feat!: remove rocrate column from Entity table

BREAKING CHANGE: The rocrate JSON column has been removed from the Entity table.
RO-Crate metadata is now served exclusively via the roCrateHandler from external
sources (filesystem, S3, etc.) rather than being stored in the database.

Changes:
- Remove rocrate column from Prisma schema
- Add migration to drop the column
- Update seed script and loadEntities to not store rocrate
- Remove rocrate from OpenSearch index mapping
- Update all test files to remove rocrate from mock data
- Update documentation to clarify rocrate is not stored in DB

The /entity/:id/rocrate endpoint continues to work using the configured
roCrateHandler to serve RO-Crate data from external storage.

Author: johnf

CLAUDE.md

@@ -1386,36 +1386,7 @@ type HeadRoCrateHandler = (
 
 #### RO-Crate Handler Examples
 
-**Stream from Database**:
-
-```typescript
-roCrateHandler: {
-  get: async (entity) => {
-    const rocrate = entity.rocrate;
-    const jsonString = JSON.stringify(rocrate, null, 2);
-
-    return {
-      type: 'stream',
-      stream: Readable.from([jsonString]),
-      metadata: {
-        contentType: 'application/ld+json',
-        contentLength: Buffer.byteLength(jsonString),
-        etag: `"${entity.id}-rocrate"`,
-        lastModified: entity.updatedAt,
-      },
-    };
-  },
-  head: async (entity) => {
-    const jsonString = JSON.stringify(entity.rocrate);
-    return {
-      contentType: 'application/ld+json',
-      contentLength: Buffer.byteLength(jsonString),
-      etag: `"${entity.id}-rocrate"`,
-      lastModified: entity.updatedAt,
-    };
-  },
-}
-```
+**Note**: RO-Crate metadata is NOT stored in the database. The `roCrateHandler` must serve RO-Crate data from external sources (filesystem, S3, etc.).
 
 **Serve from Filesystem**:
 

example/data/README.md

@@ -277,7 +277,9 @@ Entities are stored in the `Entity` table with:
 - `rootCollection`: Top-level collection reference
 - `metadataLicenseId`: License for metadata
 - `contentLicenseId`: License for content
-- `rocrate`: Full RO-Crate metadata (JSON)
+- `meta`: Storage metadata (JSON, optional)
+
+Note: RO-Crate metadata is served via the `/entity/:id/rocrate` endpoint using the configured `roCrateHandler`, not stored in the database.
 
 ### OpenSearch Mappings
 

example/data/seed.ts

@@ -74,7 +74,6 @@ const processCollection = async (
       collectionRoot.contentLicense?.['@id'] ||
       collectionRoot.license?.['@id'] ||
       'https://creativecommons.org/licenses/by/4.0/',
-    rocrate: collectionCrate,
     meta: {
       storagePath: collectionPath,
     },
@@ -115,7 +114,6 @@ const processItem = async (
       itemRoot.metadataLicense?.['@id'] || itemRoot.license?.['@id'] || 'https://creativecommons.org/licenses/by/4.0/',
     contentLicenseId:
       itemRoot.contentLicense?.['@id'] || itemRoot.license?.['@id'] || 'https://creativecommons.org/licenses/by/4.0/',
-    rocrate: itemCrate,
     meta: {
       storagePath: itemPath,
     },
@@ -167,7 +165,6 @@ const processItem = async (
             itemRoot.contentLicense?.['@id'] ||
             itemRoot.license?.['@id'] ||
             'https://creativecommons.org/licenses/by/4.0/',
-          rocrate: fileNode,
           meta: {
             remapRootTo: fileRocrateId,
             storagePath: itemPath,

prisma/migrations/20251211025510_remove_rocrate_column/migration.sql

@@ -0,0 +1,8 @@
+/*
+  Warnings:
+
+  - You are about to drop the column `rocrate` on the `Entity` table. All the data in the column will be lost.
+
+*/
+-- AlterTable
+ALTER TABLE `Entity` DROP COLUMN `rocrate`;

prisma/models/entity.prisma

@@ -13,7 +13,6 @@ model Entity {
   contentLicenseId    String     @db.VarChar(2048)
   fileId              String?    @db.VarChar(2048)
 
-  rocrate             Json
   meta                Json?
 
   createdAt DateTime @default(now())

scripts/loadEntities.ts

@@ -48,19 +48,6 @@ const fetchEntities = async (limit: number, offset: number) => {
   return result;
 };
 
-const fetchEntityRocrate = async (entityId: string) => {
-  const url = `${API_URL}/entity/${encodeURIComponent(entityId)}`;
-  console.log(`Fetching ROCrate for ${entityId}`);
-
-  const response = await fetch(url);
-  if (!response.ok) {
-    throw new Error(`ROCrate API request failed with status ${response.status} for entity ${entityId}`);
-  }
-
-  const rocrate = await response.json();
-  return rocrate;
-};
-
 const generateDummyData = () => {
   const languages = ['English', 'Japanese', 'French', 'Spanish', 'German', 'Italian'];
   const mediaTypes = ['audio/wav', 'audio/mp3', 'video/mp4', 'text/plain', 'image/jpeg'];
@@ -114,7 +101,6 @@ const createIndex = async () => {
             location: { type: 'geo_point' },
             createdAt: { type: 'date' },
             updatedAt: { type: 'date' },
-            rocrate: { type: 'text' },
           },
         },
       },
@@ -129,7 +115,7 @@ const createIndex = async () => {
 
 // Index entity to OpenSearch
 // biome-ignore lint/suspicious/noExplicitAny: FIXME
-const indexEntity = async (entity: any, dummyData: any, rocrate: any) => {
+const indexEntity = async (entity: any, dummyData: any) => {
   const document = {
     rocrateId: entity.id,
     name: entity.name,
@@ -144,7 +130,6 @@ const indexEntity = async (entity: any, dummyData: any, rocrate: any) => {
     location: dummyData.location,
     createdAt: new Date().toISOString(),
     updatedAt: new Date().toISOString(),
-    rocrate: JSON.stringify(rocrate),
   };
 
   await opensearch.index({
@@ -174,9 +159,6 @@ const loadEntities = async (): Promise<void> => {
       for (const entity of data.entities) {
         const dummyData = generateDummyData();
 
-        // Fetch ROCrate data for this entity
-        const rocrate = await fetchEntityRocrate(entity.id);
-
         await prisma.entity.create({
           data: {
             rocrateId: entity.id,
@@ -185,11 +167,12 @@ const loadEntities = async (): Promise<void> => {
             entityType: entity.entityType,
             memberOf: entity.memberOf || null,
             rootCollection: entity.root || null,
-            rocrate: rocrate,
+            metadataLicenseId: 'https://creativecommons.org/licenses/by/4.0/',
+            contentLicenseId: 'https://creativecommons.org/licenses/by/4.0/',
           },
         });
 
-        await indexEntity(entity, dummyData, rocrate);
+        await indexEntity(entity, dummyData);
       }
 
       processedCount += data.entities.length;

src/routes/crate.test.ts

@@ -39,22 +39,6 @@ describe('Crate Route', () => {
     contentLicenseId: 'https://creativecommons.org/licenses/by/4.0/',
     createdAt: new Date(),
     updatedAt: new Date(),
-    rocrate: {
-      '@context': 'https://w3id.org/ro/crate/1.1/context',
-      '@graph': [
-        {
-          '@id': 'ro-crate-metadata.json',
-          '@type': 'CreativeWork',
-          conformsTo: { '@id': 'https://w3id.org/ro/crate/1.1' },
-          about: { '@id': './' },
-        },
-        {
-          '@id': './',
-          '@type': 'Dataset',
-          name: 'Test RO-Crate',
-        },
-      ],
-    },
     meta: {},
   };
 

src/routes/entities.test.ts

@@ -30,7 +30,6 @@ describe('Entities Route', () => {
           contentLicenseId: 'https://creativecommons.org/licenses/by/4.0/',
           createdAt: new Date(),
           updatedAt: new Date(),
-          rocrate: {},
           meta: {},
         },
         {
@@ -46,7 +45,6 @@ describe('Entities Route', () => {
           contentLicenseId: 'https://creativecommons.org/licenses/by/4.0/',
           createdAt: new Date(),
           updatedAt: new Date(),
-          rocrate: {},
           meta: {},
         },
       ];
@@ -296,7 +294,6 @@ describe('Entities Route', () => {
           contentLicenseId: 'https://creativecommons.org/licenses/by/4.0/',
           createdAt: new Date(),
           updatedAt: new Date(),
-          rocrate: {},
           meta: {},
         },
       ];

src/routes/entity.test.ts

@@ -30,7 +30,6 @@ describe('Entity Route', () => {
         contentLicenseId: 'https://creativecommons.org/licenses/by/4.0/',
         createdAt: new Date(),
         updatedAt: new Date(),
-        rocrate: {},
         meta: {},
       };
 
@@ -143,7 +142,6 @@ describe('Entity Route', () => {
         contentLicenseId: 'https://creativecommons.org/licenses/by/4.0/',
         createdAt: new Date(),
         updatedAt: new Date(),
-        rocrate: {},
         meta: {},
       };
 

src/routes/search.test.ts

@@ -810,7 +810,6 @@ describe('Search Route', () => {
           id: 1,
           fileId: null,
           meta: {},
-          rocrate: '',
           rocrateId: 'http://example.com/entity/1',
           name: 'Test Entity 1',
           description: 'Entity with missing parent',