I'd like to process all JPEG thumbnails generated with easy-thumbnail via PIL thru jpegoptim .
Using PIL's optimization: image.save(..,optimize=1,...)
isn't optimizing much at all.
For example:
Can anyone point me to Python examples or libraries that integrate jpegoptim?
You can use thumbnail_created
signal and call external app via subporecess.Popen
. I just realize this in my project. You can even optimize images when they uploaded using saved_file
signal!
Here is my code:
import subprocess
from os.path import splitext
from django.dispatch import receiver
from easy_thumbnails.signals import saved_file, thumbnail_created
@receiver(saved_file)
def optimize_file(sender, fieldfile, **kwargs):
optimize(fieldfile.path)
@receiver(thumbnail_created)
def optimize_thumbnail(sender, **kwargs):
optimize(sender.path)
def optimize(path):
runString = {
".jpeg": u"jpegoptim -f --strip-all '%(file)s'",
".jpg": u"jpegoptim -f --strip-all '%(file)s'",
".png": u"optipng -force -o7 '%(file)s' && advpng -z4 '%(file)s' && pngcrush -rem gAMA -rem alla -rem cHRM -rem iCCP -rem sRGB -rem time '%(file)s' '%(file)s.bak' && mv '%(file)s.bak' '%(file)s'"
}
ext = splitext(path)[1].lower()
if ext in runString:
subprocess.Popen(runString[ext] % {'file': path}, shell=True)
runString
taken from trimage . On Debian, you need to install following packages: jpegoptim optipng pngcrush advancecomp
. Or just use another tools, such as smush.py
.
I also found this project which encapsulates code above, has gif support and better filetype recognition.
I found https://github.com/thebeansgroup/smush.py which is a lossless image optimiser in Python >=2.7. I went with https://github.com/beatak/smush.py which is a fork that works for Python >= 2.5, since we are using debian stable on our server.
It uses:
Hopefully using pngnq multiple times on files doesn't degrade quality, we plan to run this script on all uploaded media weekly.
I doubt that there are any python bindings to jpegoptim. The options I can think of are:
The technical post webpages of this site follow the CC BY-SA 4.0 protocol. If you need to reprint, please indicate the site URL or the original address.Any question please contact:yoyou2525@163.com.