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 6.7KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263
  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/models/db"
  15. "code.gitea.io/gitea/modules/git"
  16. "code.gitea.io/gitea/modules/graceful"
  17. "code.gitea.io/gitea/modules/log"
  18. "code.gitea.io/gitea/modules/queue"
  19. "code.gitea.io/gitea/modules/setting"
  20. "code.gitea.io/gitea/modules/storage"
  21. )
  22. // ArchiveRequest defines the parameters of an archive request, which notably
  23. // includes the specific repository being archived as well as the commit, the
  24. // name by which it was requested, and the kind of archive being requested.
  25. // This is entirely opaque to external entities, though, and mostly used as a
  26. // handle elsewhere.
  27. type ArchiveRequest struct {
  28. RepoID int64
  29. refName string
  30. Type git.ArchiveType
  31. CommitID string
  32. }
  33. // SHA1 hashes will only go up to 40 characters, but SHA256 hashes will go all
  34. // the way to 64.
  35. var shaRegex = regexp.MustCompile(`^[0-9a-f]{4,64}$`)
  36. // NewRequest creates an archival request, based on the URI. The
  37. // resulting ArchiveRequest is suitable for being passed to ArchiveRepository()
  38. // if it's determined that the request still needs to be satisfied.
  39. func NewRequest(repoID int64, repo *git.Repository, uri string) (*ArchiveRequest, error) {
  40. r := &ArchiveRequest{
  41. RepoID: repoID,
  42. }
  43. var ext string
  44. switch {
  45. case strings.HasSuffix(uri, ".zip"):
  46. ext = ".zip"
  47. r.Type = git.ZIP
  48. case strings.HasSuffix(uri, ".tar.gz"):
  49. ext = ".tar.gz"
  50. r.Type = git.TARGZ
  51. case strings.HasSuffix(uri, ".bundle"):
  52. ext = ".bundle"
  53. r.Type = git.BUNDLE
  54. default:
  55. return nil, fmt.Errorf("Unknown format: %s", uri)
  56. }
  57. r.refName = strings.TrimSuffix(uri, ext)
  58. var err error
  59. // Get corresponding commit.
  60. if repo.IsBranchExist(r.refName) {
  61. r.CommitID, err = repo.GetBranchCommitID(r.refName)
  62. if err != nil {
  63. return nil, err
  64. }
  65. } else if repo.IsTagExist(r.refName) {
  66. r.CommitID, err = repo.GetTagCommitID(r.refName)
  67. if err != nil {
  68. return nil, err
  69. }
  70. } else if shaRegex.MatchString(r.refName) {
  71. if repo.IsCommitExist(r.refName) {
  72. r.CommitID = r.refName
  73. } else {
  74. return nil, git.ErrNotExist{
  75. ID: r.refName,
  76. }
  77. }
  78. } else {
  79. return nil, fmt.Errorf("Unknow ref %s type", r.refName)
  80. }
  81. return r, nil
  82. }
  83. // GetArchiveName returns the name of the caller, based on the ref used by the
  84. // caller to create this request.
  85. func (aReq *ArchiveRequest) GetArchiveName() string {
  86. return strings.ReplaceAll(aReq.refName, "/", "-") + "." + aReq.Type.String()
  87. }
  88. func doArchive(r *ArchiveRequest) (*models.RepoArchiver, error) {
  89. ctx, committer, err := db.TxContext()
  90. if err != nil {
  91. return nil, err
  92. }
  93. defer committer.Close()
  94. archiver, err := models.GetRepoArchiver(ctx, r.RepoID, r.Type, r.CommitID)
  95. if err != nil {
  96. return nil, err
  97. }
  98. if archiver != nil {
  99. // FIXME: If another process are generating it, we think it's not ready and just return
  100. // Or we should wait until the archive generated.
  101. if archiver.Status == models.RepoArchiverGenerating {
  102. return nil, nil
  103. }
  104. } else {
  105. archiver = &models.RepoArchiver{
  106. RepoID: r.RepoID,
  107. Type: r.Type,
  108. CommitID: r.CommitID,
  109. Status: models.RepoArchiverGenerating,
  110. }
  111. if err := models.AddRepoArchiver(ctx, archiver); err != nil {
  112. return nil, err
  113. }
  114. }
  115. rPath, err := archiver.RelativePath()
  116. if err != nil {
  117. return nil, err
  118. }
  119. _, err = storage.RepoArchives.Stat(rPath)
  120. if err == nil {
  121. if archiver.Status == models.RepoArchiverGenerating {
  122. archiver.Status = models.RepoArchiverReady
  123. if err = models.UpdateRepoArchiverStatus(ctx, archiver); err != nil {
  124. return nil, err
  125. }
  126. }
  127. return archiver, committer.Commit()
  128. }
  129. if !errors.Is(err, os.ErrNotExist) {
  130. return nil, fmt.Errorf("unable to stat archive: %v", err)
  131. }
  132. rd, w := io.Pipe()
  133. defer func() {
  134. w.Close()
  135. rd.Close()
  136. }()
  137. var done = make(chan error)
  138. repo, err := archiver.LoadRepo()
  139. if err != nil {
  140. return nil, fmt.Errorf("archiver.LoadRepo failed: %v", err)
  141. }
  142. gitRepo, err := git.OpenRepository(repo.RepoPath())
  143. if err != nil {
  144. return nil, err
  145. }
  146. defer gitRepo.Close()
  147. go func(done chan error, w *io.PipeWriter, archiver *models.RepoArchiver, gitRepo *git.Repository) {
  148. defer func() {
  149. if r := recover(); r != nil {
  150. done <- fmt.Errorf("%v", r)
  151. }
  152. }()
  153. if archiver.Type == git.BUNDLE {
  154. err = gitRepo.CreateBundle(
  155. graceful.GetManager().ShutdownContext(),
  156. archiver.CommitID,
  157. w,
  158. )
  159. } else {
  160. err = gitRepo.CreateArchive(
  161. graceful.GetManager().ShutdownContext(),
  162. archiver.Type,
  163. w,
  164. setting.Repository.PrefixArchiveFiles,
  165. archiver.CommitID,
  166. )
  167. }
  168. _ = w.CloseWithError(err)
  169. done <- err
  170. }(done, w, archiver, gitRepo)
  171. // TODO: add lfs data to zip
  172. // TODO: add submodule data to zip
  173. if _, err := storage.RepoArchives.Save(rPath, rd, -1); err != nil {
  174. return nil, fmt.Errorf("unable to write archive: %v", err)
  175. }
  176. err = <-done
  177. if err != nil {
  178. return nil, err
  179. }
  180. if archiver.Status == models.RepoArchiverGenerating {
  181. archiver.Status = models.RepoArchiverReady
  182. if err = models.UpdateRepoArchiverStatus(ctx, archiver); err != nil {
  183. return nil, err
  184. }
  185. }
  186. return archiver, committer.Commit()
  187. }
  188. // ArchiveRepository satisfies the ArchiveRequest being passed in. Processing
  189. // will occur in a separate goroutine, as this phase may take a while to
  190. // complete. If the archive already exists, ArchiveRepository will not do
  191. // anything. In all cases, the caller should be examining the *ArchiveRequest
  192. // being returned for completion, as it may be different than the one they passed
  193. // in.
  194. func ArchiveRepository(request *ArchiveRequest) (*models.RepoArchiver, error) {
  195. return doArchive(request)
  196. }
  197. var archiverQueue queue.UniqueQueue
  198. // Init initlize archive
  199. func Init() error {
  200. handler := func(data ...queue.Data) {
  201. for _, datum := range data {
  202. archiveReq, ok := datum.(*ArchiveRequest)
  203. if !ok {
  204. log.Error("Unable to process provided datum: %v - not possible to cast to IndexerData", datum)
  205. continue
  206. }
  207. log.Trace("ArchiverData Process: %#v", archiveReq)
  208. if _, err := doArchive(archiveReq); err != nil {
  209. log.Error("Archive %v faild: %v", datum, err)
  210. }
  211. }
  212. }
  213. archiverQueue = queue.CreateUniqueQueue("repo-archive", handler, new(ArchiveRequest))
  214. if archiverQueue == nil {
  215. return errors.New("unable to create codes indexer queue")
  216. }
  217. go graceful.GetManager().RunWithShutdownFns(archiverQueue.Run)
  218. return nil
  219. }
  220. // StartArchive push the archive request to the queue
  221. func StartArchive(request *ArchiveRequest) error {
  222. has, err := archiverQueue.Has(request)
  223. if err != nil {
  224. return err
  225. }
  226. if has {
  227. return nil
  228. }
  229. return archiverQueue.Push(request)
  230. }