summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
-rw-r--r--api-bank.rst1
-rw-r--r--api-merchant.rst13
2 files changed, 5 insertions, 9 deletions
diff --git a/api-bank.rst b/api-bank.rst
index f038d6b4..41d739bf 100644
--- a/api-bank.rst
+++ b/api-bank.rst
@@ -167,6 +167,7 @@ request.
:query direction: argument taking values `debit` or `credit`, according to the caller willing to receive both incoming and outgoing, only outgoing, or only incoming records. Use `both` to return both directions.
:query cancelled: argument taking values `omit` or `show` to filter out rejected transactions
:query account_number: bank account whose history is to be returned. *Currently ignored*, as multiple bank accounts per user are not implemented yet.
+ :query ordering: can be `descending` or `ascending` and regulates whether the row are returned youger-to-older or vice versa. Defaults to `descending`.
**Response**
diff --git a/api-merchant.rst b/api-merchant.rst
index aa435b75..5c797038 100644
--- a/api-merchant.rst
+++ b/api-merchant.rst
@@ -550,16 +550,11 @@ Transaction history
**Request**
- :query date: only transactions *older* than this parameter will be returned. It's a timestamp, given in seconds.
- Being optional, it defaults to the current time if not given.
- :query start: only transactions having `row_id` less than `start` will be returned. Being optional, it defaults to the
- highest `row_id` contained in the DB (namely, the youngest entry).
- :query delta: at most `delta` entries will be returned. Being optional, it defaults to 20.
+ :query date: time threshold, see `delta` for its interpretation.
+ :query start: row number threshold, see `delta` for its interpretation. Defaults to `UINT64_MAX`, namely the biggest row id possible in the database.
+ :query delta: takes value of the form `N (-N)`, so that at most `N` values strictly younger (older) than `start` and `date` are returned. Defaults to `-20`.
:query instance: on behalf of which merchant instance the query should be accomplished.
-
- A typical usage is to firstly call this API without `start` and `date` parameter, then fetch the oldest
- `row_id` from the results, and then keep calling the API by using the oldest row ID as `start` parameter.
- This way we simply "scroll" results from the youngest to the oldest, `delta` entries at time.
+ :query ordering: takes value `descending` or `ascending` according to the results wanted from younger to older or vice versa. Defaults to `descending`.
**Response**