Allow get_product_trades to be paginated

This implements the pagination of trades, as per the docs (https://docs.gdax.com/#get-trades).

gdax/public_client.py

@@ -119,30 +119,35 @@ def get_product_ticker(self, product_id):
         """
         return self._get('/products/{}/ticker'.format(str(product_id)))
 
-    def get_product_trades(self, product_id):
-        """List the latest trades for a product.
+    def get_product_trades(self, product_id, before='', after='', limit='', result=[]):
+        url = self.url + '/products/{}/trades'.format(str(product_id))
+        params = {}
 
-        Args:
-            product_id (str): Product
+        if before:
+            params['before'] = str(before)
+        if after:
+            params['after'] = str(after)
+        if limit and limit < 100:
+            # the default limit is 100
+            # we only add it if the limit is less than 100
+            params['limit'] = limit
 
-        Returns:
-            list: Latest trades. Example::
-                [{
-                    "time": "2014-11-07T22:19:28.578544Z",
-                    "trade_id": 74,
-                    "price": "10.00000000",
-                    "size": "0.01000000",
-                    "side": "buy"
-                }, {
-                    "time": "2014-11-07T01:08:43.642366Z",
-                    "trade_id": 73,
-                    "price": "100.00000000",
-                    "size": "0.01000000",
-                    "side": "sell"
-                }]
+        r = requests.get(url, params=params)
+        r.raise_for_status()
 
-        """
-        return self._get('/products/{}/trades'.format(str(product_id)))
+        result.extend(r.json())
+
+        if 'cb-after' in r.headers and limit is not len(result):
+            # update limit
+            limit -= len(result)
+            if limit <= 0:
+                return result
+
+            # ensure that we don't get rate-limited/blocked
+            time.sleep(0.4)
+            return self.get_product_trades(product_id=product_id, after=r.headers['cb-after'], limit=limit, result=result)
+
+        return result
 
     def get_product_historic_rates(self, product_id, start=None, end=None,
                                    granularity=None):