Merge pull request #8281 from Yay295/eps_test

Tests/test_file_eps.py

@@ -8,6 +8,7 @@ import pytest
 from PIL import EpsImagePlugin, Image, UnidentifiedImageError, features
 
 from .helper import (
+    assert_image_equal_tofile,
     assert_image_similar,
     assert_image_similar_tofile,
     hopper,
@@ -19,18 +20,18 @@ from .helper import (
 HAS_GHOSTSCRIPT = EpsImagePlugin.has_ghostscript()
 
 # Our two EPS test files (they are identical except for their bounding boxes)
-FILE1 = "Tests/images/zero_bb.eps"
+FILE1 = "Tests/images/eps/zero_bb.eps"
-FILE2 = "Tests/images/non_zero_bb.eps"
+FILE2 = "Tests/images/eps/non_zero_bb.eps"
 
 # Due to palletization, we'll need to convert these to RGB after load
-FILE1_COMPARE = "Tests/images/zero_bb.png"
+FILE1_COMPARE = "Tests/images/eps/zero_bb.png"
-FILE1_COMPARE_SCALE2 = "Tests/images/zero_bb_scale2.png"
+FILE1_COMPARE_SCALE2 = "Tests/images/eps/zero_bb_scale2.png"
 
-FILE2_COMPARE = "Tests/images/non_zero_bb.png"
+FILE2_COMPARE = "Tests/images/eps/non_zero_bb.png"
-FILE2_COMPARE_SCALE2 = "Tests/images/non_zero_bb_scale2.png"
+FILE2_COMPARE_SCALE2 = "Tests/images/eps/non_zero_bb_scale2.png"
 
 # EPS test files with binary preview
-FILE3 = "Tests/images/binary_preview_map.eps"
+FILE3 = "Tests/images/eps/binary_preview_map.eps"
 
 # Three unsigned 32bit little-endian values:
 #   0xC6D3D0C5 magic number
@@ -126,6 +127,15 @@ def test_binary_header_only() -> None:
         EpsImagePlugin.EpsImageFile(data)
 
 
+@pytest.mark.parametrize("prefix", (b"", simple_binary_header))
+def test_simple_eps_file(prefix: bytes) -> None:
+    data = io.BytesIO(prefix + b"\n".join(simple_eps_file))
+    with Image.open(data) as img:
+        assert img.mode == "RGB"
+        assert img.size == (100, 100)
+        assert img.format == "EPS"
+
+
 @pytest.mark.parametrize("prefix", (b"", simple_binary_header))
 def test_missing_version_comment(prefix: bytes) -> None:
     data = io.BytesIO(prefix + b"\n".join(simple_eps_file_without_version))
@@ -141,23 +151,21 @@ def test_missing_boundingbox_comment(prefix: bytes) -> None:
 
 
 @pytest.mark.parametrize("prefix", (b"", simple_binary_header))
-def test_invalid_boundingbox_comment(prefix: bytes) -> None:
+@pytest.mark.parametrize(
-    data = io.BytesIO(prefix + b"\n".join(simple_eps_file_with_invalid_boundingbox))
+    "file_lines",
+    (
+        simple_eps_file_with_invalid_boundingbox,
+        simple_eps_file_with_invalid_boundingbox_valid_imagedata,
+    ),
+)
+def test_invalid_boundingbox_comment(
+    prefix: bytes, file_lines: tuple[bytes, ...]
+) -> None:
+    data = io.BytesIO(prefix + b"\n".join(file_lines))
     with pytest.raises(OSError, match="cannot determine EPS bounding box"):
         EpsImagePlugin.EpsImageFile(data)
 
 
-@pytest.mark.parametrize("prefix", (b"", simple_binary_header))
-def test_invalid_boundingbox_comment_valid_imagedata_comment(prefix: bytes) -> None:
-    data = io.BytesIO(
-        prefix + b"\n".join(simple_eps_file_with_invalid_boundingbox_valid_imagedata)
-    )
-    with Image.open(data) as img:
-        assert img.mode == "RGB"
-        assert img.size == (100, 100)
-        assert img.format == "EPS"
-
-
 @pytest.mark.parametrize("prefix", (b"", simple_binary_header))
 def test_ascii_comment_too_long(prefix: bytes) -> None:
     data = io.BytesIO(prefix + b"\n".join(simple_eps_file_with_long_ascii_comment))
@@ -177,7 +185,7 @@ def test_load_long_binary_data(prefix: bytes) -> None:
     data = io.BytesIO(prefix + b"\n".join(simple_eps_file_with_long_binary_data))
     with Image.open(data) as img:
         img.load()
-        assert img.mode == "RGB"
+        assert img.mode == "1"
         assert img.size == (100, 100)
         assert img.format == "EPS"
 
@@ -187,7 +195,7 @@ def test_load_long_binary_data(prefix: bytes) -> None:
 )
 @pytest.mark.skipif(not HAS_GHOSTSCRIPT, reason="Ghostscript not available")
 def test_cmyk() -> None:
-    with Image.open("Tests/images/pil_sample_cmyk.eps") as cmyk_image:
+    with Image.open("Tests/images/eps/pil_sample_cmyk.eps") as cmyk_image:
         assert cmyk_image.mode == "CMYK"
         assert cmyk_image.size == (100, 100)
         assert cmyk_image.format == "EPS"
@@ -204,8 +212,8 @@ def test_cmyk() -> None:
 @pytest.mark.skipif(not HAS_GHOSTSCRIPT, reason="Ghostscript not available")
 def test_showpage() -> None:
     # See https://github.com/python-pillow/Pillow/issues/2615
-    with Image.open("Tests/images/reqd_showpage.eps") as plot_image:
+    with Image.open("Tests/images/eps/reqd_showpage.eps") as plot_image:
-        with Image.open("Tests/images/reqd_showpage.png") as target:
+        with Image.open("Tests/images/eps/reqd_showpage.png") as target:
             # should not crash/hang
             plot_image.load()
             # fonts could be slightly different
@@ -214,11 +222,11 @@ def test_showpage() -> None:
 
 @pytest.mark.skipif(not HAS_GHOSTSCRIPT, reason="Ghostscript not available")
 def test_transparency() -> None:
-    with Image.open("Tests/images/reqd_showpage.eps") as plot_image:
+    with Image.open("Tests/images/eps/reqd_showpage.eps") as plot_image:
         plot_image.load(transparency=True)
         assert plot_image.mode == "RGBA"
 
-        with Image.open("Tests/images/reqd_showpage_transparency.png") as target:
+        with Image.open("Tests/images/eps/reqd_showpage_transparency.png") as target:
             # fonts could be slightly different
             assert_image_similar(plot_image, target, 6)
 
@@ -245,9 +253,19 @@ def test_bytesio_object() -> None:
         assert_image_similar(img, image1_scale1_compare, 5)
 
 
-def test_1_mode() -> None:
+@pytest.mark.skipif(not HAS_GHOSTSCRIPT, reason="Ghostscript not available")
-    with Image.open("Tests/images/1.eps") as im:
+@pytest.mark.parametrize(
-        assert im.mode == "1"
+    # These images have an "ImageData" descriptor.
+    "filename",
+    (
+        "Tests/images/eps/1.eps",
+        "Tests/images/eps/1_boundingbox_after_imagedata.eps",
+        "Tests/images/eps/1_second_imagedata.eps",
+    ),
+)
+def test_1(filename: str) -> None:
+    with Image.open(filename) as im:
+        assert_image_equal_tofile(im, "Tests/images/eps/1.bmp")
 
 
 def test_image_mode_not_supported(tmp_path: Path) -> None:
@@ -302,7 +320,9 @@ def test_render_scale2() -> None:
 
 
 @pytest.mark.skipif(not HAS_GHOSTSCRIPT, reason="Ghostscript not available")
-@pytest.mark.parametrize("filename", (FILE1, FILE2, "Tests/images/illu10_preview.eps"))
+@pytest.mark.parametrize(
+    "filename", (FILE1, FILE2, "Tests/images/eps/illu10_preview.eps")
+)
 def test_resize(filename: str) -> None:
     with Image.open(filename) as im:
         new_size = (100, 100)
@@ -344,10 +364,10 @@ def test_readline(prefix: bytes, line_ending: bytes) -> None:
 @pytest.mark.parametrize(
     "filename",
     (
-        "Tests/images/illu10_no_preview.eps",
+        "Tests/images/eps/illu10_no_preview.eps",
-        "Tests/images/illu10_preview.eps",
+        "Tests/images/eps/illu10_preview.eps",
-        "Tests/images/illuCS6_no_preview.eps",
+        "Tests/images/eps/illuCS6_no_preview.eps",
-        "Tests/images/illuCS6_preview.eps",
+        "Tests/images/eps/illuCS6_preview.eps",
     ),
 )
 def test_open_eps(filename: str) -> None:
@@ -359,7 +379,7 @@ def test_open_eps(filename: str) -> None:
 @pytest.mark.skipif(not HAS_GHOSTSCRIPT, reason="Ghostscript not available")
 def test_emptyline() -> None:
     # Test file includes an empty line in the header data
-    emptyline_file = "Tests/images/zero_bb_emptyline.eps"
+    emptyline_file = "Tests/images/eps/zero_bb_emptyline.eps"
 
     with Image.open(emptyline_file) as image:
         image.load()
@@ -371,7 +391,7 @@ def test_emptyline() -> None:
 @pytest.mark.timeout(timeout=5)
 @pytest.mark.parametrize(
     "test_file",
-    ["Tests/images/timeout-d675703545fee17acab56e5fec644c19979175de.eps"],
+    ["Tests/images/eps/timeout-d675703545fee17acab56e5fec644c19979175de.eps"],
 )
 def test_timeout(test_file: str) -> None:
     with open(test_file, "rb") as f:
@@ -384,7 +404,7 @@ def test_bounding_box_in_trailer() -> None:
     # Check bounding boxes are parsed in the same way
     # when specified in the header and the trailer
     with (
-        Image.open("Tests/images/zero_bb_trailer.eps") as trailer_image,
+        Image.open("Tests/images/eps/zero_bb_trailer.eps") as trailer_image,
         Image.open(FILE1) as header_image,
     ):
         assert trailer_image.size == header_image.size
@@ -392,12 +412,12 @@ def test_bounding_box_in_trailer() -> None:
 
 def test_eof_before_bounding_box() -> None:
     with pytest.raises(OSError):
-        with Image.open("Tests/images/zero_bb_eof_before_boundingbox.eps"):
+        with Image.open("Tests/images/eps/zero_bb_eof_before_boundingbox.eps"):
             pass
 
 
 def test_invalid_data_after_eof() -> None:
-    with open("Tests/images/illuCS6_preview.eps", "rb") as f:
+    with open("Tests/images/eps/illuCS6_preview.eps", "rb") as f:
         img_bytes = io.BytesIO(f.read() + b"\r\n%" + (b" " * 255))
 
     with Image.open(img_bytes) as img:

Tests/test_pickle.py

@@ -56,10 +56,10 @@ def helper_pickle_string(protocol: int, test_file: str, mode: str | None) -> Non
         ),
         ("Tests/images/hopper.tif", None),
         ("Tests/images/test-card.png", None),
-        ("Tests/images/zero_bb.png", None),
+        ("Tests/images/eps/zero_bb.png", None),
-        ("Tests/images/zero_bb_scale2.png", None),
+        ("Tests/images/eps/zero_bb_scale2.png", None),
-        ("Tests/images/non_zero_bb.png", None),
+        ("Tests/images/eps/non_zero_bb.png", None),
-        ("Tests/images/non_zero_bb_scale2.png", None),
+        ("Tests/images/eps/non_zero_bb_scale2.png", None),
         ("Tests/images/p_trns_single.png", None),
         ("Tests/images/pil123p.png", None),
         ("Tests/images/itxt_chunks.png", None),

src/PIL/EpsImagePlugin.py

@@ -121,7 +121,13 @@ def Ghostscript(
                 lengthfile -= len(s)
                 f.write(s)
 
-    device = "pngalpha" if transparency else "ppmraw"
+    if transparency:
+        # "RGBA"
+        device = "pngalpha"
+    else:
+        # "pnmraw" automatically chooses between
+        # PBM ("1"), PGM ("L"), and PPM ("RGB").
+        device = "pnmraw"
 
     # Build Ghostscript command
     command = [
@@ -151,8 +157,9 @@ def Ghostscript(
             startupinfo = subprocess.STARTUPINFO()
             startupinfo.dwFlags |= subprocess.STARTF_USESHOWWINDOW
         subprocess.check_call(command, startupinfo=startupinfo)
-        out_im = Image.open(outfile)
+        with Image.open(outfile) as out_im:
-        out_im.load()
+            out_im.load()
+            return out_im.im.copy()
     finally:
         try:
             os.unlink(outfile)
@@ -161,10 +168,6 @@ def Ghostscript(
         except OSError:
             pass
 
-    im = out_im.im.copy()
-    out_im.close()
-    return im
-
 
 def _accept(prefix: bytes) -> bool:
     return prefix[:4] == b"%!PS" or (len(prefix) >= 4 and i32(prefix) == 0xC6D3D0C5)
@@ -191,6 +194,11 @@ class EpsImageFile(ImageFile.ImageFile):
 
         self._mode = "RGB"
 
+        # When reading header comments, the first comment is used.
+        # When reading trailer comments, the last comment is used.
+        bounding_box: list[int] | None = None
+        imagedata_size: tuple[int, int] | None = None
+
         byte_arr = bytearray(255)
         bytes_mv = memoryview(byte_arr)
         bytes_read = 0
@@ -211,8 +219,8 @@ class EpsImageFile(ImageFile.ImageFile):
                 msg = 'EPS header missing "%%BoundingBox" comment'
                 raise SyntaxError(msg)
 
-        def _read_comment(s: str) -> bool:
+        def read_comment(s: str) -> bool:
-            nonlocal reading_trailer_comments
+            nonlocal bounding_box, reading_trailer_comments
             try:
                 m = split.match(s)
             except re.error as e:
@@ -227,18 +235,12 @@ class EpsImageFile(ImageFile.ImageFile):
             if k == "BoundingBox":
                 if v == "(atend)":
                     reading_trailer_comments = True
-                elif not self.tile or (trailer_reached and reading_trailer_comments):
+                elif not bounding_box or (trailer_reached and reading_trailer_comments):
                     try:
                         # Note: The DSC spec says that BoundingBox
                         # fields should be integers, but some drivers
                         # put floating point values there anyway.
-                        box = [int(float(i)) for i in v.split()]
+                        bounding_box = [int(float(i)) for i in v.split()]
-                        self._size = box[2] - box[0], box[3] - box[1]
-                        self.tile = [
-                            ImageFile._Tile(
-                                "eps", (0, 0) + self.size, offset, (length, box)
-                            )
-                        ]
                     except Exception:
                         pass
             return True
@@ -289,7 +291,7 @@ class EpsImageFile(ImageFile.ImageFile):
                     continue
 
                 s = str(bytes_mv[:bytes_read], "latin-1")
-                if not _read_comment(s):
+                if not read_comment(s):
                     m = field.match(s)
                     if m:
                         k = m.group(1)
@@ -308,6 +310,12 @@ class EpsImageFile(ImageFile.ImageFile):
                 # Check for an "ImageData" descriptor
                 # https://www.adobe.com/devnet-apps/photoshop/fileformatashtml/#50577413_pgfId-1035096
 
+                # If we've already read an "ImageData" descriptor,
+                # don't read another one.
+                if imagedata_size:
+                    bytes_read = 0
+                    continue
+
                 # Values:
                 # columns
                 # rows
@@ -333,22 +341,35 @@ class EpsImageFile(ImageFile.ImageFile):
                 else:
                     break
 
-                self._size = columns, rows
+                # Parse the columns and rows after checking the bit depth and mode
-                return
+                # in case the bit depth and/or mode are invalid.
+                imagedata_size = columns, rows
             elif bytes_mv[:5] == b"%%EOF":
                 break
             elif trailer_reached and reading_trailer_comments:
                 # Load EPS trailer
                 s = str(bytes_mv[:bytes_read], "latin-1")
-                _read_comment(s)
+                read_comment(s)
             elif bytes_mv[:9] == b"%%Trailer":
                 trailer_reached = True
             bytes_read = 0
 
-        if not self.tile:
+        # A "BoundingBox" is always required,
+        # even if an "ImageData" descriptor size exists.
+        if not bounding_box:
             msg = "cannot determine EPS bounding box"
             raise OSError(msg)
 
+        # An "ImageData" size takes precedence over the "BoundingBox".
+        self._size = imagedata_size or (
+            bounding_box[2] - bounding_box[0],
+            bounding_box[3] - bounding_box[1],
+        )
+
+        self.tile = [
+            ImageFile._Tile("eps", (0, 0) + self.size, offset, (length, bounding_box))
+        ]
+
     def _find_offset(self, fp: IO[bytes]) -> tuple[int, int]:
         s = fp.read(4)