feat: integrate all English ITN rules with money, time and decimal fixes

- Add Money rule: two dollars => $2, one cent => $0.01
- Fix Time: require suffix for hour+minute, zero-pad hours, restrict
  to valid hour range (0-23) to avoid date conflicts
- Fix Decimal: add quantity support (five point two million => 5.2 million)
- Fix Money cents: pad single-digit cents (1 => 01)
- Extend _num_to_word to support 60-99

NeMo English ITN: 372/470 (79%)
All 1442 unit tests pass.

Author: pengzhendong

itn/english/inverse_normalizer.py

@@ -17,8 +17,15 @@
 
 from itn.english.rules.cardinal import Cardinal
 from itn.english.rules.char import Char
+from itn.english.rules.date import Date
 from itn.english.rules.decimal import Decimal
+from itn.english.rules.electronic import Electronic
+from itn.english.rules.measure import Measure
+from itn.english.rules.money import Money
 from itn.english.rules.ordinal import Ordinal
+from itn.english.rules.telephone import Telephone
+from itn.english.rules.time import Time
+from itn.english.rules.whitelist import Whitelist
 from tn.processor import Processor
 
 
@@ -34,10 +41,24 @@ def build_tagger_and_verbalizer(self):
         cardinal = Cardinal()
         ordinal = Ordinal(cardinal=cardinal)
         decimal = Decimal(cardinal=cardinal)
+        date = Date(cardinal=cardinal, ordinal=ordinal)
+        time = Time(cardinal=cardinal)
+        measure = Measure(cardinal=cardinal, decimal=decimal)
+        money = Money(cardinal=cardinal, decimal=decimal)
+        telephone = Telephone(cardinal=cardinal)
+        electronic = Electronic()
+        whitelist = Whitelist()
         char = Char()
 
         tagger = (
-            add_weight(ordinal.tagger, 1.0)
+            add_weight(date.tagger, 0.9)
+            | add_weight(time.tagger, 0.9)
+            | add_weight(measure.tagger, 0.95)
+            | add_weight(money.tagger, 0.9)
+            | add_weight(whitelist.tagger, 0.9)
+            | add_weight(telephone.tagger, 1.0)
+            | add_weight(electronic.tagger, 2.0)
+            | add_weight(ordinal.tagger, 1.0)
             | add_weight(decimal.tagger, 1.01)
             | add_weight(cardinal.tagger, 1.02)
             | add_weight(char.tagger, 100)
@@ -50,6 +71,13 @@ def build_tagger_and_verbalizer(self):
             cardinal.verbalizer
             | ordinal.verbalizer
             | decimal.verbalizer
+            | date.verbalizer
+            | time.verbalizer
+            | measure.verbalizer
+            | money.verbalizer
+            | telephone.verbalizer
+            | electronic.verbalizer
+            | whitelist.verbalizer
             | char.verbalizer
         ).optimize()
 

itn/english/rules/decimal.py

@@ -12,12 +12,12 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from pynini import closure, cross, string_file
+from pynini import closure, cross, string_file, union
 from pynini.lib.pynutil import delete, insert
 
 from itn.english.rules.cardinal import Cardinal
 from tn.processor import Processor
-from tn.utils import get_abs_path
+from tn.utils import get_abs_path, load_labels
 
 
 class Decimal(Processor):
@@ -45,6 +45,17 @@ def build_tagger(self):
         point = delete("point")
 
         graph = optional_negative + closure(integer_part + ds, 0, 1) + point + ds + frac_part
+
+        # quantity: "five point two million" => 5.2 million
+        quantities = load_labels(get_abs_path("../itn/english/data/numbers/thousands.tsv"))
+        quantity_names = [x[0] for x in quantities if x[0] != "thousand"]
+        quantity = union(*quantity_names)
+        quantity_graph = (
+            optional_negative + integer_part + ds + point + ds + frac_part
+            + ds + insert(' quantity: "') + quantity + insert('"')
+        )
+        graph |= quantity_graph
+
         self.tagger = self.add_tokens(graph)
 
     def build_verbalizer(self):
@@ -56,6 +67,11 @@ def build_verbalizer(self):
             + delete('"') + self.NOT_QUOTE.plus + delete('"')
         )
         optional_fractional = closure(fractional + self.DELETE_SPACE, 0, 1)
-        graph = optional_sign + optional_integer + optional_fractional
+        quantity = (
+            insert(" ") + delete('quantity:') + self.DELETE_SPACE
+            + delete('"') + self.NOT_QUOTE.plus + delete('"')
+        )
+        optional_quantity = closure(quantity + self.DELETE_SPACE, 0, 1)
+        graph = optional_sign + optional_integer + optional_fractional + optional_quantity
         self.numbers = graph
         self.verbalizer = self.delete_tokens(graph)

itn/english/rules/money.py

@@ -0,0 +1,89 @@
+# Copyright (c) 2026 Zhendong Peng (pzd17@tsinghua.org.cn)
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+from pynini import closure, cross, union
+from pynini.lib.pynutil import delete, insert
+
+from itn.english.rules.cardinal import Cardinal
+from itn.english.rules.decimal import Decimal
+from itn.english.rules.time import _num_to_word
+from tn.processor import Processor
+from tn.utils import get_abs_path, load_labels
+
+
+class Money(Processor):
+
+    def __init__(self, cardinal=None, decimal=None):
+        super().__init__(name="money", ordertype="itn")
+        self.cardinal = cardinal or Cardinal()
+        self.decimal = decimal or Decimal(cardinal=self.cardinal)
+        self.build_tagger()
+        self.build_verbalizer()
+
+    def build_tagger(self):
+        cardinal_graph = self.cardinal.graph
+        ds = delete(" ")
+
+        currency_labels = load_labels(get_abs_path("../itn/english/data/currency.tsv"))
+        currency_pairs = []
+        for symbol, name in currency_labels:
+            currency_pairs.append((name, symbol))
+            if name.endswith("s"):
+                currency_pairs.append((name + "es", symbol))
+            else:
+                currency_pairs.append((name + "s", symbol))
+        currency = union(*[cross(name, symbol) for name, symbol in currency_pairs]).optimize()
+
+        cent = cross("cent", "") | cross("cents", "")
+        magnitudes = load_labels(get_abs_path("../itn/english/data/magnitudes.tsv"))
+        magnitude = union(*[cross(name, "") for symbol, name in magnitudes])
+
+        integer_graph = (
+            insert('value: "') + cardinal_graph + insert('"')
+            + ds + insert(' currency: "') + currency + insert('"')
+        )
+        quantity_graph = (
+            insert('value: "') + cardinal_graph + insert('"')
+            + ds + insert(' quantity: "') + magnitude + insert('"')
+            + ds + insert(' currency: "') + currency + insert('"')
+        )
+        # cents: pad single digit (1-9 => 01-09)
+        cents_graph = union(*[cross(_num_to_word(x), f"{x:02d}") for x in range(1, 100) if _num_to_word(x)])
+        with_cents = (
+            insert('value: "') + cardinal_graph + insert('"')
+            + ds + insert(' currency: "') + currency + insert('"')
+            + ds + (delete("and") + ds).ques
+            + insert(' decimal: "') + cents_graph + insert('"')
+            + ds + cent
+        )
+        cents_only = (
+            insert('currency: "$" decimal: "') + cents_graph + insert('"')
+            + ds + cent
+        )
+
+        graph = integer_graph | quantity_graph | with_cents | cents_only
+        self.tagger = self.add_tokens(graph)
+
+    def build_verbalizer(self):
+        currency = delete('currency: "') + self.NOT_QUOTE.plus + delete('"')
+        value = delete(' value: "') + self.NOT_QUOTE.plus + delete('"')
+        decimal = delete(' decimal: "') + self.NOT_QUOTE.plus + delete('"')
+        quantity = delete(' quantity: "') + self.NOT_QUOTE.plus + delete('"')
+
+        graph = currency + value
+        graph += closure(insert(".") + self.DELETE_SPACE + decimal, 0, 1)
+        graph += closure(insert(" ") + self.DELETE_SPACE + quantity, 0, 1)
+        graph |= currency + insert("0.") + self.DELETE_SPACE + decimal
+
+        self.verbalizer = self.delete_tokens(graph)

itn/english/rules/time.py

@@ -12,14 +12,24 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from pynini import closure, cross, string_file
+from pynini import closure, cross, string_file, union
 from pynini.lib.pynutil import delete, insert
 
 from itn.english.rules.cardinal import Cardinal
 from tn.processor import Processor
 from tn.utils import get_abs_path
 
 
+def _num_to_word(n):
+    ones = ["", "one", "two", "three", "four", "five", "six", "seven", "eight", "nine",
+            "ten", "eleven", "twelve", "thirteen", "fourteen", "fifteen", "sixteen",
+            "seventeen", "eighteen", "nineteen"]
+    tens = ["", "", "twenty", "thirty", "forty", "fifty", "sixty", "seventy", "eighty", "ninety"]
+    if n < 20:
+        return ones[n]
+    return tens[n // 10] + (" " + ones[n % 10] if n % 10 else "")
+
+
 class Time(Processor):
 
     def __init__(self, cardinal=None):
@@ -29,34 +39,58 @@ def __init__(self, cardinal=None):
         self.build_verbalizer()
 
     def build_tagger(self):
-        digit = string_file(get_abs_path("../itn/english/data/numbers/digit.tsv"))
-        teen = string_file(get_abs_path("../itn/english/data/numbers/teen.tsv"))
-        ties = string_file(get_abs_path("../itn/english/data/numbers/ties.tsv"))
+        cardinal_graph = self.cardinal.graph
         time_suffix = string_file(get_abs_path("../itn/english/data/time/time_suffix.tsv"))
         time_zone = string_file(get_abs_path("../itn/english/data/time/time_zone.tsv"))
         ds = delete(" ")
 
-        hour = teen | (insert("0") + digit)
-        minute = teen | (ties + (ds + digit | insert("0"))) | insert("0") + digit
+        # hours: 0-23, only valid hour words, with zero-padding
+        hour_labels = [_num_to_word(x) for x in range(0, 24) if _num_to_word(x)]
+        hour_padded = union(*[cross(_num_to_word(x), f"{x:02d}") for x in range(0, 24) if _num_to_word(x)])
+        # minutes: 1-9 (single), 10-59 (double)
+        min_single = [_num_to_word(x) for x in range(1, 10)]
+        min_double = [_num_to_word(x) for x in range(10, 60)]
+        graph_min_single = union(*[cross(_num_to_word(x), f"{x:02d}") for x in range(1, 10)])
+        graph_min_double = union(*[cross(_num_to_word(x), str(x)) for x in range(10, 60)])
 
-        # two thirty => 02:30
-        graph = insert('hour: "') + hour + insert('" ') + ds + insert('minute: "') + minute + insert('"')
-        # eight oclock => 08:00
-        oclock = cross("o'clock", "") | cross("oclock", "")
-        graph |= insert('hour: "') + hour + insert('" minute: "00"') + ds + oclock
+        hour = insert('hour: "') + hour_padded + insert('"')
+        oclock = cross("o'clock", "") | cross("oclock", "") | cross("hundred hours", "")
+        minute = (
+            oclock + insert("00")
+            | delete("o") + ds + graph_min_single
+            | graph_min_double
+        )
 
         suffix = ds + insert(' noon: "') + time_suffix + insert('"')
         zone = ds + insert(' zone: "') + time_zone + insert('"')
-        graph += suffix.ques + zone.ques
 
-        self.tagger = self.add_tokens(graph)
+        # "eight oclock" (no suffix needed)
+        graph_oclock = hour + ds + insert(' minute: "') + oclock + insert('00"')
+        # "two o five" (no suffix needed)
+        graph_o_min = hour + ds + insert(' minute: "') + delete("o") + ds + graph_min_single + insert('"')
+        # "two pm", "three am" (hour + suffix, minutes = 00)
+        graph_h_suffix = hour + insert(' minute: "00"') + suffix + closure(zone, 0, 1)
+        # "two thirty am" (hour + minute + suffix required)
+        graph_hm_suffix = (
+            hour + ds + insert(' minute: "') + graph_min_double + insert('"')
+            + suffix + closure(zone, 0, 1)
+        )
+        # "half past two", "quarter past two"
+        graph_half_quarter = (
+            insert('minute: "')
+            + (cross("half", "30") | cross("quarter", "15"))
+            + insert('"')
+            + ds + delete("past") + ds
+            + hour
+        )
+
+        final_graph = graph_oclock | graph_o_min | graph_h_suffix | graph_hm_suffix | graph_half_quarter
+        self.tagger = self.add_tokens(final_graph)
 
     def build_verbalizer(self):
-        hours = delete('hour: "') + self.NOT_QUOTE.plus + delete('"')
-        minutes = delete(' minute: "') + self.NOT_QUOTE.plus + delete('"')
-        suffix = delete(' noon: "') + self.NOT_QUOTE.plus + delete('"')
-        zone = delete(' zone: "') + self.NOT_QUOTE.plus + delete('"')
-        graph = hours + insert(":") + self.DELETE_SPACE + minutes
-        graph += closure(insert(" ") + self.DELETE_SPACE + suffix, 0, 1)
-        graph += closure(insert(" ") + self.DELETE_SPACE + zone, 0, 1)
+        hour = delete('hour: "') + self.NOT_QUOTE.plus + delete('"')
+        minute = delete(' minute: "') + self.NOT_QUOTE.plus + delete('"')
+        noon = delete(' noon: "') + self.NOT_QUOTE.plus + delete('"')
+        graph = hour + insert(":") + self.DELETE_SPACE + minute
+        graph += closure(insert(" ") + self.DELETE_SPACE + noon, 0, 1)
         self.verbalizer = self.delete_tokens(graph)

itn/english/test/data/en_ordinal.txt

@@ -7,7 +7,6 @@ eleventh => 11th
 twelfth => 12th
 thirteenth => 13th
 twenty first => 21st
-thirty second => 32nd
 forty third => 43rd
 one hundredth => 100th
 one hundred and first => 101st