Normalize text before fuzzy matching.

    Profiles:
    - general: standard text normalization
    - company: removes legal company suffixes
    - product: standardizes product units and model names
    - address: standardizes common address terms

    Args:
        value: Text to normalize.
        profile: Type of normalization to apply.
    

    Compare two strings using fuzzy matching.

    Strategies:
    - ratio
    - partial
    - token_sort
    - token_set
    - weighted
    - strict

    Args:
        first: First text value.
        second: Second text value.
        threshold: Minimum score required for a match.
        profile: Normalization profile.
        strategy: Score-selection strategy.
    

    Find and rank candidate strings most similar to a query.

    Args:
        query: Text to search for.
        choices: Candidate values.
        limit: Maximum number of matches.
        threshold: Minimum similarity score.
        profile: Normalization profile.
        strategy: Score-selection strategy.
    

    Group strings that probably represent the same entity.

    Useful for:
    - company-name deduplication
    - customer-name cleanup
    - product catalogue cleanup
    - contact-list cleanup

    Args:
        values: Values to examine.
        threshold: Minimum score for joining a group.
        profile: Normalization profile.
        strategy: Score-selection strategy.
    

    Compare two strings and explain why they match or differ.

    Args:
        first: First text value.
        second: Second text value.
        threshold: Minimum score required for a match.
        profile: Normalization profile.
        strategy: Score-selection strategy.