Read bounding box information from the trailer of EPS files if specified

This commit is contained in:
nopperl 2023-09-07 20:20:35 +02:00
parent 52c6d68613
commit 12d8b439d5
3 changed files with 61 additions and 16 deletions

Binary file not shown.

View File

@ -404,3 +404,14 @@ def test_timeout(test_file):
with pytest.raises(Image.UnidentifiedImageError): with pytest.raises(Image.UnidentifiedImageError):
with Image.open(f): with Image.open(f):
pass pass
@pytest.mark.parametrize(
("trailer_image_file", "header_image_file"),
[("Tests/images/zero_bb_trailer.eps", FILE1)],
)
def test_boundary_box_in_trailer(trailer_image_file, header_image_file):
# Check whether boundary boxes which are specified in the trailer are parsed
# just like boundary boxes specified in the header
with Image.open(trailer_image_file) as trailer_image, Image.open(header_image_file) as header_image:
assert trailer_image.size == header_image.size

View File

@ -233,7 +233,9 @@ class EpsImageFile(ImageFile.ImageFile):
byte_arr = bytearray(255) byte_arr = bytearray(255)
bytes_mv = memoryview(byte_arr) bytes_mv = memoryview(byte_arr)
bytes_read = 0 bytes_read = 0
reading_comments = True reading_header_comments = True
reading_trailer_comments = False
trailer_reached = False
def check_required_header_comments(): def check_required_header_comments():
if "PS-Adobe" not in self.info: if "PS-Adobe" not in self.info:
@ -265,9 +267,9 @@ class EpsImageFile(ImageFile.ImageFile):
msg = "not an EPS file" msg = "not an EPS file"
raise SyntaxError(msg) raise SyntaxError(msg)
else: else:
if reading_comments: if reading_header_comments:
check_required_header_comments() check_required_header_comments()
reading_comments = False reading_header_comments = False
# reset bytes_read so we can keep reading # reset bytes_read so we can keep reading
# data until the end of the line # data until the end of the line
bytes_read = 0 bytes_read = 0
@ -275,7 +277,7 @@ class EpsImageFile(ImageFile.ImageFile):
bytes_read += 1 bytes_read += 1
continue continue
if reading_comments: if reading_header_comments:
# Load EPS header # Load EPS header
# if this line doesn't start with a "%", # if this line doesn't start with a "%",
@ -283,7 +285,7 @@ class EpsImageFile(ImageFile.ImageFile):
# then we've reached the end of the header/comments # then we've reached the end of the header/comments
if byte_arr[0] != ord("%") or bytes_mv[:13] == b"%%EndComments": if byte_arr[0] != ord("%") or bytes_mv[:13] == b"%%EndComments":
check_required_header_comments() check_required_header_comments()
reading_comments = False reading_header_comments = False
continue continue
s = str(bytes_mv[:bytes_read], "latin-1") s = str(bytes_mv[:bytes_read], "latin-1")
@ -298,17 +300,10 @@ class EpsImageFile(ImageFile.ImageFile):
k, v = m.group(1, 2) k, v = m.group(1, 2)
self.info[k] = v self.info[k] = v
if k == "BoundingBox": if k == "BoundingBox":
try: if v == "(atend)":
# Note: The DSC spec says that BoundingBox reading_trailer_comments = True
# fields should be integers, but some drivers else:
# put floating point values there anyway. self._read_boundary_box(v, offset, length)
box = [int(float(i)) for i in v.split()]
self._size = box[2] - box[0], box[3] - box[1]
self.tile = [
("eps", (0, 0) + self.size, offset, (length, box))
]
except Exception:
pass
else: else:
m = field.match(s) m = field.match(s)
if m: if m:
@ -355,7 +350,32 @@ class EpsImageFile(ImageFile.ImageFile):
self._size = columns, rows self._size = columns, rows
return return
elif trailer_reached and reading_trailer_comments:
# Load EPS trailer
# if this line starts with "%%EOF",
# then we've reached the end of the trailer
if bytes_mv[:5] == b"%%EOF":
reading_trailer_comments = False
continue
s = str(bytes_mv[:bytes_read], "latin-1")
try:
m = split.match(s)
except re.error as e:
msg = "not an EPS file"
raise SyntaxError(msg) from e
if m:
k, v = m.group(1, 2)
self.info[k] = v
if k == "BoundingBox":
if not self._size:
self._read_boundary_box(v, offset, length)
if bytes_mv[:9] == b"%%Trailer":
trailer_reached = True
bytes_read = 0 bytes_read = 0
check_required_header_comments() check_required_header_comments()
@ -387,6 +407,19 @@ class EpsImageFile(ImageFile.ImageFile):
return length, offset return length, offset
def _read_boundary_box(self, v, offset, length):
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()]
self._size = box[2] - box[0], box[3] - box[1]
self.tile = [
("eps", (0, 0) + self.size, offset, (length, box))
]
except Exception:
pass
def load(self, scale=1, transparency=False): def load(self, scale=1, transparency=False):
# Load EPS via Ghostscript # Load EPS via Ghostscript
if self.tile: if self.tile:
@ -402,6 +435,7 @@ class EpsImageFile(ImageFile.ImageFile):
pass pass
# -------------------------------------------------------------------- # --------------------------------------------------------------------