Fix bug with missing downloadURL for docs/sheets

Google Docs + Sheets do not populate 'DownloadUrl', but
instead provide an map of exportLinks ie csv, docx, pptx,
txt etc. Simple heuristic to determine the url to export
as well as create the target extension to avoid clobbering
of the original file on a re-sync.

README.md

@@ -49,6 +49,9 @@ Background sync is not just hard, it's stupid. My technical and philosophical ra
 * Probably, it doesn't work on Windows.
 * Google Drive allows a directory to contain files/directories with the same name. Client doesn't handle these cases yet. We don't recommend you to use `drive` if you have such files/directories to avoid data loss.
 * Racing conditions occur if remote is being modified while we're trying to update the file. Google Drive provides resource versioning with ETags, use Etags to avoid racy cases.
+* Google Docs + Sheets + Presentations data  cannot be downloaded raw but only
+as exported to different forms e.g docx, xlsx, csv etc hence doing a pull of
+these types will result in a exported document.
 
 ## License
 Copyright 2013 Google Inc. All Rights Reserved.

pull.go

@@ -18,6 +18,7 @@ import (
 	"fmt"
 	"io"
 	"os"
+	"strings"
 	"path/filepath"
 	"sync"
 )
@@ -26,6 +27,22 @@ const (
 	maxNumOfConcPullTasks = 4
 )
 
+func docExportsMap() *map[string][]string {
+	return &map[string][]string {
+		"text/plain": []string{"text/plain", "txt",},
+		"application/vnd.google-apps.drawing": []string{"image/svg+xml", "svg+xml",},
+		"application/vnd.google-apps.spreadsheet": []string{
+		"application/vnd.openxmlformats-officedocument.spreadsheetml.sheet", "xlsx",
+		},
+		"application/vnd.google-apps.document": []string{
+			"application/vnd.openxmlformats-officedocument.wordprocessingml.document", "docx",
+		},
+		"application/vnd.google-apps.presentation": []string{
+			"application/vnd.openxmlformats-officedocument.presentationml.presentation", "pptx",
+		},
+	}
+}
+
 // Pull from remote if remote path exists and in a god context. If path is a
 // directory, it recursively pulls from the remote if there are remote changes.
 // It doesn't check if there are remote changes if isForce is set.
@@ -90,7 +107,8 @@ func (g *Commands) localMod(wg *sync.WaitGroup, change *Change) (err error) {
 	defer g.taskDone()
 	defer wg.Done()
 	destAbsPath := g.context.AbsPathOf(change.Path)
-	if change.Src.BlobAt != "" {
+
+	if change.Src.BlobAt != "" || change.Src.ExportLinks != nil {
 		// download and replace
 		if err = g.download(change); err != nil {
 			return
@@ -108,7 +126,7 @@ func (g *Commands) localAdd(wg *sync.WaitGroup, change *Change) (err error) {
 	if change.Src.IsDir {
 		return os.Mkdir(destAbsPath, os.ModeDir|0755)
 	}
-	if change.Src.BlobAt != "" {
+	if change.Src.BlobAt != "" || change.Src.ExportLinks != nil {
 		// download and create
 		if err = g.download(change); err != nil {
 			return
@@ -124,7 +142,31 @@ func (g *Commands) localDelete(wg *sync.WaitGroup, change *Change) (err error) {
 }
 
 func (g *Commands) download(change *Change) (err error) {
-	destAbsPath := g.context.AbsPathOf(change.Path)
+	exportUrl := ""
+	baseName := change.Path
+
+	// If BlobAt is not set, we are most likely dealing with
+	// Document/SpreadSheet/Image. In this case we'll use the target
+	// exportable type since we cannot directly download the raw data.
+	// We also need to pay attention and add the exported extension
+	// to avoid overriding the original file on re-syncing.
+	if len(change.Src.BlobAt) < 1 {
+		var ok bool
+		var mimeKeyExtList[]string
+
+		exportsMap := *docExportsMap()
+		mimeKeyExtList, ok = exportsMap[change.Src.MimeType]
+		if !ok {
+			mimeKeyExtList = []string{"text/plain", "txt"}
+		}
+
+		exportUrl = change.Src.ExportLinks[mimeKeyExtList[0]]
+		fmt.Print("Exported ", baseName)
+		baseName = strings.Join([]string{baseName, mimeKeyExtList[1]}, ".")
+		fmt.Println(" to: ", baseName)
+	}
+
+	destAbsPath := g.context.AbsPathOf(baseName)
 	var fo *os.File
 	fo, err = os.Create(destAbsPath)
 	if err != nil {
@@ -144,7 +186,7 @@ func (g *Commands) download(change *Change) (err error) {
 			blob.Close()
 		}
 	}()
-	blob, err = g.rem.Download(change.Src.Id)
+	blob, err = g.rem.Download(change.Src.Id, exportUrl)
 	if err != nil {
 		return err
 	}

remote.go

@@ -120,8 +120,14 @@ func (r *Remote) Publish(id string) (string, error) {
 	return "https://googledrive.com/host/" + id, nil
 }
 
-func (r *Remote) Download(id string) (io.ReadCloser, error) {
-	resp, err := r.transport.Client().Get("https://googledrive.com/host/" + id)
+func (r *Remote) Download(id string, exportUrl string) (io.ReadCloser, error) {
+	var url string
+	if len(exportUrl) < 1 {
+		url = "https://googledrive.com/host/" + id
+	} else {
+		url = exportUrl
+	}
+	resp, err := r.transport.Client().Get(url)
 	if err != nil || resp.StatusCode < 200 || resp.StatusCode > 299 {
 		return resp.Body, err
 	}

types.go

@@ -38,7 +38,9 @@ type File struct {
 	ModTime     time.Time
 	Size        int64
 	BlobAt      string
+	MimeType    string
 	Md5Checksum string
+	ExportLinks map[string]string
 }
 
 func NewRemoteFile(f *drive.File) *File {
@@ -50,8 +52,10 @@ func NewRemoteFile(f *drive.File) *File {
 		IsDir:       f.MimeType == "application/vnd.google-apps.folder",
 		ModTime:     mtime,
 		Size:        f.FileSize,
+		MimeType:    f.MimeType,
 		BlobAt:      f.DownloadUrl,
 		Md5Checksum: f.Md5Checksum,
+		ExportLinks: f.ExportLinks,
 	}
 }