Merge branch 'main' into iss2138

CHANGELOG.md

@@ -1,5 +1,26 @@
 # CHANGELOG
 
+## Version 3.16.0, 2023-09-10
+
+### Security (SEC)
+-  Infinite recursion caused by IndirectObject clone (#2156)
+
+### New Features (ENH)
+-  Ease access to ViewerPreferences (#2144)
+
+### Bug Fixes (BUG)
+-  Catch the case where w[0] is an IndirectObject instead of an int (#2154)
+-  Cope with indirect objects in filters and remove deprecated code (#2177)
+-  Accept tabs in cmaps (#2174) / cope with extra space (#2151)
+-  Merge pages without resources (#2150)
+-  getcontents() shall return None if contents is NullObject (#2161)
+-  Fix conversion from 1 to LA (#2175)
+
+### Robustness (ROB)
+-  Accept XYZ with no arguments (#2178)
+
+[Full Changelog](https://github.com/py-pdf/pypdf/compare/3.15.5...3.16.0)
+
 ## Version 3.15.5, 2023-09-03
 
 ### Bug Fixes (BUG)

pypdf/_cmap.py

@@ -410,7 +410,7 @@ def compute_space_width(
         else:
             w = []
         while len(w) > 0:
-            st = w[0]
+            st = w[0] if isinstance(w[0], int) else w[0].get_object()
             second = w[1].get_object()
             if isinstance(second, int):
                 for x in range(st, second):

pypdf/_page.py

@@ -1183,8 +1183,13 @@ def _merge_page_writer(
         pdf = self.indirect_reference.pdf
 
         rename = {}
+        if PG.RESOURCES not in self:
+            self[NameObject(PG.RESOURCES)] = DictionaryObject()
         original_resources = cast(DictionaryObject, self[PG.RESOURCES].get_object())
-        page2resources = cast(DictionaryObject, page2[PG.RESOURCES].get_object())
+        if PG.RESOURCES not in page2:
+            page2resources = DictionaryObject()
+        else:
+            page2resources = cast(DictionaryObject, page2[PG.RESOURCES].get_object())
 
         for res in (
             RES.EXT_G_STATE,

pypdf/_version.py

@@ -1 +1 @@
-__version__ = "3.15.5"
+__version__ = "3.16.0"

pypdf/_writer.py

@@ -862,6 +862,7 @@ def _update_text_field(self, field: DictionaryObject) -> None:
         # Extract font information
         da = cast(str, field[AA.DA])
         font_properties = da.replace("\n", " ").replace("\r", " ").split(" ")
+        font_properties = [x for x in font_properties if x != ""]
         font_name = font_properties[font_properties.index("Tf") - 2]
         font_height = float(font_properties[font_properties.index("Tf") - 1])
         if font_height == 0:

pypdf/filters.py

@@ -43,6 +43,7 @@
 from ._utils import (
     b_,
     deprecate_with_replacement,
+    deprecation_no_replacement,
     logger_warning,
     ord_,
 )
@@ -53,7 +54,7 @@
 from .constants import ImageAttributes as IA
 from .constants import LzwFilterParameters as LZW
 from .constants import StreamAttributes as SA
-from .errors import PdfReadError, PdfStreamError
+from .errors import DeprecationError, PdfReadError, PdfStreamError
 from .generic import (
     ArrayObject,
     DictionaryObject,
@@ -93,7 +94,7 @@ class FlateDecode:
     @staticmethod
     def decode(
         data: bytes,
-        decode_parms: Union[None, ArrayObject, DictionaryObject] = None,
+        decode_parms: Optional[DictionaryObject] = None,
         **kwargs: Any,
     ) -> bytes:
         """
@@ -113,42 +114,37 @@ def decode(
         if "decodeParms" in kwargs:  # deprecated
             deprecate_with_replacement("decodeParms", "parameters", "4.0.0")
             decode_parms = kwargs["decodeParms"]
+        if isinstance(decode_parms, ArrayObject):  # type: ignore
+            raise DeprecationError("decode_parms as ArrayObject is depreciated")
+
         str_data = decompress(data)
         predictor = 1
 
         if decode_parms:
             try:
-                if isinstance(decode_parms, ArrayObject):
-                    for decode_parm in decode_parms:
-                        if "/Predictor" in decode_parm:
-                            predictor = decode_parm["/Predictor"]
-                else:
-                    predictor = decode_parms.get("/Predictor", 1)
+                predictor = decode_parms.get("/Predictor", 1)
             except (AttributeError, TypeError):  # Type Error is NullObject
                 pass  # Usually an array with a null object was read
         # predictor 1 == no predictor
         if predictor != 1:
             # The /Columns param. has 1 as the default value; see ISO 32000,
             # §7.4.4.3 LZWDecode and FlateDecode Parameters, Table 8
             DEFAULT_BITS_PER_COMPONENT = 8
-            if isinstance(decode_parms, ArrayObject):
+            try:
+                columns = cast(int, decode_parms[LZW.COLUMNS].get_object())  # type: ignore
+            except (TypeError, KeyError):
                 columns = 1
-                bits_per_component = DEFAULT_BITS_PER_COMPONENT
-                for decode_parm in decode_parms:
-                    if "/Columns" in decode_parm:
-                        columns = decode_parm["/Columns"]
-                    if LZW.BITS_PER_COMPONENT in decode_parm:
-                        bits_per_component = decode_parm[LZW.BITS_PER_COMPONENT]
-            else:
-                columns = (
-                    1 if decode_parms is None else decode_parms.get(LZW.COLUMNS, 1)
-                )
-                colors = 1 if decode_parms is None else decode_parms.get(LZW.COLORS, 1)
-                bits_per_component = (
-                    decode_parms.get(LZW.BITS_PER_COMPONENT, DEFAULT_BITS_PER_COMPONENT)
-                    if decode_parms
-                    else DEFAULT_BITS_PER_COMPONENT
+            try:
+                colors = cast(int, decode_parms[LZW.COLORS].get_object())  # type: ignore
+            except (TypeError, KeyError):
+                colors = 1
+            try:
+                bits_per_component = cast(
+                    int,
+                    decode_parms[LZW.BITS_PER_COMPONENT].get_object(),  # type: ignore
                 )
+            except (TypeError, KeyError):
+                bits_per_component = DEFAULT_BITS_PER_COMPONENT
 
             # PNG predictor can vary by row and so is the lead byte on each row
             rowlength = (
@@ -259,7 +255,7 @@ class ASCIIHexDecode:
     @staticmethod
     def decode(
         data: Union[str, bytes],
-        decode_parms: Union[None, ArrayObject, DictionaryObject] = None,
+        decode_parms: Optional[DictionaryObject] = None,
         **kwargs: Any,
     ) -> bytes:
         """
@@ -278,9 +274,8 @@ def decode(
         Raises:
           PdfStreamError:
         """
-        if "decodeParms" in kwargs:  # deprecated
-            deprecate_with_replacement("decodeParms", "parameters", "4.0.0")
-            decode_parms = kwargs["decodeParms"]  # noqa: F841
+        # decode_parms is unused here
+
         if isinstance(data, str):
             data = data.encode()
         retval = b""
@@ -321,7 +316,7 @@ class RunLengthDecode:
     @staticmethod
     def decode(
         data: bytes,
-        decode_parms: Union[None, ArrayObject, DictionaryObject] = None,
+        decode_parms: Optional[DictionaryObject] = None,
         **kwargs: Any,
     ) -> bytes:
         """
@@ -337,9 +332,8 @@ def decode(
         Raises:
           PdfStreamError:
         """
-        if "decodeParms" in kwargs:  # deprecated
-            deprecate_with_replacement("decodeParms", "parameters", "4.0.0")
-            decode_parms = kwargs["decodeParms"]  # noqa: F841
+        # decode_parms is unused here
+
         lst = []
         index = 0
         while True:
@@ -453,7 +447,7 @@ def decode(self) -> str:
     @staticmethod
     def decode(
         data: bytes,
-        decode_parms: Union[None, ArrayObject, DictionaryObject] = None,
+        decode_parms: Optional[DictionaryObject] = None,
         **kwargs: Any,
     ) -> str:
         """
@@ -466,9 +460,8 @@ def decode(
         Returns:
           decoded data.
         """
-        if "decodeParms" in kwargs:  # deprecated
-            deprecate_with_replacement("decodeParms", "parameters", "4.0.0")
-            decode_parms = kwargs["decodeParms"]  # noqa: F841
+        # decode_parms is unused here
+
         return LZWDecode.Decoder(data).decode()
 
 
@@ -478,12 +471,11 @@ class ASCII85Decode:
     @staticmethod
     def decode(
         data: Union[str, bytes],
-        decode_parms: Union[None, ArrayObject, DictionaryObject] = None,
+        decode_parms: Optional[DictionaryObject] = None,
         **kwargs: Any,
     ) -> bytes:
-        if "decodeParms" in kwargs:  # deprecated
-            deprecate_with_replacement("decodeParms", "parameters", "4.0.0")
-            decode_parms = kwargs["decodeParms"]  # noqa: F841
+        # decode_parms is unused here
+
         if isinstance(data, str):
             data = data.encode("ascii")
         group_index = b = 0
@@ -511,25 +503,21 @@ class DCTDecode:
     @staticmethod
     def decode(
         data: bytes,
-        decode_parms: Union[None, ArrayObject, DictionaryObject] = None,
+        decode_parms: Optional[DictionaryObject] = None,
         **kwargs: Any,
     ) -> bytes:
-        if "decodeParms" in kwargs:  # deprecated
-            deprecate_with_replacement("decodeParms", "parameters", "4.0.0")
-            decode_parms = kwargs["decodeParms"]  # noqa: F841
+        # decode_parms is unused here
         return data
 
 
 class JPXDecode:
     @staticmethod
     def decode(
         data: bytes,
-        decode_parms: Union[None, ArrayObject, DictionaryObject] = None,
+        decode_parms: Optional[DictionaryObject] = None,
         **kwargs: Any,
     ) -> bytes:
-        if "decodeParms" in kwargs:  # deprecated
-            deprecate_with_replacement("decodeParms", "parameters", "4.0.0")
-            decode_parms = kwargs["decodeParms"]  # noqa: F841
+        # decode_parms is unused here
         return data
 
 
@@ -591,13 +579,18 @@ def _get_parameters(
     @staticmethod
     def decode(
         data: bytes,
-        decode_parms: Union[None, ArrayObject, DictionaryObject] = None,
+        decode_parms: Optional[DictionaryObject] = None,
         height: int = 0,
         **kwargs: Any,
     ) -> bytes:
+        # decode_parms is unused here
         if "decodeParms" in kwargs:  # deprecated
             deprecate_with_replacement("decodeParms", "parameters", "4.0.0")
             decode_parms = kwargs["decodeParms"]
+        if isinstance(decode_parms, ArrayObject):  # deprecated
+            deprecation_no_replacement(
+                "decode_parms being an ArrayObject", removed_in="3.15.5"
+            )
         parms = CCITTFaxDecode._get_parameters(decode_parms, height)
 
         img_size = len(data)

tests/test_cmap.py

@@ -191,6 +191,18 @@ def test_unixxx_glyphs():
         assert pat in txt
 
 
+@pytest.mark.enable_socket()
+def test_cmap_compute_space_width():
+    # issue 2137
+    # original file URL:
+    url = "https://arxiv.org/pdf/2005.05909.pdf"
+    # URL from github issue is too long to pass code stype check, use original arxiv URL instead
+    # url = "https://github.com/py-pdf/pypdf/files/12489914/Morris.et.al.-.2020.-.TextAttack.A.Framework.for.Adversarial.Attacks.Data.Augmentation.and.Adversarial.Training.in.NLP.pdf"
+    name = "TextAttack_paper.pdf"
+    reader = PdfReader(BytesIO(get_data_from_url(url, name=name)))
+    reader.pages[0].extract_text()  # no error
+
+
 @pytest.mark.enable_socket()
 def test_tabs_in_cmap():
     """Issue #2173"""

tests/test_filters.py

@@ -10,7 +10,7 @@
 from PIL import Image
 
 from pypdf import PdfReader
-from pypdf.errors import PdfReadError, PdfStreamError
+from pypdf.errors import DeprecationError, PdfReadError, PdfStreamError
 from pypdf.filters import (
     ASCII85Decode,
     ASCIIHexDecode,
@@ -69,16 +69,15 @@ def test_flatedecode_unsupported_predictor():
             codec.decode(codec.encode(s), DictionaryObject({"/Predictor": predictor}))
 
 
-@pytest.mark.parametrize(
-    "params", [ArrayObject([]), ArrayObject([{"/Predictor": 1}]), "a"]
-)
+@pytest.mark.parametrize("params", [ArrayObject([]), ArrayObject([{"/Predictor": 1}])])
 def test_flate_decode_decompress_with_array_params(params):
     """FlateDecode decode() method works correctly with array parameters."""
     codec = FlateDecode()
     s = ""
     s = s.encode()
     encoded = codec.encode(s)
-    assert codec.decode(encoded, params) == s
+    with pytest.raises(DeprecationError):
+        assert codec.decode(encoded, params) == s
 
 
 @pytest.mark.parametrize(

tests/test_page.py

@@ -1227,6 +1227,16 @@ def create_stamp_pdf() -> BytesIO:
     )
 
 
+def test_merge_with_no_resources():
+    """Test for issue #2147"""
+    writer = PdfWriter()
+    p0 = writer.add_blank_page(900, 1200)
+    del p0["/Resources"]
+    p1 = writer.add_blank_page(900, 1200)
+    del p1["/Resources"]
+    writer.pages[0].merge_page(p1)
+
+
 def test_get_contents_from_nullobject():
     """Issue #2157"""
     writer = PdfWriter()

tests/test_writer.py

@@ -1797,6 +1797,20 @@ def test_viewerpreferences():
     assert writer.viewer_preferences is None
 
 
+def test_extra_spaces_in_da_text(caplog):
+    writer = PdfWriter(clone_from=RESOURCE_ROOT / "form.pdf")
+    t = writer.pages[0]["/Annots"][0].get_object()["/DA"]
+    t = t.replace("/Helv", "/Helv   ")
+    writer.pages[0]["/Annots"][0].get_object()[NameObject("/DA")] = TextStringObject(t)
+    writer.update_page_form_field_values(
+        writer.pages[0], {"foo": "abcd"}, auto_regenerate=False
+    )
+    t = writer.pages[0]["/Annots"][0].get_object()["/AP"]["/N"].get_data()
+    assert "Font dictionary for  not found." not in caplog.text
+    assert b"/Helv" in t
+    assert b"(abcd)" in t
+
+
 @pytest.mark.enable_socket()
 def test_object_contains_indirect_reference_to_self():
     url = "https://github.com/py-pdf/pypdf/files/12389243/testbook.pdf"