diff --git a/src/elicznik/elicznik.py b/src/elicznik/elicznik.py
index a20b00b..e59a199 100755
--- a/src/elicznik/elicznik.py
+++ b/src/elicznik/elicznik.py
@@ -1,15 +1,13 @@
 #!/usr/bin/env python3
 
+import csv
 import datetime
 
 from .session import Session
-from collections import defaultdict
-import csv
+
 
 class ELicznik:
     LOGIN_URL = "https://logowanie.tauron-dystrybucja.pl/login"
-    CHART_URL = "https://elicznik.tauron-dystrybucja.pl/energia/api"
-    READINGS_URL = "https://elicznik.tauron-dystrybucja.pl/odczyty/api"
     DATA_URL = "https://elicznik.tauron-dystrybucja.pl/energia/do/dane"
 
     def __init__(self, username, password):
@@ -35,90 +33,60 @@ class ELicznik:
     def __exit__(self, exc_type, exc_val, exc_tb):
         pass
 
-    def _get_raw_readings(self, type_, start_date, end_date=None):
-        end_date = end_date or start_date
-        data = self.session.post(
-            self.CHART_URL,
-            data={
-                "type": type_,
-                "from": start_date.strftime("%d.%m.%Y"),
-                "to": end_date.strftime("%d.%m.%Y"),
-                "profile": "full time",
-            },
-        ).json()
-
-        data = data.get("data", {}).get("allData", {})
-        for element in data:
-            date = element.get("Date")
-            hour = int(element.get("Hour"))
-            # TODO: There's also an "Extra" field, which seems to be set to be set to "T" only for the one extra hour
-            # when switching from CEST to CET (e.g. 3 AM on 2021-10-31)
-            timestamp = datetime.datetime.strptime(date, "%Y-%m-%d")
-            timestamp += datetime.timedelta(hours=hour)
-            value = element.get("EC")
-            yield timestamp, value
-
-    def get_readings_production(self, start_date, end_date=None):
-        return dict(self._get_raw_readings("oze", start_date, end_date))
-
-    def get_readings_consumption(self, start_date, end_date=None):
-        return dict(self._get_raw_readings("consum", start_date, end_date))
-
-    def get_readings(self, start_date, end_date=None):
-        consumed = self.get_readings_consumption(start_date, end_date)
-        produced = self.get_readings_production(start_date, end_date)
-        return sorted(
-            (timestamp, float(consumed.get(timestamp)), float(produced.get(timestamp)))
-            for timestamp in set(consumed) | set(produced)
-        )
-
     def get_raw_data(self, start_date, end_date=None):
         end_date = end_date or start_date
-        if start_date == end_date:
-            start_date -= datetime.timedelta(days=1)
         return self.session.post(
             self.DATA_URL,
             data={
                 "form[from]": start_date.strftime("%d.%m.%Y"),
-                "form[to]" : end_date.strftime("%d.%m.%Y"),
-                "form[type]": "godzin", # or "dzien"
+                "form[to]": end_date.strftime("%d.%m.%Y"),
+                "form[type]": "godzin",  # or "dzien"
                 "form[consum]": 1,
                 "form[oze]": 1,
-                "form[fileType]": "CSV", # or "XLS"
+                "form[fileType]": "CSV",  # or "XLS"
             },
-        ).content.decode().split('\n')
+        ).text.splitlines()
 
-    def get_data(self, start_date, end_date=None):
+    @staticmethod
+    def parse_timestamp(timespec):
+        date, time = timespec.split(None, 1)
+        hour = int(time.split(":")[0]) - 1
+        return datetime.datetime.strptime(date, "%Y-%m-%d") + datetime.timedelta(
+            hours=hour
+        )
+
+    def get_readings(self, start_date, end_date=None):
         end_date = end_date or start_date
-        data = csv.reader(self.get_raw_data(start_date, end_date)[1:], delimiter=';')
-        cons = defaultdict(float)
-        prod = defaultdict(float)
-        for rec in data:
-            try :
-                t, v, r, *_ = rec
-            except ValueError:
-                # print('ValueError:', rec)
-                continue
-            date, hour = t.split()
-            h, m = hour.split(':')
-            timestamp = datetime.datetime.strptime(date, "%d.%m.%Y")
-            timestamp += datetime.timedelta(hours=int(h), minutes=int(m))
-            # Skip records outside a single day block
-            if start_date == end_date and timestamp.day != start_date.day:
-                print(f'Skip {timestamp} not within {start_date}.')
-                continue
-            v = v.replace(',','.')
-            if r=='pobór':
-                cons[timestamp] = v
-            elif r=='oddanie':
-                prod[timestamp] = v
-            else :
-                print('Unknown data format:', l)
+        data = self.get_raw_data(start_date, end_date)
+
+        records = [
+            {
+                "timestamp": self.parse_timestamp(rec["Data"]),
+                "value": float(rec[" Wartość kWh"].replace(",", ".")),
+                "type": rec["Rodzaj"],
+            }
+            for rec in csv.DictReader(data, delimiter=";")
+        ]
+
+        # skip records which are outside the requested date range
+        # TODO: is this really needed?
+        records = [
+            rec for rec in records if start_date <= rec["timestamp"].date() <= end_date
+        ]
+
+        prod = {
+            rec["timestamp"]: rec["value"] for rec in records if rec["type"] == "pobór"
+        }
+        cons = {
+            rec["timestamp"]: rec["value"]
+            for rec in records
+            if rec["type"] == "oddanie"
+        }
+
         # TODO
         # This probably drops the data from the double hour during DST change
         # Needs to be investigated and fixed
         return sorted(
-            (timestamp, float(cons[timestamp]), float(prod[timestamp]))
+            (timestamp, cons.get(timestamp), prod.get(timestamp))
             for timestamp in set(cons) | set(prod)
         )
-