[#50] Re-add titles support for multi-sheet spreadsheets

Author: Bjwebb

flattentool/input.py

@@ -36,15 +36,17 @@ class SpreadsheetInput(object):
     or csv).
 
     """
-    def convert_dict_titles(self, dicts):
+    def convert_dict_titles(self, dicts, title_lookup=None):
         """
         Replace titles with field names in the given list of dictionaries
         (``dicts``) using the titles lookup in the schema parser.
 
         """
+        if self.parser:
+            title_lookup = title_lookup or self.parser.title_lookup
         for d in dicts:
-            if self.parser:
-                yield { self.parser.title_lookup.lookup_header(k):v for k,v in d.items() }
+            if title_lookup:
+                yield { title_lookup.lookup_header(k):v for k,v in d.items() }
             else:
                 yield d
 
@@ -66,10 +68,8 @@ def get_main_sheet_lines(self):
     def get_sub_sheets_lines(self):
         for sub_sheet_name in self.sub_sheet_names:
             if self.convert_titles:
-                # TODO: This won't work properly any more (breaks roundtrip
-                # tests, but we should also have something more like unit
-                # tests!)
-                yield sub_sheet_name, self.convert_dict_titles(self.get_sheet_lines(sub_sheet_name))
+                yield sub_sheet_name, self.convert_dict_titles(self.get_sheet_lines(sub_sheet_name),
+                    self.parser.sub_sheets[sub_sheet_name].title_lookup if sub_sheet_name in self.parser.sub_sheets else None)
             else:
                 yield sub_sheet_name, self.get_sheet_lines(sub_sheet_name)
 

flattentool/schema.py

@@ -21,6 +21,10 @@ class TitleLookup(UserDict):
     property_name = None
 
     def lookup_header(self, title_header):
+        # Ignore titles with a / in, as they may contain types
+        # https://github.com/OpenDataServices/flatten-tool/issues/56
+        if '/' in title_header:
+            return title_header
         return self.lookup_header_list(title_header.split(':'))
 
     def lookup_header_list(self, title_header_list):
@@ -143,6 +147,7 @@ def parse_schema_dict(self, parent_name, schema_dict, parent_id_fields=None, tit
                         if sub_sheet_name not in self.sub_sheets:
                             self.sub_sheets[sub_sheet_name] = Sheet(root_id=self.root_id, name=sub_sheet_name)
                         sub_sheet = self.sub_sheets[sub_sheet_name]
+                        sub_sheet.title_lookup = title_lookup.get(title)
 
                         for field in id_fields:
                             sub_sheet.add_field(field+':'+property_name, id_field=True)