def __init__(self, cardinal: GraphFst, decimal: GraphFst):
super().__init__(name="measure", kind="classify")
cardinal_graph = cardinal.graph_no_exception
graph_digit = pynini.string_file(
get_abs_path("data/numbers/digit.tsv"))
graph_four = pynini.cross("tư", "4")
graph_one = pynini.cross("mốt", "1")
graph_half = pynini.cross("rưỡi", "5")
graph_unit = pynini.string_file(get_abs_path("data/measurements.tsv"))
graph_unit_singular = pynini.invert(graph_unit) # singular -> abbr
optional_graph_negative = pynini.closure(
pynutil.insert("negative: ") +
pynini.cross(pynini.union("âm", "trừ"), '"true"') +
delete_extra_space,
0,
1,
)
unit_singular = convert_space(graph_unit_singular)
unit_misc = pynutil.insert("/") + pynutil.delete(
"trên") + delete_space + convert_space(graph_unit_singular)
unit_singular = (pynutil.insert('units: "') +
(unit_singular | unit_misc | pynutil.add_weight(
unit_singular + delete_space + unit_misc, 0.01)) +
pynutil.insert('"'))
subgraph_decimal = (pynutil.insert("decimal { ") +
optional_graph_negative +
decimal.final_graph_wo_negative +
pynutil.insert(" }") + delete_extra_space +
unit_singular)
subgraph_cardinal = (pynutil.insert("cardinal { ") +
optional_graph_negative +
pynutil.insert('integer: "') + cardinal_graph +
pynutil.insert('"') + pynutil.insert(" }") +
delete_extra_space + unit_singular)
fraction_graph = (delete_extra_space +
pynutil.insert('fractional_part: "') +
(graph_digit | graph_half | graph_one | graph_four) +
pynutil.insert('"'))
subgraph_cardinal |= (pynutil.insert("cardinal { ") +
optional_graph_negative +
pynutil.insert('integer: "') + cardinal_graph +
pynutil.insert('" }') + delete_extra_space +
unit_singular + fraction_graph)
final_graph = subgraph_decimal | subgraph_cardinal
final_graph = self.add_tokens(final_graph)
self.fst = final_graph.optimize()
def __init__(self, cardinal: GraphFst, decimal: GraphFst):
super().__init__(name="money", kind="classify")
# quantity, integer_part, fractional_part, currency
cardinal_graph = cardinal.graph_no_exception
graph_decimal_final = decimal.final_graph_wo_negative
graph_half = pynini.cross("rưỡi", "5")
unit = pynini.string_file(get_abs_path("data/currency.tsv"))
unit_singular = pynini.invert(unit)
graph_unit_singular = pynutil.insert("currency: \"") + convert_space(
unit_singular) + pynutil.insert("\"")
add_leading_zero_to_double_digit = (NEMO_DIGIT + NEMO_DIGIT) | (
pynutil.insert("0") + NEMO_DIGIT)
# twelve dollars fifty, only after integer
optional_cents_suffix = pynini.closure(
delete_extra_space + pynutil.insert("fractional_part: \"") +
(pynutil.add_weight(
cardinal_graph @ add_leading_zero_to_double_digit, -0.7)
| graph_half) + pynutil.insert("\""),
0,
1,
)
graph_integer = (pynutil.insert("integer_part: \"") + cardinal_graph +
pynutil.insert("\"") + delete_extra_space +
graph_unit_singular + optional_cents_suffix)
graph_decimal = graph_decimal_final + delete_extra_space + graph_unit_singular + optional_cents_suffix
final_graph = graph_integer | graph_decimal
final_graph = self.add_tokens(final_graph)
self.fst = final_graph.optimize()
def __init__(self):
super().__init__(name="whitelist", kind="classify")
whitelist = pynini.string_file(
get_abs_path("data/whitelist.tsv")).invert()
graph = pynutil.insert('name: "') + convert_space(
whitelist) + pynutil.insert('"')
self.fst = graph.optimize()
def __init__(self):
super().__init__(name="time", kind="classify")
# hours, minutes, seconds, suffix, zone, style, speak_period
graph_hours_to = pynini.string_file(
get_abs_path("data/time/hours_to.tsv"))
graph_minutes_to = pynini.string_file(
get_abs_path("data/time/minutes_to.tsv"))
graph_hours = pynini.string_file(get_abs_path("data/time/hours.tsv"))
graph_minutes = pynini.string_file(
get_abs_path("data/time/minutes.tsv"))
time_zone_graph = pynini.invert(
pynini.string_file(get_abs_path("data/time/time_zone.tsv")))
graph_half = pynini.cross("rưỡi", "30")
oclock = pynini.cross("giờ", "")
minute = pynini.cross("phút", "")
optional_minute = pynini.closure(delete_space + minute, 0, 1)
second = pynini.cross("giây", "")
final_graph_hour = pynutil.insert(
"hours: \"") + graph_hours + pynutil.insert(
"\"") + delete_space + oclock
graph_minute = graph_minutes + optional_minute
graph_second = graph_minute + delete_space + second
final_time_zone_optional = pynini.closure(
delete_space + insert_space + pynutil.insert("zone: \"") +
convert_space(time_zone_graph) + pynutil.insert("\""),
0,
1,
)
graph_hm = (final_graph_hour + delete_extra_space +
pynutil.insert("minutes: \"") +
(graph_minute | graph_half) + pynutil.insert("\""))
graph_hms = graph_hm + delete_extra_space + pynutil.insert(
"seconds: \"") + graph_second + pynutil.insert("\"")
graph_ms = (pynutil.insert("minutes: \"") + graph_minute +
pynutil.insert("\"") + delete_extra_space +
pynutil.insert("seconds: \"") +
(graph_second | graph_half) + pynutil.insert("\""))
graph_hours_to_component = graph_hours @ graph_hours_to
graph_minutes_to_component = graph_minutes @ graph_minutes_to
graph_time_to = (pynutil.insert("hours: \"") +
graph_hours_to_component + pynutil.insert("\"") +
delete_space + oclock + delete_space +
pynutil.delete("kém") + delete_extra_space +
pynutil.insert("minutes: \"") +
graph_minutes_to_component + optional_minute +
pynutil.insert("\""))
final_graph = (final_graph_hour | graph_hm
| graph_hms) + final_time_zone_optional
final_graph |= graph_ms
final_graph |= graph_time_to
final_graph = self.add_tokens(final_graph)
self.fst = final_graph.optimize()