表記に揺らぎがあるリストの項目をPythonで名寄せして対応づける

from collections.abc import Sequence
import numpy as np
import pandas as pd
import Levenshtein


def get_ratios(ser1: Sequence, ser2: Sequence):
    ratios = []
    for t1 in ser1:
        row = []
        for t2 in ser2:
            row.append(Levenshtein.ratio(t1, t2))
        ratios.append(row)
    return pd.DataFrame(ratios, index=ser1, columns=ser2)


def make_series(arr, name=None):
    if type(arr) == pd.core.series.Series:
        return arr
    else:
        return pd.Series(arr, name=name)


def sort_key(arr):
    lst = list(pd.concat([ser1, ser2]))
    return [lst.index(t) if t in lst else None for t in arr]


def align_lists(ser1, ser2, threshold=0):
    ser1 = make_series(ser1, name="series1")
    ser2 = make_series(ser2, name="series2")

    if len(ser1) < len(ser2):
        ser_s, ser_l = ser1, ser2
    else:
        ser_s, ser_l = ser2, ser1

    ratios = get_ratios(ser_l, ser_s)
    common = pd.DataFrame(
        np.vstack([ser_l.iloc[np.argmax(ratios, axis=0)], ser_s]),
        index=[ser_l.name, ser_s.name],
    ).T
    ser_s_in_pairs = ratios.max(axis=0)[ratios.max(axis=0) > threshold]

    pairs = (
        pd.DataFrame(ser_s_in_pairs)
        .assign(
            **{
                ser_l.name: ser_l.iloc[
                    np.argmax(ratios.loc[:, ser_s_in_pairs.index], axis=0)
                ].values
            }
        )
        .reset_index()
        .rename(columns={0: "score"})
    )
    table = (
        pd.concat(
            [
                pairs,
                pd.DataFrame(ser_l[~ser_l.isin(pairs[ser_l.name])]).assign(score=0),
                pd.DataFrame(ser_s[~ser_s.isin(pairs[ser_s.name])]).assign(score=0),
            ],
            ignore_index=True,
        )
        .sort_values([ser1.name, ser2.name], key=sort_key)
        .fillna("")
        .reset_index(drop=True)[[ser1.name, ser2.name, "score"]]
    )
    return table

list1 = pd.Series(
    [
        "Black Panther: Wakanda Forever",
        "Barbie",
        "The Super Mario Brothers Movie",
        "Spider-Man: Across the Spider-Verse",
        "Guardians of the Galaxy Vol. 3",
        "Oppenheimer",
        "The Little Mermaid",
        "Ant-Man and the Wasp: Quantumania",
        "John Wick: Chapter 4",
        "Sound of Freedom",
        "Taylor Swift: The Eras Tour",
    ],
    name="List 1",
)

list2 = pd.Series(
    [
        "The Little Mermaid",
        "Taylor Swift's Eras Tour",
        "Ant-Man and the Wasp: Quantumania",
        "The Super Mario Bros. Movie",
        "Top Gun: Maverick",
        "Oppenheimer",
        "Barbecue",
        "Guardians of the Galaxy Volume Three",
        "John Wick 4.0",
        "sound of freedom",
        "Avatar: The Way of Water",
    ],
    name="List 2",
)

align_lists(list1,list2,threshold=0.6)