def pdf_to_image(input_file, output_file):
input_file = os.path.abspath(input_file)
output_file = os.path.abspath(output_file)
# Read headphone model from the PDF
f = open(input_file, 'rb')
text = PyPDF2.PdfFileReader(f).getPage(0).extractText()
if 'crinacle' in text.lower():
raise ValueError('Measured by Crinacle')
# Convert to image with ghostscript
# Using temporary paths with Ghostscript because it seems to be unable to work with non-ascii characters
tmp_in = os.path.join(os.path.split(input_file)[0], '__tmp.pdf')
tmp_out = os.path.join(os.path.split(output_file)[0], '__tmp.png')
if tmp_in == input_file or tmp_out == output_file:
# Skip tmp files in case it was passed as input
raise ValueError('tmp file')
shutil.copy(input_file, tmp_in)
gs = Ghostscript(b'pdf2png', b'-dNOPAUSE', b'-sDEVICE=png16m',
b'-dBATCH', b'-r600', b'-dUseCropBox',
f'-sOutputFile={tmp_out}'.encode('utf-8'),
tmp_in.encode('utf-8'))
gs.exit()
shutil.copy(tmp_out, output_file)
print('\nSaved image to "{}"\n'.format(output_file))
f.close()
return Image.open(output_file)
def pdf_to_image(input_file, output_dir):
input_file = os.path.abspath(input_file)
output_dir = os.path.abspath(output_dir)
# Read headphone model from the PDF
f = open(input_file, 'rb')
pdf = PyPDF2.PdfFileReader(f)
page = pdf.getPage(0)
try:
t = page.extractText()
start_ind = t.index('All rights reserved.') + len(
'All rights reserved.')
end_ind = t.index('%THD+noise')
name = t[start_ind:end_ind]
print('Read "{name}" in "{fp}"'.format(name=name, fp=input_file))
except:
print('Fail to read "{}"'.format(input_file))
return
# Convert to image with ghostscript
output_file_path = '{}.png'.format(os.path.join(output_dir, name))
Ghostscript(b'pdf2png', b'-dNOPAUSE', b'-sDEVICE=png16m', b'-dBATCH',
b'-r600', b'-dUseCropBox',
'-sOutputFile={}'.format(output_file_path).encode('utf-8'),
input_file.encode('utf-8'))
print('\nSaved image to "{}"\n'.format(output_file_path))
f.close()
def pdf_to_image(input_file, output_file):
input_file = os.path.abspath(input_file)
output_file = os.path.abspath(output_file)
# Read headphone model from the PDF
f = open(input_file, 'rb')
# Convert to image with ghostscript
# Using temporary paths with Ghostscript because it seems to be unable to work with non-ascii characters
tmp_in = os.path.join(os.path.split(input_file)[0], '__tmp.pdf')
tmp_out = os.path.join(os.path.split(output_file)[0], '__tmp.png')
if tmp_in == input_file or tmp_out == output_file:
return
shutil.copy(input_file, tmp_in)
Ghostscript(
b'pdf2png',
b'-dNOPAUSE',
b'-sDEVICE=png16m',
b'-dBATCH',
b'-r600',
b'-dUseCropBox',
f'-sOutputFile={tmp_out}'.encode('utf-8'),
tmp_in.encode('utf-8')
)
shutil.copy(tmp_out, output_file)
print('\nSaved image to "{}"\n'.format(output_file))
f.close()
return Image.open(output_file)
def pdf2png(pdf, size):
"""Transform a PDF to a PNG thumbnail."""
temporary_name = os.path.join(gettempdir(), uuid4().hex)
# Dump it to a temp file so that we can feed it to ghostscript
# It would be cool to use gs.run_file, but it almost never works
real_file = NamedTemporaryFile(delete=False)
pdf.seek(0)
real_file.write(pdf.read())
real_file.close()
args = [s.encode('utf-8') for s in [
"",
"-sstdout=/dev/null",
"-dNOPAUSE", "-dBATCH", "-dSAFER",
"-dFirstPage=1", "-dLastPage=1",
"-dTextAlphaBits=4", "-dGraphicsAlphaBits=4",
"-sDEVICE=png16m",
"-r42",
"-sOutputFile=%s" % temporary_name,
"-f%s" % real_file.name]]
Ghostscript(*args)
os.remove(real_file.name)
with open(temporary_name, 'rb') as f:
img = BytesIO(f.read())
os.remove(temporary_name)
return image2png(img, size)
def pdf_to_image(input_file, output_file):
input_file = os.path.abspath(input_file)
output_file = os.path.abspath(output_file)
# Read headphone model from the PDF
f = open(input_file, 'rb')
# Convert to image with ghostscript
Ghostscript(b'pdf2png', b'-dNOPAUSE', b'-sDEVICE=png16m', b'-dBATCH',
b'-r600', b'-dUseCropBox',
'-sOutputFile={}'.format(output_file).encode('utf-8'),
input_file.encode('utf-8'))
print('\nSaved image to "{}"\n'.format(output_file))
f.close()
return Image.open(output_file)