omerh/fcompare

Simple file comparison written in go

fcompare's Issues

Code review

fcompare/main.go

Line 18 in 9985c4b

type filesInfo struct {

What is the filesInfo type for? You don't use it!

fcompare/main.go

Line 53 in 9985c4b

filesInformation := []*fileInfo{}

Why pointer? The fileInfo struct occupies only a handful of words, copying it is cheaper than dereferencing the pointer.
Why not allocate the []fileInfo properly - you know its probable size:
filesInformation := make([]fileInfo, 0, len(files))

fcompare/main.go

Line 56 in 9985c4b

if !file.IsDir() {

if !file.Mode().IsRegular() { continue }

fcompare/main.go

Line 68 in 9985c4b

for i := 0; i < len(filesInformation); i++ {

for _, s := range filesInformation

fcompare/main.go

Line 84 in 9985c4b

filesInformation = append(filesInformation[:i], filesInformation[i+1:]...)

Why do you remove it? Your outer iteration is over it, your inner compareFilesSlice is not , but this line is here for its shadowing side effect, not what's written!

By the way, this algorithm is quite inefficient: you hash the files at least twice - at least cache the hash value into fileInfo (which can be a pointer in this case, or re-add it to the map after filling the cache),
and use a map[int64][]fileInfo to group fileInfos by size.