Allows you to define a maximum file size, and will split zip file output based on that size. Each zip file produced is a completely functional archive, i.e. it doesn't generate a single huge zip file then split that into pieces.

Given a folder with a set of files of various sizes

sources/10mb-32.txt	10000000
sources/11mb-32.txt	11000000
sources/1mb-32.txt	1000000
sources/2mb-32.txt	2000000
sources/3mb-32.txt	3000000
sources/4mb-32.txt	4000000
sources/5mb-32.txt	5000000
sources/6mb-32.txt	6000000
sources/7mb-32.txt	7000000
sources/8mb-32.txt	8000000
sources/9mb-32.txt	9000000

the following program will archive all the text files above

package main

import (
	"fmt"
	"github.com/jar-o/splitzip"
)

func main() {
	fmt.Printf("Zip(), any error? %+v\n", splitzip.Zip("sources", "sources/zip", "test", 10000000, .4))
}

and will produce a numbered sequence of zip files smaller than the defined "max size" of 10MB:

sources/zip/test0.zip	9444579
sources/zip/test1.zip	9444588
sources/zip/test2.zip	9444622
sources/zip/test3.zip	9444513
sources/zip/test4.zip	3778674

The final parameter .4 in the call to splitzip.Zip() above is the expected average space savings for this set of files. Because the density of files will vary, and hence their compressibility, you can use this number to optimize for your set of file types. If you're not certain, just leave it as 0.