fix: update vfs/zip implementation to ensure minimum range requests for go1.17

Using `gitlab.com/gitlab-org/golang-archive-zip` instead of the `archive/zip`
package from the stdlib.

Changelog: changed

4 files changed, 88 insertions, 1 deletions

diff --git a/go.mod b/go.mod
index 2cc393a5..f947d3cc 100644
--- a/go.mod
+++ b/go.mod
@@ -25,6 +25,7 @@ require (
 	github.com/tj/assert v0.0.3 // indirect
 	github.com/tj/go-redirects v0.0.0-20180508180010-5c02ead0bbc5
 	gitlab.com/gitlab-org/go-mimedb v1.45.0
+	gitlab.com/gitlab-org/golang-archive-zip v0.1.1
 	gitlab.com/gitlab-org/labkit v1.11.0
 	golang.org/x/crypto v0.0.0-20200622213623-75b288015ac9
 	golang.org/x/net v0.0.0-20210503060351-7fd8e65b6420
diff --git a/go.sum b/go.sum
index 842c41f2..3df973d5 100644
--- a/go.sum
+++ b/go.sum
@@ -443,6 +443,8 @@ github.com/yuin/goldmark v1.2.1/go.mod h1:3hX8gzYuyVAZsxl0MRgGTJEmQBFcNTphYh9dec
 github.com/yuin/goldmark v1.3.5/go.mod h1:mwnBkeHKe2W/ZEtQ+71ViKU8L12m81fl3OWwC1Zlc8k=
 gitlab.com/gitlab-org/go-mimedb v1.45.0 h1:PO8dx6HEWzPYU6MQTYnCbpQEJzhJLW/Bh43+2VUHTgc=
 gitlab.com/gitlab-org/go-mimedb v1.45.0/go.mod h1:wa9y/zOSFKmTXLyBs4clz2FNVhZQmmEQM9TxslPAjZ0=
+gitlab.com/gitlab-org/golang-archive-zip v0.1.1 h1:35k9giivbxwF03+8A05Cm8YoxoakU8FBCj5gysjCTCE=
+gitlab.com/gitlab-org/golang-archive-zip v0.1.1/go.mod h1:ZDtqpWPGPB9qBuZnZDrKQjIdJtkN7ZAoVwhT6H2o2kE=
 gitlab.com/gitlab-org/labkit v1.11.0 h1:oUi02wj097Xr8NBkyDEHQGtCThI/oocoYSAv9sT8BLU=
 gitlab.com/gitlab-org/labkit v1.11.0/go.mod h1:tnvyKiPF/Y0MeBy+ACYWRWexOEQk6PTRGUc9GstcnXc=
 go.opencensus.io v0.21.0/go.mod h1:mSImk1erAIZhrmZN+AvHh14ztQfjbGwt4TtuofqLduU=
diff --git a/internal/vfs/zip/archive.go b/internal/vfs/zip/archive.go
index 588fb76d..551d10c0 100644
--- a/internal/vfs/zip/archive.go
+++ b/internal/vfs/zip/archive.go
@@ -1,7 +1,6 @@
 package zip
 
 import (
-	"archive/zip"
 	"context"
 	"errors"
 	"fmt"
@@ -14,6 +13,9 @@ import (
 	"sync/atomic"
 	"time"
 
+	// TODO: Revert back to zip/archive once we no longer support go1.17
+	// https://gitlab.com/gitlab-org/gitlab-pages/-/issues/673
+	zip "gitlab.com/gitlab-org/golang-archive-zip"
 	"gitlab.com/gitlab-org/labkit/log"
 
 	"gitlab.com/gitlab-org/gitlab-pages/internal/httprange"
diff --git a/internal/vfs/zip/archive_test.go b/internal/vfs/zip/archive_test.go
index 423d9eaf..37529e22 100644
--- a/internal/vfs/zip/archive_test.go
+++ b/internal/vfs/zip/archive_test.go
@@ -5,11 +5,14 @@ import (
 	"bytes"
 	"context"
 	"crypto/rand"
+	"fmt"
 	"io"
 	"net/http"
 	"net/http/httptest"
 	"os"
+	"path/filepath"
 	"strconv"
+	"strings"
 	"sync/atomic"
 	"testing"
 	"time"
@@ -382,6 +385,85 @@ func TestReadArchiveFails(t *testing.T) {
 	require.EqualError(t, err, os.ErrNotExist.Error())
 }
 
+func createArchive(t *testing.T, dir string) (map[string][]byte, int64) {
+	t.Helper()
+
+	f, err := os.Create(filepath.Join(dir, "public.zip"))
+	require.NoError(t, err)
+	defer f.Close()
+	zw := zip.NewWriter(f)
+
+	entries := make(map[string][]byte)
+	for _, size := range []int{0, 32 * 1024, 128 * 1024, 5 * 1024 * 1024} {
+		entryName := fmt.Sprintf("public/file_%d", size)
+		entries[entryName] = bytes.Repeat([]byte{'z'}, size)
+
+		w, err := zw.Create(entryName)
+		require.NoError(t, err)
+
+		_, err = w.Write(entries[entryName])
+		require.NoError(t, err)
+	}
+
+	require.NoError(t, zw.Close())
+
+	fi, err := f.Stat()
+	require.NoError(t, err)
+	require.NoError(t, f.Close())
+
+	return entries, fi.Size()
+}
+
+func TestMinimalRangeRequests(t *testing.T) {
+	dir := t.TempDir()
+	entries, size := createArchive(t, dir)
+
+	mux := http.NewServeMux()
+
+	var ranges []string
+	mux.HandleFunc("/", func(rw http.ResponseWriter, r *http.Request) {
+		rangeHdr := r.Header.Get("Range")
+		if rangeHdr == "" {
+			rw.Header().Add("Content-Length", fmt.Sprintf("%d", size))
+			return
+		}
+
+		ranges = append(ranges, rangeHdr)
+
+		http.FileServer(http.Dir(dir)).ServeHTTP(rw, r)
+	})
+
+	srv := httptest.NewServer(mux)
+	defer srv.Close()
+
+	ctx, cancel := context.WithCancel(context.Background())
+	defer cancel()
+
+	fs := New(&zipCfg).(*zipVFS)
+	zip := newArchive(fs, time.Second)
+	err := zip.openArchive(ctx, srv.URL+"/public.zip")
+	require.NoError(t, err)
+
+	require.Len(t, zip.files, len(entries))
+	require.Len(t, ranges, 3, "range requests should be minimal")
+
+	for _, zf := range zip.files {
+		if !zf.Mode().IsRegular() {
+			continue
+		}
+
+		f, err := zip.Open(context.Background(), strings.TrimPrefix(zf.Name, "public/"))
+		require.NoError(t, err)
+
+		io.Copy(io.Discard, f)
+
+		require.NoError(t, f.Close())
+	}
+
+	// ensure minimal requests: https://gitlab.com/gitlab-org/gitlab-pages/-/issues/625
+	require.Len(t, ranges, 11, "range requests should be minimal")
+}
+
 func openZipArchive(t *testing.T, requests *int64, fromDisk bool) (*zipArchive, func()) {
 	t.Helper()