Feat: support emit on window close for window join and interval join

[chenzl25]

Is your feature request related to a problem? Please describe.

Currently, interval join and window join emit rows for each update. We can support an emit-on-window-close version for them like Aggregation.

CREATE SOURCE s1 (
 id int,
 value int,
 ts TIMESTAMP,
 WATERMARK FOR ts AS ts - INTERVAL '20' SECOND
) WITH (
    connector = 'datagen',
    fields.ts.kind = 'random',
    fields.ts.max_past = '20 seconds',
    fields.ts.max_past_mode = 'relative',
    datagen.rows.per.second='1',
);

CREATE SOURCE s2 (
 id int,
 value int,
 ts TIMESTAMP,
 WATERMARK FOR ts AS ts - INTERVAL '20' SECOND
) WITH (
    connector = 'datagen',
    fields.ts.kind = 'random',
    fields.ts.max_past = '20 seconds',
    fields.ts.max_past_mode = 'relative',
    datagen.rows.per.second='1',
);

Window join:

dev=> explain CREATE MATERIALIZED VIEW mv1 AS
SELECT s1.window_start as window_start, s1.id as id, s2.value AS value1, s2.value AS value2
FROM tumble(s1, s1.ts, interval '1 minute'),
     tumble(s2, s2.ts, interval '1 minute')
WHERE s1.window_start = s2.window_start and s1.id = s2.id
emit on window close;
NOTICE:  EMIT ON WINDOW CLOSE is currently an experimental feature. Please use it with caution.
ERROR:  Failed to run the query

Caused by:
  Not supported: The query cannot be executed in Emit-On-Window-Close mode.
HINT: Try define a watermark column in the source, or avoid aggregation without GROUP BY

Interval Join:

dev=> CREATE MATERIALIZED VIEW interval_join AS
SELECT s1.id AS id1,
       s1.value AS value1,
       s2.id AS id2,
       s2.value AS value2
FROM s1 JOIN s2
ON s1.id = s2.id and s1.ts between s2.ts and s2.ts + INTERVAL '1' MINUTE
EMIT ON WINDOW CLOSE;
NOTICE:  EMIT ON WINDOW CLOSE is currently an experimental feature. Please use it with caution.
ERROR:  Failed to run the query

Caused by:
  Not supported: The query cannot be executed in Emit-On-Window-Close mode.
HINT: Try define a watermark column in the source, or avoid aggregation without GROUP BY

Describe the solution you'd like

No response

Describe alternatives you've considered

No response

Additional context

No response