You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

archiver.go 6.6KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260
  1. // Copyright 2020 The Gitea Authors.
  2. // All rights reserved.
  3. // Use of this source code is governed by a MIT-style
  4. // license that can be found in the LICENSE file.
  5. package archiver
  6. import (
  7. "errors"
  8. "fmt"
  9. "io"
  10. "os"
  11. "regexp"
  12. "strings"
  13. "code.gitea.io/gitea/models"
  14. "code.gitea.io/gitea/modules/git"
  15. "code.gitea.io/gitea/modules/graceful"
  16. "code.gitea.io/gitea/modules/log"
  17. "code.gitea.io/gitea/modules/queue"
  18. "code.gitea.io/gitea/modules/setting"
  19. "code.gitea.io/gitea/modules/storage"
  20. )
  21. // ArchiveRequest defines the parameters of an archive request, which notably
  22. // includes the specific repository being archived as well as the commit, the
  23. // name by which it was requested, and the kind of archive being requested.
  24. // This is entirely opaque to external entities, though, and mostly used as a
  25. // handle elsewhere.
  26. type ArchiveRequest struct {
  27. RepoID int64
  28. refName string
  29. Type git.ArchiveType
  30. CommitID string
  31. }
  32. // SHA1 hashes will only go up to 40 characters, but SHA256 hashes will go all
  33. // the way to 64.
  34. var shaRegex = regexp.MustCompile(`^[0-9a-f]{4,64}$`)
  35. // NewRequest creates an archival request, based on the URI. The
  36. // resulting ArchiveRequest is suitable for being passed to ArchiveRepository()
  37. // if it's determined that the request still needs to be satisfied.
  38. func NewRequest(repoID int64, repo *git.Repository, uri string) (*ArchiveRequest, error) {
  39. r := &ArchiveRequest{
  40. RepoID: repoID,
  41. }
  42. var ext string
  43. switch {
  44. case strings.HasSuffix(uri, ".zip"):
  45. ext = ".zip"
  46. r.Type = git.ZIP
  47. case strings.HasSuffix(uri, ".tar.gz"):
  48. ext = ".tar.gz"
  49. r.Type = git.TARGZ
  50. case strings.HasSuffix(uri, ".bundle"):
  51. ext = ".bundle"
  52. r.Type = git.BUNDLE
  53. default:
  54. return nil, fmt.Errorf("Unknown format: %s", uri)
  55. }
  56. r.refName = strings.TrimSuffix(uri, ext)
  57. var err error
  58. // Get corresponding commit.
  59. if repo.IsBranchExist(r.refName) {
  60. r.CommitID, err = repo.GetBranchCommitID(r.refName)
  61. if err != nil {
  62. return nil, err
  63. }
  64. } else if repo.IsTagExist(r.refName) {
  65. r.CommitID, err = repo.GetTagCommitID(r.refName)
  66. if err != nil {
  67. return nil, err
  68. }
  69. } else if shaRegex.MatchString(r.refName) {
  70. if repo.IsCommitExist(r.refName) {
  71. r.CommitID = r.refName
  72. } else {
  73. return nil, git.ErrNotExist{
  74. ID: r.refName,
  75. }
  76. }
  77. } else {
  78. return nil, fmt.Errorf("Unknow ref %s type", r.refName)
  79. }
  80. return r, nil
  81. }
  82. // GetArchiveName returns the name of the caller, based on the ref used by the
  83. // caller to create this request.
  84. func (aReq *ArchiveRequest) GetArchiveName() string {
  85. return strings.ReplaceAll(aReq.refName, "/", "-") + "." + aReq.Type.String()
  86. }
  87. func doArchive(r *ArchiveRequest) (*models.RepoArchiver, error) {
  88. ctx, commiter, err := models.TxDBContext()
  89. if err != nil {
  90. return nil, err
  91. }
  92. defer commiter.Close()
  93. archiver, err := models.GetRepoArchiver(ctx, r.RepoID, r.Type, r.CommitID)
  94. if err != nil {
  95. return nil, err
  96. }
  97. if archiver != nil {
  98. // FIXME: If another process are generating it, we think it's not ready and just return
  99. // Or we should wait until the archive generated.
  100. if archiver.Status == models.RepoArchiverGenerating {
  101. return nil, nil
  102. }
  103. } else {
  104. archiver = &models.RepoArchiver{
  105. RepoID: r.RepoID,
  106. Type: r.Type,
  107. CommitID: r.CommitID,
  108. Status: models.RepoArchiverGenerating,
  109. }
  110. if err := models.AddRepoArchiver(ctx, archiver); err != nil {
  111. return nil, err
  112. }
  113. }
  114. rPath, err := archiver.RelativePath()
  115. if err != nil {
  116. return nil, err
  117. }
  118. _, err = storage.RepoArchives.Stat(rPath)
  119. if err == nil {
  120. if archiver.Status == models.RepoArchiverGenerating {
  121. archiver.Status = models.RepoArchiverReady
  122. return archiver, models.UpdateRepoArchiverStatus(ctx, archiver)
  123. }
  124. return archiver, nil
  125. }
  126. if !errors.Is(err, os.ErrNotExist) {
  127. return nil, fmt.Errorf("unable to stat archive: %v", err)
  128. }
  129. rd, w := io.Pipe()
  130. defer func() {
  131. w.Close()
  132. rd.Close()
  133. }()
  134. var done = make(chan error)
  135. repo, err := archiver.LoadRepo()
  136. if err != nil {
  137. return nil, fmt.Errorf("archiver.LoadRepo failed: %v", err)
  138. }
  139. gitRepo, err := git.OpenRepository(repo.RepoPath())
  140. if err != nil {
  141. return nil, err
  142. }
  143. defer gitRepo.Close()
  144. go func(done chan error, w *io.PipeWriter, archiver *models.RepoArchiver, gitRepo *git.Repository) {
  145. defer func() {
  146. if r := recover(); r != nil {
  147. done <- fmt.Errorf("%v", r)
  148. }
  149. }()
  150. if archiver.Type == git.BUNDLE {
  151. err = gitRepo.CreateBundle(
  152. graceful.GetManager().ShutdownContext(),
  153. archiver.CommitID,
  154. w,
  155. )
  156. } else {
  157. err = gitRepo.CreateArchive(
  158. graceful.GetManager().ShutdownContext(),
  159. archiver.Type,
  160. w,
  161. setting.Repository.PrefixArchiveFiles,
  162. archiver.CommitID,
  163. )
  164. }
  165. _ = w.CloseWithError(err)
  166. done <- err
  167. }(done, w, archiver, gitRepo)
  168. // TODO: add lfs data to zip
  169. // TODO: add submodule data to zip
  170. if _, err := storage.RepoArchives.Save(rPath, rd, -1); err != nil {
  171. return nil, fmt.Errorf("unable to write archive: %v", err)
  172. }
  173. err = <-done
  174. if err != nil {
  175. return nil, err
  176. }
  177. if archiver.Status == models.RepoArchiverGenerating {
  178. archiver.Status = models.RepoArchiverReady
  179. if err = models.UpdateRepoArchiverStatus(ctx, archiver); err != nil {
  180. return nil, err
  181. }
  182. }
  183. return archiver, commiter.Commit()
  184. }
  185. // ArchiveRepository satisfies the ArchiveRequest being passed in. Processing
  186. // will occur in a separate goroutine, as this phase may take a while to
  187. // complete. If the archive already exists, ArchiveRepository will not do
  188. // anything. In all cases, the caller should be examining the *ArchiveRequest
  189. // being returned for completion, as it may be different than the one they passed
  190. // in.
  191. func ArchiveRepository(request *ArchiveRequest) (*models.RepoArchiver, error) {
  192. return doArchive(request)
  193. }
  194. var archiverQueue queue.UniqueQueue
  195. // Init initlize archive
  196. func Init() error {
  197. handler := func(data ...queue.Data) {
  198. for _, datum := range data {
  199. archiveReq, ok := datum.(*ArchiveRequest)
  200. if !ok {
  201. log.Error("Unable to process provided datum: %v - not possible to cast to IndexerData", datum)
  202. continue
  203. }
  204. log.Trace("ArchiverData Process: %#v", archiveReq)
  205. if _, err := doArchive(archiveReq); err != nil {
  206. log.Error("Archive %v faild: %v", datum, err)
  207. }
  208. }
  209. }
  210. archiverQueue = queue.CreateUniqueQueue("repo-archive", handler, new(ArchiveRequest))
  211. if archiverQueue == nil {
  212. return errors.New("unable to create codes indexer queue")
  213. }
  214. go graceful.GetManager().RunWithShutdownFns(archiverQueue.Run)
  215. return nil
  216. }
  217. // StartArchive push the archive request to the queue
  218. func StartArchive(request *ArchiveRequest) error {
  219. has, err := archiverQueue.Has(request)
  220. if err != nil {
  221. return err
  222. }
  223. if has {
  224. return nil
  225. }
  226. return archiverQueue.Push(request)
  227. }