don't update the cursor in read_records unless the sync mode is incre…

…mental

(and other formatting changes)

airbyte-integrations/connectors/source-kobotoolbox/source_kobotoolbox/source.py

@@ -8,12 +8,13 @@
 from typing import Any, Iterable, List, Mapping, MutableMapping, Optional, Tuple
 from abc import ABC
 from urllib.parse import parse_qsl, urlparse
-from datetime import datetime, timedelta
+from datetime import datetime, timedelta, timezone
 
 import requests
 from airbyte_cdk.sources import AbstractSource
 from airbyte_cdk.sources.streams import IncrementalMixin, Stream
 from airbyte_cdk.sources.streams.http import HttpStream
+from airbyte_cdk.models import SyncMode
 
 stream_json_schema = {
     "$schema": "http://json-schema.org/draft-07/schema#",
@@ -86,39 +87,45 @@ def get_json_schema(self):
     @property
     def state(self) -> Mapping[str, Any]:
         """State will be a dict : {cursor_field: '2023-03-15T00:00:00.000+05:30'}"""
+        retval = {}
+
         if self._cursor_value:
-            return {self.cursor_field: self._cursor_value}
+            retval[self.cursor_field] = self._cursor_value
+        else:
+            retval[self.cursor_field] = self.start_time
 
-        return {self.cursor_field: self.start_time}
+        return retval
 
     @state.setter
     def state(self, value: Mapping[str, Any]):
         """setter for state"""
         if self.cursor_field in value:
             self._cursor_value = value[self.cursor_field]
 
-    def mk_tzaware_utc(self, dt):
+    def mk_tzaware_utc(self, dt: datetime):
         """
         add a utc-tzinfo object to the dt if it doesn't have tzinfo
         if it has a tzinfo, convert to utc
         """
-        from datetime import timezone
-
         if dt.tzinfo is None:
             return dt.replace(tzinfo=timezone.utc)
         return dt.astimezone(timezone.utc)
 
     def mk_query(self):
         """query using endtime"""
+        retval = {}
         if self.cursor_field == "_submission_time":
-            return {self.cursor_field: {"$gte": self.state[self.cursor_field]}}
+            retval[self.cursor_field] = {"$gte": self.state[self.cursor_field]}
+
         else:
             start_sub_time = datetime.fromisoformat(self.state[self.cursor_field])
             start_sub_time -= timedelta(days=self.max_days_to_close)
             start_sub_time = self.mk_tzaware_utc(start_sub_time)
             tzaware_start_time = self.mk_tzaware_utc(datetime.fromisoformat(self.start_time))
             start_sub_time = max(start_sub_time, tzaware_start_time)
-            return {"_submission_time": {"$gte": start_sub_time.isoformat()}, self.cursor_field: {"$gte": self.state[self.cursor_field]}}
+            retval[self.cursor_field] = {"$gte": self.state[self.cursor_field]}
+            retval["_submission_time"] = {"$gte": start_sub_time.isoformat()}
+        return retval
 
     def request_params(
         self,
@@ -127,7 +134,9 @@ def request_params(
         next_page_token: Mapping[str, Any] = None,
     ) -> MutableMapping[str, Any]:
         """build the query request params"""
-        params = {"start": 0, "limit": self.PAGINATION_LIMIT, "sort": json.dumps({self.cursor_field: 1})}
+        sort_params = {}
+        sort_params[self.cursor_field] = 1
+        params = {"start": 0, "limit": self.PAGINATION_LIMIT, "sort": json.dumps(sort_params)}
 
         query = self.mk_query()
 
@@ -180,11 +189,19 @@ def parse_response(self, response: requests.Response, **kwargs) -> Iterable[Mapp
                 retval["endtime"] = endtime.isoformat()
             yield retval
 
-    def read_records(self, *args, **kwargs) -> Iterable[Mapping[str, Any]]:
+    def read_records(
+        self,
+        sync_mode: SyncMode,
+        cursor_field: List[str] | None = None,
+        stream_slice: Mapping[str, Any] | None = None,
+        stream_state: Mapping[str, Any] | None = None,
+        **kwargs,
+    ) -> Iterable[Mapping[str, Any]]:
         """read the records from the stream"""
-        for record in super().read_records(*args, **kwargs):
-            self._cursor_value = record[self.cursor_field]
+        for record in super().read_records(sync_mode, cursor_field, stream_slice, stream_state, **kwargs):
             yield record
+            if sync_mode == SyncMode.incremental:
+                self._cursor_value = max(record[self.cursor_field], self._cursor_value) if self._cursor_value else record[self.cursor_field]
 
 
 class KoboStreamSubmissionTime(KoboToolStream):