data file validator: allow data to be passed (needed by hepdata-conve…

…rter)

Signed-off-by: Graeme Watt <[email protected]>

README.rst

@@ -64,7 +64,10 @@ To validate files, you need to instantiate a validator (I love OO).
     # if there are any error messages, they are retrievable through this call
     submission_file_validator.get_messages()
 
-    
+    # the error messages can be printed
+    submission_file_validator.print_errors(submission_file_path)
+
+
 Data file validation is exactly the same.
 
 .. code:: python
@@ -79,6 +82,9 @@ Data file validation is exactly the same.
     # if there are any error messages, they are retrievable through this call
     data_file_validator.get_messages()
 
+    # the error messages can be printed
+    data_file_validator.print_errors('data.yaml')
+
 
 Optionally, if you have already loaded the YAML object, then you can pass it through
 as a data object. You must also pass through the file_path since this is used as a key

hepdata_validator/data_file_validator.py

@@ -66,41 +66,62 @@ def load_custom_schema(self, type, schema_file_path=None):
             raise UnsupportedDataSchemaException(
                 message="There is no schema defined for the '{0}' data type.".format(type))
 
-    def validate(self, file_path):
-        try:
-            default_data_schema = json.load(
-                open(self.default_schema_file, 'r'))
+    def validate(self, **kwargs):
+        """
+        Validates a data file
+
+        :param file_path: path to file to be loaded.
+        :param data: pre loaded YAML object (optional).
+        :return: Bool to indicate the validity of the file.
+        """
+
+        default_data_schema = json.load(open(self.default_schema_file, 'r'))
+
+        # even though we are using the yaml package to load,
+        # it supports JSON and YAML
+        data = kwargs.pop("data", None)
+        file_path = kwargs.pop("file_path", None)
+
+        if file_path is None:
+            raise LookupError("file_path argument must be supplied")
+
+        if data is None:
 
             try:
-                data = yaml.load_all(open(file_path, 'r'), Loader=yaml.CLoader)
+                # We try to load using the CLoader for speed improvements.
+                try:
+                    data = yaml.load(open(file_path, 'r'), Loader=yaml.CLoader)
+                except ScannerError as se:
+                    self.add_validation_message(ValidationMessage(file=file_path, message=
+                    'There was a problem parsing the file.\n' + str(se)))
+                    return False
             except: #pragma: no cover
-                data = yaml.load_all(open(file_path, 'r')) #pragma: no cover
+                try:  # pragma: no cover
+                    data = yaml.load(open(file_path, 'r'))  # pragma: no cover
+                except ScannerError as se:  # pragma: no cover
+                    self.add_validation_message(
+                        ValidationMessage(file=file_path, message=
+                    'There was a problem parsing the file.\n' + str(se))) # pragma: no cover
+                    return False
 
-            for data_item in data:
-                if data_item is None:
-                    continue
-                try:
-                    if 'type' in data_item:
-                        custom_schema = self.load_custom_schema(data_item['type'])
-                        json_validate(data_item, custom_schema)
-                    else:
-                        json_validate(data_item, default_data_schema)
+        try:
 
-                except ValidationError as ve:
-                    self.add_validation_message(
-                        ValidationMessage(file=file_path,
-                                          message=ve.message + ' in ' + str(ve.instance)))
-            if self.has_errors(file_path):
-                return False
+            if 'type' in data:
+                custom_schema = self.load_custom_schema(data['type'])
+                json_validate(data, custom_schema)
             else:
-                return True
-        except ScannerError as se:
+                json_validate(data, default_data_schema)
+
+        except ValidationError as ve:
+
             self.add_validation_message(
                 ValidationMessage(file=file_path,
-                                  message='There was a problem parsing the file. '
-                                          'This can be because you forgot spaces '
-                                          'after colons in your YAML file for instance.\n{0}'.format(se.__repr__()))
-            )
+                                    message=ve.message + ' in ' + str(ve.instance)))
+
+        if self.has_errors(file_path):
+            return False
+        else:
+            return True
 
 
 class UnsupportedDataSchemaException(Exception):

hepdata_validator/submission_file_validator.py

@@ -14,21 +14,20 @@ class SubmissionFileValidator(Validator):
     """
     base_path = os.path.dirname(__file__)
     default_schema_file = base_path + '/schemas/submission_schema.json'
-    additonal_info_schema = base_path + '/schemas/additional_info_schema.json'
+    additional_info_schema = base_path + '/schemas/additional_info_schema.json'
 
     def validate(self, **kwargs):
         """
         Validates a submission file
+
         :param file_path: path to file to be loaded.
         :param data: pre loaded YAML object (optional).
         :return: Bool to indicate the validity of the file.
         """
         try:
-            submission_file_schema = json.load(
-                    open(self.default_schema_file, 'r'))
+            submission_file_schema = json.load(open(self.default_schema_file, 'r'))
 
-            additional_file_section_schema = json.load(
-                    open(self.additonal_info_schema, 'r'))
+            additional_file_section_schema = json.load(open(self.additional_info_schema, 'r'))
 
             # even though we are using the yaml package to load,
             # it supports JSON and YAML
@@ -57,18 +56,18 @@ def validate(self, **kwargs):
                 except ValidationError as ve:
                     self.add_validation_message(
                             ValidationMessage(file=file_path,
-                                              message=ve.message + ' in ' + str(ve.instance)))
+                                                message=ve.message + ' in ' + str(ve.instance)))
 
             if self.has_errors(file_path):
                 return False
             else:
                 return True
-        except ScannerError as se:
-            self.add_validation_message(
-                    ValidationMessage(file=file_path,
-                                      message='There was a problem parsing the file.  '
-                                              'This can be because you forgot spaces '
-                                              'after colons in your YAML file for instance.  '
-                                              'Diagnostic information follows.\n' + str(se))
-            )
-            return False
+
+        except ScannerError as se:  # pragma: no cover
+            self.add_validation_message(  # pragma: no cover
+                ValidationMessage(file=file_path, message=
+                    'There was a problem parsing the file.  '
+                    'This can be because you forgot spaces '
+                    'after colons in your YAML file for instance.  '
+                    'Diagnostic information follows.\n' + str(se)))
+            return False

hepdata_validator/version.py

@@ -27,4 +27,4 @@
 
 from __future__ import absolute_import, print_function
 
-__version__ = "0.1.13"
+__version__ = "0.1.14"

testsuite/test_data/invalid_data_file.yaml

@@ -3,6 +3,4 @@ type: "different"
 some_variables:
   - header:{name: SQRT(S), units: GEV}
     values:
-      - value: 7000
-
----
+      - value: 7000

testsuite/test_data/valid_file_custom.yaml

@@ -3,6 +3,4 @@ type: "different"
 some_variables:
   - header: {name: SQRT(S), units: GEV}
     values:
-      - value: 7000
-
----
+      - value: 7000