zipstream

Package zipstream is an on-the-fly extractor/reader for zip archive similar as Java's java.util.zip.ZipInputStream, there is no need to provide io.ReaderAt and archive size, that is, just only one io.Reader parameter.

Implementation

Most code in this package is directly copied from Golang standard archive/zip library, you can read the ZIP specification from here.

Usage

go get github.com/zhyee/zipstream

Examples

package main

import (
	"io"
	"log"
	"net/http"

	"github.com/zhyee/zipstream"
)

func main() {

	resp, err := http.Get("https://github.com/golang/go/archive/refs/tags/go1.16.10.zip")
	if err != nil {
		log.Fatal(err)
	}
	defer resp.Body.Close()

	zr := zipstream.NewReader(resp.Body)

	for zr.Next() {
		e, err := zr.Entry()
		if err != nil {
			log.Fatalf("unable to get next entry: %s", err)
		}

		log.Println("entry name: ", e.Name)
		log.Println("entry comment: ", e.Comment)
		log.Println("entry reader version: ", e.ReaderVersion)
		log.Println("entry modify time: ", e.Modified)
		log.Println("entry compressed size: ", e.CompressedSize64)
		log.Println("entry uncompressed size: ", e.UncompressedSize64)
		log.Println("entry is a dir: ", e.IsDir())

		if !e.IsDir() {
			rc, err := e.Open()
			if err != nil {
				log.Fatalf("unable to open zip file: %s", err)
			}
			content, err := io.ReadAll(rc)
			if err != nil {
				log.Fatalf("read zip file content fail: %s", err)
			}

			log.Println("file length:", len(content))

			if uint64(len(content)) != e.UncompressedSize64 {
				log.Fatalf("read zip file length not equal with UncompressedSize64")
			}
			if err := rc.Close(); err != nil {
				log.Fatalf("close zip entry reader fail: %s", err)
			}
		}
	}
	if err = zr.Err(); err != nil {
		log.Fatalf("encounter error: %v", err)
    }
}

Limitations

• Repeatable read is not supported, each file in the zip archive can be read only once per Reader instance.
• Some central directory header fields are unresolved — including version made by, internal file attributes, external file attributes and relative offset of local header. Additionally, certain other central directory header fields (e.g., extra field) may differ from their counterparts in the local file header.
• Concurrent reading of multiple files is not supported.