PyMuPDF 1.25.4__tar.gz → 1.25.5__tar.gz

{pymupdf-1.25.4 → pymupdf-1.25.5}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: PyMuPDF
-Version: 1.25.4
+Version: 1.25.5
 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-1.25.4 → pymupdf-1.25.5}/changes.txt

@@ -2,8 +2,27 @@ Change Log
 ==========
 
 
+**Changes in version 1.25.5 (2025-03-31)**
+
+* Fixed issues:
+
+  * **Fixed** `4372 <https://github.com/pymupdf/PyMuPDF/issues/4372>`_: Text insertion fails due to missing /Resources object
+  * **Fixed** `4400 <https://github.com/pymupdf/PyMuPDF/issues/4400>`_: Infinite loop in fill_textbox
+  * **Fixed** `4403 <https://github.com/pymupdf/PyMuPDF/issues/4403>`_: Unable to get_text() - layer/clip nesting too deep
+  * **Fixed** `4415 <https://github.com/pymupdf/PyMuPDF/issues/4415>`_: PDF page is mirrored, origin is at bottom-left
+
+* Other:
+
+  * Use MuPDF-1.25.6.
+  * Fixed MuPDF SEGV on MacOS with particular fonts.
+  * Fixed `Annot.get_textpage()`'s `clip` arg.
+  * Fixed Python-3.14 (pre-release) build error.
+
+
 **Changes in version 1.25.4 (2025-03-14)**
 
+* Use MuPDF-1.25.5.
+
 * Fixed issues:
 
   * **Fixed** `4079 <https://github.com/pymupdf/PyMuPDF/issues/4079>`_: Unexpected result for apply_redactions()

{pymupdf-1.25.4 → pymupdf-1.25.5}/scripts/test.py

@@ -94,6 +94,8 @@ Options:
         Whether to rebuild mupdf when we build PyMuPDF. Default is 1.
     --gdb 0|1
         Run tests under gdb. Requires user interaction.
+    --pytest-prefix <command>
+        Use specified prefix when running pytest. E.g. `gdb --args`.
     --pybind 0|1
         Experimental, for investigating
         https://github.com/pymupdf/PyMuPDF/issues/3869. Runs run basic code
@@ -185,6 +187,7 @@ def main(argv):
     implementations = 'r'
     test_names = list()
     venv = 2
+    pytest_prefix = None
     pybind = False
     pytest_options = None
     timeout = None
@@ -222,7 +225,7 @@ def main(argv):
             implementations = next(args)
         elif arg in ('--mupdf', '-m'):
             mupdf = next(args)
-            if not mupdf.startswith('git:') and mupdf != '-':
+            if not mupdf.startswith('git:') and '://' not in mupdf and mupdf != '-':
                 assert os.path.isdir(mupdf), f'Not a directory: {mupdf=}.'
                 mupdf = os.path.abspath(mupdf)
             os.environ['PYMUPDF_SETUP_MUPDF_BUILD'] = mupdf
@@ -236,6 +239,8 @@ def main(argv):
             value = next(args)
             assert value in ('0', '1'), f'`-s` must be followed by `0` or `1`, not {value=}.'
             os.environ['PYMUPDF_SETUP_PY_LIMITED_API'] = value
+        elif arg == '--pytest-prefix':
+            pytest_prefix = next(args)
         elif arg == '--pybind':
             pybind = int(next(args))
         elif arg == '--system-site-packages':
@@ -312,6 +317,7 @@ def main(argv):
                 pytest_options=pytest_options,
                 timeout=timeout,
                 gdb=gdb,
+                pytest_prefix=pytest_prefix,
                 test_fitz=test_fitz,
                 pytest_k=pytest_k,
                 pybind=pybind,
@@ -603,6 +609,7 @@ def test(
         pytest_options=None,
         timeout=None,
         gdb=False,
+        pytest_prefix=None,
         test_fitz=True,
         pytest_k=None,
         pybind=False,
@@ -624,6 +631,8 @@ def test(
             See top-level option `-p`.
         gdb:
             See top-level option `--gdb`.
+        pytest_prefix:
+            See top-level option `--pytest-prefix`.
         test_fitz:
             See top-level option `-f`.
     '''
@@ -732,6 +741,8 @@ def test(
                     )
         elif gdb:
             command = f'{python} {pymupdf_dir_rel}/tests/run_compound.py{run_compound_args} gdb --args {python} -m pytest {pytest_options} {pytest_arg}'
+        elif pytest_prefix:
+            command = f'{python} {pymupdf_dir_rel}/tests/run_compound.py{run_compound_args} {pytest_prefix} {python} -m pytest {pytest_options} {pytest_arg}'
         elif platform.system() == 'Windows':
             # `python -m pytest` doesn't seem to work.
            command = f'{python} {pymupdf_dir_rel}/tests/run_compound.py{run_compound_args} pytest {pytest_options} {pytest_arg}'

{pymupdf-1.25.4 → pymupdf-1.25.5}/setup.py

@@ -1242,7 +1242,7 @@ classifier = [
 #
 
 # PyMuPDF version.
-version_p = '1.25.4'
+version_p = '1.25.5'
 
 # PyMuPDFb version. This is the PyMuPDF version whose PyMuPDFb wheels we will
 # (re)use if generating separate PyMuPDFb wheels. Though as of PyMuPDF-1.24.11
@@ -1250,7 +1250,7 @@ version_p = '1.25.4'
 #
 version_b = '1.25.3'
 
-version_mupdf = '1.25.5'
+version_mupdf = '1.25.6'
 
 if os.path.exists(f'{g_root}/{g_pymupdfb_sdist_marker}'):
     

{pymupdf-1.25.4 → pymupdf-1.25.5}/src/__init__.py

@@ -373,9 +373,9 @@ def _int_rc(text):
 
 # Basic version information.
 #
-pymupdf_version = "1.25.4"
+pymupdf_version = "1.25.5"
 mupdf_version = mupdf.FZ_VERSION
-pymupdf_date = "2025-03-14 00:00:01"
+pymupdf_date = "2025-03-31 00:00:01"
 
 # Versions as tuples; useful when comparing versions.
 #
@@ -419,8 +419,15 @@ def _format_g(value, *, fmt='%g'):
         
 format_g = _format_g
 
+# ByteString is gone from typing in 3.14.
+# collections.abc.Buffer available from 3.12 only
+try:
+    ByteString = typing.ByteString
+except AttributeError:
+    ByteString = bytes | bytearray | memoryview
+
 # Names required by class method typing annotations.
-OptBytes = typing.Optional[typing.ByteString]
+OptBytes = typing.Optional[ByteString]
 OptDict = typing.Optional[dict]
 OptFloat = typing.Optional[float]
 OptInt = typing.Union[int, None]
@@ -1022,8 +1029,11 @@ class Annot:
     def get_textpage(self, clip=None, flags=0):
         """Make annotation TextPage."""
         CheckParent(self)
-        options = mupdf.FzStextOptions()
-        options.flags = flags
+        options = mupdf.FzStextOptions(flags)
+        if clip:
+            clip2 = JM_rect_from_py(clip)
+            options.clip = clip2.internal()
+            options.flags |= mupdf.FZ_STEXT_CLIP_RECT
         annot = self.this
         stextpage = mupdf.FzStextPage(annot, options)
         ret = TextPage(stextpage)
@@ -3979,7 +3989,7 @@ class Document:
 
     def embfile_add(self,
             name: str,
-            buffer_: typing.ByteString,
+            buffer_: ByteString,
             filename: OptStr =None,
             ufilename: OptStr =None,
             desc: OptStr =None,
@@ -8096,7 +8106,9 @@ class Page:
 
         value = JM_insert_font(pdf, bfname, fontfile,fontbuffer, set_simple, idx, wmode, serif, encoding, ordering)
         # get the objects /Resources, /Resources/Font
-        resources = mupdf.pdf_dict_get_inheritable( page.obj(), PDF_NAME('Resources'))
+        resources = mupdf.pdf_dict_get_inheritable(page.obj(), PDF_NAME('Resources'))
+        if not resources.pdf_is_dict():
+            resources = mupdf.pdf_dict_put_dict(page.obj(), PDF_NAME("Resources"), 5)
         fonts = mupdf.pdf_dict_get(resources, PDF_NAME('Font'))
         if not fonts.m_internal:    # page has no fonts yet
             fonts = mupdf.pdf_new_dict(pdf, 5)
@@ -8282,7 +8294,7 @@ class Page:
     def add_file_annot(
             self,
             point: point_like,
-            buffer_: typing.ByteString,
+            buffer_: ByteString,
             filename: str,
             ufilename: OptStr =None,
             desc: OptStr =None,
@@ -18379,7 +18391,7 @@ def get_text_length(text: str, fontname: str ="helv", fontsize: float =11, encod
     raise ValueError("Font '%s' is unsupported" % fontname)
 
 
-def image_profile(img: typing.ByteString) -> dict:
+def image_profile(img: ByteString) -> dict:
     """ Return basic properties of an image.
 
     Args:

{pymupdf-1.25.4 → pymupdf-1.25.5}/src/extra.i

@@ -4113,7 +4113,7 @@ PyObject* ll_JM_color_count(fz_pixmap *pm, PyObject *clip)
     unsigned char* s = pm->samples + stride * (irect.y0 - pm->y) + n * (irect.x0 - pm->x);
     // Cache previous pixel.
     char oldpix[10];
-    assert(n < = size(oldpix));
+    assert(n <= sizeof(oldpix));
     memcpy(oldpix, s, n);
     long cnt = 0;
     for (size_t i = 0; i < height; i++)

{pymupdf-1.25.4 → pymupdf-1.25.5}/src/table.py

@@ -1883,7 +1883,7 @@ page information themselves.
 # -----------------------------------------------------------------------------
 def make_chars(page, clip=None):
     """Extract text as "rawdict" to fill CHARS."""
-    global CHARS, TEXTPAGE
+    global TEXTPAGE
     page_number = page.number + 1
     page_height = page.rect.height
     ctm = page.transformation_matrix
@@ -1943,7 +1943,6 @@ def make_chars(page, clip=None):
 # else to lines.
 # ------------------------------------------------------------------------
 def make_edges(page, clip=None, tset=None, add_lines=None):
-    global EDGES
     snap_x = tset.snap_x_tolerance
     snap_y = tset.snap_y_tolerance
     min_length = tset.edge_min_length

{pymupdf-1.25.4 → pymupdf-1.25.5}/src/utils.py

@@ -29,12 +29,21 @@ point_like = "point_like"
 rect_like = "rect_like"
 matrix_like = "matrix_like"
 quad_like = "quad_like"
+
+# ByteString is gone from typing in 3.14.
+# collections.abc.Buffer available from 3.12 only
+try:
+    ByteString = typing.ByteString
+except AttributeError:
+    # pylint: disable=unsupported-binary-operation
+    ByteString = bytes | bytearray | memoryview
+
 AnyType = typing.Any
 OptInt = typing.Union[int, None]
 OptFloat = typing.Optional[float]
 OptStr = typing.Optional[str]
 OptDict = typing.Optional[dict]
-OptBytes = typing.Optional[typing.ByteString]
+OptBytes = typing.Optional[ByteString]
 OptSeq = typing.Optional[typing.Sequence]
 
 """
@@ -4601,12 +4610,12 @@ def fill_textbox(
         words = line.split(" ")  # the words in the line
 
         # cut in parts any words that are longer than rect width
-        words, word_lengths = norm_words(std_width, words)
+        words, word_lengths = norm_words(width, words)
 
         n = len(words)
         while True:
             line0 = " ".join(words[:n])
-            wl = sum(word_lengths[:n]) + space_len * (len(word_lengths[:n]) - 1)
+            wl = sum(word_lengths[:n]) + space_len * (n - 1)
             if wl <= width:
                 new_lines.append((line0, wl))
                 words = words[n:]
@@ -4618,6 +4627,7 @@ def fill_textbox(
 
             if len(words) == 0:
                 break
+            assert n
 
     # -------------------------------------------------------------------------
     # List of lines created. Each item is (text, tl), where 'tl' is the PDF

{pymupdf-1.25.4 → pymupdf-1.25.5}/src_classic/fitz_old.i

@@ -321,12 +321,20 @@ point_like = "point_like"
 rect_like = "rect_like"
 matrix_like = "matrix_like"
 quad_like = "quad_like"
+
+# ByteString is gone from typing in 3.14.
+# collections.abc.Buffer available from 3.12 only
+try:
+    ByteString = typing.ByteString
+except AttributeError:
+    ByteString = bytes | bytearray | memoryview
+
 AnyType = typing.Any
 OptInt = typing.Union[int, None]
 OptFloat = typing.Optional[float]
 OptStr = typing.Optional[str]
 OptDict = typing.Optional[dict]
-OptBytes = typing.Optional[typing.ByteString]
+OptBytes = typing.Optional[ByteString]
 OptSeq = typing.Optional[typing.Sequence]
 
 try:
@@ -1342,7 +1350,7 @@ struct Document
             self.xref_set_key(xref, "Params/ModDate", get_pdf_str(date))
             return xref
 
-        def embfile_add(self, name: str, buffer: typing.ByteString,
+        def embfile_add(self, name: str, buffer: ByteString,
                                   filename: OptStr =None,
                                   ufilename: OptStr =None,
                                   desc: OptStr =None,) -> None:
@@ -5964,7 +5972,7 @@ struct Page {
 
 
         def add_file_annot(self, point: point_like,
-            buffer: typing.ByteString,
+            buffer: ByteString,
             filename: str,
             ufilename: OptStr =None,
             desc: OptStr =None,

{pymupdf-1.25.4 → pymupdf-1.25.5}/src_classic/helper-python.i

@@ -1245,7 +1245,7 @@ def planish_line(p1: point_like, p2: point_like) -> Matrix:
     return Matrix(util_hor_matrix(p1, p2))
 
 
-def image_profile(img: typing.ByteString) -> dict:
+def image_profile(img: ByteString) -> dict:
     """ Return basic properties of an image.
 
     Args:

{pymupdf-1.25.4 → pymupdf-1.25.5}/src_classic/utils.py

@@ -23,12 +23,20 @@ point_like = "point_like"
 rect_like = "rect_like"
 matrix_like = "matrix_like"
 quad_like = "quad_like"
+
+# ByteString is gone from typing in 3.14.
+# collections.abc.Buffer available from 3.12 only
+try:
+    ByteString = typing.ByteString
+except AttributeError:
+    ByteString = bytes | bytearray | memoryview
+
 AnyType = typing.Any
 OptInt = typing.Union[int, None]
 OptFloat = typing.Optional[float]
 OptStr = typing.Optional[str]
 OptDict = typing.Optional[dict]
-OptBytes = typing.Optional[typing.ByteString]
+OptBytes = typing.Optional[ByteString]
 OptSeq = typing.Optional[typing.Sequence]
 
 """

{pymupdf-1.25.4 → pymupdf-1.25.5}/tests/test_annots.py

@@ -276,6 +276,7 @@ def test_2270():
         for page_number, page in enumerate(document):
             for textBox in page.annots(types=(pymupdf.PDF_ANNOT_FREE_TEXT,pymupdf.PDF_ANNOT_TEXT)):
                 print("textBox.type :", textBox.type)
+                print(f"{textBox.rect=}")
                 print("textBox.get_text('words') : ", textBox.get_text('words'))
                 print("textBox.get_text('text') : ", textBox.get_text('text'))
                 print("textBox.get_textbox(textBox.rect) : ", textBox.get_textbox(textBox.rect))
@@ -286,16 +287,27 @@ def test_2270():
                 assert textBox.get_textbox(textBox.rect) == 'abc123'
                 assert textBox.info['content'] == 'abc123'
 
-                if hasattr(pymupdf, 'mupdf'):
-                    # Additional check that Annot.get_textpage() returns a
-                    # TextPage that works with page.get_text() - prior to
-                    # 2024-01-30 the TextPage had no `.parent` member.
-                    textpage = textBox.get_textpage()
-                    text = page.get_text()
-                    print(f'{text=}')
-                    text = page.get_text(textpage=textpage)
-                    print(f'{text=}')
-                    print(f'{getattr(textpage, "parent")=}')
+                # Additional check that Annot.get_textpage() returns a
+                # TextPage that works with page.get_text() - prior to
+                # 2024-01-30 the TextPage had no `.parent` member.
+                textpage = textBox.get_textpage()
+                text = page.get_text()
+                print(f'{text=}')
+                text = page.get_text(textpage=textpage)
+                print(f'{text=}')
+                print(f'{getattr(textpage, "parent")=}')
+
+                if pymupdf.mupdf_version_tuple >= (1, 26):
+                    # Check Annotation.get_textpage()'s <clip> arg.
+                    clip = textBox.rect
+                    clip.x1 = clip.x0 + (clip.x1 - clip.x0) / 3
+                    textpage2 = textBox.get_textpage(clip=clip)
+                    text = textpage2.extractText()
+                    print(f'With {clip=}: {text=}')
+                    assert text == 'ab\n'
+                else:
+                    assert not hasattr(pymupdf.mupdf, 'FZ_STEXT_CLIP_RECT')
+                    
 
 def test_2934_add_redact_annot():
     '''

{pymupdf-1.25.4 → pymupdf-1.25.5}/tests/test_general.py

@@ -1669,4 +1669,19 @@ def test_3886():
     rms_0 = gentle_compare.pixmaps_rms(pixmap, pixmap_clean0)
     rms_1 = gentle_compare.pixmaps_rms(pixmap, pixmap_clean1)
     print(f'test_3886(): {rms_0=} {rms_1=}')
-        
+
+def test_4415():
+    path = os.path.normpath(f'{__file__}/../../tests/resources/test_4415.pdf')
+    path_out = os.path.normpath(f'{__file__}/../../tests/resources/test_4415_out.png')
+    path_out_expected = os.path.normpath(f'{__file__}/../../tests/resources/test_4415_out_expected.png')
+    with pymupdf.open(path) as document:
+        page = document[0]
+        rot = page.rotation
+        orig = pymupdf.Point(100, 100)  # apparent insertion point
+        text = 'Text at Top-Left'
+        mrot = page.derotation_matrix  # matrix annihilating page rotation
+        page.insert_text(orig * mrot, text, fontsize=60, rotate=rot)
+        pixmap = page.get_pixmap()
+        pixmap.save(path_out)
+        rms = gentle_compare.pixmaps_rms(path_out_expected, path_out)
+        assert rms == 0, f'{rms=}'

{pymupdf-1.25.4 → pymupdf-1.25.5}/tests/test_textbox.py

@@ -280,3 +280,12 @@ def test_3916():
     page = doc.new_page()
     spare_height, scale = page.insert_htmlbox(rect, "Hello, World!", scale_low=0.5)
     assert spare_height == -1
+
+
+def test_4400():
+    with pymupdf.open() as document:
+        page = document.new_page()
+        writer = pymupdf.TextWriter(page.rect)
+        text = '111111111'
+        print(f'Calling writer.fill_textbox().', flush=1)
+        writer.fill_textbox(rect=pymupdf.Rect(0, 0, 100, 20), pos=(80, 0), text=text, fontsize=8)