Jules - misc updates for next release.

changes.txt

@@ -2,6 +2,24 @@ Change Log
 ==========
 
 
+**Changes in version 1.26.1 ()**
+
+* Use MuPDF-1.26.2.
+
+* Fixed issues:
+
+  * **Fixed** `4520 <https://github.com/pymupdf/PyMuPDF/issues/4520>`_: show_pdf_page does not like empty pages created by new_page
+  * **Fixed** `4524 <https://github.com/pymupdf/PyMuPDF/issues/4524>`_: fitz.get_text ignores 'pages' kwarg
+  * **Fixed** `4412 <https://github.com/pymupdf/PyMuPDF/issues/4412>`_: Regression? Spurious error? in insert_pdf in v1.25.4
+
+* Other:
+
+  * Partial fix for `4503 <https://github.com/pymupdf/PyMuPDF/issues/4503>`_: Undetected character styles
+  * New method `Document.rewrite_images()`, useful for reducing file size, changing image formats, or converting color spaces.
+  * `Page.get_text()`: restrict positional args to match docs.
+  * Removed bogus definition of class `Shape`.
+
+
 **Changes in version 1.26.0 (2025-05-22)**
 
 * Use MuPDF-1.26.1.

scripts/test.py

@@ -91,6 +91,7 @@
     --help
     -h
         Show help.
+
     -i <implementations>
         Set PyMuPDF implementations to test.
         <implementations> must contain only these individual characters:
@@ -109,7 +110,6 @@
         specified PyMuPDF will download its default mupdf .tgz.]
 
     -M 0|1
-
     --build-mupdf 0|1
         Whether to rebuild mupdf when we build PyMuPDF. Default is 1.
 
@@ -233,8 +233,6 @@
         Is prepended to command line args.
 '''
 
-import gh_release
-
 import glob
 import os
 import platform
@@ -253,6 +251,13 @@
 finally:
     del sys.path[0]
 
+try:
+    sys.path.insert(0, f'{pymupdf_dir_abs}/scripts')
+    import gh_release
+finally:
+    del sys.path[0]
+
+
 pymupdf_dir = pipcl.relpath(pymupdf_dir_abs)
 
 log = pipcl.log0
@@ -491,7 +496,7 @@ def main(argv):
             cibuildwheel(env_extra, cibw_name, cibw_pyodide)
 
         elif command.startswith('install.'):
-            name = command.lstrip('install.')
+            name = command[len('install.'):]
             run(f'pip install --force-reinstall {name}')
             have_installed = True
 

setup.py

@@ -575,6 +575,12 @@ def get_mupdf(path=None, sha=None):
 darwin = sys.platform.startswith( 'darwin')
 windows = platform.system() == 'Windows' or platform.system().startswith('CYGWIN')
 msys2 = platform.system().startswith('MSYS_NT-')
+
+if os.environ.get('PYODIDE') == '1':
+    if os.environ.get('OS') != 'pyodide':
+        log('PYODIDE=1, setting OS=pyodide.')
+        os.environ['OS'] = 'pyodide'
+
 pyodide = os.environ.get('OS') == 'pyodide'
 
 

src/__init__.py

@@ -19424,18 +19424,24 @@ def handle_args():
             arg = args[0]
             if isinstance( arg, (list, tuple)) and len( arg) == 2:
                 p1, p2 = arg
-                return *p1, *p2
+                ret = *p1, *p2
+                assert len(ret) == 4
+                return ret
             if isinstance( arg, (list, tuple)) and len( arg) == 3:
                 a, b, c = arg
                 a = make_tuple(a)
                 b = make_tuple(b)
                 c = make_tuple(c)
                 ret = *a, *b, *c
+                assert len(ret) == 4
                 return ret
-            arg = make_tuple( arg)
-            return arg
+            ret = make_tuple( arg)
+            assert len(ret) == 4, f'{arg=} {ret=}'
+            return ret
         elif len(args) == 2:
-            return get_xy( args[0]) + get_xy( args[1])
+            ret = get_xy( args[0]) + get_xy( args[1])
+            assert len(ret) == 4
+            return ret
         elif len(args) == 3:
             x0, y0 = get_xy( args[0])
             if (x0, y0) != (None, None):

src/utils.py

@@ -900,6 +900,7 @@ def get_image_rects(page: pymupdf.Page, name, transform=False) -> list:
 def get_text(
     page: pymupdf.Page,
     option: str = "text",
+    *,
     clip: rect_like = None,
     flags: OptInt = None,
     textpage: pymupdf.TextPage = None,

tests/test_general.py

@@ -19,6 +19,7 @@
 import sys
 import textwrap
 import time
+import util
 
 import gentle_compare
 
@@ -419,8 +420,8 @@ def test_2238():
         wt_expected += 'trying to repair broken xref\n'
         wt_expected += 'repairing PDF document'
         assert wt == wt_expected, f'{wt=}'
-    first_page = doc.load_page(0).get_text('text', pymupdf.INFINITE_RECT())
-    last_page = doc.load_page(-1).get_text('text', pymupdf.INFINITE_RECT())
+    first_page = doc.load_page(0).get_text('text', clip=pymupdf.INFINITE_RECT())
+    last_page = doc.load_page(-1).get_text('text', clip=pymupdf.INFINITE_RECT())
 
     print(f'first_page={first_page!r}')
     print(f'last_page={last_page!r}')
@@ -1710,9 +1711,14 @@ def test_3624():
         print(f'Saving to {path_png=}.')
         pixmap.save(path_png)
         rms = gentle_compare.pixmaps_rms(path_png_expected, path_png)
+        print(f'{rms=}')
         # We get small differences in sysinstall tests, where some thirdparty
         # libraries can differ.
-        assert rms < 1
+        if rms > 1:
+            pixmap_diff = gentle_compare.pixmaps_diff(path_png_expected, path_png)
+            path_png_diff = os.path.normpath(f'{__file__}/../../tests/test_3624_diff.png')
+            pixmap_diff.save(path_png_diff)
+            assert 0, f'{rms=}'
 
 
 def test_4043():
@@ -1874,3 +1880,17 @@ def show(items):
                 {'depth': 0, 'locked': 0, 'number': 7, 'on': 1, 'text': 'layer_7', 'type': 'checkbox'},
                 ]
 
+
+def test_4533():
+    if 1:
+        print(f'test_4533(): doing nothing because known to segv.')
+        return
+    path = util.download(
+            'https://github.com/user-attachments/files/20497146/NineData_user_manual_V3.0.5.pdf',
+            'test_4533.pdf',
+            size=16864501,
+            )
+    print(f'Opening {path=}.', flush=1)
+    with pymupdf.open(path) as document:
+        print(f'Have opened {path=}.', flush=1)
+        print(f'{len(document)=}', flush=1)

tests/test_textextract.py

@@ -823,3 +823,82 @@ def test_4363():
         print(f'Found:\n    {text!r}')
         assert 0
 
+
+def test_4546():
+    # This issue will not be fixed (in mupdf) because the test input is faulty.
+    #
+    path = os.path.normpath(f'{__file__}/../../tests/resources/test_4546.pdf')
+    with pymupdf.open(path) as document:
+        page = document[0]
+        text = page.get_text()[:200]
+
+    # We can't actually test with 1.23.5 because it uses `fitz.` not `pymupdf.`.
+    expected_1_23_5 = b'JOB No.: \nShipper (complete name and address) \xe5\x8f\x91\xe8\xb4\xa7\xe4\xba\xba(\xe5\x90\x8d\xe7\xa7\xb0\xe5\x8f\x8a\xe5\x9c\xb0\n\xe5\x9d\x80) \nSINORICH TRANSPORT LIMITED\nADD:7C,WEST BLDG.,ZHONGQU\nMANSION,211 ZHONGSHAN\nRD. SHANTOU,515041 CN\nTEL:0754-88570001 FAX:0754-88572709\nS/O No. '.decode()
+
+    # This output is different from expected_1_23_5.
+    expected_mupdf_1_26_1 = b'JOB No.: Shipper (complete name and address) \xe5\x8f\x91\xe8\xb4\xa7\xe4\xba\xba(\xe5\x90\x8d\xe7\xa7\xb0\xe5\x8f\x8a\xe5\x9c\xb0\xe5\x9d\x80)  Tel:                                  Fax: \n \nS/O No. \xe6\x89\x98\xe8\xbf\x90\xe5\x8d\x95\xe5\x8f\xb7\xe7\xa0\x81     \nSINORICH TRANSPORT LIMITED \nSHIPPING ORDER \n\xe6\x89\x98\xe8\xbf\x90\xe5\x8d\x95 \n \xe5\xb8\x82\xe5\x9c\xba\xe9\x83\xa8: \n88570009 \n88577019 \n88'.decode()
+
+    print(f'expected_1_23_5\n{textwrap.indent(expected_1_23_5, "    ")}')
+    print(f'expected_mupdf_1_26_1\n{textwrap.indent(expected_mupdf_1_26_1, "    ")}')
+
+    print(f'{pymupdf.version=}')
+    print(f'text is:\n{textwrap.indent(text, "    ")}')
+    print(f'{text=}')
+    print(f'{text.encode()=}')
+
+    if pymupdf.mupdf_version_tuple >= (1, 26, 1):
+        assert text == expected_mupdf_1_26_1
+    else:
+        print(f'No expected output for {pymupdf.mupdf_version_tuple=}')
+
+
+def test_4503():
+    # Check detection of strikeout text. Behaviour is improved with
+    # mupdf>=1.26.2, but not perfect.
+    #
+    path = os.path.normpath(f'{__file__}/../../tests/resources/test_4503.pdf')
+    span_0 = None
+    text_0 = None
+    print()
+    print(f'{pymupdf.mupdf_version_tuple=}')
+    with pymupdf.open(path) as document:
+        page = document[0]
+        # Specify TEXT_COLLECT_STYLES so we collect char_flags, which contains
+        # FZ_STEXT_STRIKEOUT etc.
+        #
+        text = page.get_text('rawdict', flags=pymupdf.TEXTFLAGS_RAWDICT | pymupdf.TEXT_COLLECT_STYLES)
+        for i, block in enumerate(text['blocks']):
+            print(f'block {i}:')
+            for j, line in enumerate(block['lines']):
+                print(f'    line {j}:')
+                for k, span in enumerate(line['spans']):
+                    text = ''
+                    for char in span['chars']:
+                        text += char['c']
+                    print(f'        span {k}: {span["flags"]=:#x} {span["char_flags"]=:#x}: {text!r}')
+                    if 'the right to request the state to review' in text:
+                        span_0 = span
+                        text_0 = text
+    assert span_0
+    #print(f'{span_0=}')
+    print(f'{span_0["flags"]=:#x}')
+    print(f'{span_0["char_flags"]=:#x}')
+    print(f'{text_0=}')
+    strikeout = span_0['char_flags'] & pymupdf.mupdf.FZ_STEXT_STRIKEOUT
+    print(f'{strikeout=}')
+
+    if pymupdf.mupdf_version_tuple >= (1, 26, 2):
+        # 2025-06-09: This is still incorrect - the span should include the
+        # following text 'and, if appropriate,'. It looks like following spans
+        # are:
+        #   strikeout=0: 'and, '
+        #   strikeout=1: 'if '
+        #   strikeout=0: 'appropri' 
+        #   strikeout=1: 'ate,'
+        #
+        assert strikeout, f'Expected bit 0 (FZ_STEXT_STRIKEOUT) to be set in {span_0["char_flags"]=:#x}.'
+        assert text_0 == 'the right to request the state to review '
+    else:
+        # Expecting the bug.
+        assert not strikeout, f'Expected bit 0 (FZ_STEXT_STRIKEOUT) to be unset in {span_0["char_flags"]=:#x}.'
+        assert text_0 == 'notice the right to request the state to review and, if appropriate,'