def test_porter_stem():
data = [("chocolates", "chocol"), ("chocolatey", "chocolatei"),
("choco", "choco"), (None, None)]
df = spark.createDataFrame(data, ["word", "expected"])
actual_df = df.withColumn("word_porter_stem",
ceja.porter_stem(col("word")))
assert_column_equality(actual_df, "word_porter_stem", "expected")
def test_match_rating_codex():
data = [("jellyfish", "JLYFSH"), ("li", "L"), ("luisa", "LS"),
(None, None)]
df = spark.createDataFrame(data, ["word", "expected"])
actual_df = df.withColumn("word_match_rating_codex",
ceja.match_rating_codex(col("word")))
assert_column_equality(actual_df, "word_match_rating_codex", "expected")
def test_hamming_distance():
data = [("jellyfish", "smellyfish", 9), ("li", "lee", 2),
("luisa", "bruna", 4), (None, None, None)]
df = spark.createDataFrame(data, ["word1", "word2", "expected"])
actual_df = df.withColumn(
"word_hamming_distance",
ceja.hamming_distance(col("word1"), col("word2")))
assert_column_equality(actual_df, "word_hamming_distance", "expected")
def test_match_rating_comparison():
data = [("mat", "matt", True), ("there", "their", True),
("luisa", "bruna", False), (None, None, None)]
df = spark.createDataFrame(data, ["word1", "word2", "expected"])
actual_df = df.withColumn(
"word_match_rating_comparison",
ceja.match_rating_comparison(col("word1"), col("word2")))
assert_column_equality(actual_df, "word_match_rating_comparison",
"expected")
def test_damerau_levenshtein_distance():
data = [("jellyfish", "smellyfish", 2), ("li", "lee", 2),
("luisa", "bruna", 4), (None, None, None)]
df = spark.create_df(data, [("word1", StringType(), True),
("word2", StringType(), True),
("expected", IntegerType(), True)])
actual_df = df.withColumn(
"word_damerau_levenshtein_distance",
ceja.damerau_levenshtein_distance(col("word1"), col("word2")))
assert_column_equality(actual_df, "word_damerau_levenshtein_distance",
"expected")
def test_remove_non_word_characters(spark_session):
occ_pr = OccupancyProcessing(spark_session)
data = [("jo&&se", "jose"), ("**li**", "li"), ("77,990", "77990"),
(None, None)]
df = spark_session.createDataFrame(data,
["name", "expected_name"]).withColumn(
"clean_name",
occ_pr.remove_non_word_characters(
F.col("name")))
assert_column_equality(df, "clean_name", "expected_name")
def test_remove_special_letter_native(spark):
source_data = [
("toto@123", "*****@*****.**", "toto123"),
("titi$___", "*****@*****.**", "titi___"),
("!!##tata", "*****@*****.**", "tata"),
]
source_df = spark.createDataFrame(source_data,
["name", "mail", "expected_name"])
actual_df = remove_special_letter_native(source_df, "name")
actual_df.show()
assert_column_equality(actual_df, "cleaned_name", "expected_name")
def test_metaphone():
data = [("jellyfish", "JLFX"), ("li", "L"), ("luisa", "LS"),
("Klumpz", "KLMPS"), ("Clumps", "KLMPS"), (None, None)]
df = spark.createDataFrame(data, ["word", "expected"])
actual_df = df.withColumn("word_metaphone", ceja.metaphone(col("word")))
assert_column_equality(actual_df, "word_metaphone", "expected")
def test_nysiis():
data = [("jellyfish", "JALYF"), ("li", "L"), ("luisa", "LAS"),
(None, None)]
df = spark.createDataFrame(data, ["word", "expected"])
actual_df = df.withColumn("word_nysiis", ceja.nysiis(col("word")))
assert_column_equality(actual_df, "word_nysiis", "expected")
