Merge branch 'master' into patch-2

.github/scripts/check_facility_has_institution_id/main.py

@@ -0,0 +1,64 @@
+# Run through all the FACILITY.yaml files and check that they have an institution_id field.
+
+import yaml
+import sys
+import glob
+import requests
+
+export_mapping = False
+
+# Get the list of valid institution ids
+response = requests.get("https://topology-institutions.osg-htc.org/api/institution_ids")
+topology_institutions = response.json()
+topology_institutions_by_id = {x['id']: x for x in topology_institutions}
+topology_institution_ids = {x['id'] for x in topology_institutions}
+
+def check_facility_institution_id(yaml_string: str):
+
+    facility = yaml.load(yaml_string, Loader=yaml.Loader)
+    if 'InstitutionID' not in facility:
+        raise Exception("FACILITY.yaml does not have an InstitutionID field")
+
+    if facility['InstitutionID'] not in topology_institution_ids and facility['InstitutionID'] is not None:
+        raise Exception(f"Invalid InstitutionID: {facility['InstitutionID']}")
+
+def provide_human_check_interface(facility_files: list):
+
+    facility_institution_mapping = {}
+    for file in facility_files:
+        with open(file, 'r') as f:
+            facility_name = file.split('/')[-2]
+            facility = yaml.load(f, Loader=yaml.Loader)
+            facility_institution_mapping[facility_name] = topology_institutions_by_id.get(facility['InstitutionID'], {}).get('name', None)
+
+    if export_mapping:
+        with open("facility_institution_mapping.yaml", 'w') as f:
+            yaml.dump(facility_institution_mapping, f)
+
+    else:
+        print(facility_institution_mapping)
+
+def main():
+
+    facility_files = glob.glob("../../../topology/**/FACILITY.yaml")
+
+    # Check the files
+    errors = []
+    for file in facility_files:
+        with open(file, 'r') as f:
+            try:
+                check_facility_institution_id(f)
+            except Exception as e:
+                errors.append((file.split("/")[-2], e))
+
+    # Print the errors and exit if needed
+    if errors:
+        for error in errors:
+            print(f"Error in {error[0]}: \n\t {error[1]}")
+        sys.exit(1)
+
+    provide_human_check_interface(facility_files)
+
+
+if __name__ == "__main__":
+    main()

.github/scripts/check_facility_has_institution_id/requirements.txt

@@ -0,0 +1,2 @@
+PyYAML==6.0.1
+requests==2.32.0

.github/workflows/check_facility_institution_id.yml

@@ -0,0 +1,20 @@
+name: Check Project FOS Precision
+on:
+  pull_request:
+    branches:
+      - main
+
+jobs:
+  check:
+    name: Check
+    runs-on: ubuntu-latest
+    if: startsWith(github.repository, 'opensciencegrid/')
+    steps:
+      - uses: actions/checkout@v3
+      - name: Set up Python
+        uses: actions/setup-python@v4
+        with:
+          python-version: 3.9.15
+          cache: 'pip' # caching pip dependencies
+      - run: pip install -r ./.github/scripts/check_facility_has_institution_id/requirements.txt
+      - run: python ./.github/scripts/check_facility_has_institution_id/main.py

.github/workflows/check_project_fos_precision.yml

@@ -2,7 +2,7 @@ name: Check Project FOS Precision
 on:
   pull_request:
     branches:
-      - main
+      - master
   schedule:
     - cron: '0 0 * * *'
 
@@ -19,4 +19,4 @@ jobs:
           python-version: 3.9.15
           cache: 'pip' # caching pip dependencies
       - run: pip install -r ./.github/scripts/check_project_fos_precision/requirements.txt
-      - run: python ./.github/scripts/check_project_fos_precision/main.py
+      - run: python ./.github/scripts/check_project_fos_precision/main.py

projects/CSU_Vogeler.yaml

@@ -0,0 +1,7 @@
+Department: Natural Resource Ecology Lab
+Description: Distributed geospatial data access and processing
+FieldOfScience: Ecological and Environmental Sciences
+FieldOfScienceID: '03.0101'
+InstitutionID: https://osg-htc.org/iid/2aj5pa9etoc7
+Organization: Colorado State University
+PIName: Jody Vogeler

projects/Illinois_Park.yaml

@@ -0,0 +1,12 @@
+Department: Statistics
+Description: 'My research broadly focuses on (a) causal inference under interference
+  and non-i.i.d. settings, (b) causal inference under unmeasured confounding, and
+  (c) optimal treatment regimes and policy learning. A common theme in my research
+  is to use non/semiparametric theory and optimization methods to develop efficient
+  and robust estimators of causal quantities in (a)-(c).
+  Website: https://www.chanpark.net/'
+FieldOfScience: Statistics
+FieldOfScienceID: '27.0503'
+InstitutionID: https://osg-htc.org/iid/10izzs5e7v1r
+Organization: University of Illinois Urbana-Champaign
+PIName: Chan Park

projects/SLU_Ahn.yaml

@@ -0,0 +1,7 @@
+Department: Computer Science
+Description: Studying multimodal ML applications in biomedical science.
+FieldOfScience: Computer Science
+FieldOfScienceID: '11.0701'
+InstitutionID: https://osg-htc.org/iid/p5mk8p3jv6wa
+Organization: Saint Louis University
+PIName: Tae-Hyuk Ahn

projects/UWMadison_Travers.yaml

@@ -0,0 +1,8 @@
+Department: Waisman Center
+Description: Study of whole brain white matter development in autistic and non-autistic
+  children
+FieldOfScience: Neuroscience
+FieldOfScienceID: '26.1501'
+InstitutionID: https://osg-htc.org/iid/fq8thqsj99zh
+Organization: University of Wisconsin-Madison
+PIName: Brittany Travers

src/schema/rgsummary.xsd

@@ -30,6 +30,7 @@
                 <xsd:complexType>
                   <xsd:sequence>
                     <xsd:element name="ID" type="xsd:int" />
+                    <xsd:element name="InstitutionID" type="xsd:string" />
                     <xsd:element name="Name" type="xsd:string" />
                     <xsd:element name="IsCCStar" type="boolean" minOccurs="0" />
                   </xsd:sequence>