _read_standard_xref_table(self, stream: StreamType) -> None:
                 except Exception:
                     # if something wrong occurred
                     if hasattr(stream, "getbuffer"):
-                        buf = bytes(stream.getbuffer())  # type: ignore
+                        buf = bytes(stream.getbuffer())
                     else:
                         p = stream.tell()
                         stream.seek(0, 0)

pypdf/_utils.py

@@ -53,7 +53,7 @@
 
 try:
     # Python 3.10+: https://www.python.org/dev/peps/pep-0484/
-    from typing import TypeAlias  # type: ignore[attr-defined]
+    from typing import TypeAlias
 except ImportError:
     from typing_extensions import TypeAlias
 
@@ -461,9 +461,9 @@ def deprecation_bookmark(**aliases: str) -> Callable:
         outline = a collection of outline items.
     """
 
-    def decoration(func: Callable) -> Any:  # type: ignore
+    def decoration(func: Callable) -> Any:
         @functools.wraps(func)
-        def wrapper(*args: Any, **kwargs: Any) -> Any:  # type: ignore
+        def wrapper(*args: Any, **kwargs: Any) -> Any:
             rename_kwargs(func.__name__, kwargs, aliases, fail=True)
             return func(*args, **kwargs)
 
@@ -472,7 +472,7 @@ def wrapper(*args: Any, **kwargs: Any) -> Any:  # type: ignore
     return decoration
 
 
-def rename_kwargs(  # type: ignore
+def rename_kwargs(
     func_name: str, kwargs: Dict[str, Any], aliases: Dict[str, str], fail: bool = False
 ) -> None:
     """

pypdf/_writer.py

@@ -278,7 +278,7 @@ def get_object(
             return self._objects[indirect_reference - 1]
         if indirect_reference.pdf != self:
             raise ValueError("pdf must be self")
-        return self._objects[indirect_reference.idnum - 1]  # type: ignore
+        return self._objects[indirect_reference.idnum - 1]
 
     def getObject(self, ido: Union[int, IndirectObject]) -> PdfObject:  # deprecated
         """
@@ -301,7 +301,7 @@ def _replace_object(
         return self._objects[indirect_reference - 1]
         if indirect_reference.pdf != self:
             raise ValueError("pdf must be self")
-        return self._objects[indirect_reference.idnum - 1]  # type: ignore
+        return self._objects[indirect_reference.idnum - 1]
 
     def _add_page(
         self,
@@ -327,8 +327,8 @@ def _add_page(
         if page_org.pdf is not None:
             other = page_org.pdf.pdf_header
             if isinstance(other, str):
-                other = other.encode()  # type: ignore
-            self.pdf_header = _get_max_pdf_version_header(self.pdf_header, other)  # type: ignore
+                other = other.encode()
+            self.pdf_header = _get_max_pdf_version_header(self.pdf_header, other)
         page[NameObject(PA.PARENT)] = self._pages
         pages = cast(DictionaryObject, self.get_object(self._pages))
         assert page.indirect_reference is not None
@@ -372,7 +372,7 @@ def set_need_appearances_writer(self, state: bool = True) -> None:
     @property
     def viewer_preferences(self) -> Optional[ViewerPreferences]:
         """Returns the existing ViewerPreferences as an overloaded dictionary."""
-        o = cast(DictionaryObject, self._root_object).get(CD.VIEWER_PREFERENCES, None)
+        o = self._root_object.get(CD.VIEWER_PREFERENCES, None)
         if o is None:
             return None
         o = o.get_object()
@@ -624,7 +624,7 @@ def open_destination(
             return create_string_object(str(oa))
         elif isinstance(oa, ArrayObject):
             try:
-                page, typ = oa[0:2]  # type: ignore
+                page, typ = oa[0:2]
                 array = oa[2:]
                 fit = Fit(typ, tuple(array))
                 return Destination("OpenAction", page, fit)
@@ -1153,7 +1153,7 @@ def _flatten(
             for attr in inheritable_page_attributes:
                 if attr in pages:
                     inherit[attr] = pages[attr]
-            for page in cast(ArrayObject, cast(DictionaryObject, pages)[PA.KIDS]):
+            for page in cast(ArrayObject, pages[PA.KIDS]):
                 addt = {}
                 if isinstance(page, IndirectObject):
                     addt["indirect_reference"] = page
@@ -1192,7 +1192,7 @@ def clone_document_from_reader(
         if TK.INFO in reader.trailer:
             self._info = reader.trailer[TK.INFO].clone(self).indirect_reference  # type: ignore
         try:
-            self._ID = cast(ArrayObject, reader.trailer[TK.ID].clone(self))  # type: ignore
+            self._ID = cast(ArrayObject, reader.trailer[TK.ID].clone(self))
         except KeyError:
             pass
         if callable(after_page_append):
@@ -1333,7 +1333,7 @@ def encrypt(
     def write_stream(self, stream: StreamType) -> None:
         if hasattr(stream, "mode") and "b" not in stream.mode:
             logger_warning(
-                f"File <{stream.name}> to write to is not in binary mode. "  # type: ignore
+                f"File <{stream.name}> to write to is not in binary mode. "
                 "It may not be written to correctly.",
                 __name__,
             )
@@ -2205,9 +2205,9 @@ def clean(content: ContentStream, images: List[str], forms: List[str]) -> None:
                     del content.operations[i]
                 elif operator == b"Do":
                     if (
-                        cast(ObjectDeletionFlag, to_delete) & ObjectDeletionFlag.IMAGES
+                        to_delete & ObjectDeletionFlag.IMAGES
                         and operands[0] in images
-                        or cast(ObjectDeletionFlag, to_delete) & ObjectDeletionFlag.TEXT
+                        or to_delete & ObjectDeletionFlag.TEXT
                         and operands[0] in forms
                     ):
                         del content.operations[i]
@@ -2234,7 +2234,7 @@ def clean_forms(
                 try:
                     content: Any = None
                     if (
-                        cast(ObjectDeletionFlag, to_delete) & ObjectDeletionFlag.IMAGES
+                        to_delete & ObjectDeletionFlag.IMAGES
                         and o["/Subtype"] == "/Image"
                     ):
                         content = NullObject()
@@ -2276,9 +2276,9 @@ def clean_forms(
                 content = ContentStream(content, page)
             images, forms = clean_forms(page, [])
 
-            clean(cast(ContentStream, content), images, forms)
+            clean(content, images, forms)
             if isinstance(page["/Contents"], ArrayObject):
-                for o in cast(ArrayObject, page["/Contents"]):
+                for o in page["/Contents"]:
                     self._objects[o.idnum - 1] = NullObject()
             try:
                 self._objects[
@@ -3244,7 +3244,6 @@ def _insert_filtered_annotations(
                     p = self._get_cloned_page(d[0], pages, reader)
                     if p is not None:
                         anc = ano.clone(self, ignore_fields=("/D",))
-                        anc = cast("DictionaryObject", anc)
                         cast("DictionaryObject", anc["/A"])[
                             NameObject("/D")
                         ] = ArrayObject([p] + d[1:])

pypdf/_xobj_image_helpers.py

@@ -15,11 +15,11 @@
 )
 
 try:
-    from typing import Literal, TypeAlias  # type: ignore[attr-defined]
+    from typing import Literal, TypeAlias
 except ImportError:
     # PEP 586 introduced typing.Literal with Python 3.8
     # For older Python versions, the backport typing_extensions is necessary:
-    from typing_extensions import Literal, TypeAlias  # type: ignore[misc, assignment]
+    from typing_extensions import Literal, TypeAlias  # type: ignore[assignment]
 
 
 try:
@@ -94,7 +94,7 @@ def _get_imagemode(
         mode_map.get(color_space)  # type: ignore
         or list(mode_map.values())[color_components]
         or prev_mode
-    )  # type: ignore
+    )
     return mode, mode == "CMYK"
 
 

pypdf/annotations/_markup_annotations.py

@@ -15,11 +15,11 @@
 from ._base import NO_FLAGS, AnnotationDictionary
 
 try:
-    from typing import TypeAlias  # type: ignore[attr-defined]
+    from typing import TypeAlias
 except ImportError:
     # PEP 613 introduced typing.TypeAlias with Python 3.10
     # For older Python versions, the backport typing_extensions is necessary:
-    from typing_extensions import TypeAlias  # type: ignore[misc]
+    from typing_extensions import TypeAlias
 
 
 Vertex: TypeAlias = Tuple[float, float]