Prevent users from creating columns with python keyword names

src/omero_metadata/cli.py

@@ -242,6 +242,9 @@ def _configure(self, parser):
         populate.add_argument("--allow_nan", action="store_true", help=(
             "Allow empty values to become Nan in Long or Double columns"))
 
+        populate.add_argument("--force_cols", action="store_true", help=(
+            "Force creation of column names which are forbidden by default"))
+
         populateroi.add_argument(
             "--measurement", type=int, default=None,
             help="Index of the measurement to populate. By default, all")
@@ -533,7 +536,8 @@ def populate(self, args):
                             cfg=args.cfg, cfgid=cfgid, attach=args.attach,
                             options=localcfg, batch_size=args.batch,
                             loops=loops, ms=ms, dry_run=args.dry_run,
-                            allow_nan=args.allow_nan)
+                            allow_nan=args.allow_nan,
+                            force_cols=args.force_cols)
         ctx.parse()
 
     def rois(self, args):

src/omero_metadata/populate.py

@@ -36,6 +36,7 @@
 import csv
 import re
 import json
+import keyword
 from getpass import getpass
 from getopt import getopt, GetoptError
 
@@ -178,11 +179,13 @@ class HeaderResolver(object):
         'plate': PlateColumn,
     }, **plate_keys)
 
-    def __init__(self, target_object, headers, column_types=None):
+    def __init__(self, target_object, headers, column_types=None,
+                 force_cols=False):
         self.target_object = target_object
         self.headers = headers
         self.headers_as_lower = [v.lower() for v in self.headers]
         self.types = column_types
+        self.force_cols = force_cols
 
     @staticmethod
     def is_row_column_types(row):
@@ -225,6 +228,20 @@ def create_columns(self):
     def columns_sanity_check(self, columns):
         column_types = [column.__class__ for column in columns]
         column_names = [column.name for column in columns]
+        # Check for column names which are python keywords or contain spaces
+        if not self.force_cols:
+            lower_case_kws = [kw.lower() for kw in keyword.kwlist]
+            for col_name in column_names:
+                if col_name.lower() in lower_case_kws:
+                    raise MetadataError(
+                        ('Cannot use column name "' + col_name +
+                         '" because it is a reserved python keyword'))
+        omero_reserved_col_names = ['Image Name',
+                                    'Dataset Name']
+        for col_name in column_names:
+            if ' ' in col_name and col_name not in omero_reserved_col_names:
+                log.warn('Column name "' + col_name +
+                         '" contains a space and cannot be used for querying')
         if WellColumn in column_types and ImageColumn in column_types:
             log.debug(column_types)
             raise MetadataError(
@@ -982,7 +999,7 @@ class ParsingContext(object):
     def __init__(self, client, target_object, file=None, fileid=None,
                  cfg=None, cfgid=None, attach=False, column_types=None,
                  options=None, batch_size=1000, loops=10, ms=500,
-                 dry_run=False, allow_nan=False):
+                 dry_run=False, allow_nan=False, force_cols=False):
         '''
         This lines should be handled outside of the constructor:
 
@@ -1006,6 +1023,7 @@ def __init__(self, client, target_object, file=None, fileid=None,
                                                        target_object,
                                                        self.value_resolver)
         self.dry_run = dry_run
+        self.force_cols = force_cols
 
     def create_annotation_link(self):
         self.target_class = self.target_object.__class__
@@ -1050,7 +1068,7 @@ def preprocess_from_handle(self, data):
         log.debug('Column types: %r' % self.column_types)
         self.header_resolver = HeaderResolver(
             self.target_object, header_row,
-            column_types=self.column_types)
+            column_types=self.column_types, force_cols=self.force_cols)
         self.columns = self.header_resolver.create_columns()
         log.debug('Columns: %r' % self.columns)
         if len(self.columns) > MAX_COLUMN_COUNT: