Merge pull request #10543 from IQSS/10169-JSON-schema-validation

Improved JSON Schema validation for datasets

Author: landreev

doc/release-notes/10169-JSON-schema-validation.md

@@ -0,0 +1,3 @@
+### Improved JSON Schema validation for datasets
+
+Enhanced JSON schema validation with checks for required and allowed child objects, type checking for field types including `primitive`, `compound` and `controlledVocabulary`. More user-friendly error messages to help pinpoint the issues in the dataset JSON. See [Retrieve a Dataset JSON Schema for a Collection](https://guides.dataverse.org/en/6.3/api/native-api.html#retrieve-a-dataset-json-schema-for-a-collection) in the API Guide and PR #10543.

doc/sphinx-guides/source/api/native-api.rst

@@ -566,9 +566,7 @@ The fully expanded example above (without environment variables) looks like this
 Retrieve a Dataset JSON Schema for a Collection
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
-Retrieves a JSON schema customized for a given collection in order to validate a dataset JSON file prior to creating the dataset. This
-first version of the schema only includes required elements and fields. In the future we plan to improve the schema by adding controlled
-vocabulary and more robust dataset field format testing:
+Retrieves a JSON schema customized for a given collection in order to validate a dataset JSON file prior to creating the dataset:
 
 .. code-block:: bash
 
@@ -593,8 +591,22 @@ While it is recommended to download a copy of the JSON Schema from the collectio
 Validate Dataset JSON File for a Collection
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
-Validates a dataset JSON file customized for a given collection prior to creating the dataset. The validation only tests for json formatting
-and the presence of required elements:
+Validates a dataset JSON file customized for a given collection prior to creating the dataset.
+
+The validation tests for:
+
+- JSON formatting
+- required fields
+- typeClass must follow these rules:
+
+  - if multiple = true then value must be a list
+  - if typeClass = ``primitive`` the value object is a String or a List of Strings depending on the multiple flag
+  - if typeClass = ``compound`` the value object is a FieldDTO or a List of FieldDTOs depending on the multiple flag
+  - if typeClass = ``controlledVocabulary`` the values are checked against the list of allowed values stored in the database
+  - typeName validations (child objects with their required and allowed typeNames are configured automatically by the database schema). Examples include:
+
+    - dsDescription validation includes checks for typeName = ``dsDescriptionValue`` (required) and ``dsDescriptionDate`` (optional)
+    - datasetContact validation includes checks for typeName = ``datasetContactName`` (required) and ``datasetContactEmail``; ``datasetContactAffiliation`` (optional)
 
 .. code-block:: bash
 

scripts/search/tests/data/dataset-finch3.json

@@ -0,0 +1,102 @@
+{
+  "datasetVersion": {
+    "license": {
+      "name": "CC0 1.0",
+      "uri": "http://creativecommons.org/publicdomain/zero/1.0"
+    },
+    "metadataBlocks": {
+      "citation": {
+        "fields": [
+          {
+            "value": "HTML & More",
+            "typeClass": "primitive",
+            "multiple": false,
+            "typeName": "title"
+          },
+          {
+            "value": [
+              {
+                "authorName": {
+                  "value": "Markup, Marty",
+                  "typeClass": "primitive",
+                  "multiple": false,
+                  "typeName": "authorName"
+                },
+                "authorAffiliation": {
+                  "value": "W4C",
+                  "typeClass": "primitive",
+                  "multiple": false,
+                  "typeName": "authorAffiliation"
+                }
+              }
+            ],
+            "typeClass": "compound",
+            "multiple": true,
+            "typeName": "author"
+          },
+          {
+            "value": [
+              {
+                "datasetContactEmail": {
+                  "typeClass": "primitive",
+                  "multiple": false,
+                  "typeName": "datasetContactEmail",
+                  "value": "markup@mailinator.com"
+                },
+                "datasetContactName": {
+                  "typeClass": "primitive",
+                  "multiple": false,
+                  "typeName": "datasetContactName",
+                  "value": "Markup, Marty"
+                }
+              }
+            ],
+            "typeClass": "compound",
+            "multiple": true,
+            "typeName": "datasetContact"
+          },
+          {
+            "value": [
+              {
+                "dsDescriptionValue": {
+                  "value": "BEGIN<br></br>END",
+                  "multiple": false,
+                  "typeClass": "primitive",
+                  "typeName": "dsDescriptionValue"
+                },
+                "dsDescriptionDate": {
+                  "typeName": "dsDescriptionDate",
+                  "multiple": false,
+                  "typeClass": "primitive",
+                  "value": "2021-07-13"
+                }
+              }
+            ],
+            "typeClass": "compound",
+            "multiple": true,
+            "typeName": "dsDescription"
+          },
+          {
+            "value": [
+              "Medicine, Health and Life Sciences"
+            ],
+            "typeClass": "controlledVocabulary",
+            "multiple": true,
+            "typeName": "subject"
+          },
+          {
+            "typeName": "language",
+            "multiple": true,
+            "typeClass": "controlledVocabulary",
+            "value": [
+              "English",
+              "Afar",
+              "aar"
+            ]
+          }
+        ],
+        "displayName": "Citation Metadata"
+      }
+    }
+  }
+}

src/main/java/edu/harvard/iq/dataverse/DataverseServiceBean.java

@@ -22,7 +22,7 @@
 import edu.harvard.iq.dataverse.storageuse.StorageQuota;
 import edu.harvard.iq.dataverse.util.StringUtil;
 import edu.harvard.iq.dataverse.util.SystemConfig;
-import edu.harvard.iq.dataverse.util.json.JsonUtil;
+
 import java.io.File;
 import java.io.IOException;
 import java.sql.Timestamp;
@@ -34,6 +34,7 @@
 import java.util.logging.Logger;
 import java.util.Properties;
 
+import edu.harvard.iq.dataverse.validation.JSONDataValidation;
 import jakarta.ejb.EJB;
 import jakarta.ejb.Stateless;
 import jakarta.inject.Inject;
@@ -888,14 +889,16 @@ public List<Object[]> getDatasetTitlesWithinDataverse(Long dataverseId) {
         return em.createNativeQuery(cqString).getResultList();
     }
 
-        
     public  String getCollectionDatasetSchema(String dataverseAlias) {
+        return getCollectionDatasetSchema(dataverseAlias, null);
+    }
+    public  String getCollectionDatasetSchema(String dataverseAlias, Map<String, Map<String,List<String>>> schemaChildMap) {
 
         Dataverse testDV = this.findByAlias(dataverseAlias);
 
         while (!testDV.isMetadataBlockRoot()) {
             if (testDV.getOwner() == null) {
-                break; // we are at the root; which by defintion is metadata blcok root, regarldess of the value
+                break; // we are at the root; which by definition is metadata block root, regardless of the value
             }
             testDV = testDV.getOwner();
         }
@@ -932,6 +935,8 @@ public  String getCollectionDatasetSchema(String dataverseAlias) {
                         dsft.setRequiredDV(dsft.isRequired());
                         dsft.setInclude(true);
                     }
+                    List<String> childrenRequired = new ArrayList<>();
+                    List<String> childrenAllowed = new ArrayList<>();
                     if (dsft.isHasChildren()) {
                         for (DatasetFieldType child : dsft.getChildDatasetFieldTypes()) {
                             DataverseFieldTypeInputLevel dsfIlChild = dataverseFieldTypeInputLevelService.findByDataverseIdDatasetFieldTypeId(testDV.getId(), child.getId());
@@ -944,8 +949,18 @@ public  String getCollectionDatasetSchema(String dataverseAlias) {
                                 child.setRequiredDV(child.isRequired() && dsft.isRequired());
                                 child.setInclude(true);
                             }
+                            if (child.isRequired()) {
+                                childrenRequired.add(child.getName());
+                            }
+                            childrenAllowed.add(child.getName());
                         }
                     }
+                    if (schemaChildMap != null) {
+                        Map<String, List<String>> map = new HashMap<>();
+                        map.put("required", childrenRequired);
+                        map.put("allowed", childrenAllowed);
+                        schemaChildMap.put(dsft.getName(), map);
+                    }
                     if(dsft.isRequiredDV()){
                         requiredDSFT.add(dsft);
                     }
@@ -1021,11 +1036,13 @@ private String getCustomMDBSchema (MetadataBlock mdb, List<DatasetFieldType> req
     }
 
     public String isDatasetJsonValid(String dataverseAlias, String jsonInput) {
-        JSONObject rawSchema = new JSONObject(new JSONTokener(getCollectionDatasetSchema(dataverseAlias)));
+        Map<String, Map<String,List<String>>> schemaChildMap = new HashMap<>();
+        JSONObject rawSchema = new JSONObject(new JSONTokener(getCollectionDatasetSchema(dataverseAlias, schemaChildMap)));
 
-        try {               
+        try {
             Schema schema = SchemaLoader.load(rawSchema);
             schema.validate(new JSONObject(jsonInput)); // throws a ValidationException if this object is invalid
+            JSONDataValidation.validate(schema, schemaChildMap, jsonInput); // throws a ValidationException if any objects are invalid
         } catch (ValidationException vx) {
             logger.info(BundleUtil.getStringFromBundle("dataverses.api.validate.json.failed") + " " + vx.getErrorMessage()); 
             String accumulatedexceptions = "";

src/main/java/edu/harvard/iq/dataverse/api/Datasets.java

@@ -1,7 +1,6 @@
 package edu.harvard.iq.dataverse.api;
 
 import com.amazonaws.services.s3.model.PartETag;
-
 import edu.harvard.iq.dataverse.*;
 import edu.harvard.iq.dataverse.DatasetLock.Reason;
 import edu.harvard.iq.dataverse.actionlogging.ActionLogRecord;