[ENH] Implemented new endpoint to return known node APIs

app/api/crud.py

@@ -60,9 +60,13 @@ async def get(
     if image_modal:
         params["image_modal"] = image_modal
 
-    for node_url in util.parse_nodes_as_list(util.NEUROBAGEL_NODES):
+    nodes_dict = util.parse_nodes_as_dict(util.NEUROBAGEL_NODES)
+    for node_url, node_name in nodes_dict.items():
         response = util.send_get_request(node_url + "query/", params)
 
+        for result in response:
+            result["node_name"] = node_name
+
         cross_node_results += response
 
     return cross_node_results
@@ -85,16 +89,17 @@ async def get_terms(data_element_URI: str):
     cross_node_results = []
     params = {data_element_URI: data_element_URI}
 
-    for node_url in util.parse_nodes_as_list(util.NEUROBAGEL_NODES):
+    for node_url in util.parse_nodes_as_dict(util.NEUROBAGEL_NODES).keys():
         response = util.send_get_request(
             node_url + "attributes/" + data_element_URI, params
         )
 
         cross_node_results.append(response)
 
-    unique_terms = set(
-        term
-        for list_of_terms in cross_node_results
-        for term in list_of_terms[data_element_URI]
-    )
-    return {data_element_URI: list(unique_terms)}
+    unique_terms_dict = {}
+
+    for list_of_terms in cross_node_results:
+        for term in list_of_terms[data_element_URI]:
+            unique_terms_dict[term["TermURL"]] = term
+
+    return {data_element_URI: list(unique_terms_dict.values())}

app/api/models.py

@@ -1,4 +1,7 @@
 """Data models."""
+
+from typing import Optional, Union
+
 from pydantic import BaseModel
 
 CONTROLLED_TERM_REGEX = r"^[a-zA-Z]+[:]\S+$"
@@ -15,3 +18,18 @@ class QueryModel(BaseModel):
     min_num_sessions: int = None
     assessment: str = None
     image_modal: str = None
+
+
+class CohortQueryResponse(BaseModel):
+    """Data model for query results for one matching dataset (i.e., a cohort)."""
+
+    node_name: str
+    dataset_uuid: str
+    # dataset_file_path: str  # TODO: Revisit this field once we have datasets without imaging info/sessions.
+    dataset_name: str
+    dataset_portal_uri: Optional[str]
+    dataset_total_subjects: int
+    records_protected: bool
+    num_matching_subjects: int
+    subject_data: Union[list[dict], str]
+    image_modals: list

app/api/routers/nodes.py

@@ -0,0 +1,14 @@
+from fastapi import APIRouter
+
+from .. import utility as util
+
+router = APIRouter(prefix="/nodes", tags=["nodes"])
+
+
+@router.get("/")
+async def get_nodes():
+    """Returns a dict of all available nodes apis where key is node URL and value is node name."""
+    return [
+        {"NodeName": v, "ApiURL": k}
+        for k, v in util.parse_nodes_as_dict(util.NEUROBAGEL_NODES).items()
+    ]

app/api/routers/query.py

@@ -1,14 +1,16 @@
 """Router for query path operations."""
 
+from typing import List
+
 from fastapi import APIRouter, Depends
 
 from .. import crud
-from ..models import QueryModel
+from ..models import CohortQueryResponse, QueryModel
 
 router = APIRouter(prefix="/query", tags=["query"])
 
 
-@router.get("/")
+@router.get("/", response_model=List[CohortQueryResponse])
 async def get_query(query: QueryModel = Depends(QueryModel)):
     """When a GET request is sent, return list of dicts corresponding to subject-level metadata aggregated by dataset."""
     response = await crud.get(

app/api/utility.py

@@ -1,23 +1,29 @@
 """Constants for federation."""
 
 import os
+import re
 
 import httpx
 from fastapi import HTTPException
 
 #  Neurobagel nodes
-NEUROBAGEL_NODES = os.environ.get("NB_NODES", "https://api.neurobagel.org/")
+NEUROBAGEL_NODES = os.environ.get(
+    "LOCAL_NB_NODES", "(https://api.neurobagel.org/, OpenNeuro)"
+)
 
 
-def parse_nodes_as_list(nodes: str) -> list:
-    """Returns user-defined Neurobagel nodes as a list.
-    Empty strings are filtered out, because they are falsy.
+def parse_nodes_as_dict(nodes: str) -> list:
+    """Returns user-defined Neurobagel nodes as a dict.
+    It uses a regular expression to match the url, name pairs.
     Makes sure node URLs end with a slash."""
-    nodes_list = nodes.split(" ")
-    for i in range(len(nodes_list)):
-        if nodes_list[i] and not nodes_list[i].endswith("/"):
-            nodes_list[i] += "/"
-    return list(filter(None, nodes_list))
+    pattern = re.compile(r"\((?P<url>https?://[^\s]+), (?P<label>[^\)]+)\)")
+    matches = pattern.findall(nodes)
+    for i in range(len(matches)):
+        url, label = matches[i]
+        if not url.endswith("/"):
+            matches[i] = (url + "/", label)
+    nodes_dict = {url: label for url, label in matches}
+    return nodes_dict
 
 
 def send_get_request(url: str, params: list):

app/main.py

@@ -6,7 +6,7 @@
 from fastapi.openapi.docs import get_redoc_html, get_swagger_ui_html
 from fastapi.responses import ORJSONResponse, RedirectResponse
 
-from .api.routers import attributes, query
+from .api.routers import attributes, nodes, query
 
 app = FastAPI(
     default_response_class=ORJSONResponse, docs_url=None, redoc_url=None
@@ -56,6 +56,7 @@ def overridden_redoc():
 
 app.include_router(query.router)
 app.include_router(attributes.router)
+app.include_router(nodes.router)
 
 # Automatically start uvicorn server on execution of main.py
 if __name__ == "__main__":

tests/test_utility.py

@@ -9,30 +9,35 @@
     "set_nodes, expected_nodes",
     [
         (
-            "https://firstnode.neurobagel.org/query",
-            ["https://firstnode.neurobagel.org/query/"],
+            "(https://firstnode.neurobagel.org/query, firstnode)",
+            {"https://firstnode.neurobagel.org/query/": "firstnode"},
         ),
         (
-            "https://firstnode.neurobagel.org/query/ https://secondnode.neurobagel.org/query",
-            [
-                "https://firstnode.neurobagel.org/query/",
-                "https://secondnode.neurobagel.org/query/",
-            ],
+            "(https://firstnode.neurobagel.org/query/, firstnode) (https://secondnode.neurobagel.org/query, secondnode)",
+            {
+                "https://firstnode.neurobagel.org/query/": "firstnode",
+                "https://secondnode.neurobagel.org/query/": "secondnode",
+            },
         ),
         (
-            " https://firstnode.neurobagel.org/query/ https://secondnode.neurobagel.org/query/  ",
-            [
-                "https://firstnode.neurobagel.org/query/",
-                "https://secondnode.neurobagel.org/query/",
-            ],
+            "(firstnode.neurobagel.org/query/, firstnode) (https://secondnode.neurobagel.org/query, secondnode)",
+            {
+                "https://secondnode.neurobagel.org/query/": "secondnode",
+            },
+        ),
+        (
+            "( , firstnode) (https://secondnode.neurobagel.org/query, secondnode)",
+            {
+                "https://secondnode.neurobagel.org/query/": "secondnode",
+            },
         ),
     ],
 )
-def test_parse_nodes_as_list(monkeypatch, set_nodes, expected_nodes):
+def test_parse_nodes_as_dict(monkeypatch, set_nodes, expected_nodes):
     """Test that Neurobagel node URLs provided in a string environment variable are correctly parseed into a list."""
     monkeypatch.setenv("NB_NODES", set_nodes)
     assert sorted(
-        util.parse_nodes_as_list(
+        util.parse_nodes_as_dict(
             os.environ.get("NB_NODES", "https://api.neurobagel.org/query/")
         )
     ) == sorted(expected_nodes)