Now I know that size = os.stat(real_path).st_size This size is not the length and width of the so-called 1920×1080 picture The comparison of the size has already beat me.

The principle is to compare the picture size and md5 to delete duplicate pictures. The great god wrote such a script in 13 years. Later, my colleague finally found the script written by the great god on the Internet, and it ran in seconds. Use md5 to compare whether the two pictures are duplicates, but this double loop traversal, I don’t know which year it will run (I didn’t test which pixel comparison and md5 comparison is faster, anyway, if there are many pictures, it is almost equivalent to waste) Later, I was reminded to start not comparing it pixel by pixel. After comparing a picture, I deleted it from the list. I added the xml file corresponding to this picture and deleted it at the same time.

It is really slow, anyway, the code ran for a night, and more than 10,000 pictures were not traversed His script is to compare two pictures pixel by pixel in a double loop of bubbling to see if the two pictures are duplicated. Well, the beginning of Caiji, I really couldn’t believe another Caiji. Recently, there is a batch of data, about 10,000 pictures, a lot of duplicates, I started to write a script to delete duplicate pictures and xml filesīefore I started, I found that a colleague’s legacy had this type of script, and decided to modify it on this basis.

Script to delete duplicate files mac