From 31e0f2e34cfb62efc75a0497649ff4c76b181104 Mon Sep 17 00:00:00 2001
From: Kathleen Tuite <ktuite@getodk.org>
Date: Fri, 19 Apr 2024 13:07:30 -0700
Subject: [PATCH] Handle dataset name capitalization conflicts in API

---
 lib/model/query/datasets.js      | 20 +++++++-
 test/integration/api/datasets.js | 87 +++++++++++++++++++++++++++-----
 2 files changed, 92 insertions(+), 15 deletions(-)

diff --git a/lib/model/query/datasets.js b/lib/model/query/datasets.js
index 69fb4c79a..52b8c5686 100644
--- a/lib/model/query/datasets.js
+++ b/lib/model/query/datasets.js
@@ -131,7 +131,25 @@ const createOrMerge = (parsedDataset, form, fields) => async ({ one, Actees, Dat
 };
 
 // Insert a Dataset when there is no associated form and immediately publish
-const createPublishedDataset = (dataset, project) => async ({ one, Actees }) => {
+const createPublishedDataset = (dataset, project) => async ({ one, Actees, Datasets }) => {
+
+  // Check for existing dataset with the same name
+  const existingDataset = await Datasets.get(project.id, dataset.name, false);
+  if (existingDataset.isDefined()) {
+    const { publishedAt, id: datasetId } = existingDataset.get();
+    if (publishedAt)
+      throw Problem.user.uniquenessViolation({ table: 'datasets', fields: [ 'name', 'projectId' ], values: [ dataset.name, project.id ], });
+
+    // If existing dataset is only a draft, allow it to be published
+    return new Dataset(await one(sql`UPDATE datasets SET "publishedAt" = "createdAt" where id = ${datasetId} RETURNING *`));
+  }
+
+  // Check for a published dataset with a similar name but different capitalization
+  const conflictingName = await Datasets.getPublishedBySimilarName(project.id, dataset.name);
+  if (conflictingName.isDefined())
+    throw Problem.user.datasetNameConflict({ current: conflictingName.get().name, provided: dataset.name });
+
+  // Final case: publish a completely new dataset
   const actee = await Actees.provision('dataset', project);
   const dsWithId = await one(insert(dataset.with({ acteeId: actee.id })));
   return new Dataset(await one(sql`UPDATE datasets SET "publishedAt" = "createdAt" where id = ${dsWithId.id} RETURNING *`));
diff --git a/test/integration/api/datasets.js b/test/integration/api/datasets.js
index 7d7d40fa1..801e4ba78 100644
--- a/test/integration/api/datasets.js
+++ b/test/integration/api/datasets.js
@@ -98,22 +98,81 @@ describe('datasets and entities', () => {
           });
       }));
 
-      it('should reject if creating a dataset that already exists', testService(async (service) => {
-        const asAlice = await service.login('alice');
+      describe('dataset name conflicts', () => {
+        it('should reject if creating a dataset that already exists', testService(async (service) => {
+          const asAlice = await service.login('alice');
 
-        await asAlice.post('/v1/projects/1/datasets')
-          .send({
-            name: 'trees'
-          })
-          .expect(200);
+          await asAlice.post('/v1/projects/1/datasets')
+            .send({
+              name: 'trees'
+            })
+            .expect(200);
 
-        // Second time
-        await asAlice.post('/v1/projects/1/datasets')
-          .send({
-            name: 'trees'
-          })
-          .expect(409);
-      }));
+          // Second time
+          await asAlice.post('/v1/projects/1/datasets')
+            .send({
+              name: 'trees'
+            })
+            .expect(409)
+            .then(({ body }) => {
+              body.code.should.equal(409.3);
+              body.message.should.startWith('A resource already exists with name,projectId value(s) of trees,');
+            });
+        }));
+
+        it('should reject if creating a dataset that has a similar name to an existing published dataset', testService(async (service) => {
+          const asAlice = await service.login('alice');
+
+          await asAlice.post('/v1/projects/1/datasets')
+            .send({
+              name: 'trees'
+            })
+            .expect(200);
+
+          // Second time
+          await asAlice.post('/v1/projects/1/datasets')
+            .send({
+              name: 'TREES'
+            })
+            .expect(409)
+            .then(({ body }) => {
+              body.code.should.equal(409.16);
+              body.message.should.startWith("A dataset named 'trees' exists and you provided 'TREES'");
+            });
+        }));
+
+        it('should allow creating a dataset that only exists as a draft', testService(async (service) => {
+          const asAlice = await service.login('alice');
+
+          // draft "people" dataset
+          await asAlice.post('/v1/projects/1/forms')
+            .send(testData.forms.simpleEntity)
+            .set('Content-Type', 'application/xml')
+            .expect(200);
+
+          await asAlice.post('/v1/projects/1/datasets')
+            .send({
+              name: 'people'
+            })
+            .expect(200);
+        }));
+
+        it('should allow creating a dataset that has a similar name to a draft dataset', testService(async (service) => {
+          const asAlice = await service.login('alice');
+
+          // draft "people" dataset
+          await asAlice.post('/v1/projects/1/forms')
+            .send(testData.forms.simpleEntity)
+            .set('Content-Type', 'application/xml')
+            .expect(200);
+
+          await asAlice.post('/v1/projects/1/datasets')
+            .send({
+              name: 'PEOPLE'
+            })
+            .expect(200);
+        }));
+      });
 
       it('should add label-only entity to dataset, all via API', testService(async (service) => {
         const asAlice = await service.login('alice');