Added couple tests to see how the unittest package works.

cp2kparser/engines/cp2kinputengine.py

+#! /usr/bin/env python
+# -*- coding: utf-8 -*-
 from collections import defaultdict
 from cp2kparser.generics.util import *
 
@@ -54,6 +56,19 @@ class CP2KInputEngine(object):
     def get_subsection(self, path, index=0):
         return self.root_section.get_subsection(path, index)
 
+    def get_keyword(self, path, index=0):
+        split = path.rsplit('/', 1)
+        section_path = split[0]
+        keyword = split[1]
+        section = self.root_section.get_subsection(section_path, index)
+        if section is not None:
+            return section.get_keyword(keyword)
+
+    def get_parameter(self, path, index=0):
+        section = self.root_section.get_subsection(path, index)
+        if section is not None:
+            return section.get_parameter()
+
 
 #===============================================================================
 class InputSection(object):
@@ -140,4 +155,6 @@ class InputSection(object):
         return result
 
     def get_parameter(self):
+        if self.params is None:
+            print_debug("The section '{}' has no parameters set".format(self.name))
         return self.params

cp2kparser/generics/nomadparser.py

@@ -5,6 +5,7 @@ import os
 import time
 from abc import ABCMeta, abstractmethod
 from cp2kparser.generics.util import *
+from pint import UnitRegistry
 
 
 #===============================================================================
@@ -130,6 +131,8 @@ class NomadParser(object):
         else:
             print_debug("Using cached result.")
 
+        # Do the conversion to SI units based on the given units
+
         stop = time.clock()
         print_debug("Elapsed time: {} ms".format((stop-start)*1000))
         return result
@@ -149,7 +152,10 @@ class NomadParser(object):
 
     @abstractmethod
     def parse_quantity(self, name):
-        """Parse a quantity from the given files."""
+        """Parse a quantity from the given files. Should return a tuple
+        containing the result object (numeric results preferably as numpy
+        arrays) and the unit of the result (None if no unit is needed)
+        """
         pass
 
     @abstractmethod

cp2kparser/implementation/extract.py → cp2kparser/implementation/autoparser.py

-import logging
+#! /usr/bin/env python
+# -*- coding: utf-8 -*-
 import os
 import json
 from cp2kparser.implementation.parser import CP2KParser
@@ -18,7 +19,7 @@ def scan_path_for_files(path):
         extension = os.path.splitext(filename)[1]
         if extension in extensions:
             file_object = {
-                "path": filename,
+                "path": os.path.join(path, filename),
                 "file_id": "",
             }
             files.append(file_object)
@@ -26,9 +27,8 @@ def scan_path_for_files(path):
 
 
 #===============================================================================
-def extract(path):
+def get_parser(path):
     files = scan_path_for_files(path)
-    logging.basicConfig(format='%(message)s', level=logging.DEBUG)
     json_input = {
         "version": "nomadparsein.json 1.0",
         "tmpDir": "/home/lauri",
@@ -37,8 +37,4 @@ def extract(path):
         "files": files
     }
     parser = CP2KParser(json.dumps(json_input))
-    print parser.get_quantity("energy_total")
-    print parser.get_quantity("XC_functional")
-    print parser.get_quantity("particle_forces")
-    # n = len(parser.get_quantity("particle_forces"))
-    # print "Number of force configurations found: {}".format(n)
+    return parser

cp2kparser/implementation/parser.py

@@ -15,6 +15,10 @@ from cp2kparser.engines.cp2kinputengine import CP2KInputEngine
 class CP2KParser(NomadParser):
     """The interface for a NoMaD CP2K parser. All parsing actions will go
     through this class.
+
+    The CP2K version 2.6.2 was used as a reference for this basic
+    implementation. For other versions there should be classes that extend from
+    this.
     """
 
     def __init__(self, input_json_string):
@@ -92,8 +96,8 @@ class CP2KParser(NomadParser):
 
         # Now check from input what the other files are called
         self.inputengine.parse_input()
-        force_path = self.inputengine.get_subsection("FORCE_EVAL/PRINT/FORCES").get_keyword("FILENAME")
-        if force_path and force_path != "__STD_OUT__":
+        force_path = self.inputengine.get_keyword("FORCE_EVAL/PRINT/FORCES/FILENAME")
+        if force_path is not None and force_path != "__STD_OUT__":
             force_path = os.path.basename(force_path) + "-1_0"
 
         # Check against the given files
@@ -157,23 +161,68 @@ class CP2KImplementation(object):
         return self.regexengine.parse(self.regexs.energy_total, self.parser.get_file_handle("output"))
 
     def XC_functional(self):
-        """Returns the type of the XC functional based on the value of the
-        extractor xc_shortcut
+        """Returns the type of the XC functional.
+
+        Can currently only determine version if they are declared as parameters
+        for XC_FUNCTIONAL or via activating subsections of XC_FUNCTIONAL.
 
-                Returns:
-                        A string containing the final result that should
-                        belong to the list defined in NoMaD wiki.
+        Returns:
+            A string containing the final result that should
+            belong to the list defined in NoMaD wiki.
         """
-        # xc_shortcut = self.regexengine.parse(self.regexs.XC_functional, self.parser.get_file_handle("input"))
-        xc_shortcut = self.inputengine.get_subsection("FORCE_EVAL/DFT/XC/XC_FUNCTIONAL").get_parameter()
 
-        return {
-                'B3LYP': "HYB_GGA_XC_B3LYP",
-                'BEEFVDW': "",
-                'BLYP': "",
-                'PADE': "LDA_XC_TETER93",
-                'PBE': "GGA_X_PBE",
-        }.get(xc_shortcut, None)
+        # First try to look at the shortcut
+        xc_shortcut = self.inputengine.get_subsection("FORCE_EVAL/DFT/XC/XC_FUNCTIONAL").get_parameter()
+        if xc_shortcut is not None and xc_shortcut != "NONE" and xc_shortcut != "NO_SHORTCUT":
+            print_debug("Shortcut defined for XC_FUNCTIONAL")
+
+            # If PBE, check version
+            if xc_shortcut == "PBE":
+                pbe_version = self.inputengine.get_subsection("FORCE_EVAL/DFT/XC/XC_FUNCTIONAL/PBE").get_keyword("PARAMETRIZATION")
+                return {
+                        'ORIG': "GGA_X_PBE",
+                        'PBESOL': "GGA_X_PBE_SOL",
+                        'REVPBE': "GGA_X_PBE_R",
+                }.get(pbe_version, "GGA_X_PBE")
+
+            return {
+                    'B3LYP': "HYB_GGA_XC_B3LYP",
+                    'BEEFVDW': None,
+                    'BLYP': "GGA_C_LYP_GGA_X_B88",
+                    'BP': None,
+                    'HCTH120': None,
+                    'OLYP': None,
+                    'LDA': "LDA_XC_TETER93",
+                    'PADE': "LDA_XC_TETER93",
+                    'PBE0': None,
+                    'TPSS': None,
+            }.get(xc_shortcut, None)
+        else:
+            print_debug("No shortcut defined for XC_FUNCTIONAL. Looking into subsections.")
+
+        # Look at the subsections and determine what part have been activated
+
+        # Becke88
+        xc_components = []
+        becke_88 = self.inputengine.get_subsection("FORCE_EVAL/DFT/XC/XC_FUNCTIONAL/BECKE88").get_parameter()
+        if becke_88 == "TRUE":
+            xc_components.append("GGA_X_B88")
+
+        # Becke 97
+        becke_97 = self.inputengine.get_parameter("FORCE_EVAL/DFT/XC/XC_FUNCTIONAL/BECKE97")
+        if becke_97 == "TRUE":
+            becke_97_param = self.inputengine.get_keyword("FORCE_EVAL/DFT/XC/XC_FUNCTIONAL/BECKE97/PARAMETRIZATION")
+            becke_97_result = {
+                    'B97GRIMME': None,
+                    'B97_GRIMME': None,
+                    'ORIG': "GGA_XC_B97",
+                    'WB97X-V': None,
+            }.get(becke_97_param, None)
+            if becke_97_result is not None:
+                xc_components.append(becke_97_result)
+
+        # Return an alphabetically sorted and joined list of the xc components
+        return "_".join(sorted(xc_components))
 
     def particle_forces(self):
         """Return all the forces for every step found.
@@ -199,6 +248,6 @@ class CP2KImplementation(object):
 
 
 #===============================================================================
-class CP2K_240_Implementation(CP2KImplementation):
+class CP2K_262_Implementation(CP2KImplementation):
     def __init__(self, parser):
         CP2KImplementation.__init__(self, parser)

cp2kparser/tests/2.6.2/functionals/becke88/becke88.inp

+&GLOBAL
+  RUN_TYPE ENERGY_FORCE
+  PROJECT_NAME si_bulk
+  PRINT_LEVEL LOW
+&END GLOBAL
+&FORCE_EVAL
+  METHOD Quickstep
+  &DFT
+    POTENTIAL_FILE_NAME /home/lauri/GTH_POTENTIALS
+    BASIS_SET_FILE_NAME /home/lauri/BASIS_SET
+    &QS
+      EPS_DEFAULT 1e-3
+    &END QS
+    &SCF
+      MAX_SCF 100
+      EPS_SCF 1e-03
+      SCF_GUESS ATOMIC
+      &PRINT
+        &RESTART OFF
+          BACKUP_COPIES 0
+        &END RESTART
+      &END PRINT
+    &END SCF
+    &XC
+      &XC_FUNCTIONAL
+        &BECKE88 TRUE
+        &END BECKE88
+      &END XC_FUNCTIONAL
+    &END XC
+    &MGRID
+      CUTOFF 50
+      NGRIDS 2
+      REL_CUTOFF 25
+    &END MGRID
+  &END DFT
+  &SUBSYS
+    &COORD
+      Si 0.0 0.0 0.0
+    &END COORD
+    &CELL
+      A 5.4306975 0.0 0.0
+      C 0.0 0.0 5.4306975
+      B 0.0 5.4306975 0.0
+      PERIODIC XYZ
+    &END CELL
+    &KIND Si
+      POTENTIAL GTH-PADE-q4
+      BASIS_SET DZVP-GTH-PADE
+    &END KIND
+  &END SUBSYS
+&END FORCE_EVAL

cp2kparser/tests/2.6.2/functionals/blyp/blyp.inp

+&GLOBAL
+  RUN_TYPE ENERGY_FORCE
+  PROJECT_NAME si_bulk
+  PRINT_LEVEL LOW
+&END GLOBAL
+&FORCE_EVAL
+  METHOD Quickstep
+  &DFT
+    POTENTIAL_FILE_NAME /home/lauri/GTH_POTENTIALS
+    BASIS_SET_FILE_NAME /home/lauri/BASIS_SET
+    &QS
+      EPS_DEFAULT 1e-3
+    &END QS
+    &SCF
+      MAX_SCF 100
+      EPS_SCF 1e-03
+      SCF_GUESS ATOMIC
+      &PRINT
+        &RESTART OFF
+          BACKUP_COPIES 0
+        &END RESTART
+      &END PRINT
+    &END SCF
+    &XC
+      &XC_FUNCTIONAL BLYP
+      &END XC_FUNCTIONAL
+    &END XC
+    &MGRID
+      CUTOFF 50
+      NGRIDS 2
+      REL_CUTOFF 25
+    &END MGRID
+  &END DFT
+  &SUBSYS
+    &COORD
+      Si 0.0 0.0 0.0
+    &END COORD
+    &CELL
+      A 5.4306975 0.0 0.0
+      C 0.0 0.0 5.4306975
+      B 0.0 5.4306975 0.0
+      PERIODIC XYZ
+    &END CELL
+    &KIND Si
+      POTENTIAL GTH-PADE-q4
+      BASIS_SET DZVP-GTH-PADE
+    &END KIND
+  &END SUBSYS
+&END FORCE_EVAL

cp2kparser/tests/2.6.2/functionals/lda/lda.inp

+&GLOBAL
+  RUN_TYPE ENERGY_FORCE
+  PROJECT_NAME si_bulk
+  PRINT_LEVEL LOW
+&END GLOBAL
+&FORCE_EVAL
+  METHOD Quickstep
+  &DFT
+    POTENTIAL_FILE_NAME /home/lauri/GTH_POTENTIALS
+    BASIS_SET_FILE_NAME /home/lauri/BASIS_SET
+    &QS
+      EPS_DEFAULT 1e-3
+    &END QS
+    &SCF
+      MAX_SCF 100
+      EPS_SCF 1e-03
+      SCF_GUESS ATOMIC
+      &PRINT
+        &RESTART OFF
+          BACKUP_COPIES 0
+        &END RESTART
+      &END PRINT
+    &END SCF
+    &XC
+      &XC_FUNCTIONAL LDA
+      &END XC_FUNCTIONAL
+    &END XC
+    &MGRID
+      CUTOFF 50
+      NGRIDS 2
+      REL_CUTOFF 25
+    &END MGRID
+  &END DFT
+  &SUBSYS
+    &COORD
+      Si 0.0 0.0 0.0
+    &END COORD
+    &CELL
+      A 5.4306975 0.0 0.0
+      C 0.0 0.0 5.4306975
+      B 0.0 5.4306975 0.0
+      PERIODIC XYZ
+    &END CELL
+    &KIND Si
+      POTENTIAL GTH-PADE-q4
+      BASIS_SET DZVP-GTH-PADE
+    &END KIND
+  &END SUBSYS
+&END FORCE_EVAL

cp2kparser/tests/test1/si_bulk.inp → cp2kparser/tests/2.6.2/functionals/pade/pade.inp

@@ -9,46 +9,31 @@
     POTENTIAL_FILE_NAME /home/lauri/GTH_POTENTIALS
     BASIS_SET_FILE_NAME /home/lauri/BASIS_SET
     &QS
-      EPS_DEFAULT 1e-10
+      EPS_DEFAULT 1e-3
     &END QS
     &SCF
-      MAX_SCF 300
-      EPS_SCF 1e-07
+      MAX_SCF 100
+      EPS_SCF 1e-03
       SCF_GUESS ATOMIC
-      &DIAGONALIZATION ON
-        ALGORITHM STANDARD
-      &END DIAGONALIZATION
-      &MIXING TRUE
-        NBUFFER 8
-        ALPHA 0.4
-        METHOD BROYDEN_MIXING
-      &END MIXING
+      &PRINT
+        &RESTART OFF
+          BACKUP_COPIES 0
+        &END RESTART
+      &END PRINT
     &END SCF
     &XC
       &XC_FUNCTIONAL PADE
       &END XC_FUNCTIONAL
     &END XC
     &MGRID
-      CUTOFF 300
-      NGRIDS 4
-      REL_CUTOFF 60
+      CUTOFF 50
+      NGRIDS 2
+      REL_CUTOFF 25
     &END MGRID
   &END DFT
-  &PRINT
-    &FORCES ON
-      FILENAME /home/lauri/forces
-    &END FORCES
-  &END PRINT
   &SUBSYS
     &COORD
       Si 0.0 0.0 0.0
-      Si 1.357674375 1.357674375 1.357674375
-      Si 2.71534875 2.71534875 0.0
-      Si 4.073023125 4.073023125 1.357674375
-      Si 2.71534875 0.0 2.71534875
-      Si 4.073023125 1.357674375 4.073023125
-      Si 0.0 2.71534875 2.71534875
-      Si 1.357674375 4.073023125 4.073023125
     &END COORD
     &CELL
       A 5.4306975 0.0 0.0

cp2kparser/tests/2.6.2/run_tests.py

+import unittest
+import os
+from cp2kparser.implementation.autoparser import get_parser
+from cp2kparser.generics.util import *
+# import logging
+# logging.basicConfig(level=logging.DEBUG)
+
+
+#===============================================================================
+class TestFunctionals(unittest.TestCase):
+
+    def getxc(self, folder, result):
+        path = os.path.join(os.path.dirname(os.path.realpath(__file__)), "functionals", folder)
+        parser = get_parser(path)
+        xc = parser.get_quantity("XC_functional")
+        self.assertEqual(xc, result)
+
+    def test_pade(self):
+        self.getxc("pade", "LDA_XC_TETER93")
+
+    def test_lda(self):
+        self.getxc("lda", "LDA_XC_TETER93")
+
+    def test_becke88(self):
+        self.getxc("becke88", "GGA_X_B88")
+
+    def test_blyp(self):
+        self.getxc("blyp", "GGA_C_LYP_GGA_X_B88")
+
+if __name__ == '__main__':
+    unittest.main()

cp2kparser/tests/test1/run.py

-from cp2kparser.implementation.extract import extract
-import os
-
-if __name__ == "__main__":
-    path = os.path.dirname(os.path.realpath("__file__"))
-    extract(path)