chitai/backend/tests/unit/test_matching.py

"""Tests for the normalization behind book-level duplicate detection."""

import pytest

from chitai.services.matching import (
    format_author_name,
    normalize_author,
    normalize_identifier,
    normalize_title,
)
from chitai.services.metadata_extractor import parse_identifier
from chitai.services.utils import isbn10_to_isbn13


class TestNormalizeTitle:
    """Two copies of one book rarely agree on how the title is written."""

    @pytest.mark.parametrize(
        ("title", "expected"),
        [
            ("The Metamorphosis", "metamorphosis"),
            ("Metamorphosis", "metamorphosis"),
            ("METAMORPHOSIS", "metamorphosis"),
            ("A Tale of Two Cities", "tale of two cities"),
            ("An Enquiry", "enquiry"),
            # Accents, punctuation and ampersands are spelling, not identity.
            ("Les Misérables", "les miserables"),
            ("Moby Dick; Or, The Whale", "moby dick or the whale"),
            ("Sense & Sensibility", "sense and sensibility"),
            # Bracketed asides and trailing edition noise say nothing about the book.
            ("Frankenstein (Illustrated)", "frankenstein"),
            ("Frankenstein [Kindle Edition]", "frankenstein"),
            ("Frankenstein, 2nd Edition", "frankenstein"),
            # The compact forms a cover actually carries.
            ("Building Microservices, 2E", "building microservices"),
            ("Building Microservices 2e", "building microservices"),
            ("Frankenstein 3 Ed", "frankenstein"),
            ("Dungeons & Dragons 5e", "dungeons and dragons"),
            ("Frankenstein Revised Edition", "frankenstein"),
            ("Dune Deluxe Edition Illustrated", "dune"),
            ("", ""),
            (None, ""),
        ],
    )
    def test_titles_that_should_agree(self, title: str | None, expected: str) -> None:
        assert normalize_title(title) == expected

    def test_a_qualifier_that_is_the_title_survives(self) -> None:
        """A trailing qualifier is noise; the same word at the front is the book."""
        assert normalize_title("The Illustrated Man") == "illustrated man"

    def test_normalization_never_empties_a_title(self) -> None:
        """An article-only title is not improved by having no article left."""
        assert normalize_title("The") == "the"

    @pytest.mark.parametrize(
        "title",
        ["Catch 22", "Fahrenheit 451", "Blade Runner 2049", "1984", "Apollo 13"],
    )
    def test_a_number_is_not_an_edition(self, title: str) -> None:
        """Edition stripping keys on the `e`; a bare number is part of the title."""
        assert normalize_title(title) == title.casefold()


class TestNormalizeAuthor:
    """One person, written down several ways."""

    @pytest.mark.parametrize(
        ("name", "expected"),
        [
            ("Franz Kafka", "franz kafka"),
            ("Kafka, Franz", "franz kafka"),
            ("KAFKA, FRANZ", "franz kafka"),
            ("Émile Zola", "emile zola"),
            ("Doyle, Arthur Conan", "arthur conan doyle"),
            # Runs of initials are joined, so spacing them out changes nothing.
            ("J.R.R. Tolkien", "jrr tolkien"),
            ("J. R. R. Tolkien", "jrr tolkien"),
            ("JRR Tolkien", "jrr tolkien"),
            ("", ""),
            (None, ""),
        ],
    )
    def test_names_that_should_agree(self, name: str | None, expected: str) -> None:
        assert normalize_author(name) == expected

    def test_two_people_are_not_reduced_together(self) -> None:
        """Surname plus initial would collide unrelated writers; it is not used."""
        assert normalize_author("Charles Dickens") != normalize_author("Colin Dexter")

    def test_a_list_is_left_alone(self) -> None:
        """More than one comma is a list or a suffix, and guessing does more harm."""
        assert normalize_author("Smith, John, Jr.") == "smith john jr"


class TestFormatAuthorName:
    """What gets stored and shown, as opposed to what gets compared."""

    @pytest.mark.parametrize(
        ("written", "expected"),
        [
            # A leftover separator from a `DC:creator` list.
            ("Newman, Sam;", "Sam Newman"),
            ("Sam Newman ", "Sam Newman"),
            ("  Dan   Vanderkam  ", "Dan Vanderkam"),
            # `Surname, Given` is how EPUBs file a name, not how anyone reads it.
            ("Kleppmann, Martin", "Martin Kleppmann"),
            ("Huxley, Aldous", "Aldous Huxley"),
            ("Liu, Cixin", "Cixin Liu"),
            # An extension carried in from the filename the name was read out of.
            ("Sam Newman.epub", "Sam Newman"),
            ("Franz Kafka.mobi", "Franz Kafka"),
            ("Brian W. Kernighan.epub", "Brian W. Kernighan"),
            ("", ""),
            (None, ""),
        ],
    )
    def test_names_are_tidied(self, written: str | None, expected: str) -> None:
        assert format_author_name(written) == expected

    @pytest.mark.parametrize(
        "written",
        [
            # Two people in one string. Flipping it would invent a third.
            "Dave Thomas, Andy Hunt",
            "Mark Richards, Neal Ford",
            # A compound surname is not recognised, and is left alone rather than
            # rearranged wrongly.
            "García Márquez, Gabriel",
        ],
    )
    def test_an_unrecognised_form_is_left_alone(self, written: str) -> None:
        assert format_author_name(written) == written

    def test_case_and_accents_belong_to_the_author(self) -> None:
        """Tidying removes what an extractor added; it does not correct spelling."""
        assert format_author_name("Michał Płachta.epub") == "Michał Płachta"
        assert format_author_name("Steve McConnell") == "Steve McConnell"

    @pytest.mark.parametrize(
        "written", ["Newman, Sam;", "Sam Newman.epub", "Kleppmann, Martin"]
    )
    def test_tidying_is_idempotent(self, written: str) -> None:
        """`unique_filter` tidies a name that may already be tidy; it must not drift."""
        once = format_author_name(written)
        assert format_author_name(once) == once


class TestNormalizeIdentifier:
    """Identifiers only help if the same edition produces the same key."""

    def test_isbn_10_and_isbn_13_are_one_key(self) -> None:
        assert normalize_identifier("isbn-10", "0486282112") == "isbn:9780486282114"
        assert normalize_identifier("isbn-13", "9780486282114") == "isbn:9780486282114"

    @pytest.mark.parametrize(
        "written", ["978-0-486-28211-4", "978 0 486 28211 4", "9780486282114"]
    )
    def test_formatting_is_not_part_of_an_isbn(self, written: str) -> None:
        assert normalize_identifier("isbn", written) == "isbn:9780486282114"

    def test_an_isbn_that_fails_its_checksum_is_no_evidence(self) -> None:
        assert normalize_identifier("isbn-13", "9780486282115") is None

    def test_uuids_are_refused(self) -> None:
        """Generated per build, so they only re-find what the hash check catches."""
        assert (
            normalize_identifier("uuid", "3f2b1c4e-1111-2222-3333-444455556666") is None
        )
        assert (
            normalize_identifier("urn:uuid", "3f2b1c4e-1111-2222-3333-444455556666")
            is None
        )

    def test_other_schemes_keep_their_own_key(self) -> None:
        assert normalize_identifier("asin", "B000FC0PDA") == "asin:b000fc0pda"
        assert normalize_identifier("ASIN", "b000fc0pda") == "asin:b000fc0pda"

    def test_something_too_short_is_not_evidence(self) -> None:
        """A Calibre id of "42" would otherwise pair two unrelated books."""
        assert normalize_identifier("calibre", "42") is None

    @pytest.mark.parametrize(("name", "value"), [("", "1234567"), ("asin", "")])
    def test_half_an_identifier_is_no_identifier(self, name: str, value: str) -> None:
        assert normalize_identifier(name, value) is None


class TestIsbnConversion:
    def test_isbn_10_converts_to_its_isbn_13(self) -> None:
        assert isbn10_to_isbn13("0486282112") == "9780486282114"

    def test_a_trailing_x_is_a_digit(self) -> None:
        assert isbn10_to_isbn13("043942089X") == "9780439420891"

    @pytest.mark.parametrize("isbn", ["0486282113", "9780486282114", "nonsense"])
    def test_anything_that_is_not_an_isbn_10_converts_to_nothing(
        self, isbn: str
    ) -> None:
        assert isbn10_to_isbn13(isbn) is None


class TestParseIdentifier:
    """What an EPUB writes, and what is worth storing for it."""

    @pytest.mark.parametrize(
        "written",
        [
            "9780486282114",
            "978-0-486-28211-4",
            "urn:isbn:9780486282114",
            "urn:isbn:978-0-486-28211-4",
            "ISBN:978-0-486-28211-4",
        ],
    )
    def test_isbns_survive_however_they_are_written(self, written: str) -> None:
        assert parse_identifier(written) == ("isbn-13", "9780486282114")

    def test_the_scheme_attribute_is_read_too(self) -> None:
        assert parse_identifier("0-486-28211-2", "ISBN") == ("isbn-10", "0486282112")

    def test_non_isbn_identifiers_are_kept(self) -> None:
        assert parse_identifier("urn:uuid:3f2b1c4e-1111-2222-3333-444455556666") == (
            "uuid",
            "3f2b1c4e-1111-2222-3333-444455556666",
        )
        assert parse_identifier("calibre:1234") == ("calibre", "1234")
        assert parse_identifier("B000FC0PDA", "mobi-asin") == ("asin", "B000FC0PDA")

    def test_an_unrecognised_prefix_is_part_of_the_value(self) -> None:
        """ "http://example.com/book" is not an identifier called "http"."""
        assert parse_identifier("http://www.gutenberg.org/5200") == (
            "id",
            "http://www.gutenberg.org/5200",
        )

    def test_a_declared_isbn_that_is_not_one_is_dropped(self) -> None:
        assert parse_identifier("urn:isbn:not-an-isbn") is None

    @pytest.mark.parametrize("written", ["", "   ", None])
    def test_nothing_yields_nothing(self, written: str | None) -> None:
        assert parse_identifier(written) is None