Handle extra empty lines in Java manifest parsing (#687)

* Add failing test for extra empty lines in manifest

Signed-off-by: Dan Luhring <dan+github@luhrings.com>

* Handle extra empty lines in Java manifests

Signed-off-by: Dan Luhring <dan+github@luhrings.com>

syft/pkg/cataloger/java/parse_java_manifest.go

@@ -19,48 +19,60 @@ const manifestGlob = "/META-INF/MANIFEST.MF"
 // For more information: https://docs.oracle.com/en/java/javase/11/docs/specs/jar/jar.html#jar-manifest
 func parseJavaManifest(path string, reader io.Reader) (*pkg.JavaManifest, error) {
 	var manifest pkg.JavaManifest
-	sections := []map[string]string{
-		make(map[string]string),
+	var sections []map[string]string
+
+	currentSection := func() int {
+		return len(sections) - 1
 	}
-	currentSection := 0
-	scanner := bufio.NewScanner(reader)
+
 	var lastKey string
+	scanner := bufio.NewScanner(reader)
+
 	for scanner.Scan() {
 		line := scanner.Text()
 
 		// empty lines denote section separators
 		if strings.TrimSpace(line) == "" {
-			currentSection++
-			// we don't want to allocate a new section map that wont necessarily be used, do that once there is
+			// we don't want to allocate a new section map that won't necessarily be used, do that once there is
 			// a non-empty line to process
 
 			// do not process line continuations after this
 			lastKey = ""
+
 			continue
-		} else if currentSection >= len(sections) {
-			sections = append(sections, make(map[string]string))
 		}
 
 		if line[0] == ' ' {
 			// this is a continuation
+
 			if lastKey == "" {
 				return nil, fmt.Errorf("found continuation with no previous key (%s)", line)
 			}
-			sections[currentSection][lastKey] += strings.TrimSpace(line)
-		} else {
-			// this is a new key-value pair
-			idx := strings.Index(line, ":")
-			if idx == -1 {
-				return nil, fmt.Errorf("unable to split java manifest key-value pairs: %q", line)
-			}
 
-			key := strings.TrimSpace(line[0:idx])
-			value := strings.TrimSpace(line[idx+1:])
-			sections[currentSection][key] = value
+			sections[currentSection()][lastKey] += strings.TrimSpace(line)
 
-			// keep track of key for potential future continuations
-			lastKey = key
+			continue
 		}
+
+		if lastKey == "" {
+			// we're entering a new section
+
+			sections = append(sections, make(map[string]string))
+		}
+
+		// this is a new key-value pair
+		idx := strings.Index(line, ":")
+		if idx == -1 {
+			return nil, fmt.Errorf("unable to split java manifest key-value pairs: %q", line)
+		}
+
+		key := strings.TrimSpace(line[0:idx])
+		value := strings.TrimSpace(line[idx+1:])
+
+		sections[currentSection()][key] = value
+
+		// keep track of key for potential future continuations
+		lastKey = key
 	}
 
 	if err := scanner.Err(); err != nil {
@@ -75,7 +87,7 @@ func parseJavaManifest(path string, reader io.Reader) (*pkg.JavaManifest, error)
 				name, ok := s["Name"]
 				if !ok {
 					// per the manifest spec (https://docs.oracle.com/en/java/javase/11/docs/specs/jar/jar.html#jar-manifest)
-					// this should never happen. If it does we want to know about it, but not necessarily stop
+					// this should never happen. If it does, we want to know about it, but not necessarily stop
 					// cataloging entirely... for this reason we only log.
 					log.Warnf("java manifest section found without a name: %s", path)
 					name = strconv.Itoa(i)

syft/pkg/cataloger/java/parse_java_manifest_test.go

@@ -58,6 +58,21 @@ func TestParseJavaManifest(t *testing.T) {
 				},
 			},
 		},
+		{
+			fixture: "test-fixtures/manifest/extra-empty-lines",
+			expected: pkg.JavaManifest{
+				Main: map[string]string{
+					"Manifest-Version": "1.0",
+					"Archiver-Version": "Plexus Archiver",
+					"Created-By":       "Apache Maven 3.6.3",
+				},
+				NamedSections: map[string]map[string]string{
+					"thing-1": {
+						"Built-By": "?",
+					},
+				},
+			},
+		},
 		{
 			fixture: "test-fixtures/manifest/continuation",
 			expected: pkg.JavaManifest{

syft/pkg/cataloger/java/test-fixtures/manifest/extra-empty-lines

@@ -0,0 +1,7 @@
+Manifest-Version: 1.0
+Archiver-Version: Plexus Archiver
+Created-By: Apache Maven 3.6.3
+
+
+Name: thing-1
+Built-By: ?