Add minimal dtool dataset metadata schema (#4246)

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Edwin Kofler <edwin@kofler.dev>

cli.js

@@ -833,11 +833,11 @@ async function assertFileSystemIsValid() {
       for (const testDir of await fs.readdir(rootTestDir)) {
         if (isIgnoredFile(testDir)) continue
 
-        const schemaName = testDir + '.json'
-        const schemaPath = path.join(SchemaDir, schemaName)
+        const schemaPath = path.join(SchemaDir, testDir + '.json')
         if (!(await exists(schemaPath))) {
           printErrorAndExit(new Error(), [
             `Failed to find a schema file at "${schemaPath}"`,
+            `Expected schema file computed from directory at "${path.join(rootTestDir, testDir)}"`,
           ])
         }
       }

src/api/json/catalog.json

@@ -7235,6 +7235,19 @@
       "fileMatch": ["*.bitmovin.json", "*.bitmovin.yml", "*.bitmovin.yaml"],
       "url": "https://raw.githubusercontent.com/bitmovin/bitmovin-api-sdk-examples/main/bitmovin-encoding-template.json"
     },
+    {
+      "name": "dtool dataset metadata",
+      "description": "Metadata attached to a dtool dataset",
+      "fileMatch": [
+        "dtool-dataset.yml",
+        "dtool-dataset.yaml",
+        "dtool-dataset.json"
+      ],
+      "url": "https://json.schemastore.org/dtool-dataset-metadata-1.0.json",
+      "versions": {
+        "v1.0": "https://json.schemastore.org/dtool-dataset-metadata-1.0.json"
+      }
+    },
     {
       "name": "pnpm Workspace (pnpm-workspace.yaml)",
       "description": "Workspace file for pnpm",

src/schemas/json/dtool-dataset-metadata-1.0.json

@@ -0,0 +1,89 @@
+{
+  "$schema": "http://json-schema.org/draft-07/schema#",
+  "$id": "https://json.schemastore.org/dtool-dataset-metadata-1.0.json",
+  "title": "DToolDataset",
+  "description": "Schema for metadata attached to a dtool dataset",
+  "type": "object",
+  "properties": {
+    "uuid": {
+      "type": "string",
+      "format": "uuid"
+    },
+    "base_uri": {
+      "type": "string"
+    },
+    "uri": {
+      "type": "string"
+    },
+    "name": {
+      "type": "string"
+    },
+    "type": {
+      "type": "string"
+    },
+    "readme": {
+      "anyOf": [
+        { "type": "string" },
+        {
+          "type": "object",
+          "additionalProperties": true
+        }
+      ]
+    },
+    "manifest": {
+      "type": "object",
+      "properties": {
+        "items": {
+          "type": "object",
+          "additionalProperties": {
+            "type": "object",
+            "properties": {
+              "hash": {
+                "type": "string"
+              },
+              "relpath": {
+                "type": "string"
+              },
+              "size_in_bytes": {
+                "type": "integer"
+              },
+              "utc_timestamp": {
+                "type": "number"
+              }
+            }
+          }
+        },
+        "hash_function": {
+          "type": "string"
+        },
+        "dtoolcore_version": {
+          "type": "string"
+        }
+      }
+    },
+    "creator_username": {
+      "type": "string"
+    },
+    "frozen_at": {
+      "anyOf": [{ "type": "string" }, { "type": "number" }]
+    },
+    "created_at": {
+      "anyOf": [{ "type": "string" }, { "type": "number" }]
+    },
+    "annotations": {
+      "type": "object"
+    },
+    "tags": {
+      "type": "array",
+      "items": {
+        "type": "string"
+      }
+    },
+    "number_of_items": {
+      "type": "integer"
+    },
+    "size_in_bytes": {
+      "type": "integer"
+    }
+  }
+}

src/test/dtool-dataset-metadata-1.0/dtool-dataset-metadata-1.0.json

@@ -0,0 +1,28 @@
+{
+  "annotations": {
+    "software": "bowtie2"
+  },
+  "base_uri": "s3://snow-white",
+  "creator_username": "olssont",
+  "frozen_at": "1536238185.881941",
+  "manifest": {
+    "dtoolcore_version": "3.7.0",
+    "hash_function": "md5sum_hexdigest",
+    "items": {
+      "e4cc3a7dc281c3d89ed4553293c4b4b110dc9bf3": {
+        "hash": "d89117c9da2cc34586e183017cb14851",
+        "relpath": "U00096.3.rev.1.bt2",
+        "size_in_bytes": 5741810,
+        "utc_timestamp": 1536832115.0
+      }
+    }
+  },
+  "name": "my-dataset",
+  "number_of_items": 1,
+  "readme": "---\ndescription: test dataset",
+  "size_in_bytes": 5741810,
+  "tags": ["rnaseq"],
+  "type": "dataset",
+  "uri": "s3://snow-white/af6727bf-29c7-43dd-b42f-a5d7ede28337",
+  "uuid": "af6727bf-29c7-43dd-b42f-a5d7ede28337"
+}