bugfix default value in convention

CHANGELOG.md

@@ -2,6 +2,9 @@
 
 Log of changes in the versions
 
+## v1.1.1
+- bugfix: Setting a default value for toolbox validators in convention yaml file was not working. Fixed it.
+
 ## v1.1.0
 - simplified and clean up much code, especially convention sub package
 - added identifier utils

h5rdmtoolbox/convention/core.py

@@ -4,7 +4,6 @@
 import h5py
 import inspect
 import pathlib
-import pydantic
 import re
 import shutil
 import sys
@@ -13,9 +12,9 @@
 from pydoc import locate
 from typing import Union, List, Dict, Tuple
 
+from h5rdmtoolbox import errors
 from h5rdmtoolbox.repository import RepositoryInterface
 from h5rdmtoolbox.wrapper import ds_decoder
-from h5rdmtoolbox import errors
 from . import cfg
 from . import consts
 from . import errors
@@ -399,7 +398,8 @@ def register(self):
         add_convention(self)
 
     def validate(self, file_or_filename: Union[str, pathlib.Path, "File"]) -> List[Dict]:
-        """Checks a file for compliance with the convention
+        """Checks a file for compliance with the convention. It will NOT raise an error but
+        return a list of invalid attributes.
 
         Parameters
         ----------
@@ -469,6 +469,7 @@ def _validate_convention(name, node):
 
         with File(file_or_filename, 'r') as f:
             logger.debug(f'Checking file {file_or_filename} for compliance with convention {self.name}')
+            _validate_convention('/', f)
             f.visititems(_validate_convention)
 
         return failed

h5rdmtoolbox/convention/generate.py

@@ -127,26 +127,41 @@ def write_convention_module_from_yaml(yaml_filename: pathlib.Path, name=None):
         for k, v in class_definitions.items():
             f.write(f'\nclass {k[1:]}(BaseModel):')
             description = v.get('description', k[1:])
-            f.write(f'\n{INDENT}"""{description}"""\n{INDENT}')
+            f.write(f'\n{INDENT}"""{description}"""\n')
             validator = {}
+            _validator_default = {}
             for kk, vv in v.items():
-                _vv = vv.strip('$')
-                if _vv in toolbox_validators.validators:
+                # _vv = vv.strip('$')
+                splitted_validator = vv.strip('$').split('=', 1)
+                validator_name = splitted_validator[0].strip()
+                if len(splitted_validator) == 2:
+                    default_value = splitted_validator[1].strip()
+                else:
+                    default_value = None
+                _validator_default[kk] = [validator_name, default_value]
+
+                if validator_name in toolbox_validators.validators:
 
                     try:
-                        if toolbox_validators.validators[_vv].__name__ != 'Annotated':  # in py3.8 this will raise an error
-                            validator[kk] = f'toolbox_validators.{toolbox_validators.validators[_vv].__name__}'
-                            used_toolbox_validators[_vv] = f'toolbox_validators.{toolbox_validators.validators[_vv].__name__}'
+                        if toolbox_validators.validators[validator_name].__name__ != 'Annotated':  # in py3.8 this will raise an error
+                            validator[kk] = f'toolbox_validators.{toolbox_validators.validators[validator_name].__name__}'
+                            used_toolbox_validators[validator_name] = f'toolbox_validators.{toolbox_validators.validators[validator_name].__name__}'
                         else:
-                            validator[kk] = f'toolbox_validators.validators["{_vv}"]'
-                            used_toolbox_validators[_vv] = f'toolbox_validators.validators["{_vv}"]'
+                            validator[kk] = f'toolbox_validators.validators["{validator_name}"]'
+                            used_toolbox_validators[validator_name] = f'toolbox_validators.validators["{validator_name}"]'
 
                     except AttributeError:
-                        validator[kk] = f'toolbox_validators.validators["{_vv}"]'
-                        used_toolbox_validators[_vv] = f'toolbox_validators.validators["{_vv}"]'
+                        validator[kk] = f'toolbox_validators.validators["{validator_name}"]'
+                        used_toolbox_validators[validator_name] = f'toolbox_validators.validators["{validator_name}"]'
                 else:
                     validator[kk] = vv
-            f.write(f'\n{INDENT}'.join([f'{ak}: {av}' for ak, av in validator.items()]))
+            for ak, av in validator.items():
+                validator_value, validator_default = _validator_default[ak]
+                if validator_default is not None:
+                    f.write(f'{INDENT}{ak}: {used_toolbox_validators.get(validator_name, validator_value)} = {validator_default}\n')
+                else:
+                    f.write(f'{INDENT}{ak}: {av}\n')
+            # f.write(f'\n{INDENT}'.join([f'{ak}: {av}' for ak, av in validator.items()]))
             f.write('\n\n')
 
         # write standard attribute classes:
@@ -185,7 +200,7 @@ def write_convention_module_from_yaml(yaml_filename: pathlib.Path, name=None):
 
     # with open(py_filename, 'a') as f:
         f.write('\nvalidator_dict = {\n' + INDENT)
-        f.write(f'\n{INDENT}'.join(f"'{k}': {k[1:]}," for k, v in class_definitions.items()))
+        f.write(f'\n{INDENT}'.join(f"\n'{k}': {k[1:]}," for k, v in class_definitions.items()))
         f.write(f'\n{INDENT}'.join(f"'{k}': {v}," for k, v in used_toolbox_validators.items()))
         # f.write(f"\n{INDENT}'$int': IntValidator,  # see h5rdmtoolbox.convention.toolbox_validators")
         # f.write(f"\n{INDENT}'$str': StringValidator,  # see h5rdmtoolbox.convention.toolbox_validators")

tests/conventions/simple_cv.yaml

@@ -1,6 +1,17 @@
 __contact__: https://orcid.org/0000-0001-8729-0482
 __institution__: https://orcid.org/members/001G000001e5aUTIAY
 __name__: simple_cv
+
+$personOrOrganization:
+  name: str
+  orcid: $orcid=None
+
+creator:
+  validator: $personOrOrganization
+  description: Name and affiliation of the data creator.
+  target_method: __init__
+  default_value: $NONE
+
 units:
   default_value: $EMPTY
   description: The physical unit of the dataset. If dimensionless, the unit is ''.

tests/conventions/test_conventions.py

@@ -640,19 +640,38 @@ def test_engmeta_example(self):
                 pass
 
     def test_read_invalid_attribute(self):
-        cv = h5tbx.convention.Convention.from_yaml(__this_dir__ / 'simple_cv.yaml')
-        h5tbx.use(None)
-        with h5tbx.File() as h5:
-            h5.create_dataset('ds', data=[1, 2], attrs=dict(units='invalid'))
-        with h5tbx.use(cv):
-            with h5tbx.File(h5.hdf_filename) as h5:
-                with h5tbx.set_config(ignore_get_std_attr_err=True):
-                    with self.assertWarns(h5tbx.warnings.StandardAttributeValidationWarning):
-                        units = h5.ds.units
-                self.assertEqual(units, 'invalid')
-        with h5tbx.use(None):
-            with h5tbx.File(h5.hdf_filename) as h5:
+        cv = h5tbx.convention.Convention.from_yaml(__this_dir__ / 'simple_cv.yaml', overwrite=True)
+        # h5tbx.use(None)
+        # with h5tbx.File() as h5:
+        #     h5.create_dataset('ds', data=[1, 2], attrs=dict(units='invalid'))
+        # with h5tbx.use(cv):
+        #     with h5tbx.File(h5.hdf_filename) as h5:
+        #         with h5tbx.set_config(ignore_get_std_attr_err=True):
+        #             with self.assertWarns(h5tbx.warnings.StandardAttributeValidationWarning):
+        #                 units = h5.ds.units
+        #         self.assertEqual(units, 'invalid')
+        # with h5tbx.use(None):
+        #     with h5tbx.File(h5.hdf_filename) as h5:
+        #         with self.assertRaises(AttributeError):
+        #             _ = h5.ds.units
+        #         units = h5.ds.attrs['units']
+        #         self.assertEqual(units, 'invalid')
+
+        with h5tbx.use('simple_cv'):
+            with h5tbx.File(creator={'name': 'Joe'}) as h5:
+                self.assertEqual(h5.creator.name, 'Joe')
+                self.assertEqual(h5.creator.orcid, None)
+            with h5tbx.File(creator={'name': 'Joe',
+                                     'invalid': '123'}) as h5:
+                self.assertEqual(h5.creator.name, 'Joe')
+                self.assertEqual(h5.creator.orcid, None)
                 with self.assertRaises(AttributeError):
-                    _ = h5.ds.units
-                units = h5.ds.attrs['units']
-                self.assertEqual(units, 'invalid')
+                    self.assertEqual(h5.creator.invalid, '123')
+            with self.assertRaises(h5tbx.errors.StandardAttributeError):
+                with h5tbx.File(creator={'name': 'Joe',
+                                         'orcid': '123'}) as h5:
+                    pass
+            with h5tbx.File(creator={'name': 'Joe',
+                                     'orcid': h5tbx.__author_orcid__}) as h5:
+                self.assertEqual(h5.creator.name, 'Joe')
+                self.assertEqual(str(h5.creator.orcid), h5tbx.__author_orcid__)