Add detector for names with trailing slashes

Collections and data object names with trailing slashes cause
problems: irods/irods#3892

This adds a detector, and extracts the output processing logic
for name-related issues to a common inner function in the
detector.

Author: stsnel

README.md

@@ -7,6 +7,7 @@ This script checks the iRODS ICAT database for unexpected issues, specifically:
 - Object names that contain characters which are not handled correctly on XML-based clients,
   such as the python-irods-client (See https://github.com/irods/irods/issues/4132 for details).
 - Data objects with empty names
+- Collection and data object names with trailing slashes (See https://github.com/irods/irods/issues/3892)
 - Files in vaults that have a directory name which is inconsistent with the collection name
 - Hard links: multiple data objects refer to the same physical file
 - Duplicate replica: multiple replica entries for the same file

icat_tools/dbcheck_outputprocessors.py

@@ -73,6 +73,11 @@ def output_item(self, check, values):
                     "Name with characters that iRODS processes incorrectly for " +
                     values['check_name'])
                 self._print_report_column_table(values['report_columns'])
+            elif values['type'] == 'trailing_slash':
+                self._prnln(
+                    "Name with trailing slash for " +
+                    values['check_name'])
+                self._print_report_column_table(values['report_columns'])
             else:
                 self.exit_error("Error: unknown output item type for names check: {}".format(
                         values['type']))
@@ -149,7 +154,7 @@ def output_item(self, check, values):
                 [check, values['object_name'], values['number_replicas'], values['min_replicas']])
 
         elif check == 'names':
-            if values['type'] == 'empty_name' or values['type'] == 'buggy_characters':
+            if values['type'] in ['empty_name','buggy_characters','trailing_slash']:
                 self.writer.writerow([check, values['type'], values['check_name']] +
                                      self._column_value_to_list(values['report_columns']))
             else:

icat_tools/detectors/nameissue_detector.py

@@ -50,8 +50,41 @@ def _check_name_buggy_characters(self, table, name, report_columns):
         cursor.execute(query)
         return cursor
 
+    def _check_name_trailing_slash(self, table, name, report_columns):
+        query = "SELECT {} FROM {} WHERE {} != '/' AND {} LIKE '%/' {}".format( ",".join(report_columns), table, name, name, self._get_prefix_condition(table))
+        cursor = self.connection.cursor("{}._check_name_trailing_slash".format(self.get_name()))
+        cursor.execute(query)
+        return cursor
+
     def run(self):
         issue_found = False
+
+        def _do_output(type_name, report_columns, query_result):
+            """Internal function for translating query results of a check to a generic
+               output dictionary and feeding it to the output processor. Also translates
+               collection IDs to collection names for readability."""
+            nonlocal issue_found
+
+            for row in query_result:
+                output = {'type': type_name, 'check_name' : check_name, 'report_columns': {}}
+                column_num = 0
+                for report_column in report_columns:
+                    if str(report_column) == 'coll_id':
+                        coll_name = utils.get_collection_name(
+                            self.connection, str(row[column_num]))
+                        if coll_name is not None:
+                            output['report_columns']['Collection name'] = coll_name
+                    else:
+                        output['report_columns'][str(report_column)] = str(
+                            row[column_num])
+                    column_num = column_num + 1
+
+                self.output_item(output)
+                issue_found = True
+
+            query_result.close()
+            return issue_found
+
         for check_name, check_params in self._get_name_check_data():
             if self.args.v:
                 self.print_progress("Running empty name test for: " + check_name)
@@ -60,16 +93,18 @@ def run(self):
                 check_params['table'],
                 check_params['name'],
                 check_params['report_columns'])
-            for row in result_empty:
-                output = {'type': 'empty_name', 'check_name' : check_name, 'report_columns': {}}
-                column_num = 0
-                for report_column in check_params['report_columns']:
-                    output['report_columns'][str(report_column)] = str(
-                        row[column_num])
-                    column_num = column_num + 1
-                self.output_item(output)
-                issue_found = True
-            result_empty.close()
+            _do_output("empty_name", check_params['report_columns'], result_empty)
+
+
+            if check_name in ["data object", "collection"]:
+                if self.args.v:
+                    self.print_progress("Running trailing slash test for: " + check_name)
+
+                result_trailing_slash = self._check_name_trailing_slash(
+                    check_params['table'],
+                    check_params['name'],
+                    check_params['report_columns'])
+                _do_output("trailing_slash", check_params['report_columns'], result_trailing_slash)
 
             if self.args.v:
                 self.print_progress("Running problematic character name test for: " + check_name)
@@ -78,21 +113,6 @@ def run(self):
                 check_params['table'],
                 check_params['name'],
                 check_params['report_columns'])
-            for row in result_buggy_characters:
-                output = {'type': 'buggy_characters', 'check_name' : check_name, 'report_columns': {}}
-                column_num = 0
-                for report_column in check_params['report_columns']:
-                    if str(report_column) == 'coll_id':
-                        coll_name = utils.get_collection_name(
-                            self.connection, str(row[column_num]))
-                        if coll_name is not None:
-                            output['report_columns']['Collection name'] = coll_name
-                    else:
-                        output['report_columns'][str(report_column)] = str(
-                            row[column_num])
-                    column_num = column_num + 1
-                self.output_item(output)
-                issue_found = True
-            result_buggy_characters.close()
+            _do_output("buggy_characters", check_params['report_columns'], result_buggy_characters)
 
         return issue_found