Add validator test

Adds invalid recid/doi testing and related test YAML files.
HEPData · Jul 6, 2023 · 6247c64 · 6247c64
1 parent 2f73905
commit 6247c64
Show file tree

Hide file tree

Showing 3 changed files with 139 additions and 0 deletions.
diff --git a/testsuite/test_data/invalid_submission_doi.yaml b/testsuite/test_data/invalid_submission_doi.yaml
@@ -0,0 +1,58 @@
+---
+additional_resources: # additional references (e.g. experiment TWiki page for analysis)
+  - {location: "http://atlas.web.cern.ch/Atlas/GROUPS/PHYSICS/PAPERS/STDM-2012-02/", description: "web page with auxiliary material"}
+
+comment: | # preserve newlines
+  CERN-LHC.  Measurements of the cross section  for ZZ production using the 4l and 2l2nu decay channels in proton-proton collisions at a centre-of-mass energy of 7 TeV with 4.6 fb^-1 of data collected in 2011.  The final states used are 4 electrons, 4 muons, 2 electrons and 2 muons, 2 electrons and missing transverse momentum, and 2 muons and missing transverse momentum (MET).
+
+  The cross section values reported in the tables should be multiplied by a factor of 1.0141 to take into account the updated value of the integrated luminosity for the ATLAS 2011 data taking period.  The uncertainty on the global normalisation ("Lumi") remains at 1.8%.  See Eur.Phys.J. C73 (2013) 2518 for more details.
+
+  The 4l channel fiducial region is defined as:
+  - 4e, 4mu or 2e2mu
+  - Ambiguities in pairing are resolved by choosing the combination that results in the smaller value of the sum |mll - mZ| for the two pairs, where mll is the mass of the dilepton system.
+  - ptLepton > 7 GeV (at least one with ptLepton > 20 (25) GeV for muons (electrons))
+  - |etaLepton| < 3.16
+  - At least one lepton pair is required to have invariant mass between 66 and 116 GeV. If the second pair also satisfies this, the event is ZZ, otherwise if the second pair satisfies mll > 20 GeV it is ZZ*.
+  - min(DeltaR(l,l)) > 0.2.
+
+  The 2l2nu channel fiducial region is defined as:
+  - 2e+MET or 2mu+MET
+  - ptLepton > 20 GeV
+  - |etaLepton| < 2.5
+  - mll must be between 76 and 106 GeV
+  - -MET*cos(phi_METZ)>75 GeV, where phi_METZ is the angle between the Z and the MET
+  - |MET - pTZ| / pTZ < 0.4, where pTZ is the transverse momentum of the dilepton system
+  - No events with a jet for which ptJet > 25 GeV and |etaJet| < 4.5
+  - No events with a third lepton for which ptLepton > 10 GeV
+  - min(DeltaR(l,l)) > 0.3
+---
+# Start a new YAML document to indicate a new data table.
+# This is Table 1.
+name: "Table 1"
+related_to_table_dois:
+ - "10.17182/hepdata.1.v1/t1"
+ - "10.17182/hepdaata.1.v1/t1"
+location: Data from Page 17 of preprint
+description: The measured fiducial cross sections.  The first systematic uncertainty is the combined systematic uncertainty excluding luminosity, the second is the luminosity
+keywords: # used for searching, possibly multiple values for each keyword
+  - {name: reactions, values: [P P --> Z0 Z0 X]}
+  - {name: observables, values: [SIG]}
+  - {name: cmenergies, values: [7000.0]} # centre-of-mass energy in GeV
+  - {name: phrases, values: [Inclusive, Integrated Cross Section, Cross Section, Proton-Proton Scattering, Z Production, Z pair Production]}
+data_file: data1.yaml
+data_license: # (optional) you can specify a license for the data
+  name: "GPL 2"
+  url: "url for license"
+  description: "Tell me about it. This can appear in the main record display" # (optional)
+additional_resources: # (optional)
+  - location: "analysis_script.py"
+    description: "Analysis script"
+  - location: "http://github.com/HEPData/hepdata"
+    description: "Full source code for creating this data"
+  - location: "root_file.root"
+    description: "Some file"
+    license: # (optional)
+      name: 'GPL 2'
+      url: "url for license"
+      description: "Tell me about it. This can appear in the main record display" # (optional)
+---
diff --git a/testsuite/test_data/invalid_submission_recid.yaml b/testsuite/test_data/invalid_submission_recid.yaml
@@ -0,0 +1,59 @@
+---
+additional_resources: # additional references (e.g. experiment TWiki page for analysis)
+  - {location: "http://atlas.web.cern.ch/Atlas/GROUPS/PHYSICS/PAPERS/STDM-2012-02/", description: "web page with auxiliary material"}
+
+comment: | # preserve newlines
+  CERN-LHC.  Measurements of the cross section  for ZZ production using the 4l and 2l2nu decay channels in proton-proton collisions at a centre-of-mass energy of 7 TeV with 4.6 fb^-1 of data collected in 2011.  The final states used are 4 electrons, 4 muons, 2 electrons and 2 muons, 2 electrons and missing transverse momentum, and 2 muons and missing transverse momentum (MET).
+
+  The cross section values reported in the tables should be multiplied by a factor of 1.0141 to take into account the updated value of the integrated luminosity for the ATLAS 2011 data taking period.  The uncertainty on the global normalisation ("Lumi") remains at 1.8%.  See Eur.Phys.J. C73 (2013) 2518 for more details.
+
+  The 4l channel fiducial region is defined as:
+  - 4e, 4mu or 2e2mu
+  - Ambiguities in pairing are resolved by choosing the combination that results in the smaller value of the sum |mll - mZ| for the two pairs, where mll is the mass of the dilepton system.
+  - ptLepton > 7 GeV (at least one with ptLepton > 20 (25) GeV for muons (electrons))
+  - |etaLepton| < 3.16
+  - At least one lepton pair is required to have invariant mass between 66 and 116 GeV. If the second pair also satisfies this, the event is ZZ, otherwise if the second pair satisfies mll > 20 GeV it is ZZ*.
+  - min(DeltaR(l,l)) > 0.2.
+
+  The 2l2nu channel fiducial region is defined as:
+  - 2e+MET or 2mu+MET
+  - ptLepton > 20 GeV
+  - |etaLepton| < 2.5
+  - mll must be between 76 and 106 GeV
+  - -MET*cos(phi_METZ)>75 GeV, where phi_METZ is the angle between the Z and the MET
+  - |MET - pTZ| / pTZ < 0.4, where pTZ is the transverse momentum of the dilepton system
+  - No events with a jet for which ptJet > 25 GeV and |etaJet| < 4.5
+  - No events with a third lepton for which ptLepton > 10 GeV
+  - min(DeltaR(l,l)) > 0.3
+  
+related_to_hepdata_recids: 
+  # Insert some random non-integer into the list
+  - aaaaaa
+---
+# Start a new YAML document to indicate a new data table.
+# This is Table 1.
+name: "Table 1"
+location: Data from Page 17 of preprint
+description: The measured fiducial cross sections.  The first systematic uncertainty is the combined systematic uncertainty excluding luminosity, the second is the luminosity
+keywords: # used for searching, possibly multiple values for each keyword
+  - {name: reactions, values: [P P --> Z0 Z0 X]}
+  - {name: observables, values: [SIG]}
+  - {name: cmenergies, values: [7000.0]} # centre-of-mass energy in GeV
+  - {name: phrases, values: [Inclusive, Integrated Cross Section, Cross Section, Proton-Proton Scattering, Z Production, Z pair Production]}
+data_file: data1.yaml
+data_license: # (optional) you can specify a license for the data
+  name: "GPL 2"
+  url: "url for license"
+  description: "Tell me about it. This can appear in the main record display" # (optional)
+additional_resources: # (optional)
+  - location: "analysis_script.py"
+    description: "Analysis script"
+  - location: "http://github.com/HEPData/hepdata"
+    description: "Full source code for creating this data"
+  - location: "root_file.root"
+    description: "Some file"
+    license: # (optional)
+      name: 'GPL 2'
+      url: "url for license"
+      description: "Tell me about it. This can appear in the main record display" # (optional)
+---
diff --git a/testsuite/test_submission_validator.py b/testsuite/test_submission_validator.py
@@ -337,3 +337,25 @@ def test_submission_with_no_data_tables(validator_v0, validator_v1, data_path, c
 
         is_valid_v0 = validator_v0.validate(file_path=file, data=yaml_obj)
         assert is_valid_v0
+
+def test_related_submission_invalid(validator_v1, data_path, capsys):
+    """
+        Tests all of the invalid related submission files at once
+        against their expected errors.
+    """
+    yaml_files = [
+        {"file" : "invalid_submission_recid.yaml", 
+            "error" : "error - 'aaaaaa' is not of type 'integer'"},
+        {"file" : "invalid_submission_doi.yaml", 
+            "error" : "error - '10.17182/hepdaata.1.v1/t1' does not match "}]
+
+    for yf in yaml_files:
+        file = os.path.join(data_path, yf['file'])
+        with open(file, 'r') as submission:
+            yaml_obj = yaml.load_all(submission, Loader=YamlLoader)
+            is_valid = validator_v1.validate(file_path=file, data=yaml_obj)
+            validator_v1.print_errors(file)
+            assert is_valid is False
+
+            out, err = capsys.readouterr()
+            assert yf['error'] in out.strip()