1# =============================================================================2# Job·Ka — Groupe KA3# Auteur : Simon-Pierre Boucher4# Contact : contact@spboucher.ai5# Fichier : tests/test_normalize.py6# Rôle : Tests unitaires de la couche de normalisation (salaires, dates,7# mode/type d'emploi, lieux, catégories)8# Créé : 2026-08-17 Modifié : 2026-08-179# =============================================================================10import datetime as dt1112from jobka.normalize import (13 categorize,14 clean_html,15 is_quebec_location,16 parse_date,17 parse_employment_type,18 parse_location,19 parse_salary,20 parse_work_mode,21 salary_to_hourly,22 salary_to_yearly,23)2425TODAY = dt.date(2026, 8, 17)262728# --- salaires ----------------------------------------------------------------2930def test_salary_hourly_range():31 assert parse_salary("25,00 $ à 30,00 $ de l'heure") == (25.0, 30.0, "hour")323334def test_salary_hourly_english():35 assert parse_salary("$22.50 - $27.75 per hour") == (22.5, 27.75, "hour")363738def test_salary_yearly_thousands():39 assert parse_salary("70 000 $ à 90 000 $ par année") == (70000.0, 90000.0, "year")404142def test_salary_yearly_english_commas():43 assert parse_salary("$85,000 - $105,000 annually") == (85000.0, 105000.0, "year")444546def test_salary_k_notation():47 assert parse_salary("80k$ - 100k$ /an") == (80000.0, 100000.0, "year")484950def test_salary_single_value():51 lo, hi, unit = parse_salary("Salaire : 28,50 $/h")52 assert (lo, hi, unit) == (28.5, 28.5, "hour")535455def test_salary_infers_unit_from_magnitude():56 assert parse_salary("55 000 $")[2] == "year"57 assert parse_salary("26,75 $")[2] == "hour"585960def test_salary_rejects_hours_of_work():61 # « 35 heures/semaine » n'est PAS un salaire62 assert parse_salary("35 heures par semaine") == (None, None, None)636465def test_salary_rejects_empty_and_garbage():66 assert parse_salary("") == (None, None, None)67 assert parse_salary("Salaire concurrentiel") == (None, None, None)686970def test_salary_conversions():71 assert salary_to_yearly(25.0, "hour") == 52000.072 assert salary_to_hourly(52000.0, "year") == 25.073 assert salary_to_yearly(5000.0, "month") == 60000.074 assert salary_to_yearly(None, "hour") is None757677# --- dates ---------------------------------------------------------------------7879def test_date_iso_passthrough():80 assert parse_date("2026-09-01") == "2026-09-01"81 assert parse_date("2026-06-16T05:06:55-04:00") == "2026-06-16"828384def test_date_epoch_ms():85 # Lever : createdAt en millisecondes86 assert parse_date(1755000000000) == dt.date.fromtimestamp(1755000000).isoformat()878889def test_date_french():90 assert parse_date("1er juillet 2026", TODAY) == "2026-07-01"91 assert parse_date("déc. 2026", TODAY) == "2026-12-01"929394def test_date_english():95 assert parse_date("July 1st, 2026", TODAY) == "2026-07-01"969798def test_date_relative():99 assert parse_date("Posted 3 Days Ago", TODAY) == "2026-08-14"100 assert parse_date("il y a 2 semaines", TODAY) == "2026-08-03"101 assert parse_date("Posted Today", TODAY) == "2026-08-17"102 assert parse_date("Posted Yesterday", TODAY) == "2026-08-16"103104105def test_date_never_invented():106 assert parse_date("bientôt") is None107 assert parse_date("") is None108 assert parse_date(None) is None109110111# --- mode / type ---------------------------------------------------------------112113def test_work_mode():114 assert parse_work_mode("Télétravail") == "teletravail"115 assert parse_work_mode("Remote - QC") == "teletravail"116 assert parse_work_mode("Mode hybride (3 jours au bureau)") == "hybride"117 assert parse_work_mode("Sur place") == "presentiel"118 assert parse_work_mode("Montréal") is None119120121def test_work_mode_hybrid_wins():122 assert parse_work_mode("hybride avec télétravail partiel") == "hybride"123124125def test_employment_type():126 assert parse_employment_type("Temps plein") == "temps_plein"127 assert parse_employment_type("Full time") == "temps_plein"128 assert parse_employment_type("Permanent") == "temps_plein"129 assert parse_employment_type("Temps partiel") == "temps_partiel"130 assert parse_employment_type("Contrat 12 mois") == "contractuel"131 assert parse_employment_type("Stage été 2027") == "stage"132 assert parse_employment_type("Emploi saisonnier") == "saisonnier"133 assert parse_employment_type("") is None134135136# --- lieux -----------------------------------------------------------------------137138def test_location_montreal():139 loc = parse_location("Montréal, QC, Canada")140 assert loc["city"] == "Montréal"141 assert loc["region"] == "Québec"142143144def test_location_quebec_city_english():145 assert parse_location("Quebec City, Quebec")["city"] == "Québec"146147148def test_location_postal_code():149 loc = parse_location("1500 rue Exemple, Lévis G6V 4Z2")150 assert loc["postal_code"] == "G6V 4Z2"151 assert loc["city"] == "Lévis"152153154def test_is_quebec_location():155 assert is_quebec_location("Montréal, QC")156 assert is_quebec_location("Saint-Laurent, Quebec, Canada")157 assert is_quebec_location("Sherbrooke")158 assert not is_quebec_location("Toronto, ON")159 assert not is_quebec_location("Richmond, BC")160 assert not is_quebec_location("Paris, IDF, France")161 assert not is_quebec_location("")162163164# --- catégories / HTML ------------------------------------------------------------165166def test_categorize():167 assert categorize("Développeur logiciel senior") == "TI"168 assert categorize("Infirmière clinicienne") == "Santé"169 assert categorize("Conseiller en ressources humaines") == "RH"170 assert categorize("Poste mystère") == ""171172173def test_clean_html():174 html = "<p>Bonjour !</p><ul><li>Point 1</li><li>Point 2</li></ul>"175 txt = clean_html(html)176 assert "Bonjour !" in txt177 assert "Point 1" in txt and "Point 2" in txt178 assert "<" not in txt179180181# --- Phase 2 (2026-08-19) : enrichissement ------------------------------------182183def test_clean_html_double_escaped():184 # contenu sur-échappé (Greenhouse/Lever) : entités dans les entités185 html = "<p>Bonjour&nbsp;!</p><li>Point</li>"186 txt = clean_html(html)187 assert "<" not in txt and " " not in txt188 assert "Bonjour" in txt and "Point" in txt189190191def test_sanitize_salary_month_is_actually_year():192 from jobka.normalize import sanitize_salary193 # « 75 000 $/mois » : hors bornes mensuelles -> réinterprété en annuel194 assert sanitize_salary(75000, 82500, "month") == (75000, 82500, "year")195196197def test_sanitize_salary_hour_is_actually_year():198 from jobka.normalize import sanitize_salary199 assert sanitize_salary(57200, None, "hour") == (57200, None, "year")200201202def test_sanitize_salary_plausible_untouched():203 from jobka.normalize import sanitize_salary204 assert sanitize_salary(25.0, 30.0, "hour") == (25.0, 30.0, "hour")205206207def test_sanitize_salary_absurd_dropped():208 from jobka.normalize import sanitize_salary209 assert sanitize_salary(2.0, 3.0, "hour") == (None, None, None)210211212def test_salary_from_text_skips_funding():213 from jobka.normalize import salary_from_text214 lo, _, _, _ = salary_from_text(215 "Nous offrons 600 000 $ en financement de pré-amorçage aux fondateurs.")216 assert lo is None217218219def test_salary_from_text_decimal_amounts():220 from jobka.normalize import salary_from_text221 lo, hi, unit, _ = salary_from_text(222 "The annual salary range is $82,500.00 to $137,500.00 per year.")223 assert (lo, hi, unit) == (82500.0, 137500.0, "year")224225226def test_salary_from_text_skips_expense():227 from jobka.normalize import salary_from_text228 lo, _, _, _ = salary_from_text(229 "You get a 250$ BEHappy expense reimbursement to buy what you want.")230 assert lo is None231232233def test_detect_language_fr():234 from jobka.normalize import detect_language235 assert detect_language(236 "Développeur logiciel",237 "Nous cherchons une personne pour rejoindre notre équipe dans nos "238 "bureaux de Montréal. Le poste est pour vous si vous aimez le défi "239 "et les tâches variées avec des exigences claires.") == "fr"240241242def test_detect_language_en():243 from jobka.normalize import detect_language244 assert detect_language(245 "Software Developer",246 "We are looking for someone to join our team in Montreal. You will "247 "work with the best and your skills will grow as we build this "248 "product for the world.") == "en"249250251def test_detect_language_too_short():252 from jobka.normalize import detect_language253 assert detect_language("Poste", "") == ""254255256def test_extract_benefits_block():257 from jobka.normalize import extract_benefits258 desc = ("Description du poste.\n\nAvantages\n"259 "• Assurances collectives\n• REER avec cotisation de l'employeur\n"260 "• 4 semaines de vacances\n\nComment postuler\nEnvoyez votre CV.")261 ben = extract_benefits(desc)262 assert "Assurances collectives" in ben263 assert "4 semaines de vacances" in ben264 assert len(ben) == 3265266267def test_extract_benefits_none_without_block():268 from jobka.normalize import extract_benefits269 assert extract_benefits("Un poste intéressant sans liste structurée.") == []270271272def test_clean_title_requisition_code():273 from jobka.normalize import clean_title274 assert clean_title("Analyste financier - R-12345") == "Analyste financier"275 assert clean_title("[REQ-9876] Technicien réseau") == "Technicien réseau"276277278def test_clean_title_city_suffix():279 from jobka.normalize import clean_title280 assert clean_title("Développeur Python - Montréal", "Montréal") == \281 "Développeur Python"282 # ville différente : suffixe conservé283 assert clean_title("Développeur Python - Montréal", "Laval") == \284 "Développeur Python - Montréal"285286287def test_clean_title_all_caps():288 from jobka.normalize import clean_title289 t = clean_title("CONSEILLER RH ET ACQUISITION DE TALENTS")290 assert t != t.upper()291 assert "RH" in t292293294def test_region_for_city():295 from jobka.regions import region_for_city296 assert region_for_city("Montréal") == "Montréal"297 assert region_for_city("Québec") == "Capitale-Nationale"298 assert region_for_city("Lévis") == "Chaudière-Appalaches"299 assert region_for_city("Dorval") == "Montréal"300 assert region_for_city("Ville inconnue XYZ") == ""301302303def test_quality_issues():304 from jobka.quality import issues305 ok = issues(title="Développeur", description="x" * 100, city="Montréal",306 date_posted="2026-08-01", date_deadline="2026-09-01",307 salary_hour_min=25.0, salary_hour_max=30.0,308 salary_year_min=None, salary_year_max=None,309 salary_unit="hour")310 assert ok == []311 bad = issues(title="", description="court", city="Toronto",312 date_posted="2026-08-01", date_deadline="2026-07-01",313 salary_hour_min=300.0, salary_hour_max=None,314 salary_year_min=None, salary_year_max=None,315 salary_unit="hour")316 assert set(bad) == {"titre_vide", "description_courte", "ville_hors_quebec",317 "dates_incoherentes", "salaire_hors_bornes"}318319320def test_quality_high_yearly_not_hourly_capped():321 from jobka.quality import issues322 # 457 500 $/an (~ 220 $/h equiv.) : legitime, sous la borne annuelle 500 k$323 assert issues(title="Directeur", description="x" * 100, city="Montréal",324 date_posted=None, date_deadline=None,325 salary_hour_min=131.97, salary_hour_max=219.95,326 salary_year_min=274500.0, salary_year_max=457500.0,327 salary_unit="year") == []328