ENH: Adds mapping of company_name and ensures str types in AssetFinder

zipline/assets/asset_writer.py

@@ -25,13 +25,14 @@ ASSET_TABLE_FIELDS = frozenset({
     'exchange',
 })
 
-# Expected fields for an Asset's metadata
+# Expected fields for a Future's metadata
 FUTURE_TABLE_FIELDS = ASSET_TABLE_FIELDS | {
     'notice_date',
     'expiration_date',
     'contract_multiplier',
 }
 
+# Expected fields for an Equity's metadata
 EQUITY_TABLE_FIELDS = ASSET_TABLE_FIELDS
 
 EXCHANGE_TABLE_FIELDS = frozenset({
@@ -355,6 +356,11 @@ class AssetDBWriter(with_metaclass(ABCMeta)):
         # Generate equities DataFrame #
         ###############################
 
+        # HACK: If company_name is provided, map it to asset_name
+        if ('company_name' in data.equities.columns) \
+                and ('asset_name' not in data.equities.columns):
+            data.equities['asset_name'] = data.equities['company_name']
+
         equities_output = _generate_output_dataframe(
             data_subset=data.equities,
             defaults=_equities_defaults,

zipline/assets/assets.py

@@ -43,6 +43,24 @@ from zipline.assets.asset_writer import (
 
 log = Logger('assets.py')
 
+# A set of fields that need to be converted to strings before building an
+# Asset to avoid unicode fields
+_asset_str_fields = frozenset({
+    'symbol',
+    'asset_name',
+    'exchange',
+})
+
+
+def _convert_asset_str_fields(dict):
+    """
+    Takes in a dict of Asset init args and converts from unicode to string
+    where applicable
+    """
+    for key, value in dict.items():
+        if key in _asset_str_fields:
+            dict[key] = str(value)
+
 
 class AssetFinder(object):
 
@@ -201,6 +219,8 @@ class AssetFinder(object):
                 data['first_traded'] = pd.Timestamp(
                     data['first_traded'], tz='UTC')
 
+            _convert_asset_str_fields(data)
+
             equity = Equity(**data)
         else:
             equity = None
@@ -239,6 +259,8 @@ class AssetFinder(object):
                 data['expiration_date'] = pd.Timestamp(
                     data['expiration_date'], tz='UTC')
 
+            _convert_asset_str_fields(data)
+
             future = Future(**data)
         else:
             future = None