From 4bd54bd32d525ec41e23960146fbb3b0418a4715 Mon Sep 17 00:00:00 2001 From: Stefan <96178532+stefan6419846@users.noreply.github.com> Date: Sun, 28 Jul 2024 17:16:57 +0200 Subject: [PATCH 01/21] DEV: Test against Python 3.13 (#2776) * DEV: Test against Python 3.13 * fix typo * add missing setup-python * fix another typo * update Pillow version * attempt to update coverage package * update number of expected coverage files --- .github/workflows/github-ci.yaml | 10 +++++----- requirements/ci-3.11.txt | 4 ++-- 2 files changed, 7 insertions(+), 7 deletions(-) diff --git a/.github/workflows/github-ci.yaml b/.github/workflows/github-ci.yaml index 820ccdcaa..1eb3d9bd0 100644 --- a/.github/workflows/github-ci.yaml +++ b/.github/workflows/github-ci.yaml @@ -57,7 +57,7 @@ jobs: runs-on: ubuntu-20.04 strategy: matrix: - python-version: ["3.7", "3.8", "3.9", "3.10", "3.11", "3.12"] + python-version: ["3.7", "3.8", "3.9", "3.10", "3.11", "3.12", "3.13-dev"] use-crypto-lib: ["cryptography"] include: - python-version: "3.7" @@ -90,7 +90,7 @@ jobs: cache-dependency-path: '**/requirements/ci.txt' - name: Setup Python (3.11+) uses: actions/setup-python@v5 - if: matrix.python-version == '3.11' || matrix.python-version == '3.12' + if: matrix.python-version == '3.11' || matrix.python-version == '3.12' || matrix.python-version == '3.13-dev' with: python-version: ${{ matrix.python-version }} allow-prereleases: true @@ -106,7 +106,7 @@ jobs: - name: Install requirements (Python 3.11+) run: | pip install -r requirements/ci-3.11.txt - if: matrix.python-version == '3.11' || matrix.python-version == '3.12' + if: matrix.python-version == '3.11' || matrix.python-version == '3.12' || matrix.python-version == '3.13-dev' - name: Remove pycryptodome and cryptography run: | pip uninstall pycryptodome cryptography -y @@ -215,8 +215,8 @@ jobs: - name: Check Number of Downloaded Files run: | downloaded_files_count=$(find \.coverage* -type f | wc -l) - if [ $downloaded_files_count -eq 8 ]; then - echo "The expected number of files (8) were downloaded." + if [ $downloaded_files_count -eq 9 ]; then + echo "The expected number of files (9) were downloaded." else echo "ERROR: Expected 8 files, but found $downloaded_files_count files." exit 1 diff --git a/requirements/ci-3.11.txt b/requirements/ci-3.11.txt index f382fe2b9..210177118 100644 --- a/requirements/ci-3.11.txt +++ b/requirements/ci-3.11.txt @@ -6,7 +6,7 @@ # attrs==23.1.0 # via flake8-bugbear -coverage[toml]==7.3.0 +coverage[toml]==7.6.0 # via # -r requirements/ci.in # pytest-cov @@ -35,7 +35,7 @@ mypy-extensions==1.0.0 # via mypy packaging==23.1 # via pytest -pillow==10.0.1 +pillow==10.4.0 # via # -r requirements/ci.in # fpdf2 From d4df20d14cb6a2839c1ab141b51e70652fb3d1f1 Mon Sep 17 00:00:00 2001 From: j-t-1 <120829237+j-t-1@users.noreply.github.com> Date: Wed, 31 Jul 2024 10:46:08 +0100 Subject: [PATCH 02/21] STY: Remove boolean value comparison (#2779) PEP 8 recommendation. --- pypdf/annotations/_markup_annotations.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/pypdf/annotations/_markup_annotations.py b/pypdf/annotations/_markup_annotations.py index 4db8dfdbf..98a222483 100644 --- a/pypdf/annotations/_markup_annotations.py +++ b/pypdf/annotations/_markup_annotations.py @@ -104,9 +104,9 @@ def __init__( self[NameObject("/Rect")] = RectangleObject(rect) font_str = "font: " - if bold is True: + if bold: font_str = f"{font_str}bold " - if italic is True: + if italic: font_str = f"{font_str}italic " font_str = f"{font_str}{font} {font_size}" font_str = f"{font_str};text-align:left;color:#{font_color}" From 3ad9234c2ec08e7cd6a8b2ec962386eda394d76d Mon Sep 17 00:00:00 2001 From: "William G. Gagnon" Date: Fri, 2 Aug 2024 11:21:53 -0400 Subject: [PATCH 03/21] ROB: Handle images with empty data when processing an image from bytes (#2786) Closes #2783. --- CONTRIBUTORS.md | 1 + pypdf/_xobj_image_helpers.py | 9 ++++++--- pypdf/errors.py | 4 ++++ tests/test_xobject_image_helpers.py | 13 +++++++++++-- 4 files changed, 22 insertions(+), 5 deletions(-) diff --git a/CONTRIBUTORS.md b/CONTRIBUTORS.md index 84f0b6ee4..89fec3b14 100644 --- a/CONTRIBUTORS.md +++ b/CONTRIBUTORS.md @@ -19,6 +19,7 @@ history and [GitHub's 'Contributors' feature](https://github.com/py-pdf/pypdf/gr * [ediamondscience](https://github.com/ediamondscience) * [Ermeson, Felipe](https://github.com/FelipeErmeson) * [Freitag, François](https://github.com/francoisfreitag) +* [Gagnon, William G.](https://github.com/williamgagnon) * [Górny, Michał](https://github.com/mgorny) * [Grillo, Miguel](https://github.com/Ineffable22) * [Gutteridge, David H.](https://github.com/dhgutteridge) diff --git a/pypdf/_xobj_image_helpers.py b/pypdf/_xobj_image_helpers.py index 45b0c145b..5ae8894fa 100644 --- a/pypdf/_xobj_image_helpers.py +++ b/pypdf/_xobj_image_helpers.py @@ -6,7 +6,7 @@ from ._utils import check_if_whitespace_only, logger_warning from .constants import ColorSpaces -from .errors import PdfReadError +from .errors import EmptyImageDataError, PdfReadError from .generic import ( ArrayObject, DecodedStreamObject, @@ -148,9 +148,12 @@ def _extended_image_frombytes( img = Image.frombytes(mode, size, data) except ValueError as exc: nb_pix = size[0] * size[1] - if len(data) % nb_pix != 0: + data_length = len(data) + if data_length == 0: + raise EmptyImageDataError("Data is 0 bytes, cannot process an image from empty data.") from exc + if data_length % nb_pix != 0: raise exc - k = nb_pix * len(mode) / len(data) + k = nb_pix * len(mode) / data_length data = b"".join([bytes((x,) * int(k)) for x in data]) img = Image.frombytes(mode, size, data) return img diff --git a/pypdf/errors.py b/pypdf/errors.py index c962dec66..ad197ffc1 100644 --- a/pypdf/errors.py +++ b/pypdf/errors.py @@ -59,4 +59,8 @@ class EmptyFileError(PdfReadError): """Raised when a PDF file is empty or has no content.""" +class EmptyImageDataError(PyPdfError): + """Raised when trying to process an image that has no data.""" + + STREAM_TRUNCATED_PREMATURELY = "Stream has ended unexpectedly" diff --git a/tests/test_xobject_image_helpers.py b/tests/test_xobject_image_helpers.py index 63ecebd9b..39b7131fc 100644 --- a/tests/test_xobject_image_helpers.py +++ b/tests/test_xobject_image_helpers.py @@ -4,8 +4,8 @@ import pytest from pypdf import PdfReader -from pypdf._xobj_image_helpers import _handle_flate -from pypdf.errors import PdfReadError +from pypdf._xobj_image_helpers import _extended_image_frombytes, _handle_flate +from pypdf.errors import EmptyImageDataError, PdfReadError from pypdf.generic import ArrayObject, DecodedStreamObject, NameObject, NumberObject from . import get_data_from_url @@ -113,3 +113,12 @@ def test_handle_flate__image_mode_1(): colors=2, obj_as_text="dummy", ) + + +def test_extended_image_frombytes_zero_data(): + mode = "RGB" + size = (1, 1) + data = b"" + + with pytest.raises(EmptyImageDataError, match="Data is 0 bytes, cannot process an image from empty data."): + _extended_image_frombytes(mode, size, data) From 582557e09a7e658fdcb19f26eb069d87875489f0 Mon Sep 17 00:00:00 2001 From: Diogo Teles Sant'Anna Date: Fri, 2 Aug 2024 15:49:29 -0300 Subject: [PATCH 04/21] SEC: Fix GitHub workflow vulnerable to script injection (#2787) Signed-off-by: Diogo Teles Sant'Anna --- .github/workflows/release.yaml | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/.github/workflows/release.yaml b/.github/workflows/release.yaml index 9f782ec08..b1a4fb27f 100644 --- a/.github/workflows/release.yaml +++ b/.github/workflows/release.yaml @@ -12,6 +12,9 @@ on: permissions: contents: write +env: + HEAD_COMMIT_MESSAGE: ${{ github.event.head_commit.message }} + jobs: build_and_publish: name: Publish a new version @@ -24,7 +27,7 @@ jobs: - name: Extract version from commit message id: extract_version run: | - VERSION=$(echo "${{ github.event.head_commit.message }}" | grep -oP '(?<=REL: )\d+\.\d+\.\d+') + VERSION=$(echo "$HEAD_COMMIT_MESSAGE" | grep -oP '(?<=REL: )\d+\.\d+\.\d+') echo "version=$VERSION" >> $GITHUB_OUTPUT - name: Extract tag message from commit message @@ -32,7 +35,7 @@ jobs: run: | VERSION="${{ steps.extract_version.outputs.version }}" delimiter="$(openssl rand -hex 8)" - MESSAGE=$(echo "${{ github.event.head_commit.message }}" | sed "0,/REL: $VERSION/s///" ) + MESSAGE=$(echo "$HEAD_COMMIT_MESSAGE" | sed "0,/REL: $VERSION/s///" ) echo "message<<${delimiter}" >> $GITHUB_OUTPUT echo "$MESSAGE" >> $GITHUB_OUTPUT echo "${delimiter}" >> $GITHUB_OUTPUT From 38f3925502c2971ad587fb616500b6f8b6333d03 Mon Sep 17 00:00:00 2001 From: j-t-1 <120829237+j-t-1@users.noreply.github.com> Date: Mon, 5 Aug 2024 09:10:47 +0100 Subject: [PATCH 05/21] MAINT: Remove unused paeth_predictor (#2773) --- pypdf/_utils.py | 14 -------------- tests/test_utils.py | 18 ------------------ 2 files changed, 32 deletions(-) diff --git a/pypdf/_utils.py b/pypdf/_utils.py index 38c0d67d7..6569707b6 100644 --- a/pypdf/_utils.py +++ b/pypdf/_utils.py @@ -390,20 +390,6 @@ def ord_(b: Union[int, str, bytes]) -> Union[int, bytes]: WHITESPACES_AS_REGEXP = b"[" + WHITESPACES_AS_BYTES + b"]" -def paeth_predictor(left: int, up: int, up_left: int) -> int: - p = left + up - up_left - dist_left = abs(p - left) - dist_up = abs(p - up) - dist_up_left = abs(p - up_left) - - if dist_left <= dist_up and dist_left <= dist_up_left: - return left - elif dist_up <= dist_up_left: - return up - else: - return up_left - - def deprecate(msg: str, stacklevel: int = 3) -> None: warnings.warn(msg, DeprecationWarning, stacklevel=stacklevel) diff --git a/tests/test_utils.py b/tests/test_utils.py index 81fcf9fb4..856bedd86 100644 --- a/tests/test_utils.py +++ b/tests/test_utils.py @@ -132,24 +132,6 @@ def test_deprecate_no_replacement(): assert warn[0].message.args[0] == error_msg -@pytest.mark.parametrize( - ("left", "up", "upleft", "expected"), - [ - (0, 0, 0, 0), - (1, 0, 0, 1), - (0, 1, 0, 1), - (0, 0, 1, 0), - (1, 2, 3, 1), - (2, 1, 3, 1), - (1, 3, 2, 2), - (3, 1, 2, 2), - (3, 2, 1, 3), - ], -) -def test_paeth_predictor(left, up, upleft, expected): - assert pypdf._utils.paeth_predictor(left, up, upleft) == expected - - @pytest.mark.parametrize( ("dat", "pos", "to_read", "expected", "expected_pos"), [ From 09f9b7ed52193bfd9e98bdd018ccaf7cbe821687 Mon Sep 17 00:00:00 2001 From: j-t-1 <120829237+j-t-1@users.noreply.github.com> Date: Mon, 5 Aug 2024 16:49:29 +0100 Subject: [PATCH 06/21] MAINT: Remove unused AnnotationFlag --- pypdf/annotations/_non_markup_annotations.py | 3 --- 1 file changed, 3 deletions(-) diff --git a/pypdf/annotations/_non_markup_annotations.py b/pypdf/annotations/_non_markup_annotations.py index dcdb3b0ff..6272cceee 100644 --- a/pypdf/annotations/_non_markup_annotations.py +++ b/pypdf/annotations/_non_markup_annotations.py @@ -1,6 +1,5 @@ from typing import TYPE_CHECKING, Any, Optional, Tuple, Union -from ..constants import AnnotationFlag from ..generic._base import ( BooleanObject, NameObject, @@ -12,8 +11,6 @@ from ..generic._rectangle import RectangleObject from ._base import AnnotationDictionary -DEFAULT_ANNOTATION_FLAG = AnnotationFlag(0) - class Link(AnnotationDictionary): def __init__( From b2d72043ab5221b58138c7d06c181b8cbc88ea8e Mon Sep 17 00:00:00 2001 From: owurman Date: Mon, 5 Aug 2024 12:14:18 -0700 Subject: [PATCH 07/21] BUG: Handle Sequence as an IndirectObject when extracting text with layout mode (#2788) * Handle Sequence as an IndirectObject The spec allows an int or float to be an IndirectObject as well, but this commit does not address that theoretical possibility. * Update pypdf/_text_extraction/_layout_mode/_font.py Co-authored-by: Stefan <96178532+stefan6419846@users.noreply.github.com> * Address PR comments -Rename w_1 to w_next_entry -Utilize ParseError instead of PdfReadError -Write a test (both positive and negative) * Handle unlikely case of IndirectObjects for float/int width elements Also adds a comment to clarify that we don't explicitly handle the IndexError exception. Rather, we let it be raised as an IndexError. * Yoda condition I removed * Last commit was a bad patch, confused by non-committed changes * Use test files from URL rather than resources * Update tests/test_text_extraction.py Co-authored-by: pubpub-zz <4083478+pubpub-zz@users.noreply.github.com> * Fix code style warnings in range() call --------- Co-authored-by: Stefan <96178532+stefan6419846@users.noreply.github.com> Co-authored-by: pubpub-zz <4083478+pubpub-zz@users.noreply.github.com> --- pypdf/_text_extraction/_layout_mode/_font.py | 26 +++++++++++++------- tests/test_text_extraction.py | 17 +++++++++++++ 2 files changed, 34 insertions(+), 9 deletions(-) diff --git a/pypdf/_text_extraction/_layout_mode/_font.py b/pypdf/_text_extraction/_layout_mode/_font.py index a912fddb2..40655b1b2 100644 --- a/pypdf/_text_extraction/_layout_mode/_font.py +++ b/pypdf/_text_extraction/_layout_mode/_font.py @@ -1,8 +1,9 @@ """Font constants and classes for "layout" mode text operations""" from dataclasses import dataclass, field -from typing import Any, Dict, Sequence, Union +from typing import Any, Dict, Sequence, Union, cast +from ...errors import ParseError from ...generic import IndirectObject from ._font_widths import STANDARD_WIDTHS @@ -58,6 +59,7 @@ def __post_init__(self) -> None: skip_count = 0 _w = d_font.get("/W", []) for idx, w_entry in enumerate(_w): + w_entry = w_entry.get_object() if skip_count: skip_count -= 1 continue @@ -66,13 +68,14 @@ def __post_init__(self) -> None: # warning and or use reader's "strict" to force an ex??? continue # check for format (1): `int [int int int int ...]` - if isinstance(_w[idx + 1], Sequence): - start_idx, width_list = _w[idx : idx + 2] + w_next_entry = _w[idx + 1].get_object() + if isinstance(w_next_entry, Sequence): + start_idx, width_list = w_entry, w_next_entry self.width_map.update( { ord_map[_cidx]: _width for _cidx, _width in zip( - range(start_idx, start_idx + len(width_list), 1), + range(cast(int, start_idx), cast(int, start_idx) + len(width_list), 1), width_list, ) if _cidx in ord_map @@ -80,18 +83,23 @@ def __post_init__(self) -> None: ) skip_count = 1 # check for format (2): `int int int` - if not isinstance(_w[idx + 1], Sequence) and not isinstance( - _w[idx + 2], Sequence - ): - start_idx, stop_idx, const_width = _w[idx : idx + 3] + elif isinstance(w_next_entry, (int, float)) and isinstance(_w[idx + 2].get_object(), (int, float)): + start_idx, stop_idx, const_width = w_entry, w_next_entry, _w[idx + 2].get_object() self.width_map.update( { ord_map[_cidx]: const_width - for _cidx in range(start_idx, stop_idx + 1, 1) + for _cidx in range(cast(int, start_idx), cast(int, stop_idx + 1), 1) if _cidx in ord_map } ) skip_count = 2 + else: + # Note: this doesn't handle the case of out of bounds (reaching the end of the width definitions + # while expecting more elements). This raises an IndexError which is sufficient. + raise ParseError( + f"Invalid font width definition. Next elements: {w_entry}, {w_next_entry}, {_w[idx + 2]}" + ) # pragma: no cover + if not self.width_map and "/BaseFont" in self.font_dictionary: for key in STANDARD_WIDTHS: if self.font_dictionary["/BaseFont"].startswith(f"/{key}"): diff --git a/tests/test_text_extraction.py b/tests/test_text_extraction.py index 1ffa68a3e..dcd4e6cae 100644 --- a/tests/test_text_extraction.py +++ b/tests/test_text_extraction.py @@ -10,6 +10,7 @@ from pypdf import PdfReader, mult from pypdf._text_extraction import set_custom_rtl +from pypdf.errors import ParseError from . import get_data_from_url @@ -156,3 +157,19 @@ def test_layout_mode_type0_font_widths(): encoding="utf-8" ) assert expected == reader.pages[0].extract_text(extraction_mode="layout") + + +@pytest.mark.enable_socket() +def test_layout_mode_indirect_sequence_font_widths(): + # Cover the situation where the sequence for font widths is an IndirectObject + # ref https://github.com/py-pdf/pypdf/pull/2788 + url = "https://github.com/user-attachments/files/16491621/2788_example.pdf" + name ="2788_example.pdf" + reader = PdfReader(BytesIO(get_data_from_url(url, name=name))) + assert reader.pages[0].extract_text(extraction_mode="layout") == "" + url = "https://github.com/user-attachments/files/16491619/2788_example_malformed.pdf" + name = "2788_example_malformed.pdf" + reader = PdfReader(BytesIO(get_data_from_url(url, name=name))) + with pytest.raises(ParseError) as exc: + reader.pages[0].extract_text(extraction_mode="layout") + assert str(exc.value).startswith("Invalid font width definition") From 5abd590740a2718fc69b8477c656ce5515a0ab33 Mon Sep 17 00:00:00 2001 From: j-t-1 <120829237+j-t-1@users.noreply.github.com> Date: Wed, 7 Aug 2024 12:14:17 +0100 Subject: [PATCH 08/21] STY: Refactor b_ (#2772) --- pypdf/_utils.py | 11 ++++------- 1 file changed, 4 insertions(+), 7 deletions(-) diff --git a/pypdf/_utils.py b/pypdf/_utils.py index 6569707b6..5fecb38e7 100644 --- a/pypdf/_utils.py +++ b/pypdf/_utils.py @@ -347,14 +347,11 @@ def b_(s: Union[str, bytes]) -> bytes: return bc[s] try: r = s.encode("latin-1") - if len(s) < 2: - bc[s] = r - return r - except Exception: + except UnicodeEncodeError: r = s.encode("utf-8") - if len(s) < 2: - bc[s] = r - return r + if len(s) < 2: + bc[s] = r + return r def str_(b: Any) -> str: From 219eb13f7eb9c2cd9519e9a69d639250853bd823 Mon Sep 17 00:00:00 2001 From: pubpub-zz <4083478+pubpub-zz@users.noreply.github.com> Date: Mon, 12 Aug 2024 10:52:40 +0200 Subject: [PATCH 09/21] MAINT: Drop Python 3.7 support (#2793) --- .github/workflows/github-ci.yaml | 14 +++++++------- pypdf/_page.py | 8 +------- pypdf/_protocols.py | 8 +------- .../_layout_mode/_fixed_width_page.py | 8 +------- pypdf/_xobj_image_helpers.py | 13 ++++--------- pypdf/types.py | 8 +------- 6 files changed, 15 insertions(+), 44 deletions(-) diff --git a/.github/workflows/github-ci.yaml b/.github/workflows/github-ci.yaml index 1eb3d9bd0..d5d9bb4d4 100644 --- a/.github/workflows/github-ci.yaml +++ b/.github/workflows/github-ci.yaml @@ -57,12 +57,12 @@ jobs: runs-on: ubuntu-20.04 strategy: matrix: - python-version: ["3.7", "3.8", "3.9", "3.10", "3.11", "3.12", "3.13-dev"] + python-version: ["3.8", "3.9", "3.10", "3.11", "3.12", "3.13-dev"] use-crypto-lib: ["cryptography"] include: - - python-version: "3.7" + - python-version: "3.8" use-crypto-lib: "pycryptodome" - - python-version: "3.7" + - python-version: "3.8" use-crypto-lib: "none" steps: - name: Update APT packages @@ -83,7 +83,7 @@ jobs: key: cache-downloaded-files - name: Setup Python uses: actions/setup-python@v5 - if: matrix.python-version == '3.7' || matrix.python-version == '3.8' || matrix.python-version == '3.9' || matrix.python-version == '3.10' + if: matrix.python-version == '3.8' || matrix.python-version == '3.9' || matrix.python-version == '3.10' with: python-version: ${{ matrix.python-version }} cache: 'pip' @@ -102,7 +102,7 @@ jobs: - name: Install requirements (Python 3) run: | pip install -r requirements/ci.txt - if: matrix.python-version == '3.7' || matrix.python-version == '3.8' || matrix.python-version == '3.9' || matrix.python-version == '3.10' + if: matrix.python-version == '3.8' || matrix.python-version == '3.9' || matrix.python-version == '3.10' - name: Install requirements (Python 3.11+) run: | pip install -r requirements/ci-3.11.txt @@ -215,8 +215,8 @@ jobs: - name: Check Number of Downloaded Files run: | downloaded_files_count=$(find \.coverage* -type f | wc -l) - if [ $downloaded_files_count -eq 9 ]; then - echo "The expected number of files (9) were downloaded." + if [ $downloaded_files_count -eq 8 ]; then + echo "The expected number of files (8) were downloaded." else echo "ERROR: Expected 8 files, but found $downloaded_files_count files." exit 1 diff --git a/pypdf/_page.py b/pypdf/_page.py index 63038d9d0..ee1dc7f60 100644 --- a/pypdf/_page.py +++ b/pypdf/_page.py @@ -28,7 +28,6 @@ # POSSIBILITY OF SUCH DAMAGE. import math -import sys from decimal import Decimal from pathlib import Path from typing import ( @@ -38,6 +37,7 @@ Iterable, Iterator, List, + Literal, Optional, Sequence, Set, @@ -85,12 +85,6 @@ StreamObject, ) -if sys.version_info >= (3, 8): - from typing import Literal -else: - from typing_extensions import Literal - - MERGE_CROP_BOX = "cropbox" # pypdf<=3.4.0 used 'trimbox' diff --git a/pypdf/_protocols.py b/pypdf/_protocols.py index 9f413660b..b5fa14879 100644 --- a/pypdf/_protocols.py +++ b/pypdf/_protocols.py @@ -2,13 +2,7 @@ from abc import abstractmethod from pathlib import Path -from typing import IO, Any, Dict, List, Optional, Tuple, Union - -try: - # Python 3.8+: https://peps.python.org/pep-0586 - from typing import Protocol -except ImportError: - from typing_extensions import Protocol # type: ignore[assignment] +from typing import IO, Any, Dict, List, Optional, Protocol, Tuple, Union from ._utils import StrByteType, StreamType diff --git a/pypdf/_text_extraction/_layout_mode/_fixed_width_page.py b/pypdf/_text_extraction/_layout_mode/_fixed_width_page.py index 1be500959..e7af1b234 100644 --- a/pypdf/_text_extraction/_layout_mode/_fixed_width_page.py +++ b/pypdf/_text_extraction/_layout_mode/_fixed_width_page.py @@ -1,10 +1,9 @@ """Extract PDF text preserving the layout of the source PDF""" -import sys from itertools import groupby from math import ceil from pathlib import Path -from typing import Any, Dict, Iterator, List, Optional, Tuple +from typing import Any, Dict, Iterator, List, Literal, Optional, Tuple, TypedDict from ..._utils import logger_warning from .. import LAYOUT_NEW_BT_GROUP_SPACE_WIDTHS @@ -12,11 +11,6 @@ from ._text_state_manager import TextStateManager from ._text_state_params import TextStateParams -if sys.version_info >= (3, 8): - from typing import Literal, TypedDict -else: - from typing_extensions import Literal, TypedDict - class BTGroup(TypedDict): """ diff --git a/pypdf/_xobj_image_helpers.py b/pypdf/_xobj_image_helpers.py index 5ae8894fa..7a3f40d95 100644 --- a/pypdf/_xobj_image_helpers.py +++ b/pypdf/_xobj_image_helpers.py @@ -2,7 +2,7 @@ import sys from io import BytesIO -from typing import Any, List, Tuple, Union, cast +from typing import Any, List, Literal, Tuple, Union, cast from ._utils import check_if_whitespace_only, logger_warning from .constants import ColorSpaces @@ -15,13 +15,6 @@ NullObject, ) -if sys.version_info[:2] >= (3, 8): - from typing import Literal -else: - # PEP 586 introduced typing.Literal with Python 3.8 - # For older Python versions, the backport typing_extensions is necessary: - from typing_extensions import Literal - if sys.version_info[:2] >= (3, 10): from typing import TypeAlias else: @@ -150,7 +143,9 @@ def _extended_image_frombytes( nb_pix = size[0] * size[1] data_length = len(data) if data_length == 0: - raise EmptyImageDataError("Data is 0 bytes, cannot process an image from empty data.") from exc + raise EmptyImageDataError( + "Data is 0 bytes, cannot process an image from empty data." + ) from exc if data_length % nb_pix != 0: raise exc k = nb_pix * len(mode) / data_length diff --git a/pypdf/types.py b/pypdf/types.py index b8fbab92c..e383dc7b1 100644 --- a/pypdf/types.py +++ b/pypdf/types.py @@ -1,13 +1,7 @@ """Helpers for working with PDF types.""" import sys -from typing import List, Union - -if sys.version_info[:2] >= (3, 8): - # Python 3.8+: https://peps.python.org/pep-0586 - from typing import Literal -else: - from typing_extensions import Literal +from typing import List, Literal, Union if sys.version_info[:2] >= (3, 10): # Python 3.10+: https://www.python.org/dev/peps/pep-0484 From 46c89dd8e1e8641a49624f3fbc1865f9c4b41374 Mon Sep 17 00:00:00 2001 From: pubpub-zz <4083478+pubpub-zz@users.noreply.github.com> Date: Mon, 12 Aug 2024 10:57:25 +0200 Subject: [PATCH 10/21] MAINT: Remove b_ and str_ (#2792) Closes #2726. Closes #2791. --- pypdf/_cmap.py | 30 +++---- pypdf/_doc_common.py | 3 +- pypdf/_encryption.py | 6 +- pypdf/_merger.py | 10 ++- pypdf/_page.py | 16 ++-- pypdf/_reader.py | 5 +- pypdf/_utils.py | 34 ++------ pypdf/_writer.py | 6 +- pypdf/filters.py | 53 ++++++++---- pypdf/generic/_base.py | 39 +++++---- pypdf/generic/_data_structures.py | 35 ++++---- pypdf/generic/_utils.py | 77 ++++++++++-------- ..._Vicksburg_Sample_OCR-crazyones-merged.pdf | Bin 217096 -> 217093 bytes tests/test_cmap.py | 18 +++- tests/test_page.py | 13 ++- tests/test_utils.py | 16 ---- tests/test_workflows.py | 4 +- tests/test_writer.py | 2 +- 18 files changed, 193 insertions(+), 174 deletions(-) diff --git a/pypdf/_cmap.py b/pypdf/_cmap.py index 9a2d10a61..d635724d2 100644 --- a/pypdf/_cmap.py +++ b/pypdf/_cmap.py @@ -3,11 +3,10 @@ from typing import Any, Dict, List, Tuple, Union, cast from ._codecs import adobe_glyphs, charset_encoding -from ._utils import b_, logger_error, logger_warning +from ._utils import logger_error, logger_warning from .generic import ( DecodedStreamObject, DictionaryObject, - IndirectObject, NullObject, StreamObject, ) @@ -258,7 +257,7 @@ def prepare_cm(ft: DictionaryObject) -> bytes: tu = ft["/ToUnicode"] cm: bytes if isinstance(tu, StreamObject): - cm = b_(cast(DecodedStreamObject, ft["/ToUnicode"]).get_data()) + cm = cast(DecodedStreamObject, ft["/ToUnicode"]).get_data() elif isinstance(tu, str) and tu.startswith("/Identity"): # the full range 0000-FFFF will be processed cm = b"beginbfrange\n<0000> <0001> <0000>\nendbfrange" @@ -448,34 +447,27 @@ def compute_space_width( en: int = cast(int, ft["/LastChar"]) if st > space_code or en < space_code: raise Exception("Not in range") - if w[space_code - st] == 0: + if w[space_code - st].get_object() == 0: raise Exception("null width") - sp_width = w[space_code - st] + sp_width = w[space_code - st].get_object() except Exception: if "/FontDescriptor" in ft and "/MissingWidth" in cast( DictionaryObject, ft["/FontDescriptor"] ): - sp_width = ft["/FontDescriptor"]["/MissingWidth"] # type: ignore + sp_width = ft["/FontDescriptor"]["/MissingWidth"].get_object() # type: ignore else: # will consider width of char as avg(width)/2 m = 0 cpt = 0 - for x in w: - if x > 0: - m += x + for xx in w: + xx = xx.get_object() + if xx > 0: + m += xx cpt += 1 sp_width = m / max(1, cpt) / 2 - if isinstance(sp_width, IndirectObject): - # According to - # 'Table 122 - Entries common to all font descriptors (continued)' - # the MissingWidth should be a number, but according to #2286 it can - # be an indirect object - obj = sp_width.get_object() - if obj is None or isinstance(obj, NullObject): - return 0.0 - return obj # type: ignore - + if sp_width is None or isinstance(sp_width, NullObject): + sp_width = 0.0 return sp_width diff --git a/pypdf/_doc_common.py b/pypdf/_doc_common.py index d4c5c43c3..ffbdb7882 100644 --- a/pypdf/_doc_common.py +++ b/pypdf/_doc_common.py @@ -49,7 +49,6 @@ from ._page import PageObject, _VirtualList from ._page_labels import index2label as page_index2page_label from ._utils import ( - b_, deprecate_with_replacement, logger_warning, parse_iso8824_date, @@ -1258,7 +1257,7 @@ def xfa(self) -> Optional[Dict[str, Any]]: if isinstance(f, IndirectObject): field = cast(Optional[EncodedStreamObject], f.get_object()) if field: - es = zlib.decompress(b_(field._data)) + es = zlib.decompress(field._data) retval[tag] = es return retval diff --git a/pypdf/_encryption.py b/pypdf/_encryption.py index 5ddd8d0ef..e5cdd9324 100644 --- a/pypdf/_encryption.py +++ b/pypdf/_encryption.py @@ -43,7 +43,7 @@ rc4_encrypt, ) -from ._utils import b_, logger_warning +from ._utils import logger_warning from .generic import ( ArrayObject, ByteStringObject, @@ -78,7 +78,7 @@ def encrypt_object(self, obj: PdfObject) -> PdfObject: elif isinstance(obj, StreamObject): obj2 = StreamObject() obj2.update(obj) - obj2.set_data(self.stm_crypt.encrypt(b_(obj._data))) + obj2.set_data(self.stm_crypt.encrypt(obj._data)) for key, value in obj.items(): # Dont forget the Stream dict. obj2[key] = self.encrypt_object(value) obj = obj2 @@ -96,7 +96,7 @@ def decrypt_object(self, obj: PdfObject) -> PdfObject: data = self.str_crypt.decrypt(obj.original_bytes) obj = create_string_object(data) elif isinstance(obj, StreamObject): - obj._data = self.stm_crypt.decrypt(b_(obj._data)) + obj._data = self.stm_crypt.decrypt(obj._data) for key, value in obj.items(): # Dont forget the Stream dict. obj[key] = self.decrypt_object(value) elif isinstance(obj, DictionaryObject): diff --git a/pypdf/_merger.py b/pypdf/_merger.py index 7176a1adf..a52a354e3 100644 --- a/pypdf/_merger.py +++ b/pypdf/_merger.py @@ -46,7 +46,6 @@ from ._utils import ( StrByteType, deprecate_with_replacement, - str_, ) from ._writer import PdfWriter from .constants import GoToActionArguments, TypArguments, TypFitArguments @@ -82,6 +81,15 @@ def __init__(self, pagedata: PageObject, src: PdfReader, id: int) -> None: self.id = id +# transfered from _utils : as this function is only required here +# and merger will be soon deprecated +def str_(b: Any) -> str: # pragma: no cover + if isinstance(b, bytes): + return b.decode("latin-1") + else: + return str(b) # will return b.__str__() if defined + + class PdfMerger: """ Use :class:`PdfWriter` instead. diff --git a/pypdf/_page.py b/pypdf/_page.py index ee1dc7f60..48cdeb149 100644 --- a/pypdf/_page.py +++ b/pypdf/_page.py @@ -846,7 +846,7 @@ def _add_transformation_matrix( FloatObject(e), FloatObject(f), ], - " cm", + b"cm", ], ) return contents @@ -864,7 +864,7 @@ def _get_contents_as_bytes(self) -> Optional[bytes]: if isinstance(obj, list): return b"".join(x.get_object().get_data() for x in obj) else: - return cast(bytes, cast(EncodedStreamObject, obj).get_data()) + return cast(EncodedStreamObject, obj).get_data() else: return None @@ -1057,11 +1057,11 @@ def _merge_page( rect.height, ], ), - "re", + b"re", ), ) - page2content.operations.insert(1, ([], "W")) - page2content.operations.insert(2, ([], "n")) + page2content.operations.insert(1, ([], b"W")) + page2content.operations.insert(2, ([], b"n")) if page2transformation is not None: page2content = page2transformation(page2content) page2content = PageObject._content_stream_rename( @@ -1195,11 +1195,11 @@ def _merge_page_writer( rect.height, ], ), - "re", + b"re", ), ) - page2content.operations.insert(1, ([], "W")) - page2content.operations.insert(2, ([], "n")) + page2content.operations.insert(1, ([], b"W")) + page2content.operations.insert(2, ([], b"n")) if page2transformation is not None: page2content = page2transformation(page2content) page2content = PageObject._content_stream_rename( diff --git a/pypdf/_reader.py b/pypdf/_reader.py index aeababa7b..7c084107c 100644 --- a/pypdf/_reader.py +++ b/pypdf/_reader.py @@ -51,7 +51,6 @@ from ._utils import ( StrByteType, StreamType, - b_, logger_warning, read_non_whitespace, read_previous_line, @@ -328,7 +327,7 @@ def _get_object_from_stream( assert cast(str, obj_stm["/Type"]) == "/ObjStm" # /N is the number of indirect objects in the stream assert idx < obj_stm["/N"] - stream_data = BytesIO(b_(obj_stm.get_data())) + stream_data = BytesIO(obj_stm.get_data()) for i in range(obj_stm["/N"]): # type: ignore read_non_whitespace(stream_data) stream_data.seek(-1, 1) @@ -932,7 +931,7 @@ def _read_pdf15_xref_stream( xrefstream = cast(ContentStream, read_object(stream, self)) assert cast(str, xrefstream["/Type"]) == "/XRef" self.cache_indirect_object(generation, idnum, xrefstream) - stream_data = BytesIO(b_(xrefstream.get_data())) + stream_data = BytesIO(xrefstream.get_data()) # Index pairs specify the subsections in the dictionary. If # none create one subsection that spans everything. idx_pairs = xrefstream.get("/Index", [0, xrefstream.get("/Size")]) diff --git a/pypdf/_utils.py b/pypdf/_utils.py index 5fecb38e7..94d45cf6d 100644 --- a/pypdf/_utils.py +++ b/pypdf/_utils.py @@ -336,31 +336,6 @@ def mark_location(stream: StreamType) -> None: stream.seek(-radius, 1) -B_CACHE: Dict[str, bytes] = {} - - -def b_(s: Union[str, bytes]) -> bytes: - if isinstance(s, bytes): - return s - bc = B_CACHE - if s in bc: - return bc[s] - try: - r = s.encode("latin-1") - except UnicodeEncodeError: - r = s.encode("utf-8") - if len(s) < 2: - bc[s] = r - return r - - -def str_(b: Any) -> str: - if isinstance(b, bytes): - return b.decode("latin-1") - else: - return str(b) # will return b.__str__() if defined - - @overload def ord_(b: str) -> int: ... @@ -397,12 +372,17 @@ def deprecation(msg: str) -> None: def deprecate_with_replacement(old_name: str, new_name: str, removed_in: str) -> None: """Raise an exception that a feature will be removed, but has a replacement.""" - deprecate(f"{old_name} is deprecated and will be removed in pypdf {removed_in}. Use {new_name} instead.", 4) + deprecate( + f"{old_name} is deprecated and will be removed in pypdf {removed_in}. Use {new_name} instead.", + 4, + ) def deprecation_with_replacement(old_name: str, new_name: str, removed_in: str) -> None: """Raise an exception that a feature was already removed, but has a replacement.""" - deprecation(f"{old_name} is deprecated and was removed in pypdf {removed_in}. Use {new_name} instead.") + deprecation( + f"{old_name} is deprecated and was removed in pypdf {removed_in}. Use {new_name} instead." + ) def deprecate_no_replacement(name: str, removed_in: str) -> None: diff --git a/pypdf/_writer.py b/pypdf/_writer.py index 00b9d498c..d73c00e3d 100644 --- a/pypdf/_writer.py +++ b/pypdf/_writer.py @@ -62,7 +62,6 @@ StrByteType, StreamType, _get_max_pdf_version_header, - b_, deprecate_with_replacement, logger_warning, ) @@ -678,9 +677,10 @@ def add_attachment(self, filename: str, data: Union[str, bytes]) -> None: # Hello world! # endstream # endobj - + if isinstance(data, str): + data = data.encode("latin-1") file_entry = DecodedStreamObject() - file_entry.set_data(b_(data)) + file_entry.set_data(data) file_entry.update({NameObject(PA.TYPE): NameObject("/EmbeddedFile")}) # The Filespec entry diff --git a/pypdf/filters.py b/pypdf/filters.py index 137e3603a..43730cc8e 100644 --- a/pypdf/filters.py +++ b/pypdf/filters.py @@ -43,7 +43,7 @@ from ._utils import ( WHITESPACES_AS_BYTES, - b_, + deprecate, deprecate_with_replacement, deprecation_no_replacement, logger_warning, @@ -376,20 +376,18 @@ class LZWDecode: """ Taken from: - http://www.java2s.com/Open-Source/Java-Document/PDF/PDF- - Renderer/com/sun/pdfview/decode/LZWDecode.java.htm + http://www.java2s.com/Open-Source/Java-Document/PDF/PDF-Renderer/com/sun/pdfview/decode/LZWDecode.java.htm """ class Decoder: + STOP = 257 + CLEARDICT = 256 + def __init__(self, data: bytes) -> None: - self.STOP = 257 - self.CLEARDICT = 256 self.data = data self.bytepos = 0 self.bitpos = 0 - self.dict = [""] * 4096 - for i in range(256): - self.dict[i] = chr(i) + self.dict = [struct.pack("B", i) for i in range(256)] + [b""] * (4096 - 256) self.reset_dict() def reset_dict(self) -> None: @@ -416,7 +414,7 @@ def next_code(self) -> int: self.bytepos = self.bytepos + 1 return value - def decode(self) -> str: + def decode(self) -> bytes: """ TIFF 6.0 specification explains in sufficient details the steps to implement the LZW encode() and decode() algorithms. @@ -429,7 +427,7 @@ def decode(self) -> str: PdfReadError: If the stop code is missing """ cW = self.CLEARDICT - baos = "" + baos = b"" while True: pW = cW cW = self.next_code() @@ -444,11 +442,11 @@ def decode(self) -> str: else: if cW < self.dictlen: baos += self.dict[cW] - p = self.dict[pW] + self.dict[cW][0] + p = self.dict[pW] + self.dict[cW][0:1] self.dict[self.dictlen] = p self.dictlen += 1 else: - p = self.dict[pW] + self.dict[pW][0] + p = self.dict[pW] + self.dict[pW][0:1] baos += p self.dict[self.dictlen] = p self.dictlen += 1 @@ -460,11 +458,11 @@ def decode(self) -> str: return baos @staticmethod - def decode( + def _decodeb( data: bytes, decode_parms: Optional[DictionaryObject] = None, **kwargs: Any, - ) -> str: + ) -> bytes: """ Decode an LZW encoded data stream. @@ -476,9 +474,28 @@ def decode( decoded data. """ # decode_parms is unused here - return LZWDecode.Decoder(data).decode() + @staticmethod + def decode( + data: bytes, + decode_parms: Optional[DictionaryObject] = None, + **kwargs: Any, + ) -> str: # deprecated + """ + Decode an LZW encoded data stream. + + Args: + data: ``bytes`` or ``str`` text to decode. + decode_parms: a dictionary of parameter values. + + Returns: + decoded data. + """ + # decode_parms is unused here + deprecate("LZWDecode.decode will return bytes instead of str in pypdf 6.0.0") + return LZWDecode.Decoder(data).decode().decode("latin-1") + class ASCII85Decode: """Decodes string ASCII85-encoded data into a byte format.""" @@ -651,7 +668,7 @@ def decode( return tiff_header + data -def decode_stream_data(stream: Any) -> Union[bytes, str]: # utils.StreamObject +def decode_stream_data(stream: Any) -> bytes: # utils.StreamObject """ Decode the stream data based on the specified filters. @@ -678,7 +695,7 @@ def decode_stream_data(stream: Any) -> Union[bytes, str]: # utils.StreamObject decodparms = stream.get(SA.DECODE_PARMS, ({},) * len(filters)) if not isinstance(decodparms, (list, tuple)): decodparms = (decodparms,) - data: bytes = b_(stream._data) + data: bytes = stream._data # If there is not data to decode we should not try to decode the data. if data: for filter_type, params in zip(filters, decodparms): @@ -691,7 +708,7 @@ def decode_stream_data(stream: Any) -> Union[bytes, str]: # utils.StreamObject elif filter_type in (FT.RUN_LENGTH_DECODE, FTA.RL): data = RunLengthDecode.decode(data) elif filter_type in (FT.LZW_DECODE, FTA.LZW): - data = LZWDecode.decode(data, params) # type: ignore + data = LZWDecode._decodeb(data, params) elif filter_type in (FT.ASCII_85_DECODE, FTA.A85): data = ASCII85Decode.decode(data) elif filter_type == FT.DCT_DECODE: diff --git a/pypdf/generic/_base.py b/pypdf/generic/_base.py index 2d606b418..309d389cc 100644 --- a/pypdf/generic/_base.py +++ b/pypdf/generic/_base.py @@ -30,18 +30,17 @@ import re from binascii import unhexlify from math import log10 +from struct import iter_unpack from typing import Any, Callable, ClassVar, Dict, Optional, Sequence, Union, cast from .._codecs import _pdfdoc_encoding_rev from .._protocols import PdfObjectProtocol, PdfWriterProtocol from .._utils import ( StreamType, - b_, deprecate_no_replacement, logger_warning, read_non_whitespace, read_until_regex, - str_, ) from ..errors import STREAM_TRUNCATED_PREMATURELY, PdfReadError, PdfStreamError @@ -308,6 +307,10 @@ def __getitem__(self, key: Any) -> Any: # items should be extracted from pointed Object return self._get_object_with_check()[key] # type: ignore + def __float__(self) -> str: + # in this case we are looking for the pointed data + return self.get_object().__float__() # type: ignore + def __str__(self) -> str: # in this case we are looking for the pointed data return self.get_object().__str__() @@ -369,10 +372,10 @@ def read_from_stream(stream: StreamType, pdf: Any) -> "IndirectObject": # PdfRe class FloatObject(float, PdfObject): def __new__( - cls, value: Union[str, Any] = "0.0", context: Optional[Any] = None + cls, value: Any = "0.0", context: Optional[Any] = None ) -> "FloatObject": try: - value = float(str_(value)) + value = float(value) return float.__new__(cls, value) except Exception as e: # If this isn't a valid decimal (happens in malformed PDFs) @@ -599,15 +602,16 @@ def write_to_stream( ) bytearr = self.get_encoded_bytes() stream.write(b"(") - for c in bytearr: - if not chr(c).isalnum() and c != b" ": + for c_ in iter_unpack("c", bytearr): + c = cast(bytes, c_[0]) + if not c.isalnum() and c != b" ": # This: # stream.write(rf"\{c:0>3o}".encode()) # gives # https://github.com/davidhalter/parso/issues/207 - stream.write(("\\%03o" % c).encode()) + stream.write(b"\\%03o" % ord(c)) else: - stream.write(b_(chr(c))) + stream.write(c) stream.write(b")") @@ -710,12 +714,13 @@ def read_from_stream(stream: StreamType, pdf: Any) -> "NameObject": # PdfReader def encode_pdfdocencoding(unicode_string: str) -> bytes: - retval = bytearray() - for c in unicode_string: - try: - retval += b_(chr(_pdfdoc_encoding_rev[c])) - except KeyError: - raise UnicodeEncodeError( - "pdfdocencoding", c, -1, -1, "does not exist in translation table" - ) - return bytes(retval) + try: + return bytes([_pdfdoc_encoding_rev[k] for k in unicode_string]) + except KeyError: + raise UnicodeEncodeError( + "pdfdocencoding", + unicode_string, + -1, + -1, + "does not exist in translation table", + ) diff --git a/pypdf/generic/_data_structures.py b/pypdf/generic/_data_structures.py index 87d688674..399836be5 100644 --- a/pypdf/generic/_data_structures.py +++ b/pypdf/generic/_data_structures.py @@ -52,7 +52,6 @@ from .._utils import ( WHITESPACES, StreamType, - b_, deprecate_no_replacement, deprecate_with_replacement, logger_warning, @@ -843,7 +842,7 @@ def _reset_node_tree_relationship(child_obj: Any) -> None: class StreamObject(DictionaryObject): def __init__(self) -> None: - self._data: Union[bytes, str] = b"" + self._data: bytes = b"" self.decoded_self: Optional[DecodedStreamObject] = None def _clone( @@ -877,7 +876,7 @@ def _clone( pass super()._clone(src, pdf_dest, force_duplicate, ignore_fields, visited) - def get_data(self) -> Union[bytes, str]: + def get_data(self) -> bytes: return self._data def set_data(self, data: bytes) -> None: @@ -885,7 +884,7 @@ def set_data(self, data: bytes) -> None: def hash_value_data(self) -> bytes: data = super().hash_value_data() - data += b_(self._data) + data += self._data return data def write_to_stream( @@ -955,7 +954,7 @@ def flate_encode(self, level: int = -1) -> "EncodedStreamObject": retval[NameObject(SA.FILTER)] = f if params is not None: retval[NameObject(SA.DECODE_PARMS)] = params - retval._data = FlateDecode.encode(b_(self._data), level) + retval._data = FlateDecode.encode(self._data, level) return retval def decode_as_image(self) -> Any: @@ -993,7 +992,7 @@ def __init__(self) -> None: self.decoded_self: Optional[DecodedStreamObject] = None # This overrides the parent method: - def get_data(self) -> Union[bytes, str]: + def get_data(self) -> bytes: from ..filters import decode_stream_data if self.decoded_self is not None: @@ -1003,7 +1002,7 @@ def get_data(self) -> Union[bytes, str]: # create decoded object decoded = DecodedStreamObject() - decoded.set_data(b_(decode_stream_data(self))) + decoded.set_data(decode_stream_data(self)) for key, value in list(self.items()): if key not in (SA.LENGTH, SA.FILTER, SA.DECODE_PARMS): decoded[key] = value @@ -1058,7 +1057,7 @@ def __init__( # The inner list has two elements: # Element 0: List # Element 1: str - self._operations: List[Tuple[Any, Any]] = [] + self._operations: List[Tuple[Any, bytes]] = [] # stream may be a StreamObject or an ArrayObject containing # multiple StreamObjects to be cat'd together. @@ -1069,14 +1068,14 @@ def __init__( if isinstance(stream, ArrayObject): data = b"" for s in stream: - data += b_(s.get_object().get_data()) + data += s.get_object().get_data() if len(data) == 0 or data[-1] != b"\n": data += b"\n" super().set_data(bytes(data)) else: stream_data = stream.get_data() assert stream_data is not None - super().set_data(b_(stream_data)) + super().set_data(stream_data) self.forced_encoding = forced_encoding def clone( @@ -1132,7 +1131,7 @@ def _clone( ignore_fields: """ src_cs = cast("ContentStream", src) - super().set_data(b_(src_cs._data)) + super().set_data(src_cs._data) self.pdf = pdf_dest self._operations = list(src_cs._operations) self.forced_encoding = src_cs.forced_encoding @@ -1249,10 +1248,10 @@ def get_data(self) -> bytes: for op in operands: op.write_to_stream(new_data) new_data.write(b" ") - new_data.write(b_(operator)) + new_data.write(operator) new_data.write(b"\n") self._data = new_data.getvalue() - return b_(self._data) + return self._data # This overrides the parent method: def set_data(self, data: bytes) -> None: @@ -1262,21 +1261,21 @@ def set_data(self, data: bytes) -> None: @property def operations(self) -> List[Tuple[Any, Any]]: if not self._operations and self._data: - self._parse_content_stream(BytesIO(b_(self._data))) + self._parse_content_stream(BytesIO(self._data)) self._data = b"" return self._operations @operations.setter - def operations(self, operations: List[Tuple[Any, Any]]) -> None: + def operations(self, operations: List[Tuple[Any, bytes]]) -> None: self._operations = operations self._data = b"" def isolate_graphics_state(self) -> None: if self._operations: - self._operations.insert(0, ([], "q")) - self._operations.append(([], "Q")) + self._operations.insert(0, ([], b"q")) + self._operations.append(([], b"Q")) elif self._data: - self._data = b"q\n" + b_(self._data) + b"\nQ\n" + self._data = b"q\n" + self._data + b"\nQ\n" # This overrides the parent method: def write_to_stream( diff --git a/pypdf/generic/_utils.py b/pypdf/generic/_utils.py index fdcdc3339..b5ac6632a 100644 --- a/pypdf/generic/_utils.py +++ b/pypdf/generic/_utils.py @@ -2,7 +2,7 @@ from typing import Dict, List, Tuple, Union from .._codecs import _pdfdoc_encoding -from .._utils import StreamType, b_, logger_warning, read_non_whitespace +from .._utils import StreamType, logger_warning, read_non_whitespace from ..errors import STREAM_TRUNCATED_PREMATURELY, PdfStreamError from ._base import ByteStringObject, TextStringObject @@ -16,7 +16,7 @@ def read_hex_string_from_stream( forced_encoding: Union[None, str, List[str], Dict[int, str]] = None, ) -> Union["TextStringObject", "ByteStringObject"]: stream.read(1) - txt = "" + arr = [] x = b"" while True: tok = read_non_whitespace(stream) @@ -26,13 +26,37 @@ def read_hex_string_from_stream( break x += tok if len(x) == 2: - txt += chr(int(x, base=16)) + arr.append(int(x, base=16)) x = b"" if len(x) == 1: x += b"0" - if len(x) == 2: - txt += chr(int(x, base=16)) - return create_string_object(b_(txt), forced_encoding) + if x != b"": + arr.append(int(x, base=16)) + return create_string_object(bytes(arr), forced_encoding) + + +__ESPACE_DICT__ = { + b"n": ord(b"\n"), + b"r": ord(b"\r"), + b"t": ord(b"\t"), + b"b": ord(b"\b"), + b"f": ord(b"\f"), + b"(": ord(b"("), + b")": ord(b")"), + b"/": ord(b"/"), + b"\\": ord(b"\\"), + b" ": ord(b" "), + b"%": ord(b"%"), + b"<": ord(b"<"), + b">": ord(b">"), + b"[": ord(b"["), + b"]": ord(b"]"), + b"#": ord(b"#"), + b"_": ord(b"_"), + b"&": ord(b"&"), + b"$": ord(b"$"), +} +__BACKSLASH_CODE__ = 92 def read_string_from_stream( @@ -54,30 +78,9 @@ def read_string_from_stream( break elif tok == b"\\": tok = stream.read(1) - escape_dict = { - b"n": b"\n", - b"r": b"\r", - b"t": b"\t", - b"b": b"\b", - b"f": b"\f", - b"c": rb"\c", - b"(": b"(", - b")": b")", - b"/": b"/", - b"\\": b"\\", - b" ": b" ", - b"%": b"%", - b"<": b"<", - b">": b">", - b"[": b"[", - b"]": b"]", - b"#": b"#", - b"_": b"_", - b"&": b"&", - b"$": b"$", - } try: - tok = escape_dict[tok] + txt.append(__ESPACE_DICT__[tok]) + continue except KeyError: if b"0" <= tok <= b"7": # "The number ddd may consist of one, two, or three @@ -85,6 +88,7 @@ def read_string_from_stream( # Three octal digits shall be used, with leading zeros # as needed, if the next character of the string is also # a digit." (PDF reference 7.3.4.2, p 16) + sav = stream.tell() - 1 for _ in range(2): ntok = stream.read(1) if b"0" <= ntok <= b"7": @@ -92,7 +96,13 @@ def read_string_from_stream( else: stream.seek(-1, 1) # ntok has to be analyzed break - tok = b_(chr(int(tok, base=8))) + i = int(tok, base=8) + if i > 255: + txt.append(__BACKSLASH_CODE__) + stream.seek(sav) + else: + txt.append(i) + continue elif tok in b"\n\r": # This case is hit when a backslash followed by a line # break occurs. If it's a multi-char EOL, consume the @@ -102,12 +112,13 @@ def read_string_from_stream( stream.seek(-1, 1) # Then don't add anything to the actual string, since this # line break was escaped: - tok = b"" + continue else: msg = f"Unexpected escaped string: {tok.decode('utf-8','ignore')}" logger_warning(msg, __name__) - txt.append(tok) - return create_string_object(b"".join(txt), forced_encoding) + txt.append(__BACKSLASH_CODE__) + txt.append(ord(tok)) + return create_string_object(bytes(txt), forced_encoding) def create_string_object( diff --git a/resources/Seige_of_Vicksburg_Sample_OCR-crazyones-merged.pdf b/resources/Seige_of_Vicksburg_Sample_OCR-crazyones-merged.pdf index 0e9633ac16c138eeaa90d3cf13e9f7cd6e2c006d..a53f28f0be432c38a1fff33672a2170eeb5f553f 100644 GIT binary patch delta 1135 zcmZuwJ5B>Z3>Bn7L8-XFHtGEtkL~mzB#PXGI^7lsK}(f3N1%gya3N-+5Rx~UZB|eA z_*(w%S~Mtv1*EGs-~I8J>sv{zoZ`BgRCDn2^;DWAA#+qhOsy z9_nuY2iNV#)9h}noP_mNYj;O2#kxAbfL+j&#EQMi$qgIN%*aCBkgS9v3tOTYrFns* zi?DXRqIyy{%_l0Z{g4pthD0$;#wiU+yHv)c!>vmue00SNBRm4^N)hR?Vv~jK{H4mw ziIFg?VZq^h6P&9s+GXO#zoHdee$2JJpxz|Hl_r{ zY(@&UHcToGu}TFAn%96`U{@PYmtn8OwFX4O-WCKhW)JHKw%|G;shLP7tsbKDqy5qZ aX;Lhg{J|DdMd<#DhbONh0xtWZ9J{UqF&k&I6)7fzU zDDp4D(!te>XPgE={aR6UiiW`%ZfqTeiIj@xjiPk$e@q9UWn5rLy=@eWjv@^6B7z!u zVL|uoObDgBOUxb4d9AVd3tS3uCsM>HdG}UKHtl4X<1DUS`o&!~`3?a}@xUY!MBL9r zrG8xWxg)aX>X2IbX&d=ezp1ic;>(#79o%KUaWwR|p2uxbk YRyD8Bjbb46Z8kjJvC(vTJUbbE19m{nQvd(} diff --git a/tests/test_cmap.py b/tests/test_cmap.py index 9dcfb252d..69f2931f6 100644 --- a/tests/test_cmap.py +++ b/tests/test_cmap.py @@ -1,13 +1,19 @@ """Test the pypdf_cmap module.""" from io import BytesIO +from pathlib import Path import pytest -from pypdf import PdfReader +from pypdf import PdfReader, PdfWriter from pypdf._cmap import build_char_map +from pypdf.generic import ArrayObject, NameObject, NullObject from . import get_data_from_url +TESTS_ROOT = Path(__file__).parent.resolve() +PROJECT_ROOT = TESTS_ROOT.parent +RESOURCE_ROOT = PROJECT_ROOT / "resources" + @pytest.mark.enable_socket() @pytest.mark.slow() @@ -206,3 +212,13 @@ def test_eten_b5(): """Issue #2356""" reader = PdfReader(BytesIO(get_data_from_url(name="iss2290.pdf"))) reader.pages[0].extract_text().startswith("1/7 \n富邦新終身壽險") + + +def test_null_missing_width(): + """For coverage of 2792""" + writer = PdfWriter(RESOURCE_ROOT / "crazyones.pdf") + page = writer.pages[0] + ft = page["/Resources"]["/Font"]["/F1"] + ft[NameObject("/Widths")] = ArrayObject() + ft["/FontDescriptor"][NameObject("/MissingWidth")] = NullObject() + page.extract_text() diff --git a/tests/test_page.py b/tests/test_page.py index cb7b6c723..72df648e4 100644 --- a/tests/test_page.py +++ b/tests/test_page.py @@ -1131,9 +1131,9 @@ def test_merge_page_resources_smoke_test(): # use these keys for some "operations", to validate renaming # (the operand name doesn't matter) contents1 = page1[NO("/Contents")] = ContentStream(None, None) - contents1.operations = [(ArrayObject(props1.keys()), "page1-contents")] + contents1.operations = [(ArrayObject(props1.keys()), b"page1-contents")] contents2 = page2[NO("/Contents")] = ContentStream(None, None) - contents2.operations = [(ArrayObject(props2.keys()), "page2-contents")] + contents2.operations = [(ArrayObject(props2.keys()), b"page2-contents")] expected_properties = { "/just1": "/just1-value", @@ -1438,3 +1438,12 @@ def test_negative_index(): src_abs = RESOURCE_ROOT / "git.pdf" reader = PdfReader(src_abs) assert reader.pages[0] == reader.pages[-1] + + +def test_get_contents_as_bytes(): + writer = PdfWriter(RESOURCE_ROOT / "crazyones.pdf") + co = writer.pages[0]["/Contents"][0] + expected = co.get_data() + assert writer.pages[0]._get_contents_as_bytes() == expected + writer.pages[0][NameObject("/Contents")] = writer.pages[0]["/Contents"][0] + assert writer.pages[0]._get_contents_as_bytes() == expected diff --git a/tests/test_utils.py b/tests/test_utils.py index 856bedd86..a4ddff883 100644 --- a/tests/test_utils.py +++ b/tests/test_utils.py @@ -109,22 +109,6 @@ def test_mark_location(): Path("pypdf_pdfLocation.txt").unlink() # cleanup -@pytest.mark.parametrize( - ("input_str", "expected"), - [ - ("foo", b"foo"), - ("😀", "😀".encode()), - ("‰", "‰".encode()), - ("▷", "▷".encode()), - ("世", "世".encode()), - # A multi-character string example with non-latin-1 characters: - ("😀😃", "😀😃".encode()), - ], -) -def test_b(input_str: str, expected: bytes): - assert pypdf._utils.b_(input_str) == expected - - def test_deprecate_no_replacement(): with pytest.warns(DeprecationWarning) as warn: pypdf._utils.deprecate_no_replacement("foo", removed_in="3.0.0") diff --git a/tests/test_workflows.py b/tests/test_workflows.py index 93bc0c9e5..4407b8fd5 100644 --- a/tests/test_workflows.py +++ b/tests/test_workflows.py @@ -391,11 +391,11 @@ def test_merge(tmp_path, url, name): { "/Author": "Unknown", "/CreationDate": "Thursday, May 06, 1999 3:56:54 PM", - "/Creator": "C:DEBÆł8", + "/Creator": r"C:\DEB\6338", "/Keywords": "", "/Producer": "Acrobat PDFWriter 3.02 for Windows", "/Subject": "", - "/Title": "C:DEBÆł8-6R.PDF", + "/Title": r"C:\DEB\6338-6R.PDF", }, ) ], diff --git a/tests/test_writer.py b/tests/test_writer.py index 9dfeffdd8..84d84d0db 100644 --- a/tests/test_writer.py +++ b/tests/test_writer.py @@ -1290,7 +1290,7 @@ def test_attachments(): to_add = [ ("foobar.txt", b"foobarcontent"), ("foobar2.txt", b"foobarcontent2"), - ("foobar2.txt", b"2nd_foobarcontent"), + ("foobar2.txt", "2nd_foobarcontent"), ] for name, content in to_add: writer.add_attachment(name, content) From a9758ae1736adc51cc9bdc120b11a6d451a17e74 Mon Sep 17 00:00:00 2001 From: pubpub-zz <4083478+pubpub-zz@users.noreply.github.com> Date: Mon, 12 Aug 2024 21:08:21 +0200 Subject: [PATCH 11/21] MAINT: Improve test coverage (#2796) --- tests/test_writer.py | 10 ++++++++++ tests/test_xmp.py | 31 ++++++++++++++++++++++++++++++- 2 files changed, 40 insertions(+), 1 deletion(-) diff --git a/tests/test_writer.py b/tests/test_writer.py index 84d84d0db..eea1c6a48 100644 --- a/tests/test_writer.py +++ b/tests/test_writer.py @@ -2188,6 +2188,10 @@ def test_replace_object(): reader._replace_object(reader.pages[0].indirect_reference, reader.pages[0]) pg = PageObject.create_blank_page(writer, 1000, 1000) reader._replace_object(reader.pages[0].indirect_reference, pg) + pg = PageObject.create_blank_page(None, 1000, 1000) + pg[NameObject("/Contents")] = writer.pages[0]["/Contents"] + writer._add_object(pg) + writer.add_page(pg) def test_mime_jupyter(): @@ -2300,3 +2304,9 @@ def test_matrix_entry_in_field_annots(): auto_regenerate=False, ) assert "/Matrix" in writer.pages[0]["/Annots"][5].get_object()["/AP"]["/N"] + + +def test_set_need_appearances_writer(): + """Minimal test for coverage""" + writer = PdfWriter() + writer.set_need_appearances_writer() diff --git a/tests/test_xmp.py b/tests/test_xmp.py index f864a9df9..6615b93c8 100644 --- a/tests/test_xmp.py +++ b/tests/test_xmp.py @@ -7,7 +7,7 @@ import pypdf.generic import pypdf.xmp -from pypdf import PdfReader +from pypdf import PdfReader, PdfWriter from pypdf.errors import PdfReadError from . import get_data_from_url @@ -42,6 +42,35 @@ def test_read_xmp_metadata_samples(src): } +def test_writer_xmp_metadata_samples(): + writer = PdfWriter(SAMPLE_ROOT / "020-xmp/output_with_metadata_pymupdf.pdf") + xmp = writer.xmp_metadata + assert xmp + assert xmp.dc_contributor == [] + assert xmp.dc_creator == ["John Doe"] + assert xmp.dc_source == "Martin Thoma" # attribute node + assert xmp.dc_description == {"x-default": "This is a text"} + assert xmp.dc_date == [datetime(1990, 4, 28, 0, 0)] + assert xmp.dc_title == {"x-default": "Sample PDF with XMP Metadata"} + assert xmp.custom_properties == { + "Style": "FooBarStyle", + "other": "worlds", + "⏰": "time", + } + co = pypdf.generic.ContentStream(None, None) + co.set_data( + xmp.stream.get_data().replace( + b'dc:source="Martin Thoma"', b'dc:source="Pubpub-Zz"' + ) + ) + writer.xmp_metadata = pypdf.xmp.XmpInformation(co) + b = BytesIO() + writer.write(b) + reader = PdfReader(b) + xmp2 = reader.xmp_metadata + assert xmp2.dc_source == "Pubpub-Zz" + + @pytest.mark.parametrize( ("src", "has_xmp"), [ From cf7fcfd568bb96bb2a3b978a0bd031a18e6d90b7 Mon Sep 17 00:00:00 2001 From: pubpub-zz <4083478+pubpub-zz@users.noreply.github.com> Date: Tue, 13 Aug 2024 11:09:11 +0200 Subject: [PATCH 12/21] ENH: Compress PDF files merging identical objects (#2795) Add compress_identical_objects(). Discovered in #2728. Closes #2794. Closes #2768. --- docs/user/file-size.md | 20 +- pypdf/_text_extraction/_layout_mode/_font.py | 22 +- pypdf/_writer.py | 246 ++++++++++--------- pypdf/generic/_base.py | 3 + tests/test_writer.py | 23 ++ 5 files changed, 177 insertions(+), 137 deletions(-) diff --git a/docs/user/file-size.md b/docs/user/file-size.md index 0ee72e37e..d47ddcc0e 100644 --- a/docs/user/file-size.md +++ b/docs/user/file-size.md @@ -9,23 +9,17 @@ Some PDF documents contain the same object multiple times. For example, if an image appears three times in a PDF it could be embedded three times. Or it can be embedded once and referenced twice. -This can be done by reading and writing the file: +When adding data to a PdfWriter, the data is copied while respecting the original format. +For example, if two pages include the same image which is duplicated in the source document, the object will be duplicated in the PdfWriter object. -```python -from pypdf import PdfReader, PdfWriter - -reader = PdfReader("big-old-file.pdf") -writer = PdfWriter() +Additionally, when you delete objects in a document, pypdf cannot easily identify whether the objects are used elsewhere or not or if the user wants to keep them in. When writing the PDF file, these objects will be hidden within (part of the file, but not displayed). -for page in reader.pages: - writer.add_page(page) +In order to reduce the file size, use a compression call: `writer.compress_identical_objects(remove_identicals=True, remove_orphans=True)` -if reader.metadata is not None: - writer.add_metadata(reader.metadata) +* `remove_identicals` enables/disables compression merging identical objects. +* `remove_orphans` enables/disables suppression of unused objects. -with open("smaller-new-file.pdf", "wb") as fp: - writer.write(fp) -``` +It is recommended to apply this process just before writing to the file/stream. It depends on the PDF how well this works, but we have seen an 86% file reduction (from 5.7 MB to 0.8 MB) within a real PDF. diff --git a/pypdf/_text_extraction/_layout_mode/_font.py b/pypdf/_text_extraction/_layout_mode/_font.py index 40655b1b2..1d9617d74 100644 --- a/pypdf/_text_extraction/_layout_mode/_font.py +++ b/pypdf/_text_extraction/_layout_mode/_font.py @@ -44,7 +44,7 @@ def __post_init__(self) -> None: self.font_dictionary["/DescendantFonts"] ): while isinstance(d_font, IndirectObject): - d_font = d_font.get_object() # type: ignore[assignment] + d_font = d_font.get_object() self.font_dictionary["/DescendantFonts"][d_font_idx] = d_font ord_map = { ord(_target): _surrogate @@ -75,7 +75,11 @@ def __post_init__(self) -> None: { ord_map[_cidx]: _width for _cidx, _width in zip( - range(cast(int, start_idx), cast(int, start_idx) + len(width_list), 1), + range( + cast(int, start_idx), + cast(int, start_idx) + len(width_list), + 1, + ), width_list, ) if _cidx in ord_map @@ -83,12 +87,20 @@ def __post_init__(self) -> None: ) skip_count = 1 # check for format (2): `int int int` - elif isinstance(w_next_entry, (int, float)) and isinstance(_w[idx + 2].get_object(), (int, float)): - start_idx, stop_idx, const_width = w_entry, w_next_entry, _w[idx + 2].get_object() + elif isinstance(w_next_entry, (int, float)) and isinstance( + _w[idx + 2].get_object(), (int, float) + ): + start_idx, stop_idx, const_width = ( + w_entry, + w_next_entry, + _w[idx + 2].get_object(), + ) self.width_map.update( { ord_map[_cidx]: const_width - for _cidx in range(cast(int, start_idx), cast(int, stop_idx + 1), 1) + for _cidx in range( + cast(int, start_idx), cast(int, stop_idx + 1), 1 + ) if _cidx in ord_map } ) diff --git a/pypdf/_writer.py b/pypdf/_writer.py index d73c00e3d..a72e2a23d 100644 --- a/pypdf/_writer.py +++ b/pypdf/_writer.py @@ -27,20 +27,19 @@ # ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE # POSSIBILITY OF SUCH DAMAGE. -import collections import decimal import enum import hashlib import re import uuid from io import BytesIO, FileIO, IOBase +from itertools import compress from pathlib import Path from types import TracebackType from typing import ( IO, Any, Callable, - Deque, Dict, Iterable, List, @@ -62,6 +61,7 @@ StrByteType, StreamType, _get_max_pdf_version_header, + deprecate, deprecate_with_replacement, logger_warning, ) @@ -156,12 +156,17 @@ def __init__( clone_from: Union[None, PdfReader, StrByteType, Path] = None, ) -> None: self._header = b"%PDF-1.3" - self._objects: List[PdfObject] = [] + self._objects: List[Optional[PdfObject]] = [] """The indirect objects in the PDF.""" - self._idnum_hash: Dict[bytes, IndirectObject] = {} - """Maps hash values of indirect objects to their IndirectObject instances.""" + """Maps hash values of indirect objects to the list of IndirectObjects. + This is used for compression. + """ + self._idnum_hash: Dict[bytes, Tuple[IndirectObject, List[IndirectObject]]] = {} + """List of already translated IDs. + dict[id(pdf)][(idnum, generation)] + """ self._id_translated: Dict[int, Dict[int, int]] = {} # The root of our page tree node. @@ -370,10 +375,13 @@ def get_object( indirect_reference: Union[int, IndirectObject], ) -> PdfObject: if isinstance(indirect_reference, int): - return self._objects[indirect_reference - 1] - if indirect_reference.pdf != self: + obj = self._objects[indirect_reference - 1] + elif indirect_reference.pdf != self: raise ValueError("pdf must be self") - return self._objects[indirect_reference.idnum - 1] + else: + obj = self._objects[indirect_reference.idnum - 1] + assert obj is not None # clarification for mypy + return obj def _replace_object( self, @@ -392,7 +400,9 @@ def _replace_object( obj = obj.clone(self) self._objects[indirect_reference - 1] = obj obj.indirect_reference = IndirectObject(indirect_reference, gen, self) - return self._objects[indirect_reference - 1] + + assert isinstance(obj, PdfObject) # clarification for mypy + return obj def _add_page( self, @@ -1242,14 +1252,13 @@ def write_stream(self, stream: StreamType) -> None: "It may not be written to correctly.", __name__, ) + # deprecated to be removed in pypdf 6.0.0 : + # if not self._root: + # self._root = self._add_object(self._root_object) + # self._sweep_indirect_references(self._root) - if not self._root: - self._root = self._add_object(self._root_object) - - self._sweep_indirect_references(self._root) - - object_positions = self._write_pdf_structure(stream) - xref_location = self._write_xref_table(stream, object_positions) + object_positions, free_objects = self._write_pdf_structure(stream) + xref_location = self._write_xref_table(stream, object_positions, free_objects) self._write_trailer(stream, xref_location) def write(self, stream: Union[Path, StrByteType]) -> Tuple[bool, IO[Any]]: @@ -1282,8 +1291,9 @@ def write(self, stream: Union[Path, StrByteType]) -> Tuple[bool, IO[Any]]: return my_file, stream - def _write_pdf_structure(self, stream: StreamType) -> List[int]: + def _write_pdf_structure(self, stream: StreamType) -> Tuple[List[int], List[int]]: object_positions = [] + free_objects = [] # will contain list of all free entries stream.write(self.pdf_header.encode() + b"\n") stream.write(b"%\xE2\xE3\xCF\xD3\n") @@ -1296,15 +1306,26 @@ def _write_pdf_structure(self, stream: StreamType) -> List[int]: obj = self._encryption.encrypt_object(obj, idnum, 0) obj.write_to_stream(stream) stream.write(b"\nendobj\n") - return object_positions - - def _write_xref_table(self, stream: StreamType, object_positions: List[int]) -> int: + else: + object_positions.append(-1) + free_objects.append(i + 1) + free_objects.append(0) # add 0 to loop in accordance with PDF spec + return object_positions, free_objects + + def _write_xref_table( + self, stream: StreamType, object_positions: List[int], free_objects: List[int] + ) -> int: xref_location = stream.tell() stream.write(b"xref\n") stream.write(f"0 {len(self._objects) + 1}\n".encode()) - stream.write(f"{0:0>10} {65535:0>5} f \n".encode()) + stream.write(f"{free_objects[0]:0>10} {65535:0>5} f \n".encode()) + free_idx = 1 for offset in object_positions: - stream.write(f"{offset:0>10} {0:0>5} n \n".encode()) + if offset > 0: + stream.write(f"{offset:0>10} {0:0>5} n \n".encode()) + else: + stream.write(f"{free_objects[free_idx]:0>10} {1:0>5} f \n".encode()) + free_idx += 1 return xref_location def _write_trailer(self, stream: StreamType, xref_location: int) -> None: @@ -1349,6 +1370,79 @@ def add_metadata(self, infos: Dict[str, Any]) -> None: assert isinstance(self._info, DictionaryObject) self._info.update(args) + def compress_identical_objects( + self, + remove_identicals: bool = True, + remove_orphans: bool = True, + ) -> None: + """ + Parse the PDF file and merge objects that have same hash. + This will make objects common to multiple pages. + Recommended to be used just before writing output. + + Args: + remove_identicals: Remove identical objects. + remove_orphans: Remove unreferenced objects. + """ + + def replace_in_obj( + obj: PdfObject, crossref: Dict[IndirectObject, IndirectObject] + ) -> None: + if isinstance(obj, DictionaryObject): + key_val = obj.items() + elif isinstance(obj, ArrayObject): + key_val = enumerate(obj) # type: ignore + else: + return + assert isinstance(obj, (DictionaryObject, ArrayObject)) + for k, v in key_val: + if isinstance(v, IndirectObject): + orphans[v.idnum - 1] = False + if v in crossref: + obj[k] = crossref[v] + else: + """the filtering on DictionaryObject and ArrayObject only + will be performed within replace_in_obj""" + replace_in_obj(v, crossref) + + # _idnum_hash :dict[hash]=(1st_ind_obj,[other_indir_objs,...]) + self._idnum_hash = {} + orphans = [True] * len(self._objects) + # look for similar objects + for idx, obj in enumerate(self._objects): + if obj is None: + continue + assert isinstance(obj.indirect_reference, IndirectObject) + h = obj.hash_value() + if remove_identicals and h in self._idnum_hash: + self._idnum_hash[h][1].append(obj.indirect_reference) + self._objects[idx] = None + else: + self._idnum_hash[h] = (obj.indirect_reference, []) + + # generate the dict converting others to 1st + cnv = {v[0]: v[1] for v in self._idnum_hash.values() if len(v[1]) > 0} + cnv_rev: Dict[IndirectObject, IndirectObject] = {} + for k, v in cnv.items(): + cnv_rev.update(zip(v, (k,) * len(v))) + + # replace reference to merged objects + for obj in self._objects: + if isinstance(obj, (DictionaryObject, ArrayObject)): + replace_in_obj(obj, cnv_rev) + + # remove orphans (if applicable) + orphans[self.root_object.indirect_reference.idnum - 1] = False # type: ignore + + orphans[self._info.indirect_reference.idnum - 1] = False # type: ignore + + try: + orphans[self._ID.indirect_reference.idnum - 1] = False # type: ignore + except AttributeError: + pass + for i in compress(range(len(self._objects)), orphans): + self._objects[i] = None + def _sweep_indirect_references( self, root: Union[ @@ -1363,7 +1457,7 @@ def _sweep_indirect_references( TextStringObject, NullObject, ], - ) -> None: + ) -> None: # deprecated """ Resolving any circular references to Page objects. @@ -1379,73 +1473,13 @@ def _sweep_indirect_references( Args: root: The root of the PDF object tree to sweep. """ - stack: Deque[ - Tuple[ - Any, - Optional[Any], - Any, - List[PdfObject], - ] - ] = collections.deque() - discovered = [] - parent = None - grant_parents: List[PdfObject] = [] - key_or_id = None - - # Start from root - stack.append((root, parent, key_or_id, grant_parents)) - - while len(stack): - data, parent, key_or_id, grant_parents = stack.pop() - - # Build stack for a processing depth-first - if isinstance(data, (ArrayObject, DictionaryObject)): - for key, value in data.items(): - stack.append( - ( - value, - data, - key, - grant_parents + [parent] if parent is not None else [], - ) - ) - elif isinstance(data, IndirectObject) and data.pdf != self: - data = self._resolve_indirect_object(data) - - if str(data) not in discovered: - discovered.append(str(data)) - stack.append((data.get_object(), None, None, [])) - - # Check if data has a parent and if it is a dict or - # an array update the value - if isinstance(parent, (DictionaryObject, ArrayObject)): - if isinstance(data, StreamObject): - # a dictionary value is a stream; streams must be indirect - # objects, so we need to change this value. - data = self._resolve_indirect_object(self._add_object(data)) - - update_hashes = [] - - # Data changed and thus the hash value changed - if parent[key_or_id] != data: - update_hashes = [parent.hash_value()] + [ - grant_parent.hash_value() for grant_parent in grant_parents - ] - parent[key_or_id] = data - - # Update old hash value to new hash value - for old_hash in update_hashes: - indirect_reference = self._idnum_hash.pop(old_hash, None) - - if indirect_reference is not None: - indirect_reference_obj = indirect_reference.get_object() - - if indirect_reference_obj is not None: - self._idnum_hash[ - indirect_reference_obj.hash_value() - ] = indirect_reference + deprecate( + "_sweep_indirect_references has been removed, please report to dev team if this warning is observed", + ) - def _resolve_indirect_object(self, data: IndirectObject) -> IndirectObject: + def _resolve_indirect_object( + self, data: IndirectObject + ) -> IndirectObject: # deprecated """ Resolves an indirect object to an indirect object in this PDF file. @@ -1470,36 +1504,10 @@ def _resolve_indirect_object(self, data: IndirectObject) -> IndirectObject: Raises: ValueError: If the input stream is closed. """ - if hasattr(data.pdf, "stream") and data.pdf.stream.closed: - raise ValueError(f"I/O operation on closed file: {data.pdf.stream.name}") - - if data.pdf == self: - return data - - # Get real object indirect object - real_obj = data.pdf.get_object(data) - - if real_obj is None: - logger_warning( - f"Unable to resolve [{data.__class__.__name__}: {data}], " - "returning NullObject instead", - __name__, - ) - real_obj = NullObject() - - hash_value = real_obj.hash_value() - - # Check if object is handled - if hash_value in self._idnum_hash: - return self._idnum_hash[hash_value] - - if data.pdf == self: - self._idnum_hash[hash_value] = IndirectObject(data.idnum, 0, self) - # This is new object in this pdf - else: - self._idnum_hash[hash_value] = self._add_object(real_obj) - - return self._idnum_hash[hash_value] + deprecate( + "_resolve_indirect_object has been removed, please report to dev team if this warning is observed", + ) + return IndirectObject(0, 0, self) def get_reference(self, obj: PdfObject) -> IndirectObject: idnum = self._objects.index(obj) + 1 diff --git a/pypdf/generic/_base.py b/pypdf/generic/_base.py index 309d389cc..9899cb48c 100644 --- a/pypdf/generic/_base.py +++ b/pypdf/generic/_base.py @@ -239,6 +239,9 @@ def __init__(self, idnum: int, generation: int, pdf: Any) -> None: # PdfReader self.generation = generation self.pdf = pdf + def __hash__(self) -> int: + return hash((self.idnum, self.generation, id(self.pdf))) + def clone( self, pdf_dest: PdfWriterProtocol, diff --git a/tests/test_writer.py b/tests/test_writer.py index eea1c6a48..49fe58538 100644 --- a/tests/test_writer.py +++ b/tests/test_writer.py @@ -2306,6 +2306,29 @@ def test_matrix_entry_in_field_annots(): assert "/Matrix" in writer.pages[0]["/Annots"][5].get_object()["/AP"]["/N"] +@pytest.mark.enable_socket() +def test_compress_identical_objects(): + """Cf #2728 and #2794""" + url = "https://github.com/user-attachments/files/16575458/tt2.pdf" + name = "iss2794.pdf" + in_bytes = BytesIO(get_data_from_url(url, name=name)) + writer = PdfWriter(in_bytes) + writer.compress_identical_objects(remove_orphans=False) + out1 = BytesIO() + writer.write(out1) + assert 0.5 * len(in_bytes.getvalue()) > len(out1.getvalue()) + writer.remove_page( + 1 + ) # page0 contains fields which keep reference to the deleted page + out2 = BytesIO() + writer.write(out2) + assert len(out1.getvalue()) - 100 < len(out2.getvalue()) + writer.compress_identical_objects(remove_identicals=False) + out3 = BytesIO() + writer.write(out3) + assert len(out2.getvalue()) > len(out3.getvalue()) + + def test_set_need_appearances_writer(): """Minimal test for coverage""" writer = PdfWriter() From 2eb565d914f1dea5c9024aa8fb5f2332dd36f7f2 Mon Sep 17 00:00:00 2001 From: pubpub-zz <4083478+pubpub-zz@users.noreply.github.com> Date: Tue, 13 Aug 2024 14:21:19 +0200 Subject: [PATCH 13/21] ROB: Fix extract_text() issues on damaged PDFs (#2760) Closes #2702. --- pypdf/_cmap.py | 2 +- tests/test_cmap.py | 18 +++++++++++++++++- 2 files changed, 18 insertions(+), 2 deletions(-) diff --git a/pypdf/_cmap.py b/pypdf/_cmap.py index d635724d2..035850a4a 100644 --- a/pypdf/_cmap.py +++ b/pypdf/_cmap.py @@ -258,7 +258,7 @@ def prepare_cm(ft: DictionaryObject) -> bytes: cm: bytes if isinstance(tu, StreamObject): cm = cast(DecodedStreamObject, ft["/ToUnicode"]).get_data() - elif isinstance(tu, str) and tu.startswith("/Identity"): + else: # if (tu is None) or cast(str, tu).startswith("/Identity"): # the full range 0000-FFFF will be processed cm = b"beginbfrange\n<0000> <0001> <0000>\nendbfrange" if isinstance(cm, str): diff --git a/tests/test_cmap.py b/tests/test_cmap.py index 69f2931f6..9ec55723f 100644 --- a/tests/test_cmap.py +++ b/tests/test_cmap.py @@ -6,7 +6,7 @@ from pypdf import PdfReader, PdfWriter from pypdf._cmap import build_char_map -from pypdf.generic import ArrayObject, NameObject, NullObject +from pypdf.generic import ArrayObject, IndirectObject, NameObject, NullObject from . import get_data_from_url @@ -214,6 +214,22 @@ def test_eten_b5(): reader.pages[0].extract_text().startswith("1/7 \n富邦新終身壽險") +@pytest.mark.enable_socket() +def test_missing_entries_in_cmap(): + """ + Issue #2702: this issue is observed on damaged pdfs + use of this file in test has been discarded as too slow/long + we will create the same error from crazyones + """ + pdf_path = RESOURCE_ROOT / "crazyones.pdf" + reader = PdfReader(pdf_path) + p = reader.pages[0] + p["/Resources"]["/Font"]["/F1"][NameObject("/ToUnicode")] = IndirectObject( + 99999999, 0, reader + ) + p.extract_text() + + def test_null_missing_width(): """For coverage of 2792""" writer = PdfWriter(RESOURCE_ROOT / "crazyones.pdf") From 5f4a5e5fadd96074f441f04ea30938ce2734aa9a Mon Sep 17 00:00:00 2001 From: pubpub-zz <4083478+pubpub-zz@users.noreply.github.com> Date: Tue, 13 Aug 2024 14:55:51 +0200 Subject: [PATCH 14/21] MAINT: fix mypy type output closes #2798 --- pypdf/_page.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/pypdf/_page.py b/pypdf/_page.py index 48cdeb149..f917a86ea 100644 --- a/pypdf/_page.py +++ b/pypdf/_page.py @@ -495,7 +495,7 @@ def _get_image( return self._get_image(ids, cast(DictionaryObject, xobjs[id[0]])) @property - def images(self) -> List[ImageFile]: + def images(self) -> _VirtualListImages: # noqa: F821 """ Read-only property emulating a list of images on a page. @@ -532,7 +532,7 @@ def images(self) -> List[ImageFile]: Inline images are extracted and named ~0~, ~1~, ..., with the indirect_reference set to None. """ - return _VirtualListImages(self._get_ids_image, self._get_image) # type: ignore + return _VirtualListImages(self._get_ids_image, self._get_image) def _translate_value_inlineimage(self, k: str, v: PdfObject) -> PdfObject: """Translate values used in inline image""" From c253f569b49de8b220f4a1ce7324c200564c30a2 Mon Sep 17 00:00:00 2001 From: pubpub-zz <4083478+pubpub-zz@users.noreply.github.com> Date: Tue, 13 Aug 2024 15:14:22 +0200 Subject: [PATCH 15/21] fix --- pypdf/_page.py | 116 ++++++++++++++++++++++++------------------------- 1 file changed, 58 insertions(+), 58 deletions(-) diff --git a/pypdf/_page.py b/pypdf/_page.py index f917a86ea..d013890b3 100644 --- a/pypdf/_page.py +++ b/pypdf/_page.py @@ -301,6 +301,63 @@ def apply_on( return list(pt1) if isinstance(pt, list) else pt1 +class _VirtualListImages(Sequence[ImageFile]): + def __init__( + self, + ids_function: Callable[[], List[Union[str, List[str]]]], + get_function: Callable[[Union[str, List[str], Tuple[str]]], ImageFile], + ) -> None: + self.ids_function = ids_function + self.get_function = get_function + self.current = -1 + + def __len__(self) -> int: + return len(self.ids_function()) + + def keys(self) -> List[Union[str, List[str]]]: + return self.ids_function() + + def items(self) -> List[Tuple[Union[str, List[str]], ImageFile]]: + return [(x, self[x]) for x in self.ids_function()] + + @overload + def __getitem__(self, index: Union[int, str, List[str]]) -> ImageFile: + ... + + @overload + def __getitem__(self, index: slice) -> Sequence[ImageFile]: + ... + + def __getitem__( + self, index: Union[int, slice, str, List[str], Tuple[str]] + ) -> Union[ImageFile, Sequence[ImageFile]]: + lst = self.ids_function() + if isinstance(index, slice): + indices = range(*index.indices(len(self))) + lst = [lst[x] for x in indices] + cls = type(self) + return cls((lambda: lst), self.get_function) + if isinstance(index, (str, list, tuple)): + return self.get_function(index) + if not isinstance(index, int): + raise TypeError("invalid sequence indices type") + len_self = len(lst) + if index < 0: + # support negative indexes + index = len_self + index + if index < 0 or index >= len_self: + raise IndexError("sequence index out of range") + return self.get_function(lst[index]) + + def __iter__(self) -> Iterator[ImageFile]: + for i in range(len(self)): + yield self[i] + + def __str__(self) -> str: + p = [f"Image_{i}={n}" for i, n in enumerate(self.ids_function())] + return f"[{', '.join(p)}]" + + class PageObject(DictionaryObject): """ PageObject represents a single page within a PDF file. @@ -495,7 +552,7 @@ def _get_image( return self._get_image(ids, cast(DictionaryObject, xobjs[id[0]])) @property - def images(self) -> _VirtualListImages: # noqa: F821 + def images(self) -> _VirtualListImages: """ Read-only property emulating a list of images on a page. @@ -2393,60 +2450,3 @@ def process_font(f: DictionaryObject) -> None: for a in cast(DictionaryObject, cast(DictionaryObject, obj["/AP"])["/N"]): _get_fonts_walk(cast(DictionaryObject, a), fnt, emb) return fnt, emb # return the sets for each page - - -class _VirtualListImages(Sequence[ImageFile]): - def __init__( - self, - ids_function: Callable[[], List[Union[str, List[str]]]], - get_function: Callable[[Union[str, List[str], Tuple[str]]], ImageFile], - ) -> None: - self.ids_function = ids_function - self.get_function = get_function - self.current = -1 - - def __len__(self) -> int: - return len(self.ids_function()) - - def keys(self) -> List[Union[str, List[str]]]: - return self.ids_function() - - def items(self) -> List[Tuple[Union[str, List[str]], ImageFile]]: - return [(x, self[x]) for x in self.ids_function()] - - @overload - def __getitem__(self, index: Union[int, str, List[str]]) -> ImageFile: - ... - - @overload - def __getitem__(self, index: slice) -> Sequence[ImageFile]: - ... - - def __getitem__( - self, index: Union[int, slice, str, List[str], Tuple[str]] - ) -> Union[ImageFile, Sequence[ImageFile]]: - lst = self.ids_function() - if isinstance(index, slice): - indices = range(*index.indices(len(self))) - lst = [lst[x] for x in indices] - cls = type(self) - return cls((lambda: lst), self.get_function) - if isinstance(index, (str, list, tuple)): - return self.get_function(index) - if not isinstance(index, int): - raise TypeError("invalid sequence indices type") - len_self = len(lst) - if index < 0: - # support negative indexes - index = len_self + index - if index < 0 or index >= len_self: - raise IndexError("sequence index out of range") - return self.get_function(lst[index]) - - def __iter__(self) -> Iterator[ImageFile]: - for i in range(len(self)): - yield self[i] - - def __str__(self) -> str: - p = [f"Image_{i}={n}" for i, n in enumerate(self.ids_function())] - return f"[{', '.join(p)}]" From d9fceeb1f0add9aedcea20318a2d3b48345832d8 Mon Sep 17 00:00:00 2001 From: pubpub-zz <4083478+pubpub-zz@users.noreply.github.com> Date: Tue, 13 Aug 2024 15:34:05 +0200 Subject: [PATCH 16/21] docs --- docs/modules/PageObject.rst | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/docs/modules/PageObject.rst b/docs/modules/PageObject.rst index 45e81b6ab..06530ce4f 100644 --- a/docs/modules/PageObject.rst +++ b/docs/modules/PageObject.rst @@ -6,6 +6,11 @@ The PageObject Class :undoc-members: :show-inheritance: +.. autoclass:: pypdf._page._VirtualListImages + :members: + :undoc-members: + :show-inheritance: + .. autoclass:: pypdf._utils.ImageFile :members: :undoc-members: From 0039d1affa3d70606c7b2f51250d895a47362e06 Mon Sep 17 00:00:00 2001 From: pubpub-zz <4083478+pubpub-zz@users.noreply.github.com> Date: Wed, 14 Aug 2024 08:21:56 +0200 Subject: [PATCH 17/21] make VirtualListImages public + doc reworded --- docs/modules/PageObject.rst | 2 +- pypdf/_page.py | 21 +++++++++++++-------- 2 files changed, 14 insertions(+), 9 deletions(-) diff --git a/docs/modules/PageObject.rst b/docs/modules/PageObject.rst index 06530ce4f..37d3b2fbe 100644 --- a/docs/modules/PageObject.rst +++ b/docs/modules/PageObject.rst @@ -6,7 +6,7 @@ The PageObject Class :undoc-members: :show-inheritance: -.. autoclass:: pypdf._page._VirtualListImages +.. autoclass:: pypdf._page.VirtualListImages :members: :undoc-members: :show-inheritance: diff --git a/pypdf/_page.py b/pypdf/_page.py index d013890b3..e879f8fd5 100644 --- a/pypdf/_page.py +++ b/pypdf/_page.py @@ -301,7 +301,13 @@ def apply_on( return list(pt1) if isinstance(pt, list) else pt1 -class _VirtualListImages(Sequence[ImageFile]): +class VirtualListImages(Sequence[ImageFile]): + """ + Provides access to images referenced within a page. + One copy only will be returned if used many times in the same page + see #pypdf._page.PageObject.images for more details + """ + def __init__( self, ids_function: Callable[[], List[Union[str, List[str]]]], @@ -552,7 +558,7 @@ def _get_image( return self._get_image(ids, cast(DictionaryObject, xobjs[id[0]])) @property - def images(self) -> _VirtualListImages: + def images(self) -> VirtualListImages: """ Read-only property emulating a list of images on a page. @@ -562,11 +568,10 @@ def images(self) -> _VirtualListImages: - An integer Examples: - reader.pages[0].images[0] # return fist image - reader.pages[0].images['/I0'] # return image '/I0' - # return image '/Image1' within '/TP1' Xobject/Form: - reader.pages[0].images['/TP1','/Image1'] - for img in reader.pages[0].images: # loop within all objects + `reader.pages[0].images[0]` # return fist image + `reader.pages[0].images['/I0']` # return image '/I0' + `reader.pages[0].images['/TP1','/Image1']` # return image '/Image1' within '/TP1' Xobject/Form + `for img in reader.pages[0].images:` # loops through all objects images.keys() and images.items() can be used. @@ -589,7 +594,7 @@ def images(self) -> _VirtualListImages: Inline images are extracted and named ~0~, ~1~, ..., with the indirect_reference set to None. """ - return _VirtualListImages(self._get_ids_image, self._get_image) + return VirtualListImages(self._get_ids_image, self._get_image) def _translate_value_inlineimage(self, k: str, v: PdfObject) -> PdfObject: """Translate values used in inline image""" From adeb5e9a0d4938455ac41dfdba9a09abb1593bce Mon Sep 17 00:00:00 2001 From: pubpub-zz <4083478+pubpub-zz@users.noreply.github.com> Date: Wed, 14 Aug 2024 11:16:31 +0200 Subject: [PATCH 18/21] for doc update --- docs/modules/PageObject.rst | 11 +-- pypdf/_page.py | 148 +++++++++++++++++++++++++++--------- pypdf/_utils.py | 79 ++++--------------- 3 files changed, 125 insertions(+), 113 deletions(-) diff --git a/docs/modules/PageObject.rst b/docs/modules/PageObject.rst index 37d3b2fbe..b4524b443 100644 --- a/docs/modules/PageObject.rst +++ b/docs/modules/PageObject.rst @@ -11,14 +11,7 @@ The PageObject Class :undoc-members: :show-inheritance: -.. autoclass:: pypdf._utils.ImageFile +.. autoclass:: pypdf._page.ImageFile :members: + :inherited-members: File :undoc-members: - :show-inheritance: - :exclude-members: IndirectObject - -.. autoclass:: pypdf._utils.File - :members: - :undoc-members: - :show-inheritance: - :exclude-members: IndirectObject diff --git a/pypdf/_page.py b/pypdf/_page.py index e879f8fd5..509855458 100644 --- a/pypdf/_page.py +++ b/pypdf/_page.py @@ -28,8 +28,11 @@ # POSSIBILITY OF SUCH DAMAGE. import math +from dataclasses import dataclass from decimal import Decimal +from io import BytesIO from pathlib import Path +from types import ModuleType from typing import ( Any, Callable, @@ -58,9 +61,8 @@ ) from ._utils import ( CompressedTransformationMatrix, - File, - ImageFile, TransformationMatrixType, + _human_readable_bytes, logger_warning, matrix_multiply, ) @@ -85,6 +87,12 @@ StreamObject, ) +Image: Optional[ModuleType] +try: + from PIL import Image +except ImportError: + Image = None # error will be raised only when using images + MERGE_CROP_BOX = "cropbox" # pypdf<=3.4.0 used 'trimbox' @@ -301,11 +309,102 @@ def apply_on( return list(pt1) if isinstance(pt, list) else pt1 +@dataclass +class ImageFile: + """ + Image within the PDF file. *This object is not designed to be built.* + + This object should not be modified except using :func:`ImageFile:replace` to replace the image with a new one. + """ + + name: str = "" + """ + FileName as identified within the PDF file + """ + + data: bytes = b"" + """ + data as bytes + """ + + image: Optional[Image.Image] = None # type: ignore + """ + data as PIL image; + """ + + indirect_reference: Optional[IndirectObject] = None + """ + Reference to the Object storing the stream + """ + + def replace(self, new_image: Any, **kwargs: Any) -> None: + """ + Replace the Image with a new PIL image. + + Args: + new_image (PIL.Image.Image): The new PIL image to replace the existing image. + **kwargs: Additional keyword arguments to pass to `Image.Image.save()`. + + Raises: + TypeError: If the image is inline or in a PdfReader. + TypeError: If the image does not belong to a PdfWriter. + TypeError: If `new_image` is not a PIL Image. + + Note: + This method replaces the existing image with a new image. + It is not allowed for inline images or images within a PdfReader. + The `kwargs` parameter allows passing additional parameters + to `Image.Image.save()`, such as quality. + """ + if Image is None: + raise ImportError( + "pillow is required to do image extraction. " + "It can be installed via 'pip install pypdf[image]'" + ) + + from ._reader import PdfReader + + # to prevent circular import + from .filters import _xobj_to_image + from .generic import DictionaryObject, PdfObject + + if self.indirect_reference is None: + raise TypeError("Can not update an inline image") + if not hasattr(self.indirect_reference.pdf, "_id_translated"): + raise TypeError("Can not update an image not belonging to a PdfWriter") + if not isinstance(new_image, Image.Image): + raise TypeError("new_image shall be a PIL Image") + b = BytesIO() + new_image.save(b, "PDF", **kwargs) + reader = PdfReader(b) + assert reader.pages[0].images[0].indirect_reference is not None + self.indirect_reference.pdf._objects[self.indirect_reference.idnum - 1] = ( + reader.pages[0].images[0].indirect_reference.get_object() + ) + cast( + PdfObject, self.indirect_reference.get_object() + ).indirect_reference = self.indirect_reference + # change the object attributes + extension, byte_stream, img = _xobj_to_image( + cast(DictionaryObject, self.indirect_reference.get_object()) + ) + assert extension is not None + self.name = self.name[: self.name.rfind(".")] + extension + self.data = byte_stream + self.image = img + + def __str__(self) -> str: + return f"{self.__class__.__name__}(name={self.name}, data: {_human_readable_bytes(len(self.data))})" + + def __repr__(self) -> str: + return self.__str__()[:-1] + f", hash: {hash(self.data)})" + + class VirtualListImages(Sequence[ImageFile]): """ Provides access to images referenced within a page. One copy only will be returned if used many times in the same page - see #pypdf._page.PageObject.images for more details + see :func:`PageObject.images` for more details """ def __init__( @@ -454,33 +553,6 @@ def create_blank_page( return page - @property - def _old_images(self) -> List[File]: # deprecated - """ - Get a list of all images of the page. - - This requires pillow. You can install it via 'pip install pypdf[image]'. - - For the moment, this does NOT include inline images. They will be added - in future. - """ - images_extracted: List[File] = [] - if RES.XOBJECT not in self[PG.RESOURCES]: # type: ignore - return images_extracted - - x_object = self[PG.RESOURCES][RES.XOBJECT].get_object() # type: ignore - for obj in x_object: - if x_object[obj][IA.SUBTYPE] == "/Image": - extension, byte_stream, img = _xobj_to_image(x_object[obj]) - if extension is not None: - filename = f"{obj[1:]}{extension}" - images_extracted.append(File(name=filename, data=byte_stream)) - images_extracted[-1].image = img - images_extracted[-1].indirect_reference = x_object[ - obj - ].indirect_reference - return images_extracted - def _get_ids_image( self, obj: Optional[DictionaryObject] = None, @@ -568,19 +640,19 @@ def images(self) -> VirtualListImages: - An integer Examples: - `reader.pages[0].images[0]` # return fist image - `reader.pages[0].images['/I0']` # return image '/I0' - `reader.pages[0].images['/TP1','/Image1']` # return image '/Image1' within '/TP1' Xobject/Form - `for img in reader.pages[0].images:` # loops through all objects + * `reader.pages[0].images[0]` # return fist image + * `reader.pages[0].images['/I0']` # return image '/I0' + * `reader.pages[0].images['/TP1','/Image1']` # return image '/Image1' within '/TP1' Xobject/Form + * `for img in reader.pages[0].images:` # loops through all objects images.keys() and images.items() can be used. The ImageFile has the following properties: - `.name` : name of the object - `.data` : bytes of the object - `.image` : PIL Image Object - `.indirect_reference` : object reference + * `.name` : name of the object + * `.data` : bytes of the object + * `.image` : PIL Image Object + * `.indirect_reference` : object reference and the following methods: `.replace(new_image: PIL.Image.Image, **kwargs)` : diff --git a/pypdf/_utils.py b/pypdf/_utils.py index 94d45cf6d..6caf86b27 100644 --- a/pypdf/_utils.py +++ b/pypdf/_utils.py @@ -36,7 +36,7 @@ import warnings from dataclasses import dataclass from datetime import datetime, timezone -from io import DEFAULT_BUFFER_SIZE, BytesIO +from io import DEFAULT_BUFFER_SIZE from os import SEEK_CUR from typing import ( IO, @@ -47,7 +47,6 @@ Pattern, Tuple, Union, - cast, overload, ) @@ -525,10 +524,18 @@ def getter(self, method): # type: ignore # noqa: ANN001, ANN202 class File: from .generic import IndirectObject - name: str - data: bytes - image: Optional[Any] = None # optional ; direct image access - indirect_reference: Optional[IndirectObject] = None # optional ; link to PdfObject + name: str = "" + """ + FileName as identified within the PDF file + """ + data: bytes = b"" + """ + data as bytes + """ + indirect_reference: Optional[IndirectObject] = None + """ + Reference to the Object storing the stream + """ def __str__(self) -> str: return f"{self.__class__.__name__}(name={self.name}, data: {_human_readable_bytes(len(self.data))})" @@ -537,66 +544,6 @@ def __repr__(self) -> str: return self.__str__()[:-1] + f", hash: {hash(self.data)})" -@dataclass -class ImageFile(File): - from .generic import IndirectObject - - image: Optional[Any] = None # optional ; direct PIL image access - indirect_reference: Optional[IndirectObject] = None # optional ; link to PdfObject - - def replace(self, new_image: Any, **kwargs: Any) -> None: - """ - Replace the Image with a new PIL image. - - Args: - new_image (PIL.Image.Image): The new PIL image to replace the existing image. - **kwargs: Additional keyword arguments to pass to `Image.Image.save()`. - - Raises: - TypeError: If the image is inline or in a PdfReader. - TypeError: If the image does not belong to a PdfWriter. - TypeError: If `new_image` is not a PIL Image. - - Note: - This method replaces the existing image with a new image. - It is not allowed for inline images or images within a PdfReader. - The `kwargs` parameter allows passing additional parameters - to `Image.Image.save()`, such as quality. - """ - from PIL import Image - - from ._reader import PdfReader - - # to prevent circular import - from .filters import _xobj_to_image - from .generic import DictionaryObject, PdfObject - - if self.indirect_reference is None: - raise TypeError("Can not update an inline image") - if not hasattr(self.indirect_reference.pdf, "_id_translated"): - raise TypeError("Can not update an image not belonging to a PdfWriter") - if not isinstance(new_image, Image.Image): - raise TypeError("new_image shall be a PIL Image") - b = BytesIO() - new_image.save(b, "PDF", **kwargs) - reader = PdfReader(b) - assert reader.pages[0].images[0].indirect_reference is not None - self.indirect_reference.pdf._objects[self.indirect_reference.idnum - 1] = ( - reader.pages[0].images[0].indirect_reference.get_object() - ) - cast( - PdfObject, self.indirect_reference.get_object() - ).indirect_reference = self.indirect_reference - # change the object attributes - extension, byte_stream, img = _xobj_to_image( - cast(DictionaryObject, self.indirect_reference.get_object()) - ) - assert extension is not None - self.name = self.name[: self.name.rfind(".")] + extension - self.data = byte_stream - self.image = img - - @functools.total_ordering class Version: COMPONENT_PATTERN = re.compile(r"^(\d+)(.*)$") From c7319a215f0df8cf1d466f971c9c9e623a2a2759 Mon Sep 17 00:00:00 2001 From: pubpub-zz <4083478+pubpub-zz@users.noreply.github.com> Date: Wed, 14 Aug 2024 11:52:09 +0200 Subject: [PATCH 19/21] fix --- pypdf/_page.py | 22 +++++++++++----------- 1 file changed, 11 insertions(+), 11 deletions(-) diff --git a/pypdf/_page.py b/pypdf/_page.py index 509855458..f959fc69c 100644 --- a/pypdf/_page.py +++ b/pypdf/_page.py @@ -32,7 +32,6 @@ from decimal import Decimal from io import BytesIO from pathlib import Path -from types import ModuleType from typing import ( Any, Callable, @@ -87,11 +86,12 @@ StreamObject, ) -Image: Optional[ModuleType] try: - from PIL import Image + from PIL.Image import Image + + pil_not_imported = False except ImportError: - Image = None # error will be raised only when using images + pil_not_imported = True # error will be raised only when using images MERGE_CROP_BOX = "cropbox" # pypdf<=3.4.0 used 'trimbox' @@ -314,7 +314,7 @@ class ImageFile: """ Image within the PDF file. *This object is not designed to be built.* - This object should not be modified except using :func:`ImageFile:replace` to replace the image with a new one. + This object should not be modified except using :func:`ImageFile.replace` to replace the image with a new one. """ name: str = "" @@ -327,7 +327,7 @@ class ImageFile: data as bytes """ - image: Optional[Image.Image] = None # type: ignore + image: Optional[Image] = None """ data as PIL image; """ @@ -337,13 +337,13 @@ class ImageFile: Reference to the Object storing the stream """ - def replace(self, new_image: Any, **kwargs: Any) -> None: + def replace(self, new_image: Image, **kwargs: Any) -> None: """ Replace the Image with a new PIL image. Args: new_image (PIL.Image.Image): The new PIL image to replace the existing image. - **kwargs: Additional keyword arguments to pass to `Image.Image.save()`. + **kwargs: Additional keyword arguments to pass to `Image.save()`. Raises: TypeError: If the image is inline or in a PdfReader. @@ -354,9 +354,9 @@ def replace(self, new_image: Any, **kwargs: Any) -> None: This method replaces the existing image with a new image. It is not allowed for inline images or images within a PdfReader. The `kwargs` parameter allows passing additional parameters - to `Image.Image.save()`, such as quality. + to `Image.save()`, such as quality. """ - if Image is None: + if pil_not_imported: raise ImportError( "pillow is required to do image extraction. " "It can be installed via 'pip install pypdf[image]'" @@ -372,7 +372,7 @@ def replace(self, new_image: Any, **kwargs: Any) -> None: raise TypeError("Can not update an inline image") if not hasattr(self.indirect_reference.pdf, "_id_translated"): raise TypeError("Can not update an image not belonging to a PdfWriter") - if not isinstance(new_image, Image.Image): + if not isinstance(new_image, Image): raise TypeError("new_image shall be a PIL Image") b = BytesIO() new_image.save(b, "PDF", **kwargs) From deb606d0c3d011e0a6fcf1c779db5441de9ff1fd Mon Sep 17 00:00:00 2001 From: pubpub-zz <4083478+pubpub-zz@users.noreply.github.com> Date: Wed, 14 Aug 2024 12:02:18 +0200 Subject: [PATCH 20/21] fix --- pypdf/_page.py | 1 + 1 file changed, 1 insertion(+) diff --git a/pypdf/_page.py b/pypdf/_page.py index f959fc69c..fe5c9db37 100644 --- a/pypdf/_page.py +++ b/pypdf/_page.py @@ -91,6 +91,7 @@ pil_not_imported = False except ImportError: + Image = object # type: ignore pil_not_imported = True # error will be raised only when using images MERGE_CROP_BOX = "cropbox" # pypdf<=3.4.0 used 'trimbox' From 8a4481be62ef9f44aa9a85f5046c4e434f6053ee Mon Sep 17 00:00:00 2001 From: pubpub-zz <4083478+pubpub-zz@users.noreply.github.com> Date: Wed, 14 Aug 2024 13:21:06 +0200 Subject: [PATCH 21/21] coverage --- tests/test_workflows.py | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/tests/test_workflows.py b/tests/test_workflows.py index 4407b8fd5..d7dc0a058 100644 --- a/tests/test_workflows.py +++ b/tests/test_workflows.py @@ -990,6 +990,13 @@ def test_replace_image(tmp_path): i.replace(reader.pages[0].images[0].image) assert exc.value.args[0] == "Can not update an inline image" + import pypdf + + pypdf._page.pil_not_imported = True + with pytest.raises(ImportError) as exc: + i.replace(reader.pages[0].images[0].image) + pypdf._page.pil_not_imported = False + @pytest.mark.enable_socket() def test_inline_images():