Fix for issue pandas-dev#169 (pandas-dev#170)

* Fix for issue pandas-dev#169

Author: bmoscon

CHANGES.md

@@ -7,6 +7,7 @@
   * Bugfix: #147 Add get_info method to ChunkStore
   * Bugfix: Periodically re-cache the library.quota to pick up any changes
   * Bugfix: #166 Add index on SHA for ChunkStore
+  * Bugfix: #169 Dtype mismatch in chunkstore updates
 
 ### 1.25 (2016-05-23)
 

arctic/chunkstore/chunkstore.py

@@ -167,7 +167,8 @@ def write(self, symbol, item, chunk_size):
             raise Exception("Can only chunk Series and DataFrames")
 
         previous_shas = []
-        if self._get_symbol_info(symbol):
+        sym = self._get_symbol_info(symbol)
+        if sym:
             previous_shas = set([Binary(x['sha']) for x in self._collection.find({'symbol': symbol},
                                                                          projection={'sha': True, '_id': False},
                                                                          )])
@@ -177,6 +178,9 @@ def write(self, symbol, item, chunk_size):
 
         for start, end, record in self.chunker.to_chunks(item, chunk_size):
             r, dtype = serialize(record, string_max_len=self.STRING_MAX)
+            # if symbol exists, dtypes better match
+            if sym and str(dtype) != sym['dtype']:
+                raise Exception('Dtype mismatch - cannot write chunk')
             records.append(r)
             ranges.append((start, end))
 
@@ -266,6 +270,8 @@ def append(self, symbol, item):
                     sym = self._get_symbol_info(symbol)
                     continue
             r, dtype = serialize(record, string_max_len=self.STRING_MAX)
+            if str(dtype) != sym['dtype']:
+                raise Exception("Dtype mismatch - cannot append")
             records.append(r)
             ranges.append((start, end))
 
@@ -321,6 +327,7 @@ def update(self, symbol, item):
         if not sym:
             raise NoDataFoundException("Symbol does not exist. Cannot update")
 
+        
         records = []
         ranges = []
         orig_ranges = []
@@ -337,7 +344,9 @@ def update(self, symbol, item):
             else:
                 orig_ranges.append((None, None))
 
-            r, _ = serialize(record, string_max_len=self.STRING_MAX)
+            r, dtype = serialize(record, string_max_len=self.STRING_MAX)
+            if str(dtype) != sym['dtype']:
+                raise Exception('Dtype mismatch - cannot update')
             records.append(r)
             ranges.append((start, end))
 

tests/integration/chunkstore/test_chunkstore.py

@@ -616,3 +616,25 @@ def test_get_info(chunkstore_lib):
             'type': u'df',
             'size': 72}
     assert(chunkstore_lib.get_info('test_df') == info)
+
+
+def test_dtype_mismatch_error(chunkstore_lib):
+    s = pd.Series([1], index=pd.date_range('2016-01-01', '2016-01-01', name='date'), name='vals')
+
+    # Write with an int
+    chunkstore_lib.write('test', s, 'D')
+
+    # Update with a float
+    with pytest.raises(Exception) as e:
+        chunkstore_lib.update('test', s * 1.0)
+    assert('Dtype mismatch' in  str(e))
+
+    with pytest.raises(Exception) as e:
+        chunkstore_lib.write('test', s * 1.0, 'D')
+    assert('Dtype mismatch' in  str(e))
+
+    with pytest.raises(Exception) as e:
+        chunkstore_lib.append('test', s * 1.0)
+    assert('Dtype mismatch' in  str(e))
+
+    assert_series_equal(s, chunkstore_lib.read('test'))