PyMuPDF 1.23.26__tar.gz → 1.24.1__tar.gz

{PyMuPDF-1.23.26 → PyMuPDF-1.24.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: PyMuPDF
-Version: 1.23.26
+Version: 1.24.1
 Summary: A high performance Python library for data extraction, analysis, conversion & manipulation of PDF (and other) documents.
 Description-Content-Type: text/markdown
 Author: Artifex

{PyMuPDF-1.23.26 → PyMuPDF-1.24.1}/READMErb.md

@@ -1,4 +1,4 @@
-# PyMuPDF 1.23.22
+# PyMuPDF 1.24.1
 
 This wheel contains [MuPDF](https://mupdf.readthedocs.io/) shared libraries for
 use by [PyMuPDF](https://pymupdf.readthedocs.io/).

{PyMuPDF-1.23.26 → PyMuPDF-1.24.1}/changes.txt

@@ -2,6 +2,66 @@ Change Log
 ==========
 
 
+**Changes in version 1.24.1 (2024-04-02)**
+
+* Fixed issues:
+
+  * **Fixed** `3278 <https://github.com/pymupdf/PyMuPDF/issues/3278>`_: apply_redactions moves some unredacted text
+  * **Fixed** `3301 <https://github.com/pymupdf/PyMuPDF/issues/3301>`_: Be more permissive when classifying links as kind LINK_URI
+  * **Fixed** `3306 <https://github.com/pymupdf/PyMuPDF/issues/3306>`_: Text containing capital 'ET' not appearing as annotation
+
+* Other:
+
+  * Use MuPDF-1.24.1.
+  * Support ObjStm Compression.
+    Methods `Document.save()`, `Document.ez_save()` and `Document.write()`
+    now support new parameters `use_objstm`, compression_effort` and
+    `preserve_metadata`.
+
+
+**Changes in version 1.24.0 (2024-03-21)**
+
+* Fixed issues:
+
+  * **Fixed** `3281 <https://github.com/pymupdf/PyMuPDF/issues/3281>`_: Preparing metadata (pyproject.toml) did not run successfully
+  * **Fixed** `3279 <https://github.com/pymupdf/PyMuPDF/issues/3279>`_: PyMuPDF no longer builds in Alpine Linux
+  * **Fixed** `3257 <https://github.com/pymupdf/PyMuPDF/issues/3257>`_: apply_redactions() deleting text outside of annoted box
+  * **Fixed** `3216 <https://github.com/pymupdf/PyMuPDF/issues/3216>`_: AttributeError: 'Annot' object has no attribute '__del__'
+  * **Fixed** `3207 <https://github.com/pymupdf/PyMuPDF/issues/3207>`_: get_drawings's items is missing line from h path operator
+  * **Fixed** `3201 <https://github.com/pymupdf/PyMuPDF/issues/3201>`_: Memory leaks when merging PDFs
+  * **Fixed** `3197 <https://github.com/pymupdf/PyMuPDF/issues/3197>`_: page.get_text() returns hexadecimal text for some characters
+  * **Fixed** `3196 <https://github.com/pymupdf/PyMuPDF/issues/3196>`_: Remove text not working in 1.23.25 version vs 1.20.2
+  * **Fixed** `3172 <https://github.com/pymupdf/PyMuPDF/issues/3172>`_: PDF's 45º lines dissapearing in png conversion
+  * **Fixed** `3135 <https://github.com/pymupdf/PyMuPDF/issues/3135>`_: Do not log warnings to stdout
+  * **Fixed** `3125 <https://github.com/pymupdf/PyMuPDF/issues/3125>`_: get_pixmap method stuck on one page and runs forever
+  * **Fixed** `2964 <https://github.com/pymupdf/PyMuPDF/issues/2964>`_: There is an issue with the image generated by the page.get_pixmap() function
+
+* Other:
+
+  * Use MuPDF-1.24.0.
+  * Add support for redacting vector graphics.
+  * Several fixes for table module
+    
+    * Add new method for outputting the table as a markdown string.
+    
+    * Address errors in computing the table header object:
+    
+      We now allow None as the cell value, because this will be resolved where
+      needed (e.g. in the pandas DataFrame).
+
+      We previously tried to enforce rect-like tuples in all header cell
+      bboxes, however this fails for tables with all-None columns.  This fix
+      enables this and constructs an empty string in the corresponding cell
+      string.
+
+      We now correctly include start / stop points of lines in the bbox of the
+      clustered graphic.  We previously joined the line's rectangle - which had
+      no effect because this is always empty.
+
+  * Improved exception text if we fail to open document.
+  * Fixed build with new libclang 18.
+  
+
 **Changes in version 1.23.26 (2024-02-29)**
 
 * Fixed issues:

PyMuPDF-1.24.1/pyproject.toml

@@ -0,0 +1,8 @@
+[build-system]
+# We define required packages in setup.py:get_requires_for_build_wheel().
+requires = []
+
+# See pep-517.
+#
+build-backend = "setup"
+backend-path = ["."]

{PyMuPDF-1.23.26 → PyMuPDF-1.24.1}/scripts/gh_release.py

@@ -259,18 +259,29 @@ def build( platform_=None, valgrind=False):
             log(f'Not running cibuildwheel because CIBW_ARCHS_MACOS is empty string.')
             return
     
-    def env_set(name, value, pass_=False):
-        assert isinstance( value, str)
-        if not name.startswith('CIBW'):
-            assert pass_, f'{name=} {value=}'
-        env_extra[ name] = value
-        if pass_ and platform.system() == 'Linux':
+    def env_pass(name):
+        '''
+        Adds `name` to CIBW_ENVIRONMENT_PASS_LINUX if required to be available
+        when building wheel with cibuildwheel.
+        '''
+        if platform.system() == 'Linux':
             v = env_extra.get('CIBW_ENVIRONMENT_PASS_LINUX', '')
             if v:
                 v += ' '
             v += name
             env_extra['CIBW_ENVIRONMENT_PASS_LINUX'] = v
+    
+    def env_set(name, value, pass_=False):
+        assert isinstance( value, str)
+        if not name.startswith('CIBW'):
+            assert pass_, f'{name=} {value=}'
+        env_extra[ name] = value
+        if pass_:
+            env_pass(name)
 
+    if os.environ.get('PYMUPDF_SETUP_LIBCLANG'):
+        env_pass('PYMUPDF_SETUP_LIBCLANG')
+    
     env_set('PYMUPDF_SETUP_IMPLEMENTATIONS', inputs_wheels_implementations, pass_=1)
     if inputs_skeleton:
         env_set('PYMUPDF_SETUP_SKELETON', inputs_skeleton, pass_=1)
@@ -640,7 +651,8 @@ def run(command, env_extra=None, check=1, timeout=None):
             Whether to raise exception if command fails.
         timeout:
             If not None, timeout in seconds; passed directory to
-            subprocess.run().
+            subprocess.run(). Note that on MacOS subprocess.run() seems to
+            leave processes running if timeout expires.
     '''
     env = None
     if env_extra:
@@ -662,15 +674,17 @@ def platform_tag():
         assert 0, f'Unrecognised: {platform.system()=}'
 
 
-# If this has changed, need to update
-# .github/workflows/*.yml.
-#
-test_packages = 'pytest fontTools psutil pymupdf-fonts flake8'
+test_packages = 'pytest fontTools pymupdf-fonts flake8 pylint'
 if platform.system() == 'Windows' and cpu_bits() == 32:
     # No pillow wheel available, and doesn't build easily.
     pass
 else:
     test_packages += ' pillow'
+if platform.system().startswith('MSYS_NT-'):
+    # psutil not available on msys2.
+    pass
+else:
+    test_packages += ' psutil'
 
 
 if __name__ == '__main__':

{PyMuPDF-1.23.26 → PyMuPDF-1.24.1}/scripts/sysinstall.py

@@ -206,11 +206,7 @@ def main():
     if pip == 'sudo':
         print('## Installing Python packages required for building MuPDF and PyMuPDF.')
         run(f'sudo pip install --upgrade pip')
-        names = (''
-                + test_py.get_pyproject_required(os.path.abspath(f'{__file__}/../../pyproject.toml'))
-                + ' '
-                + test_py.get_pyproject_required(os.path.abspath(f'{mupdf_dir}/pyproject.toml'))
-                )
+        names = test_py.wrap_get_requires_for_build_wheel(f'{__file__}/../..')
         run(f'sudo pip install {names}')
     
     print('## Build and install MuPDF.')
@@ -272,6 +268,7 @@ def main():
         # `python -m installer` fails to overwrite existing files.
         run(f'{sudo}rm -r {p}/site-packages/fitz || true')
         run(f'{sudo}rm -r {p}/site-packages/PyMuPDF-*.dist-info || true')
+        run(f'{sudo}rm -r {root_prefix}/bin/pymupdf || true')
         if pip == 'venv':
             run(f'{sudo}{venv_name}/bin/python -m installer --destdir {root} --prefix {prefix} {wheel}')
         else:
@@ -339,8 +336,13 @@ def main():
     command = ''
     if pip == 'venv':
         command += f'. {test_venv}/bin/activate &&'
-    command += f' LD_LIBRARY_PATH={root_prefix}/lib PYTHONPATH={pythonpath}'
-    command += f' pytest -k "not test_color_count and not test_3050" {pymupdf_dir}'
+    command += f' LD_LIBRARY_PATH={root_prefix}/lib PYTHONPATH={pythonpath} PATH=$PATH:{root_prefix}/bin'
+    run(f'ls -l {root_prefix}/bin/')
+    # 2024-03-20: Not sure whether/where `pymupdf` binary is installed, so we
+    # disable the test_cli* tests.
+    excluded_tests = 'test_color_count test_3050 test_cli test_cli_out test_pylint'.split()
+    excluded_tests = ' and not '.join(excluded_tests)
+    command += f' pytest -k "not {excluded_tests}" {pymupdf_dir}'
     run(command)
 
 

{PyMuPDF-1.23.26 → PyMuPDF-1.24.1}/scripts/test.py

@@ -13,8 +13,8 @@ Examples:
     ./PyMuPDF/scripts/test.py --mupdf 'git:https://git.ghostscript.com/mupdf.git' buildtest
         Build and test with internal checkout of mupdf master.
 
-    ./PyMuPDF/scripts/test.py --mupdf 'git:--branch 1.23.x https://github.com/ArtifexSoftware/mupdf.git' buildtest
-        Build and test using internal checkout of mupdf 1.23.x branch from Github.
+    ./PyMuPDF/scripts/test.py --mupdf 'git:--branch 1.24.x https://github.com/ArtifexSoftware/mupdf.git' buildtest
+        Build and test using internal checkout of mupdf 1.24.x branch from Github.
 
 Usage:
     scripts/test.py <options> <commands>
@@ -300,12 +300,11 @@ def build(
     print(f'{build_type=}')
     print(f'{build_isolation=}')
     
-    if platform.system() == 'OpenBSD':
-        # libclang not available on pypi.org, so we need to force use of system
-        # package py3-llvm with --no-build-isolation, manually installing other
-        # required packages.
-        if build_isolation is None:
-            build_isolation = False
+    if build_isolation is None:
+        # On OpenBSD libclang is not available on pypi.org, so we need to force
+        # use of system package py3-llvm with --no-build-isolation, manually
+        # installing other required packages.
+        build_isolation = False if platform.system() == 'OpenBSD' else True
     
     if build_isolation:
         # This is the default on non-OpenBSD.
@@ -314,12 +313,11 @@ def build(
         # Not using build isolation - i.e. pip will not be using its own clean
         # venv, so we need to explicitly install required packages.  Manually
         # install required packages from pyproject.toml.
-        names = get_pyproject_required()
-        if platform.system() == 'OpenBSD':
-            # libclang not available. We require system already has py3-llvm
-            # installed.
-            log(f'OpenBSD: libclang not available; assuming system package py3-llvm is installed.')
-            names = names.replace('libclang', '')
+        sys.path.insert(0, os.path.abspath(f'{__file__}/../..'))
+        import setup
+        names = setup.get_requires_for_build_wheel()
+        del sys.path[0]
+        names = ' '.join(names)
         if venv_quick:
             log(f'{venv_quick=}: Not installing packages with pip: {names}')
         else:
@@ -442,6 +440,27 @@ def get_pyproject_required(ppt=None):
         else:
             assert 0, f'Failed to find "requires" line in {ppt}'
 
+def wrap_get_requires_for_build_wheel(dir_):
+    '''
+    Returns space-separated list of required
+    packages. Looks at `dir_`/pyproject.toml and calls
+    `dir_`/setup.py:get_requires_for_build_wheel().
+    '''
+    dir_abs = os.path.abspath(dir_)
+    ret = list()
+    ppt = os.path.join(dir_abs, 'pyproject.toml')
+    if os.path.exists(ppt):
+        ret += get_pyproject_required(ppt)
+    if os.path.exists(os.path.join(dir_abs, 'setup.py')):
+        sys.path.insert(0, dir_abs)
+        try:
+            from setup import get_requires_for_build_wheel as foo
+            for i in foo():
+                ret.append(i)
+        finally:
+            del sys.path[0]
+    return ' '.join(ret)
+
 
 def log(text):
     gh_release.log(text, caller=1)

{PyMuPDF-1.23.26 → PyMuPDF-1.24.1}/setup.py

@@ -68,6 +68,9 @@ Environmental variables:
             Build wheel called `PyMuPDFb` containing only shared libraries
             that are not specific to a particular Python version - e.g.
             on Linux this will be `libmupdf.so` and `libmupdfcpp.so`.
+    
+    PYMUPDF_SETUP_LIBCLANG
+        For internal testing.
         
     PYMUPDF_SETUP_MUPDF_BUILD
         If set, overrides location of MuPDF when building PyMuPDF:
@@ -202,12 +205,6 @@ assert g_flavour in ('p', 'b', 'pb'), \
 
 g_root = os.path.abspath( f'{__file__}/..')
 
-# setuptools seems to require current directory to be PyMuPDF.
-#
-assert os.path.abspath( os.getcwd()) == g_root, \
-        f'Current directory must be the PyMuPDF directory'
-
-
 def _fs_remove(path):
     '''
     Removes file or directory, without raising exception if it doesn't exist.
@@ -384,7 +381,7 @@ def get_mupdf_internal(out, location=None, sha=None, local_tgz=None):
     log(f'get_mupdf_internal(): {out=} {location=} {sha=}')
     assert out in ('dir', 'tgz')
     if location is None:
-        location = 'https://mupdf.com/downloads/archive/mupdf-1.23.10-source.tar.gz'
+        location = 'https://mupdf.com/downloads/archive/mupdf-1.24.1-source.tar.gz'
         #location = 'git:--branch master https://github.com/ArtifexSoftware/mupdf.git'
     
     if location == '':
@@ -498,6 +495,7 @@ openbsd = sys.platform.startswith( 'openbsd')
 freebsd = sys.platform.startswith( 'freebsd')
 darwin = sys.platform.startswith( 'darwin')
 windows = platform.system() == 'Windows' or platform.system().startswith('CYGWIN')
+msys2 = platform.system().startswith('MSYS_NT-')
 pyodide = os.environ.get('OS') == 'pyodide'
 
 
@@ -779,7 +777,13 @@ def build_mupdf_unix( mupdf_local, env, build_type):
     if build_prefix_extra:
         build_prefix += f'{build_prefix_extra}-'
     build_prefix += 'shared-'
-    if os.environ.get('PYMUPDF_SETUP_MUPDF_TESSERACT') != '0':
+    if msys2:
+        # Error in mupdf/scripts/tesseract/endianness.h:
+        # #error "I don't know what architecture this is!"
+        log(f'msys2: building MuPDF without tesseract.')
+    elif os.environ.get('PYMUPDF_SETUP_MUPDF_TESSERACT') == '0':
+        log(f'PYMUPDF_SETUP_MUPDF_TESSERACT=0 so building mupdf without tesseract.')
+    else:
         build_prefix += 'tesseract-'
     unix_build_dir = f'{mupdf_local}/build/{build_prefix}{build_type}'
     # We need MuPDF's Python bindings, so we build MuPDF with
@@ -1094,8 +1098,8 @@ with open( f'{g_root}/READMErb.md', encoding='utf-8') as f:
 # We generate different wheels depending on g_flavour.
 #
 
-version = '1.23.26'
-version_b = '1.23.22'
+version = '1.24.1'
+version_b = '1.24.1'
 
 tag_python = None
 requires_dist = None,
@@ -1158,7 +1162,32 @@ p = pipcl.Package(
 build_wheel = p.build_wheel
 build_sdist = p.build_sdist
 
+def get_requires_for_build_wheel(config_settings=None):
+    '''
+    Adds to pyproject.toml:[build-system]:requires, allowing programmatic
+    control over what packages we require.
+    '''
+    ret = list()
+    ret.append('setuptools')
+    libclang = os.environ.get('PYMUPDF_SETUP_LIBCLANG')
+    if libclang:
+        print(f'Overriding to use {libclang=}.')
+        ret.append(libclang)
+    elif openbsd:
+        print(f'OpenBSD: libclang not available via pip; assuming `pkg_add py3-llvm`.')
+    elif darwin and platform.machine() == 'arm64':
+        print(f'MacOS/arm64: forcing use of libclang 16.0.6 because 18.1.1 known to fail with `clang.cindex.TranslationUnitLoadError: Error parsing translation unit.`')
+        ret.append('libclang==16.0.6')
+    else:
+        ret.append('libclang')
+    if msys2:
+        print(f'msys2: pip install of swig does not build; assuming `pacman -S swig`.')
+    elif openbsd:
+        print(f'OpenBSD: pip install of swig does not build; assuming `pkg_add swig`.')
+    else:
+        ret.append( 'swig')
+    return ret
+
 
-import sys
 if __name__ == '__main__':
     p.handle_argv(sys.argv)