Abin Simon 4ee66c3d3d
Refactor http client creation for onedrive file download (#1736)
## Description

This way we are able to use any middlewares added in both places.

## Type of change

<!--- Please check the type of change your PR introduces: --->
- [ ] 🌻 Feature
- [ ] 🐛 Bugfix
- [ ] 🗺️ Documentation
- [ ] 🤖 Test
- [ ] 💻 CI/Deployment
- [x] 🐹 Trivial/Minor

## Issue(s)

<!-- Can reference multiple issues. Use one of the following "magic words" - "closes, fixes" to auto-close the Github issue. -->
* #<issue>

## Test Plan

<!-- How will this be tested prior to merging.-->
- [ ] 💪 Manual
- [ ]  Unit test
- [x] 💚 E2E
2022-12-09 04:59:17 +00:00

186 lines
5.5 KiB
Go

package onedrive
import (
"context"
"io"
"time"
"github.com/microsoftgraph/msgraph-sdk-go/models"
"github.com/pkg/errors"
"github.com/alcionai/corso/src/internal/connector/graph"
"github.com/alcionai/corso/src/internal/connector/support"
"github.com/alcionai/corso/src/internal/connector/uploadsession"
"github.com/alcionai/corso/src/pkg/backup/details"
"github.com/alcionai/corso/src/pkg/logger"
)
const (
// downloadUrlKey is used to find the download URL in a
// DriveItem response
downloadURLKey = "@microsoft.graph.downloadUrl"
)
// sharePointItemReader will return a io.ReadCloser for the specified item
// It crafts this by querying M365 for a download URL for the item
// and using a http client to initialize a reader
func sharePointItemReader(
ctx context.Context,
service graph.Service,
driveID, itemID string,
) (details.ItemInfo, io.ReadCloser, error) {
item, rc, err := driveItemReader(ctx, service, driveID, itemID)
if err != nil {
return details.ItemInfo{}, nil, err
}
dii := details.ItemInfo{
SharePoint: sharePointItemInfo(item, *item.GetSize()),
}
return dii, rc, nil
}
// oneDriveItemReader will return a io.ReadCloser for the specified item
// It crafts this by querying M365 for a download URL for the item
// and using a http client to initialize a reader
func oneDriveItemReader(
ctx context.Context,
service graph.Service,
driveID, itemID string,
) (details.ItemInfo, io.ReadCloser, error) {
item, rc, err := driveItemReader(ctx, service, driveID, itemID)
if err != nil {
return details.ItemInfo{}, nil, err
}
dii := details.ItemInfo{
OneDrive: oneDriveItemInfo(item, *item.GetSize()),
}
return dii, rc, nil
}
// driveItemReader will return a io.ReadCloser for the specified item
// It crafts this by querying M365 for a download URL for the item
// and using a http client to initialize a reader
func driveItemReader(
ctx context.Context,
service graph.Service,
driveID, itemID string,
) (models.DriveItemable, io.ReadCloser, error) {
logger.Ctx(ctx).Debugw("Reading Item", "id", itemID, "time", time.Now())
item, err := service.Client().DrivesById(driveID).ItemsById(itemID).Get(ctx, nil)
if err != nil {
return nil, nil, errors.Wrapf(err, "failed to get item %s", itemID)
}
// Get the download URL - https://docs.microsoft.com/en-us/graph/api/driveitem-get-content
// These URLs are pre-authenticated and can be used to download the data using the standard
// http client
if _, found := item.GetAdditionalData()[downloadURLKey]; !found {
return nil, nil, errors.Errorf("file does not have a download URL. ID: %s, %#v",
itemID, item.GetAdditionalData())
}
downloadURL := item.GetAdditionalData()[downloadURLKey].(*string)
httpClient := graph.CreateHTTPClient()
httpClient.Timeout = 0 // infinite timeout for pulling large files
resp, err := httpClient.Get(*downloadURL)
if err != nil {
return nil, nil, errors.Wrapf(err, "failed to download file from %s", *downloadURL)
}
return item, resp.Body, nil
}
// oneDriveItemInfo will populate a details.OneDriveInfo struct
// with properties from the drive item. ItemSize is specified
// separately for restore processes because the local itemable
// doesn't have its size value updated as a side effect of creation,
// and kiota drops any SetSize update.
func oneDriveItemInfo(di models.DriveItemable, itemSize int64) *details.OneDriveInfo {
ed, ok := di.GetCreatedBy().GetUser().GetAdditionalData()["email"]
email := ""
if ok {
email = *ed.(*string)
}
return &details.OneDriveInfo{
ItemType: details.OneDriveItem,
ItemName: *di.GetName(),
Created: *di.GetCreatedDateTime(),
Modified: *di.GetLastModifiedDateTime(),
Size: itemSize,
Owner: email,
}
}
// sharePointItemInfo will populate a details.SharePointInfo struct
// with properties from the drive item. ItemSize is specified
// separately for restore processes because the local itemable
// doesn't have its size value updated as a side effect of creation,
// and kiota drops any SetSize update.
func sharePointItemInfo(di models.DriveItemable, itemSize int64) *details.SharePointInfo {
var (
id string
url string
)
// TODO: we rely on this info for details/restore lookups,
// so if it's nil we have an issue, and will need an alternative
// way to source the data.
gsi := di.GetSharepointIds()
if gsi != nil {
if gsi.GetSiteId() != nil {
id = *gsi.GetSiteId()
}
if gsi.GetSiteUrl() != nil {
url = *gsi.GetSiteUrl()
}
}
return &details.SharePointInfo{
ItemType: details.OneDriveItem,
ItemName: *di.GetName(),
Created: *di.GetCreatedDateTime(),
Modified: *di.GetLastModifiedDateTime(),
Size: itemSize,
Owner: id,
WebURL: url,
}
}
// driveItemWriter is used to initialize and return an io.Writer to upload data for the specified item
// It does so by creating an upload session and using that URL to initialize an `itemWriter`
// TODO: @vkamra verify if var session is the desired input
func driveItemWriter(
ctx context.Context,
service graph.Service,
driveID, itemID string,
itemSize int64,
) (io.Writer, error) {
session := models.NewUsersItemDrivesItemItemsItemCreateUploadSessionPostRequestBody()
r, err := service.Client().DrivesById(driveID).ItemsById(itemID).CreateUploadSession().Post(ctx, session, nil)
if err != nil {
return nil, errors.Wrapf(
err,
"failed to create upload session for item %s. details: %s",
itemID,
support.ConnectorStackErrorTrace(err),
)
}
url := *r.GetUploadUrl()
logger.Ctx(ctx).Debugf("Created an upload session for item %s. URL: %s", itemID, url)
return uploadsession.NewWriter(itemID, url, itemSize), nil
}