Friday, November 6, 2020

7-zip optimization: sorting by file type

 Recently I was compressing the assets I scraped from a now-deleted Discord server and was able to save almost 600 MiB by having 7-zip sort by file type (the default behavior is to sort by filename) in addition to a few other options. This lets 7-zip take advantage of similarities between files of the same type.


The original data was 6,661,786,104 bytes, with the default compression it was 6,041,019,462 bytes (90.682%), and with "sort by file type" enabled, it was 5,427,406,392 bytes (81.471%). In this case, the optional parameters reduced the size of the output by 613,613,070 bytes (585.19 MiB).

This article was updated on February 11, 2021 because one of the parameters, "tr=on", doesn't exist and causes an error. I changed a few other parameters as well and updated the results in the previous paragraph because the new ones saved an extra 2.15 MiB.