def converter(self):
if not os.path.exists(self.path):
os.mkdir(self.path)
numero_de_paginas = len(self.arquivo.sequence)
self.detalhes.append("Iniciando importação ...")
try:
for pagina in range(numero_de_paginas):
nome_arquivo = "%s[%d]" % (self.arquivo_pdf, pagina)
img = Image(filename=nome_arquivo, resolution=150)
img.compression_quality = 99
nome_arquivo = "pagina%d.jpg" % (pagina + 1)
self.detalhes.append("\tPágina %d importada" % (pagina + 1))
img.save(filename=self.path + nome_arquivo)
self.lista_de_imagens.append(nome_arquivo)
img.close()
self.detalhes.append("Importação concluída.\n")
time.sleep(1)
self.arquivo.close()
except BlobError:
self.detalhes.append("Erro na importação!!!")
time.sleep(1)
self.detalhes.append("Fim.")
self.arquivo.close()
os.rmdir(self.path)
def convert_pdf_to_img(blob, img_type="jpg", quality=75, resolution=200):
"""
Converts PDF with multiple pages into one image.
It needs the file content NOT the filename or ioBytes and returns the image content.
Note: It has memory leak!!
http://stackoverflow.com/a/26233785/1497443
Example:
with open('my.pdf', "r") as f:
file_content = f.read()
# import ipdb
# ipdb.set_trace()
hh = convert_pdf_to_jpg(file_content)
with open('my.jpg', 'wb') as f:
f.write(hh)
"""
from wand.image import Image as WandImage
from wand.color import Color as WandColor
pdf = WandImage(blob=blob, resolution=resolution)
pages = len(pdf.sequence)
wimage = WandImage(width=pdf.width, height=pdf.height * pages, background=WandColor("white"))
for i in xrange(pages):
wimage.composite(pdf.sequence[i], top=pdf.height * i, left=0)
if img_type == "jpg":
wimage.compression_quality = quality
return wimage.make_blob(img_type)
def deepfry(img: Image):
resize(img)
img.format = "jpeg"
img.compression_quality = 1
img.modulate(saturation=700)
return img
def save_file():
event_uuid = str(uuid.uuid1())
event_path = get_event_dir(event_uuid)
if request.headers['Content-Type'].startswith('image/'):
os.makedirs(event_path)
img_filename = os.path.join(
event_path, request.headers['Content-Type'].replace('/', '.'))
with open(img_filename, 'wb') as img_file:
img_bytes = BytesIO(request.data)
img_file.write(img_bytes.read())
elif request.headers['Content-Type'] == 'application/pdf':
os.makedirs(event_path)
with Image(file=BytesIO(request.data), resolution=900) as pdfdoc:
with pdfdoc.convert('jpeg') as png_pages:
for idx, png in enumerate(png_pages.sequence):
img = Image(png)
img.compression_quality = 100
img_filename = os.path.join(event_path,
'image{}.jpg'.format(idx + 1))
img.save(filename=img_filename)
print '{} images extracted from PDF'.format(
len(png_pages.sequence))
else:
return 'Not a valid file', 400
return event_uuid, 200
def _resize(image_obj, width, height, **options):
img_format = options['format']
upscale = options['upscale']
quality = options['quality']
img_file = io.BytesIO()
img = Image(file=image_obj)
original_width, original_height = img.size
new_width, new_height = width, height
# Stretchs or fits the width with desired width
if width < original_width or upscale:
new_width = width
# Stretchs or fits the height with desired height
if (height < original_height) or upscale:
new_height = height
# Preserves the aspect ratio fit to new height
if width == 0:
new_width = int(original_width * (new_height / original_height))
# Preserves the aspect ratio fit to new width
if height == 0:
new_height = int(original_height * (new_width / original_width))
img.resize(new_width, new_height)
img.format = img_format
img.compression_quality = quality
img.save(file=img_file)
img_file.seek(0)
return img_file
def _convert_image(self, file_path: str, preview_dims: ImgDims) -> Image:
"""
refer: https://legacy.imagemagick.org/Usage/thumbnails/
like cmd: convert -layers merge -background white -thumbnail widthxheight \
-auto-orient -quality 85 -interlace plane input.jpeg output.jpeg
"""
img = Image(filename=file_path)
resize_dim = compute_resize_dims(dims_in=ImgDims(width=img.width,
height=img.height),
dims_out=preview_dims)
img.auto_orient()
img.iterator_reset()
img.background_color = Color("white")
img.merge_layers("merge")
if self.progressive:
img.interlace_scheme = "plane"
img.compression_quality = self.quality
img.thumbnail(resize_dim.width, resize_dim.height)
return img
def convert2pdf(path, images):
img = Image()
for image in images:
img.read(filename=os.path.join(path, image))
img.compression_quality = 75
img.save(filename=os.path.join(path, 'exported.pdf'))
return path
def resize_image(self, img_path, size_config, size_name, dest_path, compress=False): #open image img = Image(filename=img_path) #transform using resize config img.transform(resize=size_config) ''' exProcessor = ExifData() #rotate image if necessary if exProcessor.check_orientation(img) in [6, 7]: img.rotate(90) img.metadata['exif:Orientation'] = 1 if exProcessor.check_orientation(img) in [6, 8]: img.rotate(-90) img.metadata['exif:Orientation'] = 1 if exProcessor.check_orientation(img) in [3, 4]: img.rotate(180) img.metadata['exif:Orientation'] = 1 ''' # Detect file extention fileName, fileExtension = os.path.splitext(img_path) #save img fileName = ''.join([fileName, '_', size_name, fileExtension]) fileName = os.path.basename(fileName) fileName = os.path.abspath(os.path.join(dest_path, fileName)) # Compress option? if compress: img.compression_quality = 70 img.save(filename=fileName) return fileName
def _run_convert_all(filename, res=120):
'''把pdf所有页面转化为图片'''
# 由于每次转换的时候都需要重新将整个PDF载入内存，所以这里使用内存缓存
pdfile, f = get_pdf_reader(filename), open(filename, "rb")
# pdfile,f = get_pdf_reader2(filename)
for i in range(0, pdfile.getNumPages()):
temp_time = time.time() * 1000
pageobj = pdfile.getPage(i)
dst_pdf = PdfFileWriter()
dst_pdf.addPage(pageobj)
pdf_bytes = io.BytesIO()
dst_pdf.write(pdf_bytes)
pdf_bytes.seek(0)
img = Image(file=pdf_bytes, resolution=res)
img.format = 'png'
img.compression_quality = 90
img.background_color = Color("white")
# 保存图片
img_path = 'dest/%s_pg%d.png' % (filename[:filename.rindex('.')],
i + 1)
img.save(filename=img_path)
img.destroy()
img, pdf_bytes, dst_pdf = None, None, None
print('convert page %d cost time %dms' %
(i + 1, (time.time() * 1000 - temp_time)))
f.close()
def _run_convert(filename, page, res=120):
idx = page + 1
temp_time = time.time() * 1000
# 由于每次转换的时候都需要重新将整个PDF载入内存，所以这里使用内存缓存
pdfile = getPdfReader(filename)
pageObj = pdfile.getPage(page)
dst_pdf = PdfFileWriter()
dst_pdf.addPage(pageObj)
pdf_bytes = io.BytesIO()
dst_pdf.write(pdf_bytes)
pdf_bytes.seek(0)
img = Image(file=pdf_bytes, resolution=res)
img.format = 'png'
img.compression_quality = 90
img.background_color = Color('white')
img.alpha_channel = 'remove'
#img.channels = 'rgb_channels'
# 保存图片
#filename[:filename.rindex('.')]
img_path = './image/%s.png' % (filename[filename.rindex('/pdf/') +
5:filename.rindex('.pdf')])
#)
img.save(filename=img_path)
img.destroy()
img = None
pdf_bytes = None
dst_pdf = None
def _run_convert(filename, page, res=120):
'''把pdf指定页码转化为图片'''
pdfile, f = get_pdf_reader(filename), open(filename, "rb")
# pdfile,f = get_pdf_reader2(filename)
if page <= pdfile.getNumPages():
idx = page + 1
temp_time = time.time() * 1000
# 由于每次转换的时候都需要重新将整个PDF载入内存，所以这里使用内存缓存
pageobj = pdfile.getPage(page)
dst_pdf = PdfFileWriter()
dst_pdf.addPage(pageobj)
pdf_bytes = io.BytesIO()
dst_pdf.write(pdf_bytes)
pdf_bytes.seek(0)
img = Image(file=pdf_bytes, resolution=res)
img.format = 'png'
img.compression_quality = 90
img.background_color = Color("white")
# 保存图片
img_path = 'dest/%s_pg%d.png' % (filename[:filename.rindex('.')], idx)
img.save(filename=img_path)
img.destroy()
img, pdf_bytes, dst_pdf = None, None, None
print('convert page %d cost time %dms' %
(idx, (time.time() * 1000 - temp_time)))
else:
print("pg%r list index out of range" % page)
f.close()
def _thumbnail(image_obj, width, height, **options):
img_format = options['format']
quality = options['quality']
img_file = io.BytesIO()
img = Image(file=image_obj)
original_width, original_height = img.size
new_width, new_height = width, height
# Preserves the aspect ratio fit to new height
if width == 0:
new_width = int(original_width * (new_height / original_height))
# Preserves the aspect ratio fit to new width
if height == 0:
new_height = int(original_height * (new_width / original_width))
# If desired width and/or height is larger than original ones, shrinks it
# If not, enlarges it
if (new_width > original_width) or (new_height > original_height):
img.transform(str(new_width) + 'x' + str(new_height) + '<')
else:
img.transform(str(new_width) + 'x' + str(new_height) + '>')
img.format = img_format
img.compression_quality = quality
img.save(file=img_file)
img_file.seek(0)
return img_file
def _run_convert(filename, page, res=120):
idx = page + 1
temp_time = time.time() * 1000
# 由于每次转换的时候都需要重新将整个PDF载入内存，所以这里使用内存缓存
pdfile = getPdfReader(filename)
pageObj = pdfile.getPage(page)
dst_pdf = PdfFileWriter()
dst_pdf.addPage(pageObj)
pdf_bytes = io.BytesIO()
dst_pdf.write(pdf_bytes)
pdf_bytes.seek(0)
img = Image(file=pdf_bytes, resolution=res)
img.format = 'png'
img.compression_quality = 90
img.background_color = Color("white")
# 保存图片
img_path = '%s%d.png' % (filename[:filename.rindex('.')], idx)
img.save(filename=img_path)
img.destroy()
img = None
pdf_bytes = None
dst_pdf = None
print('convert page %d cost time %d' % (idx, (time.time() * 1000 - temp_time)))
def convertPdfToImage(file_loc):
new_dir = file_loc.split(".")[0]
if not os.path.exists(new_dir):
os.makedirs(new_dir)
doc = Image(filename=file_loc,resolution=300)
doc.compression_quality = 20
for i, page in enumerate(doc.sequence):
with Image(page) as page_image:
page_image.alpha_channel = False
page_image.background_color = Color('white')
save_location = new_dir+"/img-{}.jpg".format(str(i).zfill(4))
page_image.save(filename=save_location)
def convertPdfToImage(file_loc):
new_dir = file_loc.split(".")[0]
if not os.path.exists(new_dir):
os.makedirs(new_dir)
doc = Image(filename=file_loc, resolution=300)
doc.compression_quality = 20
for i, page in enumerate(doc.sequence):
with Image(page) as page_image:
page_image.alpha_channel = False
page_image.background_color = Color('white')
save_location = new_dir + "/img-{}.jpg".format(str(i).zfill(4))
page_image.save(filename=save_location)
def _rotate(image_obj, degree, **options):
img_format = options['format']
quality = options['quality']
img_file = io.BytesIO()
img = Image(file=image_obj)
img.rotate(degree)
img.format = img_format
img.compression_quality = quality
img.save(file=img_file)
img_file.seek(0)
return img_file
def _flip(image_obj, direction, **options):
img_format = options['format']
quality = options['quality']
img_file = io.BytesIO()
img = Image(file=image_obj)
if direction == 'v':
img.flip()
if direction == 'h':
img.flop()
img.format = img_format
img.compression_quality = quality
img.save(file=img_file)
img_file.seek(0)
return img_file
def writeImageNoWarp(pdf, out, resolution=pdfDense, trim=True):
with Image(filename=pdf, resolution=resolution) as pdf:
pages = len(pdf.sequence)
height = 0
width = 0
for j in range(pages):
## Go through the sequence and trim them individually
with pdf.sequence.index_context(j):
if trim:
pdf.trim()
height += pdf.height + 10
if (width < pdf.width):
width = pdf.width
else:
width = pdf.width
height = pdf.height
i = Image(width=width, height=height)
height = [r.height for r in pdf.sequence]
height = height[::-1]
height.append(0)
height = height[::-1]
width = [math.floor((width - r.width) / 2) for r in pdf.sequence]
for j in range(pages):
i.composite(pdf.sequence[j], top=height[j] + j * 10, left=width[j])
i.background_color = Color('white') # Set white background.
i.alpha_channel = 'remove'
if trim: ## This is necessary because of stacked charts (above)
i.trim()
i.sharpen(radius=5.0, sigma=5.0)
i.normalize()
i.quantize(16, dither=False)
i.compression_quality = 00
i.save(filename=out.replace(".png", ".png8"))
tmp = out.replace(".png", "") + "*"
#out = glob.glob(tmp)
if (os.system("rename .png8 .png %s*" % tmp)):
print("Error in saving file.")
if (os.system("optipng -quiet %s*" % tmp)):
print("Error in optipng.")
commstr = 'cwebp -quiet -lossless -z 9 -metadata exif %s -o %s' % (
out, out.replace(".png", ".webp"))
if (os.system(commstr)):
print("Error in cwebp")
def convert(filename, page=0, res=120):
reader = PdfFileReader(filename, strict=False)
page_obj = reader.getPage(page)
dst_pdf = PdfFileWriter()
dst_pdf.addPage(page_obj)
pdf_bytes = io.BytesIO()
dst_pdf.write(pdf_bytes)
pdf_bytes.seek(0)
img = Image(file=pdf_bytes, resolution=res)
img.format = 'png'
img.compression_quality = 90
img.background_color = Color('white')
img_path = filename.replace('pdf', 'png')
img.save(filename=img_path)
img.destroy()
def convert(filename, res=120):
pdf_file = PdfFileReader(open(filename, "rb"))
pageObj = pdf_file.getPage(0)
dst_pdf = PdfFileWriter()
dst_pdf.addPage(pageObj)
pdf_bytes = io.BytesIO()
dst_pdf.write(pdf_bytes)
pdf_bytes.seek(0)
img_filename = filename.replace('old', 'img').replace('pdf', 'png')
img = Image(file=pdf_bytes, resolution=res)
img.format = 'png'
img.compression_quality = 120
img.background_color = Color("white")
img.save(filename=img_filename)
img.destroy()
def _convert_pdf2img(self, file, store_path):
pdf_file = PdfFileReader(file, strict=False)
pageObj = pdf_file.getPage(0)
dst_pdf = PdfFileWriter()
dst_pdf.addPage(pageObj)
pdf_bytes = io.BytesIO()
dst_pdf.write(pdf_bytes)
pdf_bytes.seek(0)
img = Image(file=pdf_bytes, resolution=200)
img.format = 'jpg'
img.compression_quality = 90
img.background_color = Color("white")
img_path = store_path + '%s.jpg' % (uuid.uuid3(uuid.uuid1(), file))
print(img_path)
img.save(filename=img_path)
img.destroy()
return img_path
def convertPdfToImage(file_loc,doc_id):
print doc_id
new_dir, file_ending = file_loc.split(".")
if file_ending == 'pdf':
if not os.path.exists(new_dir):
os.makedirs(new_dir)
doc = Image(filename=file_loc,resolution=300)
doc.compression_quality = 20
for i, page in enumerate(doc.sequence):
with Image(page) as page_image:
page_image.alpha_channel = False
page_image.background_color = Color('white')
save_location = new_dir+"/img-{}.jpg".format(str(i).zfill(4))
page_image.save(filename=save_location)
t1 = threading.Thread(target=extractTextInImage,args=(file_loc,doc_id))
t1.start()
t2 = threading.Thread(target=updateImgMatchesForNewDoc,args=(doc_id,))
t2.start()
def _run_convert(pdfile, savedfilename, page_index, index, res=120):
pageObj = pdfile.getPage(page_index) #获取pdf的第page_index页
dst_pdf = PdfFileWriter()
dst_pdf.addPage(pageObj)
pdf_bytes = io.BytesIO()
dst_pdf.write(pdf_bytes)
pdf_bytes.seek(0)
img = Image(file=pdf_bytes, resolution=res)
img.format = 'png'
img.compression_quality = 90
img.background_color = Color("white")
img_path = '%s%04d.jpg' % (savedfilename, index)
img.save(filename=img_path)
print(img_path)
img.destroy()
def _run_convert(filename, page, res=120):
idx = page + 1
pdfile = getPdfReader(filename)
pageObj = pdfile.getPage(page)
dst_pdf = PdfFileWriter()
dst_pdf.addPage(pageObj)
pdf_bytes = io.BytesIO()
dst_pdf.write(pdf_bytes)
pdf_bytes.seek(0)
img = Image(file=pdf_bytes, resolution=res)
img.format = 'jpg'
img.compression_quality = 90
img.background_color = Color("white")
img_path = '%s%d.jpg' % (filename[:filename.rindex('.')], idx)
img.save(filename=img_path)
img.destroy()
def picture_save_api(request_file,
original_file_size=(0, 0),
original_file_type='jpeg',
compress_file_size=(0, 0),
compress_file_type='jpeg'):
pic = Picture()
pic.save()
original_image = Image(file=urlopen(request_file))
compress_image = original_image.clone()
original_image.format = original_file_type
compress_image.format = compress_file_type
w, h = original_image.size
original_width = w
compress_width = w
original_height = h
compress_height = h
if original_file_size[0] > 0:
original_width = original_file_size[0]
if original_file_size[1] > 0:
original_height = original_file_size[1]
if compress_file_size[0] > 0:
compress_width = compress_file_size[0]
if compress_file_size[1] > 0:
compress_height = compress_file_size[1]
original_image.resize(original_width, original_height)
original_image.compression_quality = 60
compress_image.resize(compress_width, compress_height)
compress_image.compression_quality = 60
pic.original_image.save(
str(pic.id) + u'_o.' + original_file_type,
ContentFile(original_image.make_blob()))
pic.compress_image.save(
str(pic.id) + u'_c.' + compress_file_type,
ContentFile(compress_image.make_blob()))
return pic
def set_image(self, binary):
self.binary = binary
with Image(blob=self.binary, resolution=150) as img:
self.extension = img.format.lower()
flattened = Image(background=Color("white"),
height=img.height, width=img.width)
flattened.composite(img, left=0, top=0)
flattened.format = "jpeg"
flattened.compression_quality = 50
thumbnail = flattened.clone()
thumbnail.transform(resize='150x200>')
self.thumbnail = thumbnail.make_blob()
preview = flattened.clone()
preview.gaussian_blur(radius=1, sigma=0.5)
preview.transform(resize='612x792>')
self.preview = preview.make_blob()
self.save()
def pdf_to_image(pdf, page=1, size=800, file_format='jpeg', quality=80):
"""Creates a image file from pdf file"""
try:
pdf.open()
except Exception as e:
raise e
else:
# do this while the pdf is open
reader = PyPDF2.PdfFileReader(pdf, strict=False)
writer = PyPDF2.PdfFileWriter()
page_content = reader.getPage(page - 1)
box = page_content.cropBox
dims = [float(a - b) for a, b in zip(box.upperRight, box.lowerLeft)]
scaleby = size / max(dims)
dims = [int(d * scaleby) for d in dims]
# resize_page(page_content, size)
writer.addPage(page_content)
outputStream = BytesIO()
writer.write(outputStream)
outputStream.seek(0)
finally:
pdf.close()
# put content of page in a new image
foreground = WandImage(
blob=outputStream,
format='pdf',
resolution=int(1.6 * 72 * scaleby),
)
# make sure the color space is correct.
# this prevents an occational bug where rgb colours are inverted
foreground.type = 'truecolormatte'
foreground.resize(*dims, 25)
# white background
background = WandImage(
width=foreground.width,
height=foreground.height,
background=Color('white')
)
background.format = file_format
background.composite(foreground, 0, 0)
background.compression_quality = quality
return background
def pages(self):
for page in range(self.file.numPages):
img = WandImage(filename=self.path + ('[%s]' % page),
resolution=self.config['wand_resolution'])
img.compression_quality = self.config['wand_compression_quality']
temp = NamedTemporaryFile(suffix='.jpg')
# Passing temp as file kwargs does not work for some reason.
# So we just pass the filename.
img.save(filename=temp.name)
# Reopen the image file as PIL object
img = Image.open(temp.name)
# Run tesseract
tr = Tesseract()
result = tr.ocr_image(img)
temp.close()
yield result
def convertPdfToImage(file_loc, doc_id):
print doc_id
new_dir, file_ending = file_loc.split(".")
if file_ending == 'pdf':
if not os.path.exists(new_dir):
os.makedirs(new_dir)
doc = Image(filename=file_loc, resolution=300)
doc.compression_quality = 20
for i, page in enumerate(doc.sequence):
with Image(page) as page_image:
page_image.alpha_channel = False
page_image.background_color = Color('white')
save_location = new_dir + "/img-{}.jpg".format(str(i).zfill(4))
page_image.save(filename=save_location)
t1 = threading.Thread(target=extractTextInImage,
args=(file_loc, doc_id))
t1.start()
t2 = threading.Thread(target=updateImgMatchesForNewDoc,
args=(doc_id, ))
t2.start()
def run_convert(filename, res):
pdfile = getPdfReader(filename)
pages = pdfile.getNumPages()
for page in range(pages):
print('正在解析%d页' % (page+1))
pageObj = pdfile.getPage(page)
dst_pdf = PdfFileWriter()
dst_pdf.addPage(pageObj)
pdf_bytes = io.BytesIO()
dst_pdf.write(pdf_bytes)
pdf_bytes.seek(0)
img = Image(file=pdf_bytes, resolution=res)
img.compression_quality = 90
img.background_color = Color("white")
# img.save(filename = 'C:\\Users\\SXL47\\Desktop\\tupian.jpg')
blob = img.make_blob('jpg')
img.destroy()
yield blob
def pdf_to_image(pdf, page=1, size=800, file_format='jpeg', quality=80):
"""Creates a image file from pdf file"""
try:
pdf.open()
except Exception as e:
raise e
else:
# do this while the pdf is open
reader = PyPDF2.PdfFileReader(pdf, strict=False)
writer = PyPDF2.PdfFileWriter()
page_content = reader.getPage(page - 1)
box = page_content.cropBox
dims = [float(a - b) for a, b in zip(box.upperRight, box.lowerLeft)]
scaleby = size / max(dims)
dims = [int(d * scaleby) for d in dims]
# resize_page(page_content, size)
writer.addPage(page_content)
outputStream = BytesIO()
writer.write(outputStream)
outputStream.seek(0)
finally:
pdf.close()
# put content of page in a new image
foreground = WandImage(
blob=outputStream,
format='pdf',
resolution=int(1.6 * 72 * scaleby),
)
# make sure the color space is correct.
# this prevents an occational bug where rgb colours are inverted
foreground.type = 'truecolormatte'
foreground.resize(*dims, 25)
# white background
background = WandImage(width=foreground.width,
height=foreground.height,
background=Color('white'))
background.format = file_format
background.composite(foreground, 0, 0)
background.compression_quality = quality
return background
def convert_scanned_pdf_to_png(self):
"""
Resolution = 300 and compression_quality = 99 are optimal to assure the image's quality.
:return: file converted to JPG
"""
img = Img(filename=self.file, resolution=300) ### Opens scanned pdf file
img.compression_quality = 99 ### Sets compression quality to 72
### Reassign self.img_file according to self.file
try:
self.img_file = self.file.split('/')[-1]
self.img_file = self.img_file.replace('.pdf', '.png')
self.img_file = os.path.join(ParseScannedPdf.temp_folder, self.img_file)
except Exception as e:
print 'Be sure to use forward slash when assigning the path of the pdf or img file.'
print e.args
sys.exit()
self.type_pdf = True
img.save(filename=self.img_file) ### Converts the file to png
time.sleep(2)
def resize_image(self,
img_path,
size_config,
size_name,
dest_path,
compress=False):
#open image
img = Image(filename=img_path)
#transform using resize config
img.transform(resize=size_config)
'''
exProcessor = ExifData()
#rotate image if necessary
if exProcessor.check_orientation(img) in [6, 7]:
img.rotate(90)
img.metadata['exif:Orientation'] = 1
if exProcessor.check_orientation(img) in [6, 8]:
img.rotate(-90)
img.metadata['exif:Orientation'] = 1
if exProcessor.check_orientation(img) in [3, 4]:
img.rotate(180)
img.metadata['exif:Orientation'] = 1
'''
# Detect file extention
fileName, fileExtension = os.path.splitext(img_path)
#save img
fileName = ''.join([fileName, '_', size_name, fileExtension])
fileName = os.path.basename(fileName)
fileName = os.path.abspath(os.path.join(dest_path, fileName))
# Compress option?
if compress:
img.compression_quality = 70
img.save(filename=fileName)
return fileName
def _run_convert(pdfile, savedfilename, page_index, index, res=240):
# http://www.imooc.com/wenda/detail/600387
# 在Python中从PDF提取图像而无需重新采样？
#
pageObj = pdfile.getPage(page_index) #获取pdf的第page_index页
dst_pdf = PdfFileWriter()
dst_pdf.addPage(pageObj)
pdf_bytes = io.BytesIO()
dst_pdf.write(pdf_bytes)
pdf_bytes.seek(0)
img = Image(file=pdf_bytes, resolution=res)
img.format = 'png'
img.compression_quality = 100
img.background_color = Color("white")
number = index + 1
img_path = '%s页面_%02d.jpg' % (savedfilename, number)
#-------------------------------------------------------------------------------------------------------------------
# Save Images
#
img.save(filename=img_path)
print("extract " + img_path)
img.destroy()
def run_convert(self, filename, page, res=120):
# idx = page + 1
pdfile = getPdfReader(filename)
pageObj = pdfile.getPage(page)
dst_pdf = PdfFileWriter()
dst_pdf.addPage(pageObj)
pdf_bytes = io.BytesIO()
dst_pdf.write(pdf_bytes)
pdf_bytes.seek(0)
img = Image(file=pdf_bytes, resolution=res)
img.format = 'png'
img.compression_quality = 100
img.background_color = Color("white")
filename = filename.split("/")[-1]
img_path = 'resource/%s%s.png' % (filename[:filename.rindex('.')],
"temp")
delete_file("temp", root='resource/')
img.save(filename=img_path)
img.destroy()
return img_path
def test_wand():
image_data = Image(blob=s)
image_data.compression_quality = 70
is_gallery_card = False
image_data_size = (image_data.width, image_data.height)
image_size = (720, 720)
if is_gallery_card:
image_size = (720, 1120)
if image_size != image_data_size:
result = image_data.resize(image_size[0], image_size[1])
with image_data.clone() as img:
result = img = crop_image(img, (720, 472))
# 공감전용 카드의 경우 댓글용 이미지를 생성하지 않는다.
if not is_gallery_card:
result = img.resize(460, 310)
result = image_data.resize(360, 360)
result = image_data.resize(132, 132)
return result
def convert_pdf_to_img(blob, img_type="jpg", quality=75, resolution=200):
"""
Converts PDF with multiple pages into one image.
It needs the file content NOT the filename or ioBytes and returns the image content.
Note: It has memory leak!!
http://stackoverflow.com/a/26233785/1497443
Example:
with open('my.pdf', "r") as f:
file_content = f.read()
# import ipdb
# ipdb.set_trace()
hh = convert_pdf_to_jpg(file_content)
with open('my.jpg', 'wb') as f:
f.write(hh)
"""
from wand.image import Image as WandImage
from wand.color import Color as WandColor
pdf = WandImage(blob=blob, resolution=resolution)
pages = len(pdf.sequence)
wimage = WandImage(width=pdf.width,
height=pdf.height * pages,
background=WandColor("white"))
for i in xrange(pages):
wimage.composite(pdf.sequence[i], top=pdf.height * i, left=0)
if img_type == "jpg":
wimage.compression_quality = quality
return wimage.make_blob(img_type)
def make_pages(request, document_pk):
document = Document.objects.get(pk=document_pk)
clear_pages(document)
pdf_file_name = document.document_file._get_path()
pdf_file = Image(filename=pdf_file_name, resolution=CONFIG['RESOLUTION'])
tmp_dir = tempfile.mkdtemp()
for page in pdf_file.sequence:
page_number = page.index + 1
page_img = Image(page)
page_img.compression_quality = CONFIG['COMPRESSION_QUALITY']
page_img_file_name = tmp_dir + \
'/page{0}-{1}.jpg'.format(document_pk, page_number)
page_img.save(filename=page_img_file_name)
create_page(document, page_number, page_img_file_name)
shutil.rmtree(tmp_dir)
messages.add_message(
request,
messages.INFO,
'Make pages for "{0}"'.format(document)
)
def picture_save_api(request_file, original_file_size=(0, 0), original_file_type='jpeg', compress_file_size=(0, 0), compress_file_type='jpeg'):
pic = Picture()
pic.save()
original_image = Image(file=urlopen(request_file))
compress_image = original_image.clone()
original_image.format = original_file_type
compress_image.format = compress_file_type
w, h = original_image.size
original_width = w
compress_width = w
original_height = h
compress_height = h
if original_file_size[0] > 0:
original_width = original_file_size[0]
if original_file_size[1] > 0:
original_height = original_file_size[1]
if compress_file_size[0] > 0:
compress_width = compress_file_size[0]
if compress_file_size[1] > 0:
compress_height = compress_file_size[1]
original_image.resize(original_width, original_height)
original_image.compression_quality = 60
compress_image.resize(compress_width, compress_height)
compress_image.compression_quality = 60
pic.original_image.save(str(pic.id) + u'_o.' + original_file_type, ContentFile(original_image.make_blob()))
pic.compress_image.save(str(pic.id) + u'_c.' + compress_file_type, ContentFile(compress_image.make_blob()))
return pic
boxWidth > minFontWidth and boxWidth < maxFontWidth: digit.resize(boxWidth, boxHeight, 'cubic') draw.composite(operator=operator, left=boxLeft, top=boxTop, width=boxWidth, height=boxHeight, image=digit) else: digit.resize(int(round(fontWidth*defaultFontScale)), int(round(fontHeight*defaultFontScale)), 'cubic') draw.composite(operator=operator, left=round((boxLeft+boxRight-fontWidth*defaultFontScale)/2.0), top=round((boxTop+boxBottom-fontHeight*defaultFontScale)/2.0), width=fontWidth*defaultFontScale, height=fontHeight*defaultFontScale, image=digit) characterIndex += 1 col += 1 if (octalDigitIndex % 6) == 4: col += 1 if (octalDigitIndex % 6) == 5: col += 7 characterIndex += 1 boxIndex += 1 # Next row, please row += 1 if (index % 4) == 3: row += 1.2 draw(img) # Create the output image. img.format = 'jpg' img.compression_quality = 25 img.save(filename=outImage) print 'output =', outImage
