working on migration script for old nidm files

Author: dbkeator

src/nidm/core/Constants.py

@@ -50,7 +50,7 @@
 SKOS = Namespace("http://www.w3.org/2004/02/skos/core#")
 FOAF = Namespace("http://xmlns.com/foaf/0.1/")
 VC = Namespace("http://www.w3.org/2006/vcard/ns#")
-DICOM = Namespace("http://neurolex.org/wiki/Category:DICOM_term/")
+DICOM = Namespace("http://neurolex.org/wiki/Category/DICOM_term/")
 DCTYPES = Namespace("http://purl.org/dc/dcmitype/")
 NCIT = Namespace("http://ncitt.ncit.nih.gov/")
 DCAT = Namespace("http://www.w3.org/ns/dcat#")

src/nidm/experiment/Query.py

@@ -660,7 +660,7 @@ def getProjectAcquisitionObjects(nidm_file_list, project_id):
         # find all the projects
         for project, _, _ in rdf_graph.triples((None, None, Constants.NIDM["Project"])):
             # check if it is our project
-            if str(project) == project_uuid:
+            if (str(project) == project_uuid) or (str(project) == str(project_uuid)):
                 for session, _, _ in rdf_graph.triples(
                     (None, isa, Constants.NIDM["Session"])
                 ):

src/nidm/experiment/tools/update_old_simple2_nidm.py

@@ -0,0 +1,103 @@
+"""Tools for working with NIDM-Experiment files"""
+
+import click
+from rdflib import Graph, Literal, Namespace, URIRef, util
+from rdflib.namespace import RDF, split_uri
+from nidm.core import Constants
+from nidm.experiment.Core import getUUID
+from nidm.experiment.Query import GetProjectsMetadata, getProjectAcquisitionObjects
+from nidm.experiment.tools.click_base import cli
+
+
+# adding click argument parsing
+@cli.command()
+@click.option(
+    "--nidm_file_list",
+    "-nl",
+    required=True,
+    help="A comma separated list of NIDM files with full path",
+)
+def update(nidm_file_list):
+    """
+    This function will update simple 2 NIDM files. It will essentially move some of the metadata in the project activity
+    to a prov:Collection entity and then include all nidm:AcquisitionObject entities as prov:hadMember entries in this
+    new prov:Collection
+    """
+    BIDS = Namespace(Constants.BIDS)
+    PROV = Namespace(Constants.PROV)
+    NIIRI = Namespace(Constants.NIIRI)
+
+    # for each nidm file provided
+    for nidm_file in nidm_file_list.split(","):
+        # create empty graph
+        graph_orig = Graph()
+
+        # bind namespaces
+        graph_orig.bind("bids", BIDS)
+        graph_orig.bind("prov", PROV)
+        graph_orig.bind("niiri", NIIRI)
+
+        # load graph
+        graph_orig.parse(nidm_file, format=util.guess_format(nidm_file))
+
+        # create a new prov:Collection, prov:Entity to store metadata about project
+        collection_uuid = getUUID()
+        graph_orig.add((NIIRI.term(collection_uuid), RDF.type, BIDS.Dataset))
+        graph_orig.add((NIIRI.term(collection_uuid), RDF.type, PROV.Collection))
+        graph_orig.add((NIIRI.term(collection_uuid), RDF.type, PROV.Entity))
+
+        # query for all metadata in nidm:Project activity
+        proj_metadata = GetProjectsMetadata([nidm_file])
+
+        # loop through projects and get associated acquisition objects
+        for proj_uuid in proj_metadata["projects"].keys():
+            # for all project metadata, only leave dcmitype:title, otherwise move to collection_uuid
+            for metadata_key in proj_metadata["projects"][proj_uuid].keys():
+                if (metadata_key != str(RDF.type)) and (
+                    metadata_key != "dctypes:title"
+                ):
+                    # get namespace and term from metadata_key
+                    nm, term = split_uri(metadata_key)
+
+                    # add RDF namespace
+                    temp_nm = Namespace(str(Constants.namespaces[nm.split(":", 1)[0]]))
+                    graph_orig.bind(nm.split(":", 1)[0], temp_nm)
+
+                    if (metadata_key == "dcat:creator") or (
+                        metadata_key == "prov:Location"
+                    ):
+                        # now add this metadata to the collection_uuid
+                        graph_orig.add(
+                            (
+                                NIIRI.collection_uuid,
+                                temp_nm.term(term),
+                                URIRef(
+                                    proj_metadata["projects"][proj_uuid][metadata_key]
+                                ),
+                            )
+                        )
+                    else:
+                        # now add this metadata to the collection_uuid
+                        graph_orig.add(
+                            (
+                                NIIRI.collection_uuid,
+                                temp_nm.term(term),
+                                Literal(
+                                    proj_metadata["projects"][proj_uuid][metadata_key]
+                                ),
+                            )
+                        )
+
+            # get acquisition objects
+            acq_objects = getProjectAcquisitionObjects(
+                [nidm_file], project_id=str(proj_uuid.split(":", 1)[1])
+            )
+
+            for acq_obj in acq_objects.keys():
+                print(acq_obj)
+
+        # graph.serialize(out_file, format="turtle")
+
+
+if __name__ == "__main__":
+    update()