Add test for checkMetadata function

datasetIngestor/checkMetadata.go

@@ -18,7 +18,10 @@ import (
 const DUMMY_TIME = "2300-01-01T11:11:11.000Z"
 const DUMMY_OWNER = "x12345"
 
+// getHost is a function that attempts to retrieve and return the fully qualified domain name (FQDN) of the current host.
+// If it encounters any error during the process, it gracefully falls back to returning the simple hostname or "unknown".
 func getHost() string {
+	// Try to get the hostname of the current machine.
 	hostname, err := os.Hostname()
 	if err != nil {
 		return "unknown"
@@ -46,27 +49,34 @@ func getHost() string {
 	return hostname
 }
 
-func CheckMetadata(client *http.Client, APIServer string, metadatafile string, user map[string]string,
-	accessGroups []string) (metaDataMap map[string]interface{}, sourceFolder string, beamlineAccount bool) {
+// CheckMetadata is a function that validates and augments metadata for a dataset.
+// It takes an HTTP client, an API server URL, a metadata file path, a user map, and a list of access groups as input.
+// It returns a map of metadata, a source folder string, and a boolean indicating whether the dataset belongs to a beamline account.
+func CheckMetadata(client *http.Client, APIServer string, metadatafile string, user map[string]string, accessGroups []string) (metaDataMap map[string]interface{}, sourceFolder string, beamlineAccount bool) {
 
-	// read full meta data
+	// Read the full metadata from the file.
 	b, err := ioutil.ReadFile(metadatafile) // just pass the file name
 	if err != nil {
 		log.Fatal(err)
 	}
 
-	var metadataObj interface{}
+	// Unmarshal the JSON metadata into an interface{} object.
+	var metadataObj interface{} // Using interface{} allows metadataObj to hold any type of data, since it has no defined methods.
 	err = json.Unmarshal(b, &metadataObj)
 	if err != nil {
 		log.Fatal(err)
 	}
-	// use type assertion to access f's underlying map
-	metaDataMap = metadataObj.(map[string]interface{})
+
+	// Use type assertion to convert the interface{} object to a map[string]interface{}.
+	metaDataMap = metadataObj.(map[string]interface{})  // `.(` is type assertion: a way to extract the underlying value of an interface and check whether it's of a specific type.
 	beamlineAccount = false
 
+	// If the user is not the ingestor, check whether any of the accessGroups equal the ownerGroup. Otherwise, check for beamline-specific accounts.
 	if user["displayName"] != "ingestor" {
-		if ownerGroup, ok := metaDataMap["ownerGroup"]; ok {
+		// Check if the metadata contains the "ownerGroup" key.
+		if ownerGroup, ok := metaDataMap["ownerGroup"]; ok {  // type assertion with a comma-ok idiom
 			validOwner := false
+			// Iterate over accessGroups to validate the owner group.
 			for _, b := range accessGroups {
 				if b == ownerGroup {
 					validOwner = true
@@ -76,13 +86,15 @@ func CheckMetadata(client *http.Client, APIServer string, metadatafile string, u
 			if validOwner {
 				log.Printf("OwnerGroup information %s verified successfully.\n", ownerGroup)
 			} else {
-				// check for beamline specific account if raw data
+				// If the owner group is not valid, check for beamline-specific accounts.
 				if creationLocation, ok := metaDataMap["creationLocation"]; ok {
+					// Split the creationLocation string to extract beamline-specific information.
 					parts := strings.Split(creationLocation.(string), "/")
 					expectedAccount := ""
 					if len(parts) == 4 {
 						expectedAccount = strings.ToLower(parts[2]) + strings.ToLower(parts[3])
 					}
+					// If the user matches the expected beamline account, grant ingest access.
 					if user["displayName"] == expectedAccount {
 						log.Printf("Beamline specific dataset %s - ingest granted.\n", expectedAccount)
 						beamlineAccount = true
@@ -91,7 +103,7 @@ func CheckMetadata(client *http.Client, APIServer string, metadatafile string, u
 					}
 				} else {
 					// for other data just check user name
-					// this is a quick and dirty test. Should be replaced by test for "globalaccess" role
+					// this is a quick and dirty test. Should be replaced by test for "globalaccess" role. TODO
 					// facilities: ["SLS", "SINQ", "SWISSFEL", "SmuS"],
 					u := user["displayName"]
 					if strings.HasPrefix(u, "sls") ||
@@ -131,7 +143,7 @@ func CheckMetadata(client *http.Client, APIServer string, metadatafile string, u
 			log.Printf("sourceFolderHost field added: %s", metaDataMap["sourceFolderHost"])
 		}
 	}
-	// far raw data add PI if missing
+	// for raw data add PI if missing
 	if val, ok := metaDataMap["type"]; ok {
 		dstype := val.(string)
 		if dstype == "raw" {

datasetIngestor/checkMetadata_test.go

@@ -0,0 +1,118 @@
+package datasetIngestor
+
+import (
+	"net/http"
+	"testing"
+	"time"
+	"reflect"
+)
+
+func TestGetHost(t *testing.T) {
+	// Call the function under test.
+	host := GetHost()
+
+	// fail the test and report an error if the returned hostname is an empty string.
+	if len(host) == 0 {
+		t.Errorf("getHost() returned an empty string")
+	}
+
+	// fail the test and report an error if the returned hostname is "unknown".
+	if host == "unknown" {
+		t.Errorf("getHost() was unable to get the hostname")
+	}
+}
+
+func TestCheckMetadata(t *testing.T) {
+	// Define mock parameters for the function
+	var TEST_API_SERVER string = "https://dacat-qa.psi.ch/api/v3" // TODO: Test Improvement. Change this to a mock server. At the moment, tests will fail if we change this to a mock server.
+	var APIServer = TEST_API_SERVER
+	var metadatafile1 = "testdata/metadata.json"
+	var metadatafile2 = "testdata/metadata-short.json"
+
+	// Mock HTTP client
+	client := &http.Client{
+		Timeout: 5 * time.Second, // Set a timeout for requests
+		Transport: &http.Transport{
+			// Customize the transport settings if needed (e.g., proxy, TLS config)
+			// For a dummy client, default settings are usually sufficient
+		},
+		CheckRedirect: func(req *http.Request, via []*http.Request) error {
+			// Customize how redirects are handled if needed
+			// For a dummy client, default behavior is usually sufficient
+			return http.ErrUseLastResponse // Use the last response for redirects
+		},
+	}
+
+	// Mock user map
+	user := map[string]string{
+		"displayName": "csaxsswissfel",
+		"mail":        "testuser@example.com",
+	}
+
+	// Mock access groups
+	accessGroups := []string{"p17880", "p17301"}
+
+	// Call the function with mock parameters
+	metaDataMap, sourceFolder, beamlineAccount := CheckMetadata(client, APIServer, metadatafile1, user, accessGroups)
+
+	// Add assertions here based on the expected behavior of the function
+	if len(metaDataMap) == 0 {
+		t.Error("Expected non-empty metadata map")
+	}
+	if sourceFolder == "" {
+		t.Error("Expected non-empty source folder")
+	}
+	if sourceFolder != "/usr/share/gnome" {
+		t.Error("sourceFolder should be '/usr/share/gnome'")
+	}
+	if reflect.TypeOf(beamlineAccount).Kind() != reflect.Bool {
+		t.Error("Expected beamlineAccount to be boolean")
+	}
+	if beamlineAccount != false {
+		t.Error("Expected beamlineAccount to be false")
+	}
+	if _, ok := metaDataMap["ownerEmail"]; !ok {
+    t.Error("metaDataMap missing required key 'ownerEmail'")
+	}
+	if _, ok := metaDataMap["principalInvestigator"]; !ok {
+    t.Error("metaDataMap missing required key 'principalInvestigator'")
+	}
+	if _, ok := metaDataMap["scientificMetadata"]; !ok {
+    t.Error("metaDataMap missing required key 'scientificMetadata'")
+	}
+	scientificMetadata, ok := metaDataMap["scientificMetadata"].([]interface{})
+	if ok {
+		firstEntry := scientificMetadata[0].(map[string]interface{})
+		sample, ok := firstEntry["sample"].(map[string]interface{})
+		if ok {
+			if _, ok := sample["name"]; !ok {
+				t.Error("Sample is missing 'name' field")
+			}
+			if _, ok := sample["description"]; !ok {
+				t.Error("Sample is missing 'description' field")
+			}
+		}
+	} else {
+		t.Error("scientificMetadata is not a list")
+	}
+
+	// test with the second metadata file
+	metaDataMap2, sourceFolder2, beamlineAccount2 := CheckMetadata(client, APIServer, metadatafile2, user, accessGroups)
+
+	// Add assertions here based on the expected behavior of the function
+	if len(metaDataMap2) == 0 {
+		t.Error("Expected non-empty metadata map")
+	}
+	if sourceFolder2 == "" {
+		t.Error("Expected non-empty source folder")
+	}
+	if sourceFolder2 != "/tmp/gnome" {
+		t.Error("sourceFolder should be '/tmp/gnome'")
+	}
+	if reflect.TypeOf(beamlineAccount2).Kind() != reflect.Bool {
+		t.Error("Expected beamlineAccount to be boolean")
+	}
+	if beamlineAccount2 != false {
+		t.Error("Expected beamlineAccount to be false")
+	}
+}

datasetIngestor/export_test.go

@@ -0,0 +1,4 @@
+// This file exports internal functions for testing purposes. Since the name of the file ends with "_test.go", it will not be included in the final build of the application.
+package datasetIngestor
+
+var GetHost = getHost

datasetIngestor/testdata/metadata-short.json

@@ -0,0 +1,8 @@
+{
+    "principalInvestigator":"egon.meier@psi.ch",
+    "creationLocation":"/PSI/SLS/CSAXS/SWISSFEL",
+    "sourceFolder": "/tmp/gnome",
+    "owner": "Andreas Menzel",
+    "type": "raw",
+    "ownerGroup": "p17880"
+}

datasetIngestor/testdata/metadata.json

@@ -0,0 +1,20 @@
+{
+    "creationLocation": "/PSI/SLS/CSAXS/SWISSFEL",
+    "datasetName": "CMakeCache",
+    "description": "",
+    "owner": "Ana Diaz",
+    "ownerEmail": "ana.diaz@psi.ch",
+    "ownerGroup": "p17301",
+    "principalInvestigator": "ana.diaz@psi.ch",
+    "scientificMetadata": [
+        {
+            "sample": {
+                "description": "",
+                "name": "",
+                "principalInvestigator": ""
+            }
+        }
+    ],
+    "sourceFolder": "/usr/share/gnome",
+    "type": "raw"
+}

go.mod

@@ -13,6 +13,7 @@ require (
 
 require (
 	github.com/creack/pty v1.1.17 // indirect
+	github.com/davecgh/go-spew v1.1.1 // indirect
 	github.com/mattn/go-colorable v0.1.9 // indirect
 	github.com/mattn/go-isatty v0.0.14 // indirect
 	golang.org/x/sys v0.2.0 // indirect

go.sum

@@ -2,8 +2,9 @@ github.com/Netflix/go-expect v0.0.0-20220104043353-73e0943537d2 h1:+vx7roKuyA63n
 github.com/Netflix/go-expect v0.0.0-20220104043353-73e0943537d2/go.mod h1:HBCaDeC1lPdgDeDbhX8XFpy1jqjK0IBG8W5K+xYqA0w=
 github.com/creack/pty v1.1.17 h1:QeVUsEDNrLBW4tMgZHvxy18sKtr6VI492kBhUfhDJNI=
 github.com/creack/pty v1.1.17/go.mod h1:MOBLtS5ELjhRRrroQr9kyvTxUAFNvYEK993ew/Vr4O4=
-github.com/davecgh/go-spew v1.1.0 h1:ZDRjVQ15GmhC3fiQ8ni8+OwkZQO4DARzQgrnXU1Liz8=
 github.com/davecgh/go-spew v1.1.0/go.mod h1:J7Y8YcW2NihsgmVo/mv3lAwl/skON4iLHjSsI+c5H38=
+github.com/davecgh/go-spew v1.1.1 h1:vj9j/u1bqnvCEfJOwUhtlOARqs3+rkHYY13jYWTU97c=
+github.com/davecgh/go-spew v1.1.1/go.mod h1:J7Y8YcW2NihsgmVo/mv3lAwl/skON4iLHjSsI+c5H38=
 github.com/fatih/color v1.13.0 h1:8LOYc1KYPPmyKMuN8QV2DNRWNbLo6LZ0iLs8+mlH53w=
 github.com/fatih/color v1.13.0/go.mod h1:kLAiJbzzSOZDVNGyDpeOxJ47H46qBXwg5ILebYFFOfk=
 github.com/kballard/go-shellquote v0.0.0-20180428030007-95032a82bc51 h1:Z9n2FFNUXsshfwJMBgNA0RU6/i7WVaAegv3PtuIHPMs=