PyMuPDF 1.24.4__tar.gz → 1.24.5__tar.gz

{PyMuPDF-1.24.4 → PyMuPDF-1.24.5}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: PyMuPDF
-Version: 1.24.4
+Version: 1.24.5
 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-1.24.4 → PyMuPDF-1.24.5}/changes.txt

@@ -2,9 +2,26 @@ Change Log
 ==========
 
 
+**Changes in version 1.24.5 (2024-05-30)**
+
+* Fixed issues:
+
+  * **Fixed** `3479 <https://github.com/pymupdf/PyMuPDF/issues/3479>`_: regression: fill_textbox: IndexError: pop from empty list
+  * **Fixed** `3488 <https://github.com/pymupdf/PyMuPDF/issues/3488>`_: set_toc method error
+
+* Other:
+
+  * Some more fixes to use MuPDF floating formatting.
+  * Removed/disabled some unnecessary diagnostics.
+  * Fixed utils.do_links() crash.
+  * Experimental new functions `pymupdf.apply_pages()` and `pymupdf.get_text()`.
+  * Addresses wrong label generation for label styles "a" and "A".
+
+
 **Changes in version 1.24.4 (2024-05-16)**
 
   * **Fixed** `3418 <https://github.com/pymupdf/PyMuPDF/issues/3418>`_: Re-introduced bug, text align add_redact_annot
+  * **Fixed** `3472 <https://github.com/pymupdf/PyMuPDF/issues/3472>`_: insert_pdf gives SystemError
 
 * Other:
 

{PyMuPDF-1.24.4 → PyMuPDF-1.24.5}/pipcl.py

@@ -237,21 +237,21 @@ class Package:
         >>> assert len(so) == 1
         >>> so = so[0]
         >>> assert os.path.getmtime(so) > t0
-    
+
     Check `entry_points` causes creation of command `foo_cli` when we install
     from our wheel using pip. [As of 2024-02-24 using pipcl's CLI interface
     directly with `setup.py install` does not support entry points.]
-    
+
         >>> print('Creating venv.', file=sys.stderr)
         >>> _ = subprocess.run(
         ...         f'cd pipcl_test && {sys.executable} -m venv pylocal',
         ...         shell=1, check=1)
-        
+
         >>> print('Installing from wheel into venv using pip.', file=sys.stderr)
         >>> _ = subprocess.run(
         ...         f'. pipcl_test/pylocal/bin/activate && pip install pipcl_test/dist/*.whl',
         ...         shell=1, check=1)
-        
+
         >>> print('Running foo_cli.', file=sys.stderr)
         >>> _ = subprocess.run(
         ...         f'. pipcl_test/pylocal/bin/activate && foo_cli',
@@ -299,7 +299,7 @@ class Package:
             requires_external = None,
             project_url = None,
             provides_extra = None,
-            
+
             entry_points = None,
 
             root = None,
@@ -374,21 +374,21 @@ class Package:
             entry_points:
                 String or dict specifying *.dist-info/entry_points.txt, for
                 example:
-                
+
                     ```
                     [console_scripts]
                     foo_cli = foo.__main__:main
                     ```
-                
+
                 or:
-                
+
                     { 'console_scripts': [
                         'foo_cli = foo.__main__:main',
                         ],
                     }
-                
+
                 See: https://packaging.python.org/en/latest/specifications/entry-points/
-            
+
             root:
                 Root of package, defaults to current directory.
 
@@ -684,7 +684,7 @@ class Package:
             # Add <name>-<version>.dist-info/COPYING.
             if self.license:
                 add_str(self.license, f'{dist_info_dir}/COPYING')
-            
+
             # Add <name>-<version>.dist-info/entry_points.txt.
             entry_points_text = self._entry_points_text()
             if entry_points_text:
@@ -735,15 +735,15 @@ class Package:
         os.makedirs(sdist_directory, exist_ok=True)
         tarpath = f'{sdist_directory}/{prefix}.tar.gz'
         log2(f'Creating sdist: {tarpath}')
-        
+
         with tarfile.open(tarpath, 'w:gz') as tar:
-            
+
             names_in_tar = list()
             def check_name(name):
                 if name in names_in_tar:
                     raise Exception(f'Name specified twice: {name}')
                 names_in_tar.append(name)
-            
+
             def add(from_, name):
                 check_name(name)
                 if isinstance(from_, str):
@@ -757,7 +757,7 @@ class Package:
                     tar.addfile(ti, io.BytesIO(from_))
                 else:
                     assert 0
-        
+
             def add_string(text, name):
                 textb = text.encode('utf8')
                 return add(textb, name)
@@ -776,7 +776,7 @@ class Package:
                     if to_rel == 'pyproject.toml':
                         found_pyproject_toml = True
                     add(from_, to_rel)
-            
+
             if not found_pyproject_toml:
                 log0(f'Warning: no pyproject.toml specified.')
 
@@ -884,10 +884,10 @@ class Package:
             add_file( from_, to_abs2, to_rel)
 
         add_str( self._metainfo(), f'{root2}/{dist_info_dir}/METADATA', f'{dist_info_dir}/METADATA')
-        
+
         if self.license:
             add_str( self.license, f'{root2}/{dist_info_dir}/COPYING', f'{dist_info_dir}/COPYING')
-        
+
         entry_points_text = self._entry_points_text()
         if entry_points_text:
             add_str(
@@ -1284,7 +1284,7 @@ class Package:
         if isinstance(p, str):
             p = p, p
         assert isinstance(p, tuple) and len(p) == 2
-        
+
         from_, to_ = p
         assert isinstance(from_, (str, bytes))
         assert isinstance(to_, str)
@@ -1798,7 +1798,7 @@ def git_items( directory, submodules=False):
     return ret
 
 
-def run( command, capture=False, check=1):
+def run( command, capture=False, check=1, verbose=1):
     '''
     Runs a command using `subprocess.run()`.
 
@@ -1818,6 +1818,8 @@ def run( command, capture=False, check=1):
         check:
             If true we raise an exception on error; otherwise we include the
             command's returncode in our return value.
+        verbose:
+            If true we show the command.
     Returns:
         check capture   Return
         --------------------------
@@ -1828,7 +1830,8 @@ def run( command, capture=False, check=1):
     '''
     lines = _command_lines( command)
     nl = '\n'
-    log2( f'Running: {nl.join(lines)}')
+    if verbose:
+        log1( f'Running: {nl.join(lines)}')
     sep = ' ' if windows() else '\\\n'
     command2 = sep.join( lines)
     cp = subprocess.run(
@@ -1924,18 +1927,18 @@ class PythonFlags:
                                 stderr=subprocess.DEVNULL,
                                 check=0,
                                 ).returncode
-                        log1(f'{e=} from {pc!r}.')
+                        log2(f'{e=} from {pc!r}.')
                         if e == 0:
                             python_config = pc
                     assert python_config, f'Cannot find python-config'
                 else:
                     python_config = f'{python_exe}-config'
-            log1(f'Using {python_config=}.')
+            log2(f'Using {python_config=}.')
             try:
-                self.includes = run( f'{python_config} --includes', capture=1).strip()
+                self.includes = run( f'{python_config} --includes', capture=1, verbose=0).strip()
             except Exception as e:
                 raise Exception('We require python development tools to be installed.') from e
-            self.ldflags = run( f'{python_config} --ldflags', capture=1).strip()
+            self.ldflags = run( f'{python_config} --ldflags', capture=1, verbose=0).strip()
             if linux():
                 # It seems that with python-3.10 on Linux, we can get an
                 # incorrect -lcrypt flag that on some systems (e.g. WSL)
@@ -2100,7 +2103,7 @@ def run_if( command, out, *prerequisites):
     if not doit:
         out_mtime = _fs_mtime( out)
         if out_mtime == 0:
-            doit = 'File does not exist: {out!e}'
+            doit = f'File does not exist: {out!r}'
 
     cmd_path = f'{out}.cmd'
     if os.path.isfile( cmd_path):
@@ -2155,7 +2158,7 @@ def run_if( command, out, *prerequisites):
             os.remove( cmd_path)
         except Exception:
             pass
-        log2( f'Running command because: {doit}')
+        log1( f'Running command because: {doit}')
 
         run( command)
 
@@ -2164,7 +2167,7 @@ def run_if( command, out, *prerequisites):
             f.write( command)
         return True
     else:
-        log2( f'Not running command because up to date: {out!r}')
+        log1( f'Not running command because up to date: {out!r}')
 
     if 0:
         log2( f'out_mtime={time.ctime(out_mtime)} pre_mtime={time.ctime(pre_mtime)}.'
@@ -2361,7 +2364,7 @@ class _Record:
             log2(f'Adding {to_}')
 
     def add_file(self, from_, to_):
-        log2(f'Adding file: {os.path.relpath(from_)} => {to_}')
+        log1(f'Adding file: {os.path.relpath(from_)} => {to_}')
         with open(from_, 'rb') as f:
             content = f.read()
         self.add_content(content, to_, verbose=False)

{PyMuPDF-1.24.4 → PyMuPDF-1.24.5}/scripts/test.py

@@ -130,7 +130,7 @@ def main(argv):
     build_type = None
     build_mupdf = True
     gdb = False
-    test_pymupdf = True
+    test_fitz = True
     implementations = None
     test_names = list()
     venv = 2
@@ -157,7 +157,7 @@ def main(argv):
         elif arg == '-d':
             build_type = 'debug'
         elif arg == '-f':
-            test_pymupdf = int(next(args))
+            test_fitz = int(next(args))
         elif arg in ('-h', '--help'):
             show_help()
             return
@@ -226,7 +226,7 @@ def main(argv):
                 pytest_options=pytest_options,
                 timeout=timeout,
                 gdb=gdb,
-                test_pymupdf=test_pymupdf,
+                test_fitz=test_fitz,
                 )
     
     for command in commands:
@@ -359,7 +359,7 @@ def test(
         pytest_options=None,
         timeout=None,
         gdb=False,
-        test_pymupdf=True,
+        test_fitz=True,
         ):
     '''
     Args:
@@ -375,7 +375,7 @@ def test(
             See top-level option `-p`.
         gdb:
             See top-level option `--gdb`.
-        test_pymupdf:
+        test_fitz:
             See top-level option `-f`.
     '''
     pymupdf_dir_rel = gh_release.relpath(pymupdf_dir)
@@ -433,7 +433,7 @@ def test(
         for p in glob.glob(f'{pymupdf_dir_rel}/tests/test_*_fitz.py'):
             print(f'Removing {p=}')
             os.remove(p)
-        if test_pymupdf:
+        if test_fitz:
             # Create copies of each test file, modified to use `pymupdf`
             # instead of `fitz`.
             for p in glob.glob(f'{pymupdf_dir_rel}/tests/test_*.py'):

{PyMuPDF-1.24.4 → PyMuPDF-1.24.5}/setup.py

@@ -571,13 +571,6 @@ def build():
             build_type,
             )
     
-    for d in (
-            mupdf_build_dir,
-            f'{g_root}/src',
-            ):
-        if d:
-            run(f'ls -l {os.path.relpath(d)}', check=0)
-    
     # Generate lists of (from, to) items to return to pipcl. We put MuPDF
     # shared libraries in a separate list so that we can build specific wheels
     # as determined by g_flavour.
@@ -613,21 +606,26 @@ def build():
 
     if path_so_leaf_b:
         # Add rebased implementation files.
-        add( ret_p, f'{g_root}/src/fitz___init__.py', 'fitz/__init__.py')   # For `fitz` module alias.
-        add( ret_p, f'{g_root}/src/fitz_table.py', 'fitz/table.py')         # For `fitz` module alias.
-        add( ret_p, f'{g_root}/src/fitz_utils.py', 'fitz/utils.py')         # For `fitz` module alias.
         to_dir = 'pymupdf/'
         add( ret_p, f'{g_root}/src/__init__.py', to_dir)
         add( ret_p, f'{g_root}/src/__main__.py', to_dir)
         add( ret_p, f'{g_root}/src/pymupdf.py', to_dir)
         add( ret_p, f'{g_root}/src/table.py', to_dir)
         add( ret_p, f'{g_root}/src/utils.py', to_dir)
-        add( ret_p, f'{g_root}/src/extra.py', to_dir)
-        add( ret_p, f'{g_root}/src/{path_so_leaf_b}', to_dir)
+        add( ret_p, f'{g_root}/src/_apply_pages.py', to_dir)
+        add( ret_p, f'{g_root}/src/build/extra.py', to_dir)
+        add( ret_p, f'{g_root}/src/build/{path_so_leaf_b}', to_dir)
+        
+        # Add support for `fitz` backwards compatibility.
+        add( ret_p, f'{g_root}/src/fitz___init__.py', 'fitz/__init__.py')
+        add( ret_p, f'{g_root}/src/fitz_table.py', 'fitz/table.py')
+        add( ret_p, f'{g_root}/src/fitz_utils.py', 'fitz/utils.py')
         
         if mupdf_local:
+            # Add MuPDF Python API.
             add( ret_p, f'{mupdf_build_dir}/mupdf.py', to_dir)
             
+            # Add MuPDF shared libraries.
             if windows:
                 wp = pipcl.wdev.WindowsPython()
                 add( ret_p, f'{mupdf_build_dir}/_mupdf.pyd', to_dir)
@@ -659,8 +657,6 @@ def build():
     else:
         add( ret, f'{g_root}/README.md', '$dist-info/README.md')
     
-    for f, t in ret:
-        log( f'build(): {f} => {t}')
     return ret
 
 
@@ -963,7 +959,7 @@ def _build_extension_rebased( mupdf_local, mupdf_build_dir, build_type):
     path_so_leaf_b = pipcl.build_extension(
             name = 'extra',
             path_i = f'{g_root}/src/extra.i',
-            outdir = f'{g_root}/src',
+            outdir = f'{g_root}/src/build',
             includes = includes,
             defines = defines,
             libpaths = libpaths,
@@ -1117,7 +1113,7 @@ classifier = [
 # We generate different wheels depending on g_flavour.
 #
 
-version = '1.24.4'
+version = '1.24.5'
 version_b = '1.24.3'
 
 if os.path.exists(f'{g_root}/{g_pymupdfb_sdist_marker}'):

{PyMuPDF-1.24.4 → PyMuPDF-1.24.5}/src/__init__.py

@@ -20,6 +20,7 @@ import re
 import string
 import sys
 import tarfile
+import time
 import typing
 import warnings
 import weakref
@@ -61,6 +62,19 @@ def _set_stream(name, default):
 _g_out_log = _set_stream('PYMUPDF_LOG', sys.stdout)
 _g_out_message = _set_stream('PYMUPDF_MESSAGE', sys.stdout)
 
+_g_log_items = list()
+_g_log_items_active = False
+
+def _log_items():
+    return _g_log_items
+
+def _log_items_active(active):
+    global _g_log_items_active
+    _g_log_items_active = active
+        
+def _log_items_clear():
+    del _g_log_items[:]
+
 
 def log( text='', caller=1):
     '''
@@ -70,7 +84,10 @@ def log( text='', caller=1):
     filename    = os.path.relpath(frame_record.filename)
     line        = frame_record.lineno
     function    = frame_record.function
-    print( f'{filename}:{line}:{function}: {text}', file=_g_out_log)
+    text = f'{filename}:{line}:{function}: {text}'
+    if _g_log_items_active:
+        _g_log_items.append(text)
+    print(text, file=_g_out_log)
     _g_out_log.flush()
 
 
@@ -85,7 +102,7 @@ def message(text=''):
 def exception_info():
     import traceback
     log(f'exception_info:')
-    traceback.print_exc(file=_g_out_log)
+    log(traceback.format_exc())
 
 
 # PDF names must not contain these characters:
@@ -183,8 +200,8 @@ def _int_rc(text):
     return int(text)
 
 VersionFitz = mupdf.FZ_VERSION # MuPDF version.
-VersionBind = "1.24.4" # PyMuPDF version.
-VersionDate = "2024-05-16 00:00:01"
+VersionBind = "1.24.5" # PyMuPDF version.
+VersionDate = "2024-05-30 00:00:01"
 VersionDate2 = VersionDate.replace('-', '').replace(' ', '').replace(':', '')
 version = (VersionBind, VersionFitz, VersionDate2)
 pymupdf_version_tuple = tuple( [_int_rc(i) for i in VersionBind.split('.')])
@@ -8994,7 +9011,8 @@ class Page:
 
         # prefix with derotation matrix
         mat = mat0 * self.derotation_matrix
-        cmd = b"%g %g %g %g %g %g cm " % tuple(mat)
+        cmd = _format_g(tuple(mat)) + ' cm '
+        cmd = cmd.encode('utf8')
         _ = TOOLS._insert_contents(self, cmd, False)  # prepend to page contents
 
         # swap x- and y-coordinates
@@ -20936,6 +20954,197 @@ def vdist(dir, a, b):
     return mupdf.fz_abs(dx * dir.y + dy * dir.x)
 
 
+def apply_pages(
+        path,
+        pagefn,
+        *,
+        pagefn_args=(),
+        pagefn_kwargs=dict(),
+        initfn=None,
+        initfn_args=(),
+        initfn_kwargs=dict(),
+        pages=None,
+        method='single',
+        concurrency=None,
+        _stats=False,
+        ):
+    '''
+    Returns list of results from `pagefn()`, optionally using concurrency for
+    speed.
+    
+    Args:
+        path:
+            Path of document.
+        pagefn:
+            Function to call for each page; is passed (page, *pagefn_args,
+            **pagefn_kwargs). Return value is added to list that we return. If
+            `method` is not 'single', must be a top-level function - nested
+            functions don't work with concurrency.
+        pagefn_args
+        pagefn_kwargs:
+            Additional args to pass to `pagefn`. Must be picklable.
+        initfn:
+            If true, called once in each worker process; is passed
+            (*initfn_args, **initfn_kwargs).
+        initfn_args
+        initfn_kwargs:
+            Args to pass to initfn. Must be picklable.
+        pages:
+            List of page numbers to process, or None to include all pages.
+        method:
+            'single'
+                Do not use concurrency.
+            'mp'
+                Operate concurrently using Python's `multiprocessing` module.
+            'fork'
+                 Operate concurrently using custom implementation with
+                 `os.fork()`. Does not work on Windows.
+        concurrency:
+            Number of worker processes to use when operating concurrently. If
+            None, we use the number of available CPUs.
+        _stats:
+            Internal, may change or be removed. If true, we output simple
+            timing diagnostics.
+    
+    Note: We require a file path rather than a Document, because Document
+    instances do not work properly after a fork - internal file descriptor
+    offsets are shared between the parent and child processes.
+    '''
+    if _stats:
+        t0 = time.time()
+    
+    if method == 'single':
+        if initfn:
+            initfn(*initfn_args, **initfn_kwargs)
+        ret = list()
+        document = Document(path)
+        for page in document:
+            r = pagefn(page, *pagefn_args, **initfn_kwargs)
+            ret.append(r)
+    
+    else:
+        # Use concurrency.
+        #
+        from . import _apply_pages
+    
+        if pages is None:
+            if _stats:
+                t = time.time()
+            with Document(path) as document:
+                num_pages = len(document)
+                pages = list(range(num_pages))
+            if _stats:
+                t = time.time() - t
+                log(f'{t:.2f}s: count pages.')
+    
+        if _stats:
+            t = time.time()
+        
+        if method == 'mp':
+            ret = _apply_pages._multiprocessing(
+                    path,
+                    pages,
+                    pagefn,
+                    pagefn_args,
+                    pagefn_kwargs,
+                    initfn,
+                    initfn_args,
+                    initfn_kwargs,
+                    concurrency,
+                    _stats,
+                    )
+    
+        elif method == 'fork':
+            ret = _apply_pages._fork(
+                    path,
+                    pages,
+                    pagefn,
+                    pagefn_args,
+                    pagefn_kwargs,
+                    initfn,
+                    initfn_args,
+                    initfn_kwargs,
+                    concurrency,
+                    _stats,
+                    )
+        
+        else:
+            assert 0, f'Unrecognised {method=}.'
+        
+        if _stats:
+            t = time.time() - t
+            log(f'{t:.2f}s: work.')
+
+    if _stats:
+        t = time.time() - t0
+        log(f'{t:.2f}s: total.')
+    return ret
+
+
+def get_text(
+        path,
+        *,
+        pages=None,
+        method='single',
+        concurrency=None,
+        
+        option='text',
+        clip=None,
+        flags=None,
+        textpage=None,
+        sort=False,
+        delimiters=None,
+        
+        _stats=False,
+        ):
+    '''
+    Returns list of results from `Page.get_text()`, optionally using
+    concurrency for speed.
+    
+    Args:
+        path:
+            Path of document.
+        pages:
+            List of page numbers to process, or None to include all pages.
+        method:
+            'single'
+                Do not use concurrency.
+            'mp'
+                Operate concurrently using Python's `multiprocessing` module.
+            'fork'
+                 Operate concurrently using custom implementation with
+                 `os.fork`. Does not work on Windows.
+        concurrency:
+            Number of worker processes to use when operating concurrently. If
+            None, we use the number of available CPUs.
+        option
+        clip
+        flags
+        textpage
+        sort
+        delimiters:
+            Passed to internal calls to `Page.get_text()`.
+    '''
+    args_dict = dict(
+            option=option,
+            clip=clip,
+            flags=flags,
+            textpage=textpage,
+            sort=sort,
+            delimiters=delimiters,
+            )
+    
+    return apply_pages(
+            path,
+            Page.get_text,
+            pagefn_kwargs=args_dict,
+            pages=pages,
+            method=method,
+            concurrency=concurrency,
+            _stats=_stats,
+            )
+
+
 class TOOLS:
     '''
     We use @staticmethod to avoid the need to create an instance of this class.