How to perform asof join (aj) with strict inequality (<) instead of default (≤)

I am using DolphinDB version 3.00.0.6 and have two in-memory tables A and B.

I need the equivalent of a Non-Equi Join with strict inequality (A.date > B.date) to find the latest record in B that is strictly before each record in A.

My Sample Data:

A = table(
    2023.01.01 2023.01.02 2023.01.03 as date, 
    `AAPL`AAPL`AAPL as stock_id, 
    10.0 20.0 30.0 as valA
)
B = table(
    2023.01.01 2023.01.02 2023.01.04 as date,
    `AAPL`AAPL`AAPL as stock_id,
    100.0 200.0 400.0 as valB
)

I first tried asof join:

select * from aj(A, B, `stock_id`date)

date       stock_id valA B_date     valB
---------- -------- ---- ---------- ----
2023.01.01 AAPL     10   2023.01.01 100 
2023.01.02 AAPL     20   2023.01.02 200 
2023.01.03 AAPL     30   2023.01.02 200 

But what I expect is:

date       stock_id valA B_date     valB
---------- -------- ---- ---------- ----
2023.01.01 AAPL     10   
2023.01.02 AAPL     20   2023.01.01 100 
2023.01.03 AAPL     30   2023.01.02 200 

I also tried to use non-equi join syntax：

select * from A join B on A.stock_id=B.stock_id and A.date > B.date

But with an syntax error:

Syntax Error: [line #14] A join filter can't involves columns in two or more tables.

DolphinDB seems not support such syntax.

Is there a built-in way to modify aj for strict inequality? How can I achieve the equivalent of A.date > B.date join condition in DolphinDB?

Since DolphinDB aj matches <=, you can subtract a minimal time unit (like 1 day for date columns, or 1 nanosecond for datetime/timestamp) from A.date before joining:

// Prepare sample data
A = table(
    2023.01.01 2023.01.02 2023.01.03 as date, 
    `AAPL`AAPL`AAPL as stock_id, 
    10.0 20.0 30.0 as valA
)
B = table(
    2023.01.01 2023.01.02 2023.01.04 as date,
    `AAPL`AAPL`AAPL as stock_id,
    100.0 200.0 400.0 as valB
)

// Use aj with adjusted date to simulate strict inequality
A1 = select date, stock_id, valA, date - 1 as date_adj from A
result = select A1.date, A1.stock_id, A1.valA, B.date as B_date, B.valB 
         from aj(A1, B, `stock_id`date_adj)

Ouput

date       stock_id valA B_date     valB
---------- -------- ---- ---------- ----
2023.01.01 AAPL     10
2023.01.02 AAPL     20   2023.01.01 100
2023.01.03 AAPL     30   2023.01.02 200