Merge pull request #261 from CanDIG/daisieh/granularity

DIG-1143: SampleDRSObjects

data/files.sql

@@ -29,6 +29,7 @@ CREATE TABLE content_object (
         id INTEGER NOT NULL,
         drs_object_id INTEGER,
         name VARCHAR,
+        contents_id VARCHAR,
         drs_uri VARCHAR,
         contents VARCHAR,
         PRIMARY KEY (id),

htsget_server/beacon_operations.py

@@ -81,6 +81,9 @@ def get_search(
         },
         "query": {
             "requestParameters": {}
+        },
+        "meta": {
+            "requestedGranularity": "record"
         }
     }
     if alternate_bases is not None:
@@ -104,7 +107,6 @@ def get_search(
     if variant_min_length is not None:
         req['query']['requestParameters']['variant_min_length'] = variant_min_length
 
-    req['requestedGranularity'] = 'record'
     try:
         result = search(req)
         return result, 200
@@ -124,8 +126,8 @@ def post_search():
     #   $ref: '#/components/schemas/Granularity'
     # testMode:
     #   $ref: '#/components/schemas/TestMode'
-    if 'requestedGranularity' not in req:
-        req['requestedGranularity'] = 'record'
+    if 'requestedGranularity' not in req['meta']:
+        req['meta']['requestedGranularity'] = 'record'
 
     params = list(req['query']['requestParameters'].keys())
     for param in params:
@@ -157,9 +159,9 @@ def search(raw_req):
     }
     if 'pagination' in raw_req:
         meta['receivedRequestSummary']['pagination'] = raw_req['pagination']
-    if 'requestedGranularity' in raw_req:
-        meta['receivedRequestSummary']['requestedGranularity'] = raw_req['requestedGranularity']
-        meta['returnedGranularity'] = raw_req['requestedGranularity']
+    if 'requestedGranularity' in raw_req['meta']:
+        meta['receivedRequestSummary']['requestedGranularity'] = raw_req['meta']['requestedGranularity']
+        meta['returnedGranularity'] = raw_req['meta']['requestedGranularity']
     ## not using includeResultsetResponses for now:
     # if 'includeResultsetResponses' in raw_req:
     #     meta['receivedRequestSummary']['includeResultsetResponses'] = raw_req['includeResultsetResponses']
@@ -278,11 +280,12 @@ def search(raw_req):
             if handover is not None:
                 handover['handoverType'] = {'id': 'CUSTOM', 'label': 'HTSGET'}
                 response['beaconHandovers'].append(handover)
-        if len(response['beaconHandovers']) > 0:
+        if len(response['beaconHandovers']) > 0 and meta['returnedGranularity'] == 'record':
             response['response'] = resultset
         else:
-            meta['returnedGranularity'] = 'count'
             response.pop('beaconHandovers')
+            if meta['returnedGranularity'] == 'boolean':
+                response['responseSummary'].pop('numTotalResults')
     else:
         response = {
             'error': {

htsget_server/database.py

@@ -3,7 +3,7 @@
 import json
 import re
 from datetime import datetime
-from config import DB_PATH, BUCKET_SIZE
+from config import DB_PATH, BUCKET_SIZE, HTSGET_URL
 
 
 engine = create_engine(DB_PATH, echo=False)
@@ -347,11 +347,13 @@ class ContentsObject(ObjectDBBase):
     drs_object_id = Column(Integer, ForeignKey('drs_object.id'))
     drs_object = relationship("DrsObject", back_populates="contents")
     name = Column(String, default='') # like a filename
+    contents_id = Column(String)
     drs_uri = Column(String, default='[]') # JSON array of strings of DRS id URIs
     contents = Column(String, default='[]') # JSON array of ContentsObject.ids
     def __repr__(self):
         result = {
             'name': self.name,
+            'id': self.contents_id,
             'drs_uri': json.loads(self.drs_uri)
         }
         if len(json.loads(self.contents)) > 0:
@@ -399,8 +401,7 @@ def create_drs_object(obj):
             new_object.name = obj['id']
 
         # optional string fields
-        if 'self_uri' in obj:
-            new_object.self_uri = obj['self_uri']
+        new_object.self_uri = f'{HTSGET_URL.replace("http://", "drs://").replace("https://", "drs://")}/{new_object.name}'
         if 'created_time' in obj:
             new_object.created_time = obj['created_time']
         if 'updated_time' in obj:
@@ -456,6 +457,8 @@ def create_drs_object(obj):
                 new_contents.drs_uri = json.dumps(contents['drs_uri'])
             if 'contents' in contents:
                 new_contents.contents = json.dumps(contents['contents'])
+            if 'id' in contents:
+                new_contents.contents_id = contents['id']
             session.add(new_contents)
         session.add(new_object)
         session.commit()

htsget_server/drs_openapi.yaml

@@ -42,6 +42,7 @@ paths:
                 type: array
                 items:
                   anyOf:
+                    - $ref: '#/components/schemas/SampleDrsObject'
                     - $ref: '#/components/schemas/GenomicDrsObject'
                     - $ref: '#/components/schemas/GenomicDataDrsObject'
                     - $ref: '#/components/schemas/GenomicIndexDrsObject'
@@ -242,8 +243,10 @@ components:
         'application/json':
           schema:
             anyOf:
+              - $ref: "#/components/schemas/SampleDrsObject"
               - $ref: "#/components/schemas/GenomicDrsObject"
               - $ref: "#/components/schemas/GenomicDataDrsObject"
+              - $ref: "#/components/schemas/GenomicIndexDrsObject"
     DatasetRequest:
       content:
         'application/json':
@@ -262,6 +265,88 @@ components:
           items:
             type: string
             description: a DRS object's self_uri
+    SampleDrsObject:
+      type: object
+      description: A DrsObject that describes the clinical sample used for genomic analysis.
+      required:
+        - id
+        - contents
+#        - self_uri
+#        - size
+#        - created_time
+#        - checksums
+      properties:
+        id:
+          type: string
+          description: The identifier for the sample, as defined in sample_registration in the MOHCCN data model.
+        self_uri:
+          type: string
+          description: |-
+            A drs:// hostname-based URI, as defined in the DRS documentation, that tells clients how to access this object.
+            The intent of this field is to make DRS objects self-contained, and therefore easier for clients to store and pass around.  For example, if you arrive at this DRS JSON by resolving a compact identifier-based DRS URI, the `self_uri` presents you with a hostname and properly encoded DRS ID for use in subsequent `access` endpoint calls.
+          example:
+            drs://drs.example.org/314159
+        size:
+          type: integer
+          format: int64
+          description: The cumulative size, in bytes, of items in the `contents` field.
+        created_time:
+          type: string
+          format: date-time
+          description: |-
+            Timestamp of content creation in RFC3339.
+            (This is the creation time of the underlying content, not of the JSON object.)
+        updated_time:
+          type: string
+          format: date-time
+          description: >-
+            Timestamp of content update in RFC3339, identical to `created_time` in systems
+            that do not support updates.
+            (This is the update time of the underlying content, not of the JSON object.)
+        version:
+          type: string
+          description: >-
+            A string representing a version.
+            (Some systems may use checksum, a RFC3339 timestamp, or an incrementing version number.)
+        checksums:
+          type: array
+          # minItems: 1
+          items:
+            $ref: '#/components/schemas/Checksum'
+          description: >-
+            The checksum of the `DrsObject`. At least one checksum must be provided.
+            For blobs, the checksum is computed over the bytes in the blob.
+            For bundles, the checksum is computed over a sorted concatenation of the
+            checksums of its top-level contained objects (not recursive, names not included).
+            The list of checksums is sorted alphabetically (hex-code) before concatenation
+            and a further checksum is performed on the concatenated checksum value.
+            For example, if a bundle contains blobs with the following checksums:
+            md5(blob1) = 72794b6d
+            md5(blob2) = 5e089d29
+            Then the checksum of the bundle is:
+            md5( concat( sort( md5(blob1), md5(blob2) ) ) )
+            = md5( concat( sort( 72794b6d, 5e089d29 ) ) )
+            = md5( concat( 5e089d29, 72794b6d ) )
+            = md5( 5e089d2972794b6d )
+            = f7a29a04
+        contents:
+          type: array
+          description: The specific genomic contents objects that were generated from this sample.
+          minItems: 1
+          items:
+            $ref: '#/components/schemas/GenomicContentsObject'
+        description:
+          type: string
+          description: A human readable description of the `DrsObject`.
+        aliases:
+          type: array
+          items:
+            type: string
+          description: >-
+            A list of strings that can be used to find other metadata
+            about this `DrsObject` from external metadata sources. These
+            aliases can be used to represent secondary
+            accession numbers or external GUIDs.
     GenomicDrsObject:
       type: object
       description: A DrsObject that describes a bundled genomic data entity. It usually will consist of a genomic data file, e.g. a variant or read file, and its associated index file.
@@ -594,6 +679,24 @@ components:
       anyOf:
         - $ref: '#/components/schemas/LocalFileAccessMethod'
         - $ref: '#/components/schemas/S3AccessMethod'
+    GenomicContentsObject:
+      type: object
+      required:
+        - name
+        - id
+      properties:
+        name:
+          type: string
+          description: The identifier of the genomic object
+        id:
+          type: string
+          enum:
+            - genomic
+        drs_uri:
+          type: array
+          description: The DRS uri(s) to the GenomicDrsObject
+          items:
+            type: string
     GenomicDataContentsObject:
       type: object
       required:
@@ -610,11 +713,7 @@ components:
             - read
         drs_uri:
           type: array
-          description: >-
-            A list of full DRS identifier URI paths
-            that may be used to obtain the object.
-            These URIs may be external to this DRS instance.
-          example: drs://drs.example.org/314159
+          description: The DRS uri(s) to the GenomicDataDrsObject
           items:
             type: string
     GenomicIndexContentsObject:
@@ -632,10 +731,6 @@ components:
             - index
         drs_uri:
           type: array
-          description: >-
-            A list of full DRS identifier URI paths
-            that may be used to obtain the object.
-            These URIs may be external to this DRS instance.
-          example: drs://drs.example.org/314159
+          description: The DRS uri(s) to the GenomicIndexDrsObject
           items:
-            type: string
+            type: string