correct behavior with invalid scan range

include/kvs.h

@@ -230,14 +230,19 @@ put(Token token, std::string_view storage_name, // NOLINT
  * @param[in] l_end If this argument is scan_endpoint::EXCLUSIVE, the interval
  * does not include the endpoint. If this argument is scan_endpoint::INCLUSIVE,
  * the interval contains the endpoint. If this is scan_endpoint::INF, there is
- * no limit on the interval in left direction. And ignore @a l_key.
+ * no limit on the interval in left direction and `l_key` is ignored.
  * @param[in] r_key An argument that specifies the right endpoint.
- * @note If r_key <l_key is specified in dictionary order, nothing will be hit.
- * @param[in] r_end If this argument is scan_endpoint :: EXCLUSIVE, the
- * interval does not include the endpoint. If this argument is
- * scan_endpoint::INCLUSIVE, the interval contains the endpoint. If this is
- * scan_endpoint::INF, there is no limit on the interval in right direction.
- * And ignore @a r_key.
+ * @param[in] r_end If this argument is scan_endpoint::EXCLUSIVE, the interval
+ * does not include the endpoint. If this argument is scan_endpoint::INCLUSIVE,
+ * the interval contains the endpoint. If this is scan_endpoint::INF, there is
+ * no limit on the interval in right direction and `r_key` is ignored.
+ * @note The interval defined by `l_key`, `l_end`, `r_key`, and `r_end` must represent a valid range
+ * and otherwise an error (Status::ERR_BAD_USAGE) will be returned. A range is invalid if any of the following
+ * conditions are satisfied.
+ * - neither `l_end` nor `r_end` is scan_endpoint::INF and `r_key` < `l_key`
+ * - neither `l_end` nor `r_end` is scan_endpoint::INF and `l_key` == `r_key` and either `l_end` or `r_end`
+ * is scan_endpoint::EXCLUSIVE
+ * - `r_key` is empty string and `r_end` is scan_endpoint::EXCLUSIVE
  * @param[out] tuple_list A set with a key, a pointer to value, and size of
  * value as a result of this function.
  * The address obtained here can be accessed safely until the Token entered at the time of address acquisition leaves.
@@ -253,17 +258,14 @@ put(Token token, std::string_view storage_name, // NOLINT
  * node_version_vec to guarantee atomicity, you can split the atomic scan.
  * Suppose you want to scan with A: C, assuming the order A <B <C. You can
  * perform an atomic scan by scanning the range A: B, B: C and using
- * node_version_vec to make sure the values ​​are not overwritten. This advantage
+ * node_version_vec to make sure the values are not overwritten. This advantage
  * is effective when the right end point is unknown but you want to scan to a
  * specific value.
  * @param[in] right_to_left If this argument is true, the scan is performed from right end.
  * When this is set to true, current implementation has following limitation: 1. max_size must be 1 so that at most
  * one entry is hit and returned as scan result 2. r_end must be scan_endpoint::INF so that the scan is performed from
  * unbounded right end. Status::ERR_BAD_USAGE is returned if these conditions are not met.
- * @return Status::ERR_BAD_USAGE The arguments is invalid. In the case1: you use
- * same l_key and r_key and one of the endpoint is exclusive. case2: one of the
- * endpoint use null key but the string size is not zero like
- * std::string_view{nullptr, non-zero-value}.
+ * @return Status::ERR_BAD_USAGE The input argument or the range given by the arguments is invalid (see note above.)
  * @return status::OK success.
  * @return Status::OK_ROOT_IS_NULL success and root is null.
  * @return status::WARN_STORAGE_NOT_EXIST The target storage of this operation