gitleaks:bugfix - updating formatter to gitleaks v8

When the v8 version of gitaleks was released, the way to do an analysis
and the output of that analysis changed. This pr updates the formatter
to the new format and updates gitleaks to v8.3.0. It also contains some
improvements in the formatter code.

Signed-off-by: Nathan Martins <nathan.martins@zup.com.br>

e2e/analysis/test_case.go

@@ -661,6 +661,7 @@ func NewTestCase() []*TestCase {
 					fmt.Sprintf("{HORUSEC_CLI} %s - %s is finished in analysisID:", tools.GitLeaks, languages.Leaks),
 					fmt.Sprintf("{HORUSEC_CLI} Running %s - %s", tools.HorusecEngine, languages.Leaks),
 					fmt.Sprintf("{HORUSEC_CLI} %s - %s is finished in analysisID:", tools.HorusecEngine, languages.Leaks),
+					fmt.Sprintf("{HORUSEC_CLI} The current path it's not a valid git repository"),
 				},
 				OutputsNotContains: []string{
 					fmt.Sprintf("{HORUSEC_CLI} Something error went wrong in %s tool", tools.GitLeaks),

internal/controllers/analyzer/analyzer.go

@@ -452,5 +452,6 @@ func (a *Analyzer) isWarning(err string) bool {
 	return strings.Contains(err, messages.MsgErrorPacketJSONNotFound) ||
 		strings.Contains(err, messages.MsgErrorYarnLockNotFound) ||
 		strings.Contains(err, messages.MsgErrorGemLockNotFound) ||
-		strings.Contains(err, messages.MsgErrorNotFoundRequirementsTxt)
+		strings.Contains(err, messages.MsgErrorNotFoundRequirementsTxt) ||
+		strings.Contains(err, messages.MsgWarnPathIsInvalidGitRepository)
 }

internal/enums/images/images.go

@@ -25,7 +25,7 @@ const (
 	Go              = "horuszup/horusec-go:v1.2.1"
 	HCL             = "horuszup/horusec-hcl:v1.1.0"
 	Javascript      = "horuszup/horusec-js:v1.2.0"
-	Leaks           = "horuszup/horusec-leaks:v1.1.0"
+	Leaks           = "horuszup/horusec-leaks:v1.2.0"
 	PHP             = "horuszup/horusec-php:v1.0.1"
 	Python          = "horuszup/horusec-python:v1.0.0"
 	Ruby            = "horuszup/horusec-ruby:v1.1.1"

internal/helpers/messages/warn.go

@@ -55,4 +55,5 @@ const (
 	// TODO: Remove MsgWarnAnalysisContainsOutdatedHash before release v2.10.0
 	MsgWarnAnalysisContainsOutdatedHash = "{HORUSEC_CLI} YOUR CONFIGURATION FILE CONTAINS SOME HASHES THAT WILL NO " +
 		"LONGER BE VALID AS OF v2.10.0 IS RELEASED. PLEASE UPDATE YOUR CONFIGURATION FILE WITH THE FOLLOWING HASHES:"
+	MsgWarnPathIsInvalidGitRepository = "{HORUSEC_CLI} The current path it's not a valid git repository"
 )

internal/services/formatters/leaks/deployments/Dockerfile

@@ -12,7 +12,7 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-FROM zricethezav/gitleaks:v8.2.7
+FROM zricethezav/gitleaks:v8.3.0
 
 COPY ./internal/services/formatters/leaks/deployments/rules.toml /rules/rules.toml
 

internal/services/formatters/leaks/gitleaks/config.go

@@ -12,14 +12,16 @@
 // See the License for the specific language governing permissions and
 // limitations under the License.
 
-//nolint
 package gitleaks
 
+// CMD contains the necessary code to execute Gitleaks inside the container. The 'git config diff.renames 0' command
+// it's necessary to avoid the 'inexact rename detection was skipped due to too many files' error in big projects.
 const CMD = `
-		{{WORK_DIR}}
-		if ! gitleaks --config-path /rules/rules.toml --path . --leaks-exit-code 0 --format json --report /tmp/leaks-result.json &> /tmp/leaks-runner-output.txt; then
-			echo /tmp/leaks-runner-output.txt
-		else
-			cat /tmp/leaks-result.json
-		fi
+	{{WORK_DIR}}
+	git config diff.renames 0
+	if ! gitleaks detect -c /rules/rules.toml -f json -r /tmp/leaks.json --exit-code 0 &> /tmp/leaks-output.txt; then
+		cat /tmp/leaks-output.txt
+	else
+		cat /tmp/leaks.json
+	fi
   `

internal/services/formatters/leaks/gitleaks/formatter.go

@@ -17,6 +17,7 @@ package gitleaks
 import (
 	"encoding/json"
 	"errors"
+	"strconv"
 	"strings"
 
 	"github.com/ZupIT/horusec-devkit/pkg/entities/vulnerability"
@@ -29,7 +30,6 @@ import (
 	"github.com/ZupIT/horusec/internal/enums/images"
 	"github.com/ZupIT/horusec/internal/helpers/messages"
 	"github.com/ZupIT/horusec/internal/services/formatters"
-	"github.com/ZupIT/horusec/internal/services/formatters/leaks/gitleaks/entities"
 	vulnhash "github.com/ZupIT/horusec/internal/utils/vuln_hash"
 )
 
@@ -62,78 +62,73 @@ func (f *Formatter) startGitLeaks(projectSubPath string) (string, error) {
 		return output, err
 	}
 
-	return output, f.formatOutputGitLeaks(output)
+	if err := f.checkOutputErrors(output); err != nil {
+		return output, err
+	}
+
+	return output, f.parseOutput(output)
 }
 
-func (f *Formatter) formatOutputGitLeaks(output string) error {
-	if output == "" || (len(output) >= 4 && output[:4] == "null") {
+func (f *Formatter) getDockerConfig(projectSubPath string) *dockerEntities.AnalysisData {
+	analysisData := &dockerEntities.AnalysisData{
+		CMD:      f.AddWorkDirInCmd(CMD, projectSubPath, tools.GitLeaks),
+		Language: languages.Leaks,
+	}
+
+	return analysisData.SetImage(f.GetCustomImageByLanguage(languages.Leaks), images.Leaks)
+}
+
+func (f *Formatter) parseOutput(output string) error {
+	issues := make([]*Issue, 0)
+
+	if err := json.Unmarshal([]byte(output), &issues); err != nil {
+		return err
+	}
+
+	if len(issues) == 0 {
 		logger.LogDebugWithLevel(messages.MsgDebugOutputEmpty,
 			map[string]interface{}{"tool": tools.GitLeaks.ToString()})
-		f.setGitLeaksOutPutInHorusecAnalysis([]entities.Issue{})
 		return nil
 	}
 
-	issues, err := f.parseOutputToIssues(output)
-	if err != nil {
-		return err
-	}
+	f.forEachIssueCreateNewVuln(issues)
 
-	f.setGitLeaksOutPutInHorusecAnalysis(issues)
 	return nil
 }
 
-func (f *Formatter) parseOutputToIssues(output string) ([]entities.Issue, error) {
-	var issues []entities.Issue
-	err := json.Unmarshal([]byte(output), &issues)
-	if err != nil && strings.Contains(err.Error(), "invalid character") {
-		err = errors.New(output)
-	}
-	return issues, err
-}
-
-func (f *Formatter) setGitLeaksOutPutInHorusecAnalysis(issues []entities.Issue) {
-	for key := range issues {
-		vuln := f.setupVulnerabilitiesSeveritiesGitLeaks(&issues[key])
-		f.AddNewVulnerabilityIntoAnalysis(vuln)
+func (f *Formatter) forEachIssueCreateNewVuln(issues []*Issue) {
+	for _, issue := range issues {
+		f.AddNewVulnerabilityIntoAnalysis(f.newVulnerability(issue))
 	}
 }
 
-func (f *Formatter) setupVulnerabilitiesSeveritiesGitLeaks(issue *entities.Issue) (
-	vulnerabilitySeverity *vulnerability.Vulnerability,
-) {
-	vulnerabilitySeverity = f.getDefaultSeverity()
-	vulnerabilitySeverity.Severity = severities.Critical
-	vulnerabilitySeverity.RuleID = vulnhash.HashRuleID(issue.Rule)
-	vulnerabilitySeverity.Details = issue.Rule
-	vulnerabilitySeverity.Code = f.GetCodeWithMaxCharacters(issue.Line, 0)
-	vulnerabilitySeverity.File = issue.File
-	vulnerabilitySeverity = vulnhash.Bind(vulnerabilitySeverity)
-	return f.setCommitAuthor(vulnerabilitySeverity, issue)
-}
-
-func (f *Formatter) setCommitAuthor(vuln *vulnerability.Vulnerability,
-	issue *entities.Issue,
-) *vulnerability.Vulnerability {
+//nolint:funlen // necessary to be long
+func (f *Formatter) newVulnerability(issue *Issue) *vulnerability.Vulnerability {
+	vuln := &vulnerability.Vulnerability{}
+
+	vuln.Language = languages.Leaks
+	vuln.SecurityTool = tools.GitLeaks
+	vuln.Severity = severities.Critical
+	vuln.RuleID = vulnhash.HashRuleID(issue.Description)
+	vuln.Details = issue.Description
+	vuln.Code = f.GetCodeWithMaxCharacters(issue.Secret, 0)
+	vuln.File = issue.File
+	vuln.Line = strconv.Itoa(issue.StartLine)
+	vuln.Column = strconv.Itoa(issue.StartColumn)
 	vuln.CommitAuthor = issue.Author
-	vuln.CommitMessage = strings.ReplaceAll(issue.CommitMessage, "\n", "")
+	vuln.CommitMessage = f.GetCodeWithMaxCharacters(issue.Message, 0)
 	vuln.CommitEmail = issue.Email
 	vuln.CommitDate = issue.Date
 	vuln.CommitHash = issue.Commit
-	return vuln
+
+	return vulnhash.Bind(vuln)
 }
 
-func (f *Formatter) getDockerConfig(projectSubPath string) *dockerEntities.AnalysisData {
-	analysisData := &dockerEntities.AnalysisData{
-		CMD:      f.AddWorkDirInCmd(CMD, projectSubPath, tools.GitLeaks),
-		Language: languages.Leaks,
+func (f *Formatter) checkOutputErrors(output string) error {
+	if strings.Contains(output, "fatal: not a git repository") ||
+		strings.Contains(output, "fatal: cannot chdir to") {
+		return errors.New(messages.MsgWarnPathIsInvalidGitRepository)
 	}
 
-	return analysisData.SetImage(f.GetCustomImageByLanguage(languages.Leaks), images.Leaks)
-}
-
-func (f *Formatter) getDefaultSeverity() *vulnerability.Vulnerability {
-	vulnerabilitySeverity := &vulnerability.Vulnerability{}
-	vulnerabilitySeverity.Language = languages.Leaks
-	vulnerabilitySeverity.SecurityTool = tools.GitLeaks
-	return vulnerabilitySeverity
+	return nil
 }

internal/services/formatters/leaks/gitleaks/entities/output.go → internal/services/formatters/leaks/gitleaks/output.go

@@ -12,18 +12,21 @@
 // See the License for the specific language governing permissions and
 // limitations under the License.
 
-package entities
+package gitleaks
 
 type Issue struct {
-	Line          string `json:"line"`
-	Offender      string `json:"offender"`
-	Commit        string `json:"commit"`
-	Repo          string `json:"repo"`
-	Rule          string `json:"rule"`
-	CommitMessage string `json:"commitMessage"`
-	Author        string `json:"author"`
-	Email         string `json:"email"`
-	File          string `json:"file"`
-	Date          string `json:"date"`
-	Tags          string `json:"tags"`
+	Description string
+	StartLine   int
+	EndLine     int
+	StartColumn int
+	EndColumn   int
+	Match       string
+	Secret      string
+	File        string
+	Commit      string
+	Entropy     int
+	Author      string
+	Email       string
+	Date        string
+	Message     string
 }