DEV: Align mypy Makefile target with strict mode

.pre-commit-config.yaml

@@ -35,3 +35,6 @@ repos:
     hooks:
       - id: mypy
         files: ^pypdf/.*
+        additional_dependencies:
+          - cryptography
+          - pycryptodome

Makefile

@@ -23,7 +23,7 @@ benchmark:
 	pytest tests/bench.py
 
 mypy:
-	mypy pypdf --ignore-missing-imports --check-untyped --strict
+	mypy pypdf
 
 ruff:
 	ruff check pypdf tests make_release.py

pypdf/_doc_common.py

@@ -92,7 +92,7 @@ def convert_to_int(d: bytes, size: int) -> Union[int, tuple[Any, ...]]:
         raise PdfReadError("Invalid size in convert_to_int")
     d = b"\x00\x00\x00\x00\x00\x00\x00\x00" + d
     d = d[-8:]
-    return struct.unpack(">q", d)[0]
+    return cast(int, struct.unpack(">q", d)[0])
 
 
 class DocumentInformation(DictionaryObject):
@@ -382,7 +382,7 @@ def recursive_call(
             node: DictionaryObject, mi: int
         ) -> tuple[Optional[PdfObject], int]:
             ma = cast(int, node.get("/Count", 1))  # default 1 for /Page types
-            if node["/Type"] == "/Page":
+            if node["/Type"] == "/Page":  # type: ignore[comparison-overlap]
                 if page_number == mi:
                     return node, -1
                 return None, mi + 1

pypdf/_page.py

@@ -540,7 +540,7 @@ def user_unit(self) -> float:
         space unit is 1/72 inch, and a value of 3 means that a user
         space unit is 3/72 inch.
         """
-        return self.get(PG.USER_UNIT, 1)
+        return cast(float, self.get(PG.USER_UNIT, 1))
 
     @staticmethod
     def create_blank_page(
@@ -947,15 +947,15 @@ def _add_transformation_matrix(
         ctm: CompressedTransformationMatrix,
     ) -> ContentStream:
         """Add transformation matrix at the beginning of the given contents stream."""
-        contents = ContentStream(contents, pdf)
-        contents.operations.insert(
+        content_stream = ContentStream(contents, pdf)
+        content_stream.operations.insert(
             0,
-            [
+            (
                 [FloatObject(x) for x in ctm],
                 b"cm",
-            ],
+            ),
         )
-        return contents
+        return content_stream
 
     def _get_contents_as_bytes(self) -> Optional[bytes]:
         """
@@ -1627,7 +1627,7 @@ def page_number(self) -> Optional[int]:
             return None
         try:
             lst = self.indirect_reference.pdf.pages
-            return lst.index(self)
+            return int(lst.index(self))
         except ValueError:
             return None
 

pypdf/_page_labels.py

@@ -156,7 +156,7 @@ def get_label_from_nums(dictionary_object: DictionaryObject, index: int) -> str:
     if not isinstance(value, dict):
         return str(index + 1)  # Fallback
     start = value.get("/St", 1)
-    prefix = value.get("/P", "")
+    prefix = cast(str, value.get("/P", ""))
     mapping_function = m[value.get("/S")]
     return prefix + mapping_function(index - start_index + start)
 

pypdf/_reader.py

@@ -730,7 +730,7 @@ def _basic_validation(self, stream: StreamType) -> None:
                     f"PDF starts with '{header_byte.decode('utf8')}', "
                     "but '%PDF-' expected"
                 )
-            logger_warning(f"invalid pdf header: {header_byte}", __name__)
+            logger_warning(f"invalid pdf header: {header_byte!r}", __name__)
         stream.seek(0, os.SEEK_END)
 
     def _find_eof_marker(self, stream: StreamType) -> None:
@@ -1005,7 +1005,7 @@ def _read_xref(self, stream: StreamType) -> Optional[int]:
                 )
             stream.seek(p, 0)
         if "/Prev" in new_trailer:
-            return new_trailer["/Prev"]
+            return cast(int, new_trailer["/Prev"])
         return None
 
     def _read_xref_other_error(

pypdf/_text_extraction/_layout_mode/_text_state_manager.py

@@ -132,7 +132,7 @@ def raw_transform(
         _d: float = 1.0,
         _e: float = 0.0,
         _f: float = 0.0,
-    ) -> dict[int, float]:
+    ) -> TextStateManagerDictType:
         """Only a/b/c/d/e/f matrix params"""
         return dict(zip(range(6), map(float, (_a, _b, _c, _d, _e, _f))))
 
@@ -148,7 +148,7 @@ def new_transform(
         is_render: bool = False,
     ) -> TextStateManagerDictType:
         """Standard a/b/c/d/e/f matrix params + 'is_text' and 'is_render' keys"""
-        result: Any = TextStateManager.raw_transform(_a, _b, _c, _d, _e, _f)
+        result = TextStateManager.raw_transform(_a, _b, _c, _d, _e, _f)
         result.update({"is_text": is_text, "is_render": is_render})
         return result
 

pypdf/_utils.py

@@ -72,6 +72,7 @@
 ]
 
 StreamType = IO[Any]
+BinaryStreamType = IO[bytes]
 StrByteType = Union[str, StreamType]
 
 
@@ -181,7 +182,7 @@ def read_until_whitespace(stream: StreamType, maxchars: Optional[int] = None) ->
     return txt
 
 
-def read_non_whitespace(stream: StreamType) -> bytes:
+def read_non_whitespace(stream: BinaryStreamType) -> bytes:
     """
     Find and read the next non-whitespace character (ignores whitespace).
 
@@ -282,7 +283,7 @@ def read_until_regex(stream: StreamType, regex: Pattern[bytes]) -> bytes:
     return b"".join(parts)
 
 
-def read_block_backwards(stream: StreamType, to_read: int) -> bytes:
+def read_block_backwards(stream: BinaryStreamType, to_read: int) -> bytes:
     """
     Given a stream at position X, read a block of size to_read ending at position X.
 

pypdf/_writer.py

@@ -3007,9 +3007,9 @@ def _insert_filtered_annotations(
         for an in annots:
             ano = cast("DictionaryObject", an.get_object())
             if (
-                ano["/Subtype"] != "/Link"
+                ano["/Subtype"] != "/Link"  # type: ignore[comparison-overlap]
                 or "/A" not in ano
-                or cast("DictionaryObject", ano["/A"])["/S"] != "/GoTo"
+                or cast("DictionaryObject", ano["/A"])["/S"] != "/GoTo"  # type: ignore[comparison-overlap]
                 or "/Dest" in ano
             ):
                 if "/Dest" not in ano:

pypdf/annotations/_base.py

@@ -19,7 +19,7 @@ def __init__(self) -> None:
 
     @property
     def flags(self) -> AnnotationFlag:
-        return self.get(NameObject("/F"), AnnotationFlag(0))
+        return AnnotationFlag(self.get(NameObject("/F"), 0))
 
     @flags.setter
     def flags(self, value: AnnotationFlag) -> None:

pypdf/filters.py

@@ -216,11 +216,11 @@ def decode(
             if predictor == 2:
                 row_length -= 1  # remove the predictor byte
                 bpp = row_length // columns
-                str_data = bytearray(str_data)
-                for i in range(len(str_data)):
+                str_data_mut = bytearray(str_data)
+                for i in range(len(str_data_mut)):
                     if i % row_length >= bpp:
-                        str_data[i] = (str_data[i] + str_data[i - bpp]) % 256
-                str_data = bytes(str_data)
+                        str_data_mut[i] = (str_data_mut[i] + str_data_mut[i - bpp]) % 256
+                str_data = bytes(str_data_mut)
             # PNG prediction:
             elif 10 <= predictor <= 15:
                 str_data = FlateDecode._decode_png_prediction(

pypdf/generic/_base.py

@@ -136,8 +136,8 @@ def clone(
         )
 
     def _reference_clone(
-        self, clone: Any, pdf_dest: PdfWriterProtocol, force_duplicate: bool = False
-    ) -> PdfObjectProtocol:
+        self, clone: "PdfObject", pdf_dest: PdfWriterProtocol, force_duplicate: bool = False
+    ) -> "PdfObject":
         """
         Reference the object within the _objects of pdf_dest only if
         indirect_reference attribute exists (which means the objects was
@@ -153,7 +153,11 @@ def _reference_clone(
 
         """
         try:
-            if not force_duplicate and clone.indirect_reference.pdf == pdf_dest:
+            if (
+                not force_duplicate
+                and clone.indirect_reference is not None
+                and clone.indirect_reference.pdf == pdf_dest
+            ):
                 return clone
         except Exception:
             pass
@@ -182,7 +186,7 @@ def _reference_clone(
                 obj = pdf_dest.get_object(
                     pdf_dest._id_translated[id(ind.pdf)][ind.idnum]
                 )
-                assert obj is not None
+                assert isinstance(obj, PdfObject), "mypy"
                 return obj
             pdf_dest._id_translated[id(ind.pdf)][ind.idnum] = i
         try:
@@ -252,6 +256,8 @@ def __hash__(self) -> int:
 
 
 class BooleanObject(PdfObject):
+    value: bool
+
     def __init__(self, value: Any) -> None:
         self.value = value
 
@@ -370,7 +376,7 @@ def clone(
             dup = pdf_dest._add_object(
                 obj.clone(pdf_dest, force_duplicate, ignore_fields)
             )
-        assert dup is not None, "mypy"
+        assert isinstance(dup, PdfObject), "mypy"
         assert dup.indirect_reference is not None, "mypy"
         return dup.indirect_reference
 
@@ -379,7 +385,8 @@ def indirect_reference(self) -> "IndirectObject":  # type: ignore[override]
         return self
 
     def get_object(self) -> Optional["PdfObject"]:
-        return self.pdf.get_object(self)
+        obj: Optional[PdfObject] = self.pdf.get_object(self)
+        return obj
 
     def __deepcopy__(self, memo: Any) -> "IndirectObject":
         return IndirectObject(self.idnum, self.generation, self.pdf)
@@ -517,7 +524,7 @@ def hash_bin(self) -> int:
         return hash((self.__class__, self.as_numeric))
 
     def myrepr(self) -> str:
-        if self == 0:
+        if self == 0:  # type: ignore[comparison-overlap]
             return "0.0"
         nb = FLOAT_WRITE_PRECISION - int(log10(abs(self)))
         return f"{self:.{max(1, nb)}f}".rstrip("0").rstrip(".")

pypdf/generic/_data_structures.py

@@ -46,6 +46,7 @@
 from .._protocols import PdfReaderProtocol, PdfWriterProtocol, XmpInformationProtocol
 from .._utils import (
     WHITESPACES,
+    BinaryStreamType,
     StreamType,
     deprecation_no_replacement,
     logger_warning,
@@ -165,20 +166,21 @@ def items(self) -> Iterable[Any]:
 
     def _to_lst(self, lst: Any) -> list[Any]:
         # Convert to list, internal
+        result: list[Any]
         if isinstance(lst, (list, tuple, set)):
-            pass
+            result = list(lst)
         elif isinstance(lst, PdfObject):
-            lst = [lst]
+            result = [lst]
         elif isinstance(lst, str):
             if lst[0] == "/":
-                lst = [NameObject(lst)]
+                result = [NameObject(lst)]
             else:
-                lst = [TextStringObject(lst)]
+                result = [TextStringObject(lst)]
         elif isinstance(lst, bytes):
-            lst = [ByteStringObject(lst)]
+            result = [ByteStringObject(lst)]
         else:  # for numbers,...
-            lst = [lst]
-        return lst
+            result = [lst]
+        return result
 
     def __add__(self, lst: Any) -> "ArrayObject":
         """
@@ -476,7 +478,7 @@ def setdefault(self, key: Any, value: Optional[Any] = None) -> Any:
         return dict.setdefault(self, key, value)
 
     def __getitem__(self, key: Any) -> PdfObject:
-        return dict.__getitem__(self, key).get_object()
+        return cast(PdfObject, dict.__getitem__(self, key).get_object())
 
     @property
     def xmp_metadata(self) -> Optional[XmpInformationProtocol]:
@@ -532,7 +534,7 @@ def _get_next_object_position(
 
     @classmethod
     def _read_unsized_from_stream(
-            cls, stream: StreamType, pdf: PdfReaderProtocol
+            cls, stream: BinaryStreamType, pdf: PdfReaderProtocol
     ) -> bytes:
         object_position = cls._get_next_object_position(
             position_before=stream.tell(), position_end=2 ** 32, generations=list(pdf.xref), pdf=pdf
@@ -768,45 +770,46 @@ def insert_child(
         if inc_parent_counter is None:
             inc_parent_counter = self.inc_parent_counter_default
         child_obj = child.get_object()
-        child = child.indirect_reference  # get_reference(child_obj)
+        assert child.indirect_reference is not None, "mypy"
+        child_reference: IndirectObject = child.indirect_reference
 
         prev: Optional[DictionaryObject]
         if "/First" not in self:  # no child yet
-            self[NameObject("/First")] = child
+            self[NameObject("/First")] = child_reference
             self[NameObject("/Count")] = NumberObject(0)
-            self[NameObject("/Last")] = child
+            self[NameObject("/Last")] = child_reference
             child_obj[NameObject("/Parent")] = self.indirect_reference
             inc_parent_counter(self, child_obj.get("/Count", 1))
             if "/Next" in child_obj:
                 del child_obj["/Next"]
             if "/Prev" in child_obj:
                 del child_obj["/Prev"]
-            return child
+            return child_reference
         prev = cast("DictionaryObject", self["/Last"])
 
         while prev.indirect_reference != before:
             if "/Next" in prev:
                 prev = cast("TreeObject", prev["/Next"])
             else:  # append at the end
-                prev[NameObject("/Next")] = cast("TreeObject", child)
+                prev[NameObject("/Next")] = cast("TreeObject", child_reference)
                 child_obj[NameObject("/Prev")] = prev.indirect_reference
                 child_obj[NameObject("/Parent")] = self.indirect_reference
                 if "/Next" in child_obj:
                     del child_obj["/Next"]
-                self[NameObject("/Last")] = child
+                self[NameObject("/Last")] = child_reference
                 inc_parent_counter(self, child_obj.get("/Count", 1))
-                return child
+                return child_reference
         try:  # insert as first or in the middle
             assert isinstance(prev["/Prev"], DictionaryObject)
-            prev["/Prev"][NameObject("/Next")] = child
+            prev["/Prev"][NameObject("/Next")] = child_reference
             child_obj[NameObject("/Prev")] = prev["/Prev"]
         except Exception:  # it means we are inserting in first position
             del child_obj["/Next"]
         child_obj[NameObject("/Next")] = prev
-        prev[NameObject("/Prev")] = child
+        prev[NameObject("/Prev")] = child_reference
         child_obj[NameObject("/Parent")] = self.indirect_reference
         inc_parent_counter(self, child_obj.get("/Count", 1))
-        return child
+        return child_reference
 
     def _remove_node_from_tree(
         self, prev: Any, prev_ref: Any, cur: Any, last: Any
@@ -1211,7 +1214,7 @@ def __init__(
                                 f"{MAX_ARRAY_BASED_STREAM_OUTPUT_LENGTH} output bytes."
                             )
                         data += new_data
-                    if len(data) == 0 or data[-1] != b"\n":
+                    if len(data) == 0 or data[-1:] != b"\n":
                         # There should be no direct need to check for a change of one byte.
                         length += 1
                         data += b"\n"
@@ -1775,8 +1778,9 @@ def bottom(self) -> Optional[FloatObject]:
     @property
     def color(self) -> Optional["ArrayObject"]:
         """Read-only property accessing the color in (R, G, B) with values 0.0-1.0."""
-        return self.get(
-            "/C", ArrayObject([FloatObject(0), FloatObject(0), FloatObject(0)])
+        return cast(
+            "ArrayObject",
+            self.get("/C", ArrayObject([FloatObject(0), FloatObject(0), FloatObject(0)])),
         )
 
     @property
@@ -1786,7 +1790,7 @@ def font_format(self) -> Optional[OutlineFontFlag]:
 
         1=italic, 2=bold, 3=both
         """
-        return self.get("/F", 0)
+        return OutlineFontFlag(self.get("/F", 0))
 
     @property
     def outline_count(self) -> Optional[int]: