fixing json schema validation

schemautils/schemautils.go

@@ -1,21 +1,27 @@
 package schemautils
 
 import (
-	"commonmeta/metadata"
-	"encoding/json"
 	"fmt"
 	"log"
+	"slices"
 
 	"github.com/xeipuuv/gojsonschema"
 )
 
-func JSONSchemaErrors(document metadata.Metadata) *gojsonschema.Result {
-	documentJSON, err := json.Marshal(document)
-	if err != nil {
-		log.Fatal(err)
+func JSONSchemaErrors(document []byte, schema ...string) *gojsonschema.Result {
+	// If no schema is provided, default to commonmeta_v0.12
+	if len(schema) == 0 {
+		schema = append(schema, "commonmeta_v0.12")
+	}
+	s := schema[len(schema)-1]
+	// JSON Schema files stored locally to validate against
+	schemata := []string{"commonmeta_v0.12", "datacite-v4.5", "crossref-v0.2", "csl-data", "cff_v1.2.0"}
+	if !slices.Contains(schemata, s) {
+		log.Fatalf("Schema %s not found", s)
 	}
-	schemaLoader := gojsonschema.NewReferenceLoader("file://../resources/commonmeta_v0.12.json")
-	documentLoader := gojsonschema.NewBytesLoader(documentJSON)
+	schemaPath := "file://../resources/" + s + ".json"
+	schemaLoader := gojsonschema.NewReferenceLoader(schemaPath)
+	documentLoader := gojsonschema.NewBytesLoader(document)
 	result, err := gojsonschema.Validate(schemaLoader, documentLoader)
 	if err != nil {
 		panic(err.Error())
@@ -24,7 +30,7 @@ func JSONSchemaErrors(document metadata.Metadata) *gojsonschema.Result {
 	if result.Valid() {
 		fmt.Printf("The document is valid\n")
 	} else {
-		fmt.Printf("Input: %v\n", string(documentJSON))
+		fmt.Printf("Input: %v\n", string(document))
 		fmt.Printf("The document is not valid. see errors :\n")
 		for _, desc := range result.Errors() {
 			fmt.Printf("- %s\n", desc)

schemautils/schemautils_test.go

@@ -3,20 +3,75 @@ package schemautils_test
 import (
 	"commonmeta/metadata"
 	"commonmeta/schemautils"
+	"encoding/json"
+	"fmt"
+	"log"
+	"os"
 	"testing"
 )
 
 func TestJSONSchemaErrors(t *testing.T) {
 	t.Parallel()
+	type testCase struct {
+		meta metadata.Metadata
+		want int
+	}
 	m := metadata.Metadata{
 		ID:   "https://doi.org/10.7554/elife.01567",
 		Type: "JournalArticle",
 		Url:  "https://elifesciences.org/articles/01567",
 	}
-	want := 0
-	result := schemautils.JSONSchemaErrors(m)
-	got := len(result.Errors())
-	if want != got {
-		t.Errorf("want %d, got %d", want, got)
+
+	// missing required ID, defaults to empty string
+	n := metadata.Metadata{
+		Type: "JournalArticle",
+	}
+
+	// Type is not supported
+	o := metadata.Metadata{
+		ID:   "https://doi.org/10.7554/elife.01567",
+		Type: "Umbrella",
+	}
+
+	testCases := []testCase{
+		{meta: m, want: 0},
+		{meta: n, want: 2},
+		{meta: o, want: 1},
+	}
+	for _, tc := range testCases {
+		documentJSON, err := json.Marshal(tc.meta)
+		if err != nil {
+			log.Fatal(err)
+		}
+		result := schemautils.JSONSchemaErrors(documentJSON)
+		got := len(result.Errors())
+		if tc.want != got {
+			t.Errorf("want %d, got %d", tc.want, got)
+		}
+	}
+}
+
+func TestJSONSchemaErrorsTestdata(t *testing.T) {
+	t.Parallel()
+	type testCase struct {
+		meta   string
+		schema string
+		want   int
+	}
+
+	testCases := []testCase{
+		{meta: "citeproc.json", schema: "csl-data", want: 0},
+		{meta: "datacite.json", schema: "datacite-v4.5", want: 3},
+	}
+	for _, tc := range testCases {
+		data, err := os.ReadFile("../testdata/" + tc.meta)
+		if err != nil {
+			fmt.Print(err)
+		}
+		result := schemautils.JSONSchemaErrors(data, tc.schema)
+		got := len(result.Errors())
+		if tc.want != got {
+			t.Errorf("want %d, got %d", tc.want, got)
+		}
 	}
 }

testdata/citeproc.json

@@ -1,28 +1,30 @@
-{
-  "type": "post-weblog",
-  "id": "https://doi.org/10.5438/4k3m-nyvg",
-  "DOI": "10.5438/4k3m-nyvg",
-  "URL": "https://blog.datacite.org/eating-your-own-dog-food",
-  "title": "Eating your own Dog Food",
-  "container-title": "DataCite Blog",
-  "publisher": { "name": "DataCite" },
-  "abstract": "Eating your own dog food is a slang term to describe that an organization should itself use the products and services it provides. For DataCite this means that we should use DOIs with appropriate metadata and strategies for long-term preservation for...",
-  "categories": [
-    "Phylogeny",
-    "Malaria",
-    "Parasites",
-    "Taxonomy",
-    "Mitochondrial genome",
-    "Africa",
-    "Plasmodium"
-  ],
-  "issued": {
-    "date-parts": [[2016, 12, 20]]
-  },
-  "author": [
-    {
-      "family": "Fenner",
-      "given": "Martin"
-    }
-  ]
-}
+[
+  {
+    "type": "post-weblog",
+    "id": "https://doi.org/10.5438/4k3m-nyvg",
+    "DOI": "10.5438/4k3m-nyvg",
+    "URL": "https://blog.datacite.org/eating-your-own-dog-food",
+    "title": "Eating your own Dog Food",
+    "container-title": "DataCite Blog",
+    "publisher": "DataCite",
+    "abstract": "Eating your own dog food is a slang term to describe that an organization should itself use the products and services it provides. For DataCite this means that we should use DOIs with appropriate metadata and strategies for long-term preservation for...",
+    "categories": [
+      "Phylogeny",
+      "Malaria",
+      "Parasites",
+      "Taxonomy",
+      "Mitochondrial genome",
+      "Africa",
+      "Plasmodium"
+    ],
+    "issued": {
+      "date-parts": [[2016, 12, 20]]
+    },
+    "author": [
+      {
+        "family": "Fenner",
+        "given": "Martin"
+      }
+    ]
+  }
+]