# Sphinx-generated HTML documentation is not suitable for packaging; see
# https://bugzilla.redhat.com/show_bug.cgi?id=2006555 for discussion.
#
# We can generate PDF documentation as a substitute.
%bcond_without doc_pdf

Name:           python-pdfminer
Version:        20221105
Release:        %autorelease
Summary:        Tool for extracting information from PDF documents

# The entire source is MIT except:
#
# LicenseRef-Fedora-Public-Domain:
#   pdfminer/arcfour.py
#     - If this is a bundled library, its origin is unclear
#   pdfminer/ascii85.py
#     - If this is a bundled library, its origin is unclear
#
# APAFML:
#   pdfminer/fontmetrics.py
#     - Data extracted and converted from the AFM files:
#       https://www.ctan.org/tex-archive/fonts/adobe/afm/
#
# BSD-3-Clause:
#   pdfminer/cmap/*
#     - Both the original bundled data and the data generated from the
#       adobe-mappings-cmap package are BSD-3-Clause-licensed.
#
# Apache-2.0 AND MIT:
#   pdfminer/_saslprep.py
#     - Forked from from Apache-2.0 code by MongoDB, Inc.—originally
#       pymongo/saslprep.py in mongo-python-driver (python-pymongo), with
#       additional modifications in pyHanko (not yet packaged); see
#       docs/licenses/LICENSE.pyHanko.
#
# Adobe-Glyph:
#   pdfminer/glyphlist.py
#     - Contains both code under the base MIT license and data extracted and
#       converted from
#       https://partners.adobe.com/public/developer/en/opentype/glyphlist.txt
#       under the Adobe Glyph List License
License:        MIT AND LicenseRef-Fedora-Public-Domain AND APAFML AND BSD-3-Clause AND (Apache-2.0 AND MIT) AND Adobe-Glyph
URL:            https://github.com/pdfminer/pdfminer.six
# This has the samples/ directory stripped out. While upstream claims the
# sample PDFs are “freely distributable”, they have unclear or unspecified
# licenses, which makes them unsuitable for Fedora. This applies especially,
# but not exclusively, to the contents of samples/nonfree.
#
# Generated with ./get_source.sh %%{version}
Source0:        pdfminer.six-%{version}-filtered.tar.xz
# Script to generate Source0; see comments above.
Source1:        get_source.sh
# Man pages written by hand for Fedora in groff_man(7) format using the
# command’s --help output
Source2:        dumppdf.1
Source3:        pdf2txt.1

BuildArch:      noarch

BuildRequires:  python3-devel
BuildRequires:  make

# We use the Japan1, Korea1, GB1, and CNS1 CMaps:
BuildRequires:  adobe-mappings-cmap-devel >= 20190730

%if %{with doc_pdf}
BuildRequires:  python3dist(sphinx)
BuildRequires:  python3-sphinx-latex
BuildRequires:  latexmk
%endif

# We do not generate BR’s from the “dev” extra because it includes an exact
# version requirement on mypy (and we do not intend to do typechecking), and it
# pulls in nox and black. We just want to use plain pytest.
BuildRequires:  python3dist(pytest)

%global common_description %{expand:
Pdfminer.six is a community maintained fork of the original PDFMiner. It is a
tool for extracting information from PDF documents. It focuses on getting and
analyzing text data. Pdfminer.six extracts the text from a page directly from
the sourcecode of the PDF. It can also be used to get the exact location, font
or color of the text.

It is built in a modular way such that each component of pdfminer.six can be
replaced easily. You can implement your own interpreter or rendering device
that uses the power of pdfminer.six for other purposes than text analysis.

Check out the full documentation on Read the Docs
(https://pdfminersix.readthedocs.io/).

Features:

  • Written entirely in Python.
  • Parse, analyze, and convert PDF documents.
  • PDF-1.7 specification support. (well, almost).
  • CJK languages and vertical writing scripts support.
  • Various font types (Type1, TrueType, Type3, and CID) support.
  • Support for extracting images (JPG, JBIG2, Bitmaps).
  • Support for various compressions (ASCIIHexDecode, ASCII85Decode, LZWDecode,
    FlateDecode, RunLengthDecode, CCITTFaxDecode)
  • Support for RC4 and AES encryption.
  • Support for AcroForm interactive form extraction.
  • Table of contents extraction.
  • Tagged contents extraction.
  • Automatic layout analysis.}

%description %{common_description}


%package -n     python3-pdfminer
Summary:        %{summary}

# The import name is pdfminer. The upstream project name (as specified in
# setup.py) is pdfminer.six, which results in a canonical project name of
# pdfminer-six.
%py_provides python3-pdfminer-six

# One file, pdfminer/_saslprep.py, is forked from from ASL 2.0 code by MongoDB,
# Inc.—originally pymongo/saslprep.py in mongo-python-driver
# (python-pymongo)—with additional modifications in pyHanko (not yet packaged),
# where it is pyhanko/pdf_utils/_saslprep.py.
#
# Since this is a fork of the python-pymongo module, and since the fork is not
# part of pyHanko’s public API, there is no possibility of using an unbundled
# version.
#
# The version history of the fork is not clear. We add unversioned virtual
# Provides for both libraries of origin.
Provides:       bundled(python3dist(pymongo))
Provides:       bundled(python3dist(pyhanko))

%description -n python3-pdfminer %{common_description}


%package doc
Summary:        Documentation for pdfminer
# See the base package License field for non-MIT sources; it appears that none
# of these contribute to the documentation.
License:        MIT

%description doc %{common_description}


%pyproject_extras_subpkg -n python3-pdfminer image


%prep
%autosetup -n pdfminer.six-%{version}

# Unbundle cmap data; it will be replaced in %%build.
rm -vf cmaprsrc/* pdfminer/cmap/*

# Remove shebang line in non-script source
sed -r -i '1{/^#!/d}' pdfminer/psparser.py
# Fix unversioned Python shebangs
%py3_shebang_fix tools

# Imitate the “publish” GitHub action, which sets the version metadata from the
# git tag when publishing to PyPI. See .github/workflows/actions.yml.
sed -r -i 's/__VERSION__/%{version}/g' pdfminer/__init__.py


%generate_buildrequires
%pyproject_buildrequires -x %{?with_doc_pdf:docs,}image


%build
# Symlink the unbundled CMap resources and convert to the pickled format.
for cmap in Japan1 Korea1 GB1 CNS1
do
  ln -s "%{adobe_mappings_rootpath}/${cmap}/cid2code.txt" \
      "cmaprsrc/cid2code_Adobe_${cmap}.txt"
done
%make_build cmap PYTHON='%{python3}'

%pyproject_wheel

%if %{with doc_pdf}
PYTHONPATH="${PWD}" %make_build -C docs latex SPHINXOPTS='%{?_smp_mflags}'
%make_build -C docs/build/latex LATEXMKOPTS='-quiet'
%endif


%install
%pyproject_install
%pyproject_save_files pdfminer

install -t '%{buildroot}%{_mandir}/man1' -D -p -m 0644 \
    '%{SOURCE2}' '%{SOURCE3}'
# Also, ship symlinks of the scripts without the .py extension.
for script in pdf2txt dumppdf
do
  ln -sf "${script}.py" "%{buildroot}%{_bindir}/${script}"
done


%check
# Skipped tests (and ignored files) are those that require the sample PDFs,
# which are not included in our version of the source tarball.
k="${k-}${k+ and }not TestColorSpace"
k="${k-}${k+ and }not TestDumpImages"
k="${k-}${k+ and }not TestDumpPDF"
k="${k-}${k+ and }not TestExtractPages"
k="${k-}${k+ and }not TestExtractText"
k="${k-}${k+ and }not TestOpenFilename"
k="${k-}${k+ and }not TestPdf2Txt"
k="${k-}${k+ and }not TestPdfDocument"
k="${k-}${k+ and }not TestPdfPage"
k="${k-}${k+ and }not test_font_size"
k="${k-}${k+ and }not test_paint_path_quadrilaterals"
k="${k-}${k+ and }not test_pdf_with_empty_characters_horizontal"
k="${k-}${k+ and }not test_pdf_with_empty_characters_vertical"

%pytest -k "${k-}" \
    --ignore='tests/test_tools_dumppdf.py' \
    --ignore='tests/test_tools_pdf2txt.py'


%files -n python3-pdfminer -f %{pyproject_files}
%license LICENSE docs/licenses/LICENSE.pyHanko
%{_bindir}/pdf2txt
%{_bindir}/pdf2txt.py
%{_mandir}/man1/pdf2txt.1*
%{_bindir}/dumppdf
%{_bindir}/dumppdf.py
%{_mandir}/man1/dumppdf.1*


%files doc
%license LICENSE
%doc CHANGELOG.md
%doc CONTRIBUTING.md
%doc README.md
%if %{with doc_pdf}
%doc docs/build/latex/pdfminersix.pdf
%endif


%changelog
%autochangelog