Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Allow gitlab to resume from encoded resume info #611

Merged
merged 5 commits into from
Jun 17, 2022
Merged

Allow gitlab to resume from encoded resume info #611

Show file tree
Hide file tree
Changes from 3 commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
bf8e54b
Separate repo resume functions into reusable functions in the sources…
trufflesteeeve Jun 6, 2022
da38994
Allow gitlab to resume from encoded resume info
trufflesteeeve Jun 7, 2022
dd37530
fixup - fix bug where an inaccurate number of repos would be reported
trufflesteeeve Jun 16, 2022
cd1d6bd
fixup - add fixes from review, including progress test
trufflesteeeve Jun 16, 2022
1e4688d
fixup - fix logic bug in getRepos
trufflesteeeve Jun 17, 2022
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
100 changes: 10 additions & 90 deletions pkg/sources/github/github.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -24,12 +24,11 @@ import (
"google.golang.org/protobuf/proto"
"google.golang.org/protobuf/types/known/anypb"

"github.com/trufflesecurity/trufflehog/v3/pkg/common"
"github.com/trufflesecurity/trufflehog/v3/pkg/giturl"
"github.com/trufflesecurity/trufflehog/v3/pkg/pb/credentialspb"
"github.com/trufflesecurity/trufflehog/v3/pkg/pb/source_metadatapb"
"github.com/trufflesecurity/trufflehog/v3/pkg/pb/sourcespb"

"github.com/trufflesecurity/trufflehog/v3/pkg/common"
"github.com/trufflesecurity/trufflehog/v3/pkg/giturl"
"github.com/trufflesecurity/trufflehog/v3/pkg/sanitizer"
"github.com/trufflesecurity/trufflehog/v3/pkg/sources"
"github.com/trufflesecurity/trufflehog/v3/pkg/sources/git"
Expand Down Expand Up @@ -355,7 +354,8 @@ func (s *Source) scan(ctx context.Context, installationClient *github.Client, ch
}

// If there is resume information available, limit this scan to only the repos that still need scanning.
progressIndexOffset := s.filterReposToResume(s.GetProgress().EncodedResumeInfo)
reposToScan, progressIndexOffset := sources.FilterReposToResume(s.repos, s.GetProgress().EncodedResumeInfo)
s.repos = reposToScan

for i, repoURL := range s.repos {
if err := s.jobSem.Acquire(ctx, 1); err != nil {
Expand All @@ -372,7 +372,11 @@ func (s *Source) scan(ctx context.Context, installationClient *github.Client, ch

s.setProgressCompleteWithRepo(i+progressIndexOffset, repoURL)
// Ensure the repo is removed from the resume info after being scanned.
defer s.removeRepoFromResumeInfo(repoURL)
defer func(s *Source) {
s.resumeInfoMutex.Lock()
defer s.resumeInfoMutex.Unlock()
s.resumeInfoSlice = sources.RemoveRepoFromResumeInfo(s.resumeInfoSlice, repoURL)
}(s)

if !strings.HasSuffix(repoURL, ".git") {
return
Expand Down Expand Up @@ -799,91 +803,7 @@ func (s *Source) setProgressCompleteWithRepo(index int, repoURL string) {
sort.Strings(s.resumeInfoSlice)

// Make the resume info string from the slice.
encodedResumeInfo := s.encodeResumeInfo()
encodedResumeInfo := sources.EncodeResumeInfo(s.resumeInfoSlice)

s.SetProgressComplete(index, len(s.repos), fmt.Sprintf("Repo: %s", repoURL), encodedResumeInfo)
}

// removeRepoFromResumeInfo removes the repoURL from the resume info.
func (s *Source) removeRepoFromResumeInfo(repoURL string) {
s.resumeInfoMutex.Lock()
defer s.resumeInfoMutex.Unlock()

index := -1
for i, repo := range s.resumeInfoSlice {
if repoURL == repo {
index = i
}
}

if index == -1 {
// We should never be able to be here. But if we are, it means the resume info never had the repo added.
// So log the error and do nothing.
s.log.Errorf("repoURL (%q) not found in list of encode resume info: %q", repoURL, s.EncodedResumeInfo)
return
}

// This removes the element at the given index.
s.resumeInfoSlice = append(s.resumeInfoSlice[:index], s.resumeInfoSlice[index+1:]...)
}

func (s *Source) encodeResumeInfo() string {
return strings.Join(s.resumeInfoSlice, "\t")
}

func (s *Source) decodeResumeInfo(resumeInfo string) {
// strings.Split will, for an empty string, return []string{""},
// which is an element, where as when there is no resume info we want an empty slice.
if resumeInfo == "" {
return
}
s.resumeInfoSlice = strings.Split(resumeInfo, "\t")
}

// filterReposToResume filters the existing repos down to those that are included in the encoded resume info.
// It also returns the difference between the original length of the repos and the new length to use for progress reporting.
// It is required that both the resumeInfo repos and the existing repos in s.repos are sorted.
func (s *Source) filterReposToResume(resumeInfo string) int {
if resumeInfo == "" {
return 0
}

s.resumeInfoMutex.Lock()
defer s.resumeInfoMutex.Unlock()

s.decodeResumeInfo(resumeInfo)

// Because this scanner is multithreaded, it is possible that we have scanned a range of repositories
// with some gaps of unlisted but completed repositories in between the ones in resumeInfo.
// So we know repositories that have not finished scanning are the ones included in the resumeInfo,
// and those that come after the last repository in the resumeInfo.
// However, it is possible that a resumed scan does not include all or even any of the repos within the resumeInfo.
// In this case, we must ensure we still scan all repos that come after the last found repo in the list.
reposToScan := []string{}
lastFoundRepoIndex := -1
resumeRepoIndex := 0
for i, repoURL := range s.repos {
// If the repoURL is bigger than what we're looking for, move to the next one.
if repoURL > s.resumeInfoSlice[resumeRepoIndex] {
resumeRepoIndex++
}

// If we've found all of our repositories end the filter.
if resumeRepoIndex == len(s.resumeInfoSlice) {
break
}

// If the repoURL is the one we're looking for, add it and update the lastFoundRepoIndex.
if repoURL == s.resumeInfoSlice[resumeRepoIndex] {
lastFoundRepoIndex = i
reposToScan = append(reposToScan, repoURL)
}
}

// Append all repos after the last one we've found.
reposToScan = append(reposToScan, s.repos[lastFoundRepoIndex+1:]...)

progressOffsetCount := len(s.repos) - len(reposToScan)
s.repos = reposToScan
return progressOffsetCount
}
149 changes: 0 additions & 149 deletions pkg/sources/github/github_test.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -373,152 +373,3 @@ func Test_setProgressCompleteWithRepo(t *testing.T) {
}
}
}

func Test_removeRepoFromResumeInfo(t *testing.T) {
tests := []struct {
startingResumeInfoSlice []string
repoURL string
wantResumeInfoSlice []string
}{
{
startingResumeInfoSlice: []string{"a", "b", "c"},
repoURL: "a",
wantResumeInfoSlice: []string{"b", "c"},
},
{
startingResumeInfoSlice: []string{"a", "b", "c"},
repoURL: "b",
wantResumeInfoSlice: []string{"a", "c"},
},
{ // This is the probably can't happen case of a repo not in the list.
startingResumeInfoSlice: []string{"a", "b", "c"},
repoURL: "not in the list",
wantResumeInfoSlice: []string{"a", "b", "c"},
},
}

logger := logrus.New()
logger.Out = io.Discard
s := &Source{
repos: []string{},
log: logger.WithField("no", "output"),
}

for _, tt := range tests {
s.resumeInfoSlice = tt.startingResumeInfoSlice
s.removeRepoFromResumeInfo(tt.repoURL)
if !reflect.DeepEqual(s.resumeInfoSlice, tt.wantResumeInfoSlice) {
t.Errorf("s.removeRepoFromResumeInfo() got: %v, want: %v", s.resumeInfoSlice, tt.wantResumeInfoSlice)
}
}
}

func Test_encodeResumeInfo(t *testing.T) {
tests := []struct {
startingResumeInfoSlice []string
wantEncodedResumeInfo string
}{
{
startingResumeInfoSlice: []string{"a", "b", "c"},
wantEncodedResumeInfo: "a\tb\tc",
},
{
startingResumeInfoSlice: []string{},
wantEncodedResumeInfo: "",
},
}

logger := logrus.New()
logger.Out = io.Discard
s := &Source{
repos: []string{},
log: logger.WithField("no", "output"),
}

for _, tt := range tests {
s.resumeInfoSlice = tt.startingResumeInfoSlice
gotEncodedResumeInfo := s.encodeResumeInfo()
if gotEncodedResumeInfo != tt.wantEncodedResumeInfo {
t.Errorf("s.encodeResumeInfo() got: %q, want: %q", gotEncodedResumeInfo, tt.wantEncodedResumeInfo)
}
}
}

func Test_decodeResumeInfo(t *testing.T) {
tests := []struct {
resumeInfo string
wantResumeInfoSlice []string
}{
{
resumeInfo: "a\tb\tc",
wantResumeInfoSlice: []string{"a", "b", "c"},
},
{
resumeInfo: "",
wantResumeInfoSlice: nil,
},
}

for _, tt := range tests {
s := &Source{}
s.decodeResumeInfo(tt.resumeInfo)
if !reflect.DeepEqual(s.resumeInfoSlice, tt.wantResumeInfoSlice) {
t.Errorf("s.decodeResumeInfo() got: %v, want: %v", s.resumeInfoSlice, tt.wantResumeInfoSlice)
}
}
}

func Test_filterReposToResume(t *testing.T) {
startingRepos := []string{"a", "b", "c", "d", "e", "f", "g"}

tests := map[string]struct {
resumeInfo string
wantProgressOffsetCount int
wantReposToScan []string
}{
"blank resume info": {
resumeInfo: "",
wantProgressOffsetCount: 0,
wantReposToScan: startingRepos,
},
"starting repos": {
resumeInfo: "a\tb",
wantProgressOffsetCount: 0,
wantReposToScan: startingRepos,
},
"early contiguous repos": {
resumeInfo: "b\tc",
wantProgressOffsetCount: 1,
wantReposToScan: []string{"b", "c", "d", "e", "f", "g"},
},
"non-contiguous repos": {
resumeInfo: "b\te",
wantProgressOffsetCount: 3,
wantReposToScan: []string{"b", "e", "f", "g"},
},
"no repos found in the repo list": {
resumeInfo: "not\tthere",
wantProgressOffsetCount: 0,
wantReposToScan: startingRepos,
},
"only some repos in the list": {
resumeInfo: "c\tnot\tthere",
wantProgressOffsetCount: 2,
wantReposToScan: []string{"c", "d", "e", "f", "g"},
},
}

for name, tt := range tests {
s := &Source{
repos: startingRepos,
}

gotProgressOffsetCount := s.filterReposToResume(tt.resumeInfo)
if gotProgressOffsetCount != tt.wantProgressOffsetCount {
t.Errorf("s.filterReposToResume() name: %q got: %d, want: %d", name, gotProgressOffsetCount, tt.wantProgressOffsetCount)
}
if !reflect.DeepEqual(s.repos, tt.wantReposToScan) {
t.Errorf("s.filterReposToResume() name: %q got: %v, want: %v", name, s.repos, tt.wantReposToScan)
}
}
}
Loading