time.py 1.8 KB

123456789101112131415161718192021222324252627282930313233343536373839
  1. import pynini
  2. from fun_text_processing.text_normalization.en.graph_utils import DAMO_ALPHA, DAMO_DIGIT, GraphFst, delete_space
  3. from pynini.lib import pynutil
  4. class TimeFst(GraphFst):
  5. """
  6. Finite state transducer for verbalizing time, e.g.
  7. time { hours: "8" minutes: "30" zone: "e s t" } -> 08:30 Uhr est
  8. time { hours: "8" } -> 8 Uhr
  9. time { hours: "8" minutes: "30" seconds: "10" } -> 08:30:10 Uhr
  10. """
  11. def __init__(self, deterministic: bool = True):
  12. super().__init__(name="time", kind="verbalize", deterministic=deterministic)
  13. add_leading_zero_to_double_digit = (DAMO_DIGIT + DAMO_DIGIT) | (pynutil.insert("0") + DAMO_DIGIT)
  14. hour = pynutil.delete("hours: \"") + pynini.closure(DAMO_DIGIT, 1) + pynutil.delete("\"")
  15. minute = pynutil.delete("minutes: \"") + pynini.closure(DAMO_DIGIT, 1) + pynutil.delete("\"")
  16. second = pynutil.delete("seconds: \"") + pynini.closure(DAMO_DIGIT, 1) + pynutil.delete("\"")
  17. zone = (
  18. pynutil.delete("zone: \"") + pynini.closure(DAMO_ALPHA + delete_space) + DAMO_ALPHA + pynutil.delete("\"")
  19. )
  20. optional_zone = pynini.closure(pynini.accep(" ") + zone, 0, 1)
  21. graph = (
  22. delete_space
  23. + pynutil.insert(":")
  24. + (minute @ add_leading_zero_to_double_digit)
  25. + pynini.closure(delete_space + pynutil.insert(":") + (second @ add_leading_zero_to_double_digit), 0, 1)
  26. + pynutil.insert(" Uhr")
  27. + optional_zone
  28. )
  29. graph_h = hour + pynutil.insert(" Uhr") + optional_zone
  30. graph_hm = hour @ add_leading_zero_to_double_digit + graph
  31. graph_hms = hour @ add_leading_zero_to_double_digit + graph
  32. final_graph = graph_hm | graph_hms | graph_h
  33. self.fst = self.delete_tokens(final_graph).optimize()