Function init

archivebox/main.py:308–472

Initialize a new ArchiveBox collection in the current directory

(force: bool=False, quick: bool=False, setup: bool=False, out_dir: Path=OUTPUT_DIR)

@enforce_types
def init(force: bool=False, quick: bool=False, setup: bool=False, out_dir: Path=OUTPUT_DIR) -> None:
    """Initialize a new ArchiveBox collection in the current directory"""
    
    from core.models import Snapshot

    out_dir.mkdir(exist_ok=True)
    is_empty = not len(set(os.listdir(out_dir)) - ALLOWED_IN_OUTPUT_DIR)

    if (out_dir / JSON_INDEX_FILENAME).exists():
        stderr("[!] This folder contains a JSON index. It is deprecated, and will no longer be kept up to date automatically.", color="lightyellow")
        stderr("    You can run `archivebox list --json --with-headers > static_index.json` to manually generate it.", color="lightyellow")

    existing_index = (out_dir / SQL_INDEX_FILENAME).exists()

    if is_empty and not existing_index:
        print('{green}[+] Initializing a new ArchiveBox v{} collection...{reset}'.format(VERSION, **ANSI))
        print('{green}----------------------------------------------------------------------{reset}'.format(**ANSI))
    elif existing_index:
        # TODO: properly detect and print the existing version in current index as well
        print('{green}[^] Verifying and updating existing ArchiveBox collection to v{}...{reset}'.format(VERSION, **ANSI))
        print('{green}----------------------------------------------------------------------{reset}'.format(**ANSI))
    else:
        if force:
            stderr('[!] This folder appears to already have files in it, but no index.sqlite3 is present.', color='lightyellow')
            stderr('    Because --force was passed, ArchiveBox will initialize anyway (which may overwrite existing files).')
        else:
            stderr(
                ("{red}[X] This folder appears to already have files in it, but no index.sqlite3 present.{reset}\n\n"
                "    You must run init in a completely empty directory, or an existing data folder.\n\n"
                "    {lightred}Hint:{reset} To import an existing data folder make sure to cd into the folder first, \n"
                "    then run and run 'archivebox init' to pick up where you left off.\n\n"
                "    (Always make sure your data folder is backed up first before updating ArchiveBox)"
                ).format(out_dir, **ANSI)
            )
            raise SystemExit(2)

    if existing_index:
        print('\n{green}[*] Verifying archive folder structure...{reset}'.format(**ANSI))
    else:
        print('\n{green}[+] Building archive folder structure...{reset}'.format(**ANSI))
    
    print(f'    + ./{ARCHIVE_DIR.relative_to(OUTPUT_DIR)}, ./{SOURCES_DIR.relative_to(OUTPUT_DIR)}, ./{LOGS_DIR.relative_to(OUTPUT_DIR)}...')
    Path(SOURCES_DIR).mkdir(exist_ok=True)
    Path(ARCHIVE_DIR).mkdir(exist_ok=True)
    Path(LOGS_DIR).mkdir(exist_ok=True)
    print(f'    + ./{CONFIG_FILE.relative_to(OUTPUT_DIR)}...')
    write_config_file({}, out_dir=out_dir)

    if (out_dir / SQL_INDEX_FILENAME).exists():
        print('\n{green}[*] Verifying main SQL index and running any migrations needed...{reset}'.format(**ANSI))
    else:
        print('\n{green}[+] Building main SQL index and running initial migrations...{reset}'.format(**ANSI))
    
    DATABASE_FILE = out_dir / SQL_INDEX_FILENAME
    for migration_line in apply_migrations(out_dir):
        print(f'    {migration_line}')

    assert DATABASE_FILE.exists()