Add command to create sqlite database

[hf-krechan]

Move the code from here https://github.com/Hochfrequenz/xml-migs-and-ahbs/blob/main/load_ahbs_into_sqlitedb.py and put it in a cli command for fundamend.

Goal is to improve the developer experience to just install fundamend and create a sqlite database.

[hf-kklein]

I see what you're trying to achieve, but I'm afraid, imho there's not that much added value. The script you mentioned (private repo) strongly relies on this function:

def load_ahbs_into_sqlite_db() -> None:
    """loads the files into the sqlite db 'ahb.db' and places it next to this script"""

    relevant_files = [
        (p, date(2024, 10, 1), date(2025, 6, 6)) for p in (repo_root / "FV2410").rglob("**/*AHB*.xml")
    ] + [(p, date(2025, 6, 6), None) for p in (repo_root / "FV2504").rglob("**/*AHB*.xml")]
    actual_sqlite_path = create_db_and_populate_with_ahb_view(relevant_files, drop_raw_tables=True)
    assert actual_sqlite_path.exists()
    engine = create_engine(f"sqlite:///{actual_sqlite_path}")
    with Session(bind=engine) as session:
        create_and_fill_ahb_expression_table(session)
        create_ahbtabellen_view(session)
        session.flush()
    shutil.copy(actual_sqlite_path, unencrypted_db_path)

if you'd like to encapsulated above logic in a CLI tool, you'd need to pass as CLI arguments:

• tuples of (file path + validity start + validity end)
• or directory names that are in some obscure more or less hardcoded logic are translated into validity periods (we have this in kohlrahbi and edi-energy-mirror and I have to go check our implicit requirements/assumptions and CLI args in the code each and every time I use it because the README is outdated too often).

This alone smells like a hard-to-explain CL interface to me if you do not introduce something like a TOC or config file.

The 7z+encryption aspect would lead to even more dependencies here (optional though).

The main effort is not putting AHBs into a DB and zipping it.

The main effort is manually navigating behind the bdew-mako.de paywall, manually downloading and moving files with the correct names to the correct location because both the file metadata and the API response data are ... to be optimized.

Files are missing when you try to download "alle aktuell gültigen XMLs".
Files are missing when you try to download "alle zukünftig gültigen XMLs" (returns only the future modified files).
Files seem to follow a naming convention but actually don't, because everyone just uploads their files as they think they should be named with no cross-checks.
Files are redundant with no way to automatically decide which XMLs to use (Fehlerkorrektur 1, Fehlerkorrektur 2, Konsolidierte Lesefassung 17, Fehlerkorrektur außerordentlich konsolidiert 3...").

That's where the pain is in my experience, not 10 lines of python of which you can copy 8 from the README.

If you build the desired feature, I surely won't reject it. But I feel like it's more of a burden to maintain in the long run than something useful to anyone but us (and we already built it in a way that works).