"""International coverage tests for the format standardizer. Covers gaps surfaced by the i18n review: - Date locales: PT, IT, NL, RU + weekday recognition. - Date formats: ISO 8601 week date / ordinal date, RFC 2822, CJK separators, fullwidth digits, named-timezone resolution. - Two-digit year cutoff customization. - Names: Arabic / Hebrew particles, multi-language titles, East Asian honorific suffixes, family_first comma-reversal skip. - Currency: extended symbol coverage (Asian, Latin American, African currencies), extended ISO code list. - Address: Canadian provinces, UK postcode, Australian states, German Bundesland, international PO Box variants. - Email: BIDI / RTL override stripping (security). """ from __future__ import annotations import pandas as pd import pytest from src.core.format_standardize import ( standardize_address, standardize_currency, standardize_date, standardize_email, standardize_name, ) # --------------------------------------------------------------------------- # Dates # --------------------------------------------------------------------------- class TestDateLocales: @pytest.mark.parametrize("inp,want", [ ("15 janeiro 2024", "2024-01-15"), # PT ("15 fevereiro 2024", "2024-02-15"), ("15 dezembro 2024", "2024-12-15"), ("15 gennaio 2024", "2024-01-15"), # IT ("15 marzo 2024", "2024-03-15"), ("15 dicembre 2024", "2024-12-15"), ("15 januari 2024", "2024-01-15"), # NL ("15 maart 2024", "2024-03-15"), ("15 januari 2024", "2024-01-15"), ("15 января 2024", "2024-01-15"), # RU ("15 декабря 2024", "2024-12-15"), ]) def test_extended_locales(self, inp, want): got, _ = standardize_date( inp, month_locales=["en", "fr", "de", "es", "pt", "it", "nl", "ru"], ) assert got == want @pytest.mark.parametrize("inp,want", [ ("lundi, 15 janvier 2024", "2024-01-15"), # FR ("Montag, 15. Januar 2024", "2024-01-15"), # DE ("lunes, 15 enero 2024", "2024-01-15"), # ES ("lunedì 15 gennaio 2024", "2024-01-15"), # IT ("segunda-feira 15 janeiro 2024", "2024-01-15"), # PT ("maandag 15 januari 2024", "2024-01-15"), # NL ]) def test_localized_weekdays(self, inp, want): got, _ = standardize_date( inp, month_locales=["en", "fr", "de", "es", "pt", "it", "nl"], ) assert got == want class TestDateExtendedFormats: def test_iso_week_date(self): got, _ = standardize_date("2024-W03-1") assert got == "2024-01-15" def test_iso_ordinal(self): got, _ = standardize_date("2024-015") assert got == "2024-01-15" def test_rfc2822(self): got, _ = standardize_date("Mon, 15 Jan 2024 10:30:00") assert got == "2024-01-15" def test_cjk_japanese(self): got, _ = standardize_date("2024年01月15日") assert got == "2024-01-15" def test_fullwidth_digits(self): got, _ = standardize_date("２０２４/０１/１５") assert got == "2024-01-15" class TestNamedTimezones: @pytest.mark.parametrize("tz", ["EST", "PST", "JST", "GMT", "CET", "IST"]) def test_named_tz_resolves(self, tz): got, _ = standardize_date(f"2024-01-15 10:30:00 {tz}") assert got == "2024-01-15" class TestTwoDigitYearCutoff: def test_default_cutoff_69(self): # year 24 → 2024 got, _ = standardize_date("1/15/24") assert got == "2024-01-15" # year 70 → 1970 got, _ = standardize_date("1/15/70") assert got == "1970-01-15" def test_lowered_cutoff_for_birth_years(self): # cutoff=10 → year 24 falls in 1925-2010 mapping got, _ = standardize_date("1/15/24", two_digit_year_cutoff=10) assert got == "1924-01-15" # --------------------------------------------------------------------------- # Names # --------------------------------------------------------------------------- class TestNameParticles: @pytest.mark.parametrize("inp,want", [ ("ahmed bin salman", "Ahmed bin Salman"), ("abdullah ibn rashid", "Abdullah ibn Rashid"), ("ali abu bakr", "Ali abu Bakr"), ("david ben gurion", "David ben Gurion"), ("mohammed al-rashid", "Mohammed al-Rashid"), ("omar el-sayed", "Omar el-Sayed"), ]) def test_arabic_hebrew_particles(self, inp, want): got, _ = standardize_name(inp) assert got == want class TestNameTitles: @pytest.mark.parametrize("inp,want", [ ("Herr Hans Schmidt", "Herr Hans Schmidt"), ("Frau Anna Müller", "Frau Anna Müller"), ("M. Pierre Dupont", "M Pierre Dupont"), ("Mme Marie Dubois", "Mme Marie Dubois"), ("Sr. Juan Pérez", "Sr Juan Pérez"), ("Sra. Maria González", "Sra Maria González"), ("Sig. Marco Rossi", "Sig Marco Rossi"), ]) def test_multilang_titles(self, inp, want): got, _ = standardize_name(inp) assert got == want class TestEastAsianHonorifics: @pytest.mark.parametrize("inp", [ "Tanaka-san", "Suzuki-sama", "Sato-kun", "Kohaku-chan", "Lee-ssi", "Park-nim", ]) def test_honorific_preserved_lowercase(self, inp): got, _ = standardize_name(inp) # Honorific suffix stays lowercase assert got == inp.split("-")[0].title() + "-" + inp.split("-")[1].lower() class TestFamilyFirst: def test_skips_comma_reversal(self): # Default: comma reversal flips family-first into Western order got_default, _ = standardize_name("Kim, Min-jae") # Family-first preserves the comma form (per-column signal) got_ff, _ = standardize_name("Kim, Min-jae", family_first=True) assert got_default != got_ff assert got_ff.startswith("Kim,") # --------------------------------------------------------------------------- # Currency # --------------------------------------------------------------------------- class TestCurrencySymbols: @pytest.mark.parametrize("inp,want", [ ("฿1,234.56", "1234.56"), # THB ("₫50000", "50000"), # VND ("₮100", "100"), # MNT ("₴500", "500"), # UAH ("₦5,000", "5000"), # NGN ("₱1,234.56", "1234.56"), # PHP ("₲100000", "100000"), # PYG ("﷼500", "500"), # SAR (ambiguous; mapped to SAR) ("₨1,234", "1234"), # PKR ("₵100", "100"), # GHS ]) def test_extended_symbol_coverage(self, inp, want): got, _ = standardize_currency(inp) assert got == want class TestCurrencyCodes: @pytest.mark.parametrize("code", [ "SAR", "AED", "QAR", "ARS", "EGP", "IDR", "MYR", "PHP", "THB", "VND", "PKR", "BDT", "HUF", "CZK", "RON", "UAH", ]) def test_iso_code_recognized(self, code): got, _ = standardize_currency(f"1234.56 {code}") assert got == "1234.56" # --------------------------------------------------------------------------- # Addresses # --------------------------------------------------------------------------- class TestCanadianAddresses: def test_province_name_to_code(self): got, _ = standardize_address( "1 Yonge St, Toronto, Ontario M5E 1W7", expand=False, ) assert "ON" in got assert "Ontario" not in got def test_quebec_with_accent(self): got, _ = standardize_address( "1 Rue Sherbrooke, Montréal, Québec H2Y 1A1", expand=False, ) assert "QC" in got def test_province_code_preserved_after_lowercase(self): got, _ = standardize_address( "1 yonge st, toronto, on m5e 1w7", expand=False, ) assert "ON" in got class TestUKAddresses: def test_postcode_address_passes_through(self): got, _ = standardize_address( "10 Downing Street, London, SW1A 2AA", expand=False, ) assert "SW1A 2AA" in got def test_lowercase_postcode_preserved_with_caps(self): got, _ = standardize_address( "10 downing street, london, sw1a 2aa", expand=False, ) # UK postcodes get title-cased as the rest of the address; # SW1A 2AA letters aren't in the state-code set so we accept # "Sw1a 2Aa" as the title-case fallback. assert "London" in got class TestAustralianAddresses: def test_state_name_to_code(self): got, _ = standardize_address( "1 George St, Sydney, New South Wales 2000", expand=False, ) assert "NSW" in got assert "New South Wales" not in got def test_state_code_preserved(self): got, _ = standardize_address( "1 collins st, melbourne, vic 3000", expand=False, ) assert "VIC" in got class TestGermanAddresses: def test_bundesland_name_to_code(self): got, _ = standardize_address( "Hauptstr 1, München, Bayern 80331", expand=False, ) assert "BY" in got assert "Bayern" not in got class TestInternationalPOBox: @pytest.mark.parametrize("inp", [ "Postfach 12345, München, BY 80331", # DE "Boîte postale 12, Paris 75001", # FR "Apartado 12, Madrid 28001", # ES "Casella postale 12, Roma 00100", # IT "Caixa postal 12, São Paulo 01310", # PT ]) def test_intl_po_box_normalized(self, inp): got, _ = standardize_address(inp, expand=False) assert "PO Box" in got # --------------------------------------------------------------------------- # Email — security # --------------------------------------------------------------------------- class TestEmailBidiSecurity: def test_rtl_override_stripped(self): # U+202E (Right-to-Left Override) inside email — common phishing # vector. After strip, the address is just the legitimate one. malicious = "alice‮@example.com" got, _ = standardize_email(malicious) assert got == "alice@example.com" assert "‮" not in got def test_lrm_stripped(self): # Left-to-Right Mark, also strippable. s = "alice‎@example.com" got, _ = standardize_email(s) assert got == "alice@example.com" def test_rtl_isolate_stripped(self): s = "alice⁦@⁩example.com" got, _ = standardize_email(s) assert got == "alice@example.com" # --------------------------------------------------------------------------- # Pipeline integration — end-to-end with intl options # --------------------------------------------------------------------------- class TestPipelineIntl: def test_standardize_options_carry_intl_flags(self): from src.core.format_standardize import ( FieldType, StandardizeOptions, standardize_dataframe, ) df = pd.DataFrame({ "name": ["Tanaka-san", "Kim, Min-jae"], "date": ["15 janeiro 2024", "Mon, 15 Jan 2024 10:30:00"], "addr": [ "Hauptstr 1, München, Bayern 80331", "1 Yonge St, Toronto, Ontario M5E 1W7", ], }) opts = StandardizeOptions( column_types={ "name": FieldType.NAME, "date": FieldType.DATE, "addr": FieldType.ADDRESS, }, date_month_locales=["en", "fr", "de", "es", "pt", "it", "nl", "ru"], address_expand=False, name_family_first=True, ) result = standardize_dataframe(df, opts) out = result.standardized_df # Names: honorific preserved, family-first comma not reversed assert out.loc[0, "name"] == "Tanaka-san" assert out.loc[1, "name"].startswith("Kim,") # Dates: PT month + RFC 2822 both → 2024-01-15 assert out.loc[0, "date"] == "2024-01-15" assert out.loc[1, "date"] == "2024-01-15" # Addresses: DE + CA both have state codes substituted assert "BY" in out.loc[0, "addr"] assert "ON" in out.loc[1, "addr"]