from ..tokenizer_exceptions import BASE_EXCEPTIONS from ...symbols import ORTH, NORM from ...util import update_exc _exc = {} for exc_data in [ {ORTH: "aprox.", NORM: "aproximadament"}, {ORTH: "pàg.", NORM: "pàgina"}, {ORTH: "p.ex.", NORM: "per exemple"}, {ORTH: "gen.", NORM: "gener"}, {ORTH: "feb.", NORM: "febrer"}, {ORTH: "abr.", NORM: "abril"}, {ORTH: "jul.", NORM: "juliol"}, {ORTH: "set.", NORM: "setembre"}, {ORTH: "oct.", NORM: "octubre"}, {ORTH: "nov.", NORM: "novembre"}, {ORTH: "dec.", NORM: "desembre"}, {ORTH: "Dr.", NORM: "doctor"}, {ORTH: "Sr.", NORM: "senyor"}, {ORTH: "Sra.", NORM: "senyora"}, {ORTH: "Srta.", NORM: "senyoreta"}, {ORTH: "núm", NORM: "número"}, {ORTH: "St.", NORM: "sant"}, {ORTH: "Sta.", NORM: "santa"}, {ORTH: "'l"}, {ORTH: "'ls"}, {ORTH: "'m"}, {ORTH: "'n"}, {ORTH: "'ns"}, {ORTH: "'s"}, {ORTH: "'t"}, ]: _exc[exc_data[ORTH]] = [exc_data] # Times _exc["12m."] = [{ORTH: "12"}, {ORTH: "m.", NORM: "p.m."}] for h in range(1, 12 + 1): for period in ["a.m.", "am"]: _exc[f"{h}{period}"] = [{ORTH: f"{h}"}, {ORTH: period, NORM: "a.m."}] for period in ["p.m.", "pm"]: _exc[f"{h}{period}"] = [{ORTH: f"{h}"}, {ORTH: period, NORM: "p.m."}] TOKENIZER_EXCEPTIONS = update_exc(BASE_EXCEPTIONS, _exc)