BUG: Allow value labels to be read with iterator

All value labels to be read before the iterator has been used

closes pandas-dev#16923

Author: bashtage

pandas/io/stata.py

@@ -997,6 +997,7 @@ def __init__(self, path_or_buf, convert_dates=True,
             self.path_or_buf = BytesIO(contents)
 
         self._read_header()
+        self._setup_dtype()
 
     def __enter__(self):
         """ enter context manager """
@@ -1299,6 +1300,20 @@ def _read_old_header(self, first_char):
         # necessary data to continue parsing
         self.data_location = self.path_or_buf.tell()
 
+    def _setup_dtype(self):
+        # Setup the dtype.
+        if self._dtype is None:
+            dtype = []  # Convert struct data types to numpy data type
+            for i, typ in enumerate(self.typlist):
+                if typ in self.NUMPY_TYPE_MAP:
+                    dtype.append(('s' + str(i), self.byteorder +
+                                  self.NUMPY_TYPE_MAP[typ]))
+                else:
+                    dtype.append(('s' + str(i), 'S' + str(typ)))
+            dtype = np.dtype(dtype)
+            self._dtype = dtype
+        return self._dtype
+
     def _calcsize(self, fmt):
         return (type(fmt) is int and fmt or
                 struct.calcsize(self.byteorder + fmt))
@@ -1408,6 +1423,7 @@ def _read_strls(self):
             # Wrap v_o in a string to allow uint64 values as keys on 32bit OS
             self.GSO[str(v_o)] = va
 
+
     # legacy
     @Appender(_data_method_doc)
     def data(self, **kwargs):
@@ -1476,20 +1492,9 @@ def read(self, nrows=None, convert_dates=None,
             self._can_read_value_labels = True
             self._read_strls()
 
-        # Setup the dtype.
-        if self._dtype is None:
-            dtype = []  # Convert struct data types to numpy data type
-            for i, typ in enumerate(self.typlist):
-                if typ in self.NUMPY_TYPE_MAP:
-                    dtype.append(('s' + str(i), self.byteorder +
-                                  self.NUMPY_TYPE_MAP[typ]))
-                else:
-                    dtype.append(('s' + str(i), 'S' + str(typ)))
-            dtype = np.dtype(dtype)
-            self._dtype = dtype
 
         # Read data
-        dtype = self._dtype
+        dtype = self._setup_dtype()
         max_read_len = (self.nobs - self._lines_read) * dtype.itemsize
         read_len = nrows * dtype.itemsize
         read_len = min(read_len, max_read_len)