func IndexGitRepo()

in gitindex/index.go [345:493]
123 lines of code
30 McCabe index (conditional complexity)

func IndexGitRepo(opts Options) error {
	// Set max thresholds, since we use them in this function.
	opts.BuildOptions.SetDefaults()
	if opts.RepoDir == "" {
		return fmt.Errorf("gitindex: must set RepoDir")
	}

	opts.BuildOptions.RepositoryDescription.Source = opts.RepoDir
	repo, err := git.PlainOpen(opts.RepoDir)
	if err != nil {
		return err
	}

	if err := setTemplatesFromConfig(&opts.BuildOptions.RepositoryDescription, opts.RepoDir); err != nil {
		log.Printf("setTemplatesFromConfig(%s): %s", opts.RepoDir, err)
	}

	repoCache := NewRepoCache(opts.RepoCacheDir)

	// branch => (path, sha1) => repo.
	repos := map[fileKey]BlobLocation{}

	// fileKey => branches
	branchMap := map[fileKey][]string{}

	// Branch => Repo => SHA1
	branchVersions := map[string]map[string]plumbing.Hash{}

	branches, err := expandBranches(repo, opts.Branches, opts.BranchPrefix)
	if err != nil {
		return err
	}
	for _, b := range branches {
		commit, err := getCommit(repo, opts.BranchPrefix, b)
		if err != nil {
			if opts.AllowMissingBranch && err.Error() == "reference not found" {
				continue
			}

			return err
		}

		opts.BuildOptions.RepositoryDescription.Branches = append(opts.BuildOptions.RepositoryDescription.Branches, zoekt.RepositoryBranch{
			Name:    b,
			Version: commit.Hash.String(),
		})

		tree, err := commit.Tree()
		if err != nil {
			return err
		}

		files, subVersions, err := TreeToFiles(repo, tree, opts.BuildOptions.RepositoryDescription.URL, repoCache)
		if err != nil {
			return err
		}
		for k, v := range files {
			repos[k] = v
			branchMap[k] = append(branchMap[k], b)
		}

		branchVersions[b] = subVersions
	}

	if opts.Incremental && opts.BuildOptions.IncrementalSkipIndexing() {
		return nil
	}

	reposByPath := map[string]BlobLocation{}
	for key, location := range repos {
		reposByPath[key.SubRepoPath] = location
	}

	opts.BuildOptions.SubRepositories = map[string]*zoekt.Repository{}
	for path, location := range reposByPath {
		tpl := opts.BuildOptions.RepositoryDescription
		if path != "" {
			tpl = zoekt.Repository{URL: location.URL.String()}
			if err := SetTemplatesFromOrigin(&tpl, location.URL); err != nil {
				log.Printf("setTemplatesFromOrigin(%s, %s): %s", path, location.URL, err)
			}
		}
		opts.BuildOptions.SubRepositories[path] = &tpl
	}
	for _, br := range opts.BuildOptions.RepositoryDescription.Branches {
		for path, repo := range opts.BuildOptions.SubRepositories {
			id := branchVersions[br.Name][path]
			repo.Branches = append(repo.Branches, zoekt.RepositoryBranch{
				Name:    br.Name,
				Version: id.String(),
			})
		}
	}

	builder, err := build.NewBuilder(opts.BuildOptions)
	if err != nil {
		return err
	}
	defer builder.Finish()

	var names []string
	fileKeys := map[string][]fileKey{}
	for key := range repos {
		n := key.FullPath()
		fileKeys[n] = append(fileKeys[n], key)
		names = append(names, n)
	}

	sort.Strings(names)
	names = uniq(names)

	for _, name := range names {
		keys := fileKeys[name]

		for _, key := range keys {
			brs := branchMap[key]
			blob, err := repos[key].Repo.BlobObject(key.ID)
			if err != nil {
				return err
			}

			if blob.Size > int64(opts.BuildOptions.SizeMax) && !opts.BuildOptions.IgnoreSizeMax(key.FullPath()) {
				if err := builder.Add(zoekt.Document{
					SkipReason:        fmt.Sprintf("file size %d exceeds maximum size %d", blob.Size, opts.BuildOptions.SizeMax),
					Name:              key.FullPath(),
					Branches:          brs,
					SubRepositoryPath: key.SubRepoPath,
				}); err != nil {
					return err
				}
				continue
			}

			contents, err := blobContents(blob)
			if err != nil {
				return err
			}
			if err := builder.Add(zoekt.Document{
				SubRepositoryPath: key.SubRepoPath,
				Name:              key.FullPath(),
				Content:           contents,
				Branches:          brs,
			}); err != nil {
				return err
			}
		}
	}
	return builder.Finish()
}