[fix](doc) week/weekofyear: backport zh-only examples to en; fix zh typo using 2023 twice

[boluor]

Summary

Three changes across two date-time function pages.

EN `week.md` (+1 example)

`mode 7` — weeks start on Monday and week 1 is the week containing the first Monday. Because the first Monday of 2023 is January 2, `WEEK('2023-01-01', 7)` returns 52 (last week of 2022). ZH already had this example.

EN `weekofyear.md` (+3 examples)

SQL	Returns	Why interesting
`WEEKOFYEAR('2022-01-02')`	52	The week 2021-12-27 to 2022-01-02 has only 2 days in 2022, so it falls in week 52 of 2021
`WEEKOFYEAR('2023-12-25')`	52	The final week of 2023
`weekofyear('1023-01-04')`	1	Pre-modern year — the function does not enforce a minimum year

ZH already had analogues of these.

ZH `weekofyear.md` typo fix

The cross-year example's prose described the 2021-12-27..2022-01-02 week, but the SQL was `WEEKOFYEAR('2023-01-02')` and returned 1 — wrong week for the prose, and a duplicate of the example just above. Change the SQL and alias to `'2022-01-02'` / `week_20220102` so the example actually demonstrates what the comment says (returns 52).

Verification

```
mysql> SELECT WEEK('2023-01-01', 7); -- 52
mysql> SELECT WEEKOFYEAR('2022-01-02'); -- 52
mysql> SELECT WEEKOFYEAR('2023-12-25'); -- 52
mysql> SELECT weekofyear('1023-01-04'); -- 1
```

Test plan

• All added examples verified on Apache Doris 4.1.1.
• ZH weekofyear typo no longer duplicates the previous example.

🤖 Generated with Claude Code