2 files changed, 382 insertions, 0 deletions
diff --git a/pkg/filecollector/file_collector.go b/pkg/filecollector/file_collector.go
new file mode 100644
index 0000000..8547bb7
--- /dev/null
+++ b/pkg/filecollector/file_collector.go
@@ -0,0 +1,210 @@
+package filecollector
+
+import (
+	"archive/tar"
+	"context"
+	"fmt"
+	"io"
+	"io/fs"
+	"os"
+	"path"
+	"path/filepath"
+	"strings"
+
+	git "github.com/go-git/go-git/v5"
+	"github.com/go-git/go-git/v5/plumbing/filemode"
+	"github.com/go-git/go-git/v5/plumbing/format/gitignore"
+	"github.com/go-git/go-git/v5/plumbing/format/index"
+)
+
+type Handler interface {
+	WriteFile(path string, fi fs.FileInfo, linkName string, f io.Reader) error
+}
+
+type TarCollector struct {
+	TarWriter *tar.Writer
+	UID       int
+	GID       int
+	DstDir    string
+}
+
+func (tc TarCollector) WriteFile(fpath string, fi fs.FileInfo, linkName string, f io.Reader) error {
+	// create a new dir/file header
+	header, err := tar.FileInfoHeader(fi, linkName)
+	if err != nil {
+		return err
+	}
+
+	// update the name to correctly reflect the desired destination when untaring
+	header.Name = path.Join(tc.DstDir, fpath)
+	header.Mode = int64(fi.Mode())
+	header.ModTime = fi.ModTime()
+	header.Uid = tc.UID
+	header.Gid = tc.GID
+
+	// write the header
+	if err := tc.TarWriter.WriteHeader(header); err != nil {
+		return err
+	}
+
+	// this is a symlink no reader provided
+	if f == nil {
+		return nil
+	}
+
+	// copy file data into tar writer
+	if _, err := io.Copy(tc.TarWriter, f); err != nil {
+		return err
+	}
+	return nil
+}
+
+type CopyCollector struct {
+	DstDir string
+}
+
+func (cc *CopyCollector) WriteFile(fpath string, fi fs.FileInfo, linkName string, f io.Reader) error {
+	fdestpath := filepath.Join(cc.DstDir, fpath)
+	if err := os.MkdirAll(filepath.Dir(fdestpath), 0o777); err != nil {
+		return err
+	}
+	if f == nil {
+		return os.Symlink(linkName, fdestpath)
+	}
+	df, err := os.OpenFile(fdestpath, os.O_CREATE|os.O_WRONLY, fi.Mode())
+	if err != nil {
+		return err
+	}
+	defer df.Close()
+	if _, err := io.Copy(df, f); err != nil {
+		return err
+	}
+	return nil
+}
+
+type FileCollector struct {
+	Ignorer   gitignore.Matcher
+	SrcPath   string
+	SrcPrefix string
+	Fs        Fs
+	Handler   Handler
+}
+
+type Fs interface {
+	Walk(root string, fn filepath.WalkFunc) error
+	OpenGitIndex(path string) (*index.Index, error)
+	Open(path string) (io.ReadCloser, error)
+	Readlink(path string) (string, error)
+}
+
+type DefaultFs struct {
+}
+
+func (*DefaultFs) Walk(root string, fn filepath.WalkFunc) error {
+	return filepath.Walk(root, fn)
+}
+
+func (*DefaultFs) OpenGitIndex(path string) (*index.Index, error) {
+	r, err := git.PlainOpen(path)
+	if err != nil {
+		return nil, err
+	}
+	i, err := r.Storer.Index()
+	if err != nil {
+		return nil, err
+	}
+	return i, nil
+}
+
+func (*DefaultFs) Open(path string) (io.ReadCloser, error) {
+	return os.Open(path)
+}
+
+func (*DefaultFs) Readlink(path string) (string, error) {
+	return os.Readlink(path)
+}
+
+//nolint:gocyclo
+func (fc *FileCollector) CollectFiles(ctx context.Context, submodulePath []string) filepath.WalkFunc {
+	i, _ := fc.Fs.OpenGitIndex(path.Join(fc.SrcPath, path.Join(submodulePath...)))
+	return func(file string, fi os.FileInfo, err error) error {
+		if err != nil {
+			return err
+		}
+		if ctx != nil {
+			select {
+			case <-ctx.Done():
+				return fmt.Errorf("copy cancelled")
+			default:
+			}
+		}
+
+		sansPrefix := strings.TrimPrefix(file, fc.SrcPrefix)
+		split := strings.Split(sansPrefix, string(filepath.Separator))
+		// The root folders should be skipped, submodules only have the last path component set to "." by filepath.Walk
+		if fi.IsDir() && len(split) > 0 && split[len(split)-1] == "." {
+			return nil
+		}
+		var entry *index.Entry
+		if i != nil {
+			entry, err = i.Entry(strings.Join(split[len(submodulePath):], "/"))
+		} else {
+			err = index.ErrEntryNotFound
+		}
+		if err != nil && fc.Ignorer != nil && fc.Ignorer.Match(split, fi.IsDir()) {
+			if fi.IsDir() {
+				if i != nil {
+					ms, err := i.Glob(strings.Join(append(split[len(submodulePath):], "**"), "/"))
+					if err != nil || len(ms) == 0 {
+						return filepath.SkipDir
+					}
+				} else {
+					return filepath.SkipDir
+				}
+			} else {
+				return nil
+			}
+		}
+		if err == nil && entry.Mode == filemode.Submodule {
+			err = fc.Fs.Walk(file, fc.CollectFiles(ctx, split))
+			if err != nil {
+				return err
+			}
+			return filepath.SkipDir
+		}
+		path := filepath.ToSlash(sansPrefix)
+
+		// return on non-regular files (thanks to [kumo](https://medium.com/@komuw/just-like-you-did-fbdd7df829d3) for this suggested update)
+		if fi.Mode()&os.ModeSymlink == os.ModeSymlink {
+			linkName, err := fc.Fs.Readlink(file)
+			if err != nil {
+				return fmt.Errorf("unable to readlink '%s': %w", file, err)
+			}
+			return fc.Handler.WriteFile(path, fi, linkName, nil)
+		} else if !fi.Mode().IsRegular() {
+			return nil
+		}
+
+		// open file
+		f, err := fc.Fs.Open(file)
+		if err != nil {
+			return err
+		}
+		defer f.Close()
+
+		if ctx != nil {
+			// make io.Copy cancellable by closing the file
+			cpctx, cpfinish := context.WithCancel(ctx)
+			defer cpfinish()
+			go func() {
+				select {
+				case <-cpctx.Done():
+				case <-ctx.Done():
+					f.Close()
+				}
+			}()
+		}
+
+		return fc.Handler.WriteFile(path, fi, "", f)
+	}
+}
diff --git a/pkg/filecollector/file_collector_test.go b/pkg/filecollector/file_collector_test.go
new file mode 100644
index 0000000..60a8d4d
--- /dev/null
+++ b/pkg/filecollector/file_collector_test.go
@@ -0,0 +1,172 @@
+package filecollector
+
+import (
+	"archive/tar"
+	"context"
+	"io"
+	"path/filepath"
+	"strings"
+	"testing"
+
+	"github.com/go-git/go-billy/v5"
+	"github.com/go-git/go-billy/v5/memfs"
+	git "github.com/go-git/go-git/v5"
+	"github.com/go-git/go-git/v5/plumbing/cache"
+	"github.com/go-git/go-git/v5/plumbing/format/gitignore"
+	"github.com/go-git/go-git/v5/plumbing/format/index"
+	"github.com/go-git/go-git/v5/storage/filesystem"
+	"github.com/stretchr/testify/assert"
+)
+
+type memoryFs struct {
+	billy.Filesystem
+}
+
+func (mfs *memoryFs) walk(root string, fn filepath.WalkFunc) error {
+	dir, err := mfs.ReadDir(root)
+	if err != nil {
+		return err
+	}
+	for i := 0; i < len(dir); i++ {
+		filename := filepath.Join(root, dir[i].Name())
+		err = fn(filename, dir[i], nil)
+		if dir[i].IsDir() {
+			if err == filepath.SkipDir {
+				err = nil
+			} else if err := mfs.walk(filename, fn); err != nil {
+				return err
+			}
+		}
+		if err != nil {
+			return err
+		}
+	}
+	return nil
+}
+
+func (mfs *memoryFs) Walk(root string, fn filepath.WalkFunc) error {
+	stat, err := mfs.Lstat(root)
+	if err != nil {
+		return err
+	}
+	err = fn(strings.Join([]string{root, "."}, string(filepath.Separator)), stat, nil)
+	if err != nil {
+		return err
+	}
+	return mfs.walk(root, fn)
+}
+
+func (mfs *memoryFs) OpenGitIndex(path string) (*index.Index, error) {
+	f, _ := mfs.Filesystem.Chroot(filepath.Join(path, ".git"))
+	storage := filesystem.NewStorage(f, cache.NewObjectLRUDefault())
+	i, err := storage.Index()
+	if err != nil {
+		return nil, err
+	}
+	return i, nil
+}
+
+func (mfs *memoryFs) Open(path string) (io.ReadCloser, error) {
+	return mfs.Filesystem.Open(path)
+}
+
+func (mfs *memoryFs) Readlink(path string) (string, error) {
+	return mfs.Filesystem.Readlink(path)
+}
+
+func TestIgnoredTrackedfile(t *testing.T) {
+	fs := memfs.New()
+	_ = fs.MkdirAll("mygitrepo/.git", 0o777)
+	dotgit, _ := fs.Chroot("mygitrepo/.git")
+	worktree, _ := fs.Chroot("mygitrepo")
+	repo, _ := git.Init(filesystem.NewStorage(dotgit, cache.NewObjectLRUDefault()), worktree)
+	f, _ := worktree.Create(".gitignore")
+	_, _ = f.Write([]byte(".*\n"))
+	f.Close()
+	// This file shouldn't be in the tar
+	f, _ = worktree.Create(".env")
+	_, _ = f.Write([]byte("test=val1\n"))
+	f.Close()
+	w, _ := repo.Worktree()
+	// .gitignore is in the tar after adding it to the index
+	_, _ = w.Add(".gitignore")
+
+	tmpTar, _ := fs.Create("temp.tar")
+	tw := tar.NewWriter(tmpTar)
+	ps, _ := gitignore.ReadPatterns(worktree, []string{})
+	ignorer := gitignore.NewMatcher(ps)
+	fc := &FileCollector{
+		Fs:        &memoryFs{Filesystem: fs},
+		Ignorer:   ignorer,
+		SrcPath:   "mygitrepo",
+		SrcPrefix: "mygitrepo" + string(filepath.Separator),
+		Handler: &TarCollector{
+			TarWriter: tw,
+		},
+	}
+	err := fc.Fs.Walk("mygitrepo", fc.CollectFiles(context.Background(), []string{}))
+	assert.NoError(t, err, "successfully collect files")
+	tw.Close()
+	_, _ = tmpTar.Seek(0, io.SeekStart)
+	tr := tar.NewReader(tmpTar)
+	h, err := tr.Next()
+	assert.NoError(t, err, "tar must not be empty")
+	assert.Equal(t, ".gitignore", h.Name)
+	_, err = tr.Next()
+	assert.ErrorIs(t, err, io.EOF, "tar must only contain one element")
+}
+
+func TestSymlinks(t *testing.T) {
+	fs := memfs.New()
+	_ = fs.MkdirAll("mygitrepo/.git", 0o777)
+	dotgit, _ := fs.Chroot("mygitrepo/.git")
+	worktree, _ := fs.Chroot("mygitrepo")
+	repo, _ := git.Init(filesystem.NewStorage(dotgit, cache.NewObjectLRUDefault()), worktree)
+	// This file shouldn't be in the tar
+	f, err := worktree.Create(".env")
+	assert.NoError(t, err)
+	_, err = f.Write([]byte("test=val1\n"))
+	assert.NoError(t, err)
+	f.Close()
+	err = worktree.Symlink(".env", "test.env")
+	assert.NoError(t, err)
+
+	w, err := repo.Worktree()
+	assert.NoError(t, err)
+
+	// .gitignore is in the tar after adding it to the index
+	_, err = w.Add(".env")
+	assert.NoError(t, err)
+	_, err = w.Add("test.env")
+	assert.NoError(t, err)
+
+	tmpTar, _ := fs.Create("temp.tar")
+	tw := tar.NewWriter(tmpTar)
+	ps, _ := gitignore.ReadPatterns(worktree, []string{})
+	ignorer := gitignore.NewMatcher(ps)
+	fc := &FileCollector{
+		Fs:        &memoryFs{Filesystem: fs},
+		Ignorer:   ignorer,
+		SrcPath:   "mygitrepo",
+		SrcPrefix: "mygitrepo" + string(filepath.Separator),
+		Handler: &TarCollector{
+			TarWriter: tw,
+		},
+	}
+	err = fc.Fs.Walk("mygitrepo", fc.CollectFiles(context.Background(), []string{}))
+	assert.NoError(t, err, "successfully collect files")
+	tw.Close()
+	_, _ = tmpTar.Seek(0, io.SeekStart)
+	tr := tar.NewReader(tmpTar)
+	h, err := tr.Next()
+	files := map[string]tar.Header{}
+	for err == nil {
+		files[h.Name] = *h
+		h, err = tr.Next()
+	}
+
+	assert.Equal(t, ".env", files[".env"].Name)
+	assert.Equal(t, "test.env", files["test.env"].Name)
+	assert.Equal(t, ".env", files["test.env"].Linkname)
+	assert.ErrorIs(t, err, io.EOF, "tar must be read cleanly to EOF")
+}