def convert(input_filepath, output_filepath=None, cleanup_files=False, *args, **kwargs):
size = kwargs.get('size')
file_format = kwargs.get('file_format', DEFAULT_FILE_FORMAT)
zoom = kwargs.get('zoom', DEFAULT_ZOOM_LEVEL)
rotation = kwargs.get('rotation', DEFAULT_ROTATION)
page = kwargs.get('page', DEFAULT_PAGE_NUMBER)
transformations = kwargs.get('transformations', [])
if transformations is None:
transformations = []
unoconv_output = None
if output_filepath is None:
output_filepath = create_image_cache_filename(input_filepath, *args, **kwargs)
if os.path.exists(output_filepath):
return output_filepath
path, extension = os.path.splitext(input_filepath)
if extension[1:].lower() in CONVERTER_OFFICE_FILE_EXTENSIONS:
result = convert_office_document(input_filepath)
if result:
unoconv_output = result
input_filepath = result
if size:
transformations.append(
{
'transformation': TRANSFORMATION_RESIZE,
'arguments': dict(zip([u'width', u'height'], size.split(DIMENSION_SEPARATOR)))
}
)
if zoom != 100:
transformations.append(
{
'transformation': TRANSFORMATION_ZOOM,
'arguments': {'percent': zoom}
}
)
if rotation != 0 and rotation != 360:
transformations.append(
{
'transformation': TRANSFORMATION_ROTATE,
'arguments': {'degrees': rotation}
}
)
try:
backend.convert_file(input_filepath=input_filepath, output_filepath=output_filepath, transformations=transformations, page=page, file_format=file_format)
finally:
if cleanup_files:
cleanup(input_filepath)
if unoconv_output:
cleanup(unoconv_output)
return output_filepath
def convert(input_filepath, output_filepath=None, cleanup_files=False, mimetype=None, *args, **kwargs):
size = kwargs.get('size')
file_format = kwargs.get('file_format', DEFAULT_FILE_FORMAT)
zoom = kwargs.get('zoom', DEFAULT_ZOOM_LEVEL)
rotation = kwargs.get('rotation', DEFAULT_ROTATION)
page = kwargs.get('page', DEFAULT_PAGE_NUMBER)
transformations = kwargs.get('transformations', [])
if transformations is None:
transformations = []
if output_filepath is None:
output_filepath = create_image_cache_filename(input_filepath, *args, **kwargs)
if os.path.exists(output_filepath):
return output_filepath
if office_converter:
try:
office_converter.convert(input_filepath, mimetype=mimetype)
if office_converter.exists:
input_filepath = office_converter.output_filepath
mimetype = 'application/pdf'
else:
# Recycle the already detected mimetype
mimetype = office_converter.mimetype
except OfficeConversionError:
raise UnknownFileFormat('office converter exception')
if size:
transformations.append(
{
'transformation': TRANSFORMATION_RESIZE,
'arguments': dict(zip([u'width', u'height'], size.split(DIMENSION_SEPARATOR)))
}
)
if zoom != 100:
transformations.append(
{
'transformation': TRANSFORMATION_ZOOM,
'arguments': {'percent': zoom}
}
)
if rotation != 0 and rotation != 360:
transformations.append(
{
'transformation': TRANSFORMATION_ROTATE,
'arguments': {'degrees': rotation}
}
)
try:
backend.convert_file(input_filepath=input_filepath, output_filepath=output_filepath, transformations=transformations, page=page, file_format=file_format, mimetype=mimetype)
finally:
if cleanup_files:
cleanup(input_filepath)
return output_filepath
def convert_file(self,
input_filepath,
output_filepath,
transformations=None,
page=DEFAULT_PAGE_NUMBER,
file_format=DEFAULT_FILE_FORMAT,
**kwargs):
tmpfile = None
mimetype = kwargs.get('mimetype', None)
if not mimetype:
mimetype, encoding = get_mimetype(open(input_filepath, 'rb'),
input_filepath,
mimetype_only=True)
if mimetype == 'application/pdf' and USE_GHOSTSCRIPT:
# If file is a PDF open it with ghostscript and convert it to
# TIFF
first_page_tmpl = '-dFirstPage=%d' % page
last_page_tmpl = '-dLastPage=%d' % page
fd, tmpfile = tempfile.mkstemp()
os.close(fd)
output_file_tmpl = '-sOutputFile=%s' % tmpfile
input_file_tmpl = '-f%s' % input_filepath
args = [
'gs',
'-q',
'-dQUIET',
'-dSAFER',
'-dBATCH',
'-dNOPAUSE',
'-dNOPROMPT',
first_page_tmpl,
last_page_tmpl,
'-sDEVICE=jpeg',
'-dJPEGQ=95',
'-r150',
output_file_tmpl,
input_file_tmpl,
'-c "60000000 setvmthreshold"', # use 30MB
'-dNOGC', # No garbage collection
'-dMaxBitmap=500000000',
'-dAlignToPixels=0',
'-dGridFitTT=0',
'-dTextAlphaBits=4',
'-dGraphicsAlphaBits=4',
]
ghostscript.Ghostscript(*args)
page = 1 # Don't execute the following while loop
input_filepath = tmpfile
try:
im = Image.open(input_filepath)
except Exception:
# Python Imaging Library doesn't recognize it as an image
raise UnknownFileFormat
finally:
if tmpfile:
cleanup(tmpfile)
current_page = 0
try:
while current_page == page - 1:
im.seek(im.tell() + 1)
current_page += 1
# do something to im
except EOFError:
# end of sequence
pass
try:
if transformations:
aspect = 1.0 * im.size[0] / im.size[1]
for transformation in transformations:
arguments = transformation.get('arguments')
if transformation[
'transformation'] == TRANSFORMATION_RESIZE:
width = int(arguments.get('width', 0))
height = int(
arguments.get('height', 1.0 * width * aspect))
im = self.resize(im, (width, height))
elif transformation[
'transformation'] == TRANSFORMATION_ZOOM:
decimal_value = float(arguments.get('percent',
100)) / 100
im = im.transform((int(im.size[0] * decimal_value),
int(im.size[1] * decimal_value)),
Image.EXTENT,
(0, 0, im.size[0], im.size[1]))
elif transformation[
'transformation'] == TRANSFORMATION_ROTATE:
# PIL counter degress counter-clockwise, reverse them
im = im.rotate(360 - arguments.get('degrees', 0))
except:
# Ignore all transformation error
pass
if im.mode not in ('L', 'RGB'):
im = im.convert('RGB')
im.save(output_filepath, format=file_format)
def convert_file(
self,
input_filepath,
output_filepath,
transformations=None,
page=DEFAULT_PAGE_NUMBER,
file_format=DEFAULT_FILE_FORMAT,
):
tmpfile = None
mimetype, encoding = get_mimetype(open(input_filepath, "rb"), input_filepath)
if mimetype == "application/pdf" and USE_GHOSTSCRIPT:
# If file is a PDF open it with ghostscript and convert it to
# TIFF
first_page_tmpl = "-dFirstPage=%d" % page
last_page_tmpl = "-dLastPage=%d" % page
fd, tmpfile = tempfile.mkstemp()
os.close(fd)
output_file_tmpl = "-sOutputFile=%s" % tmpfile
input_file_tmpl = "-f%s" % input_filepath
args = [
"gs",
"-q",
"-dQUIET",
"-dSAFER",
"-dBATCH",
"-dNOPAUSE",
"-dNOPROMPT",
first_page_tmpl,
last_page_tmpl,
"-sDEVICE=jpeg",
"-dJPEGQ=75",
"-r150",
output_file_tmpl,
input_file_tmpl,
'-c "60000000 setvmthreshold"', # use 30MB
"-dNOGC", # No garbage collection
"-dMaxBitmap=500000000",
"-dAlignToPixels=0",
"-dGridFitTT=0",
"-dTextAlphaBits=4",
"-dGraphicsAlphaBits=4",
]
ghostscript.Ghostscript(*args)
page = 1 # Don't execute the following while loop
input_filepath = tmpfile
try:
im = Image.open(input_filepath)
except Exception:
# Python Imaging Library doesn't recognize it as an image
raise UnknownFileFormat
finally:
if tmpfile:
cleanup(tmpfile)
current_page = 0
try:
while current_page == page - 1:
im.seek(im.tell() + 1)
current_page += 1
# do something to im
except EOFError:
# end of sequence
pass
try:
if transformations:
aspect = 1.0 * im.size[0] / im.size[1]
for transformation in transformations:
arguments = transformation.get("arguments")
if transformation["transformation"] == TRANSFORMATION_RESIZE:
width = int(arguments.get("width", 0))
height = int(arguments.get("height", 1.0 * width * aspect))
im = self.resize(im, (width, height))
elif transformation["transformation"] == TRANSFORMATION_ZOOM:
decimal_value = float(arguments.get("percent", 100)) / 100
im = im.transform(
(int(im.size[0] * decimal_value), int(im.size[1] * decimal_value)),
Image.EXTENT,
(0, 0, im.size[0], im.size[1]),
)
elif transformation["transformation"] == TRANSFORMATION_ROTATE:
# PIL counter degress counter-clockwise, reverse them
im = im.rotate(360 - arguments.get("degrees", 0))
except:
# Ignore all transformation error
pass
if im.mode not in ("L", "RGB"):
im = im.convert("RGB")
im.save(output_filepath, format=file_format)
def convert_file(self, input_filepath, output_filepath, transformations=None, page=DEFAULT_PAGE_NUMBER, file_format=DEFAULT_FILE_FORMAT, **kwargs):
tmpfile = None
mimetype = kwargs.get('mimetype', None)
if not mimetype:
mimetype, encoding = get_mimetype(open(input_filepath, 'rb'), input_filepath, mimetype_only=True)
if mimetype == 'application/pdf' and USE_GHOSTSCRIPT:
# If file is a PDF open it with ghostscript and convert it to
# TIFF
first_page_tmpl = '-dFirstPage=%d' % page
last_page_tmpl = '-dLastPage=%d' % page
fd, tmpfile = tempfile.mkstemp()
os.close(fd)
output_file_tmpl = '-sOutputFile=%s' % tmpfile
input_file_tmpl = '-f%s' % input_filepath
args = [
'gs', '-q', '-dQUIET', '-dSAFER', '-dBATCH',
'-dNOPAUSE', '-dNOPROMPT',
first_page_tmpl, last_page_tmpl,
'-sDEVICE=jpeg', '-dJPEGQ=95',
'-r150', output_file_tmpl,
input_file_tmpl,
'-c "60000000 setvmthreshold"', # use 30MB
'-dNOGC', # No garbage collection
'-dMaxBitmap=500000000',
'-dAlignToPixels=0',
'-dGridFitTT=0',
'-dTextAlphaBits=4',
'-dGraphicsAlphaBits=4',
]
ghostscript.Ghostscript(*args)
page = 1 # Don't execute the following while loop
input_filepath = tmpfile
try:
im = Image.open(input_filepath)
except Exception:
# Python Imaging Library doesn't recognize it as an image
raise UnknownFileFormat
finally:
if tmpfile:
cleanup(tmpfile)
current_page = 0
try:
while current_page == page - 1:
im.seek(im.tell() + 1)
current_page += 1
# do something to im
except EOFError:
# end of sequence
pass
try:
if transformations:
aspect = 1.0 * im.size[0] / im.size[1]
for transformation in transformations:
arguments = transformation.get('arguments')
if transformation['transformation'] == TRANSFORMATION_RESIZE:
width = int(arguments.get('width', 0))
height = int(arguments.get('height', 1.0 * width * aspect))
im = self.resize(im, (width, height))
elif transformation['transformation'] == TRANSFORMATION_ZOOM:
decimal_value = float(arguments.get('percent', 100)) / 100
im = im.transform((int(im.size[0] * decimal_value), int(im.size[1] * decimal_value)), Image.EXTENT, (0, 0, im.size[0], im.size[1]))
elif transformation['transformation'] == TRANSFORMATION_ROTATE:
# PIL counter degress counter-clockwise, reverse them
im = im.rotate(360 - arguments.get('degrees', 0))
except:
# Ignore all transformation error
pass
if im.mode not in ('L', 'RGB'):
im = im.convert('RGB')
im.save(output_filepath, format=file_format)
def convert(input_filepath,
output_filepath=None,
cleanup_files=False,
mimetype=None,
*args,
**kwargs):
size = kwargs.get('size')
file_format = kwargs.get('file_format', DEFAULT_FILE_FORMAT)
zoom = kwargs.get('zoom', DEFAULT_ZOOM_LEVEL)
rotation = kwargs.get('rotation', DEFAULT_ROTATION)
page = kwargs.get('page', DEFAULT_PAGE_NUMBER)
transformations = kwargs.get('transformations', [])
if transformations is None:
transformations = []
if output_filepath is None:
output_filepath = create_image_cache_filename(input_filepath, *args,
**kwargs)
if os.path.exists(output_filepath):
return output_filepath
if office_converter:
try:
office_converter.convert(input_filepath, mimetype=mimetype)
if office_converter.exists:
input_filepath = office_converter.output_filepath
mimetype = 'application/pdf'
else:
# Recycle the already detected mimetype
mimetype = office_converter.mimetype
except OfficeConversionError:
raise UnknownFileFormat('office converter exception')
if size:
transformations.append({
'transformation':
TRANSFORMATION_RESIZE,
'arguments':
dict(zip([u'width', u'height'], size.split(DIMENSION_SEPARATOR)))
})
if zoom != 100:
transformations.append({
'transformation': TRANSFORMATION_ZOOM,
'arguments': {
'percent': zoom
}
})
if rotation != 0 and rotation != 360:
transformations.append({
'transformation': TRANSFORMATION_ROTATE,
'arguments': {
'degrees': rotation
}
})
try:
backend.convert_file(input_filepath=input_filepath,
output_filepath=output_filepath,
transformations=transformations,
page=page,
file_format=file_format,
mimetype=mimetype)
finally:
if cleanup_files:
cleanup(input_filepath)
return output_filepath
