### Description of the issue/feature this PR addresses:
Prevent Out of Memory (OOM) crashes and drastically improve execution speed when generating the Swedish SIE4 verification export for massive datasets.

### Current behavior before PR:
When exporting a large volume of journal entries (e.g., 190,000+ account moves), the `_export_l10n_se_sie4_verification` method relies on iterating through heavy ORM recordsets and accessing relational child fields (move.line_ids) inside a loop. This triggers a severe N+1 query problem, maxing out server RAM and causing an OOM crash.

### Desired behavior after PR is merged:
The method now utilizes a hybrid data extraction approach:
- The ORM is used strictly to safely evaluate domains (multi-company rules, dates, states) and fetch a lightweight list of valid move_ids.
- A single SQL query with JOIN statements fetches all parent moves, child lines, and account codes in exactly one database query.
- itertools.groupby chunks the flat, lightweight dictionary results back into their respective journal entries.
The export now handles massive datasets in seconds with minimal memory overhead, while remaining perfectly secure.

### Benchmark:

For Memory:
| # Input Data | Before PR | After PR |
| -------- | -------- | -------- |
| ~190,000 moves | MemoryError | 407MB|
| ~200,000 moves | 1.8GB | 174.8 MB|

For Speed:
| # Input Data | Before PR | After PR |
| -------- | -------- | -------- |
| ~190,000 moves | MemoryError | 5.10s |
| ~200,000 moves | 1m29s| 5.3s|

### Reference:
opw-6067999