[pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

Author: pre-commit-ci[bot]

nemo_text_processing/text_normalization/fr/taggers/date.py

@@ -1,20 +1,23 @@
 import pynini
 from pynini.lib import pynutil
-from nemo_text_processing.text_normalization.fr.utils import get_abs_path
-
-from nemo_text_processing.text_normalization.en.graph_utils import GraphFst, NEMO_DIGIT
 
+from nemo_text_processing.text_normalization.en.graph_utils import NEMO_DIGIT, GraphFst
+from nemo_text_processing.text_normalization.fr.utils import get_abs_path
 
-# TODO: add articles? 'le...' 
+# TODO: add articles? 'le...'
 
 month_numbers = pynini.string_file(get_abs_path("data/dates/months.tsv"))
 eras = pynini.string_file(get_abs_path("data/dates/eras.tsv"))
-delete_leading_zero = (pynutil.delete("0") | (NEMO_DIGIT - "0")) + NEMO_DIGIT #reminder, NEMO_DIGIT = filter on digits
+delete_leading_zero = (
+    pynutil.delete("0") | (NEMO_DIGIT - "0")
+) + NEMO_DIGIT  # reminder, NEMO_DIGIT = filter on digits
+
 
 class DateFst(GraphFst):
     ''' Finite state transducer for classyfing dates, e.g.:
         '02.03.2003' -> date {day: 'deux' month: 'mai' year: 'deux mille trois' preserve order: true} 
     '''
+
     def __init__(self, cardinal: GraphFst, deterministic: bool = True):
         super().__init__(name="dates", kind="classify")
 
@@ -23,10 +26,10 @@ def __init__(self, cardinal: GraphFst, deterministic: bool = True):
         # 'le' -> 'le', 'les' -> 'les'
         le_determiner = pynini.accep("le ") | pynini.accep("les ")
         self.optional_le = pynini.closure(le_determiner, 0, 1)
-        
-        # '01' -> 'un' 
+
+        # '01' -> 'un'
         optional_leading_zero = delete_leading_zero | NEMO_DIGIT
-        valid_day_number = pynini.union(*[str(x) for x in range(1,32)]) 
+        valid_day_number = pynini.union(*[str(x) for x in range(1, 32)])
         premier = pynini.string_map([("1", "premier")])
         day_number_to_word = premier | cardinal_graph
 
@@ -59,7 +62,7 @@ def __init__(self, cardinal: GraphFst, deterministic: bool = True):
             )
 
         # Accepts "janvier", "février", etc
-        month_name_graph = pynutil.insert("month: \"") +  month_numbers.project("output") + pynutil.insert("\"")
+        month_name_graph = pynutil.insert("month: \"") + month_numbers.project("output") + pynutil.insert("\"")
 
         self.fst |= (
             pynutil.insert("date { ")
@@ -73,9 +76,8 @@ def __init__(self, cardinal: GraphFst, deterministic: bool = True):
         # Accepts "70s", "80s", etc
         self.fst |= pynutil.insert("date { year: \"") + eras + pynutil.insert("\" preserve_order: true }")
 
-
         # Accepts date ranges, "17-18-19 juin"  -> date { day: "17" day: "18": day: "19"}
-        for separator in ["-", "/"]: 
+        for separator in ["-", "/"]:
             day_range_graph = (
                 pynutil.insert("day: \"")
                 + pynini.closure(digit_to_day + pynutil.delete(separator) + pynutil.insert(" "), 1)
@@ -95,16 +97,17 @@ def __init__(self, cardinal: GraphFst, deterministic: bool = True):
         self.fst = self.fst.optimize()
 
 
-
 def apply_fst(text, fst):
     try:
         output = pynini.shortestpath(text @ fst).string()
         print(f"'{text}' --> '{output}'")
     except pynini.FstOpError:
         print(f"Error: No valid output with given input: '{text}'")
 
+
 if __name__ == "__main__":
     from nemo_text_processing.text_normalization.fr.taggers.cardinal import CardinalFst
+
     fst = DateFst(CardinalFst())
 
     print('DETERMINER')
@@ -132,7 +135,7 @@ def apply_fst(text, fst):
     apply_fst("02/03/2003", fst.fst)
     apply_fst("02-03-2003", fst.fst)
     apply_fst("le 02.03.2003", fst.fst)
-    
+
     apply_fst("02.03", fst.fst)
     apply_fst("17 janvier", fst.fst)
     apply_fst("10 mars 2023", fst.fst)
@@ -142,4 +145,6 @@ def apply_fst(text, fst):
     apply_fst("80s", fst.fst)
 
     print("\nDATE RANGES")
-    apply_fst("les 17/18/19 juin", fst.fst) # returns: date { day: "les dix-sept" day: "dix-huit" day: "dix-neuf" month: "juin" preserve_order: true }
+    apply_fst(
+        "les 17/18/19 juin", fst.fst
+    )  # returns: date { day: "les dix-sept" day: "dix-huit" day: "dix-neuf" month: "juin" preserve_order: true }

nemo_text_processing/text_normalization/fr/taggers/tokenize_and_classify.py

@@ -26,12 +26,12 @@
 )
 from nemo_text_processing.text_normalization.en.taggers.punctuation import PunctuationFst
 from nemo_text_processing.text_normalization.fr.taggers.cardinal import CardinalFst
+from nemo_text_processing.text_normalization.fr.taggers.date import DateFst
 from nemo_text_processing.text_normalization.fr.taggers.decimals import DecimalFst
 from nemo_text_processing.text_normalization.fr.taggers.fraction import FractionFst
 from nemo_text_processing.text_normalization.fr.taggers.ordinal import OrdinalFst
 from nemo_text_processing.text_normalization.fr.taggers.whitelist import WhiteListFst
 from nemo_text_processing.text_normalization.fr.taggers.word import WordFst
-from nemo_text_processing.text_normalization.fr.taggers.date import DateFst
 from nemo_text_processing.utils.logging import logger
 
 
@@ -86,7 +86,7 @@ def __init__(
             self.whitelist = WhiteListFst(input_case=input_case, deterministic=deterministic, input_file=whitelist)
             whitelist_graph = self.whitelist.fst
             punct_graph = PunctuationFst(deterministic=deterministic).fst
-            
+
             self.date = DateFst(self.cardinal, deterministic=deterministic)
             date_graph = self.date.fst
 

nemo_text_processing/text_normalization/fr/verbalizers/date.py

@@ -19,9 +19,10 @@
     NEMO_NOT_QUOTE,
     NEMO_SPACE,
     GraphFst,
-    delete_preserve_order
+    delete_preserve_order,
 )
 
+
 class DateFst(GraphFst):
     """
     Finite state transducer for verbalizing date, e.g.
@@ -50,13 +51,15 @@ def __init__(self, deterministic: bool = True):
         delete_tokens = self.delete_tokens(self.graph)
         self.fst = delete_tokens.optimize()
 
+
 def apply_fst(text, fst):
     try:
         output = pynini.shortestpath(text @ fst).string()
         print(f"'{text}' --> '{output}'")
     except pynini.FstOpError:
         print(f"Error: No valid output with given input: '{text}'")
 
+
 if __name__ == "__main__":
     fst = DateFst()
 

nemo_text_processing/text_normalization/fr/verbalizers/verbalize.py

@@ -14,10 +14,10 @@
 from nemo_text_processing.text_normalization.en.graph_utils import GraphFst
 from nemo_text_processing.text_normalization.en.verbalizers.whitelist import WhiteListFst
 from nemo_text_processing.text_normalization.fr.verbalizers.cardinal import CardinalFst
+from nemo_text_processing.text_normalization.fr.verbalizers.date import DateFst
 from nemo_text_processing.text_normalization.fr.verbalizers.decimals import DecimalFst
 from nemo_text_processing.text_normalization.fr.verbalizers.fraction import FractionFst
 from nemo_text_processing.text_normalization.fr.verbalizers.ordinal import OrdinalFst
-from nemo_text_processing.text_normalization.fr.verbalizers.date import DateFst
 
 
 class VerbalizeFst(GraphFst):

nemo_text_processing/text_normalization/fr_tutorial/taggers/my_test_script.py

@@ -3,47 +3,47 @@
 
 from nemo_text_processing.text_normalization.fr.utils import get_abs_path
 
+
 def apply_fst(text, fst):
-  """ Given a string input, returns the output string
+    """ Given a string input, returns the output string
   produced by traversing the path with lowest weight.
   If no valid path accepts input string, returns an
   error.
   """
-  try:
-     print(pynini.shortestpath(text @ fst).string())
-  except pynini.FstOpError:
-    print(f"Error: No valid output with given input: '{text}'")
-
-zero = pynini.string_map([("zéro","0")]) # French only pronounces zeroes as stand alone
-digits_map = pynini.string_map([ # pynini function that creates explicit input-output mappings for a WFST
-				("un","1"),
-				("une","1"),
-				("deux","2"),
-				("trois","3"),
-				("quatre","4"),
-				("cinq","5"),
-				("six","6"),
-				("sept","7"),
-				("huit","8"),
-				("neuf","9")
-])
+    try:
+        print(pynini.shortestpath(text @ fst).string())
+    except pynini.FstOpError:
+        print(f"Error: No valid output with given input: '{text}'")
+
+
+zero = pynini.string_map([("zéro", "0")])  # French only pronounces zeroes as stand alone
+digits_map = pynini.string_map(
+    [  # pynini function that creates explicit input-output mappings for a WFST
+        ("un", "1"),
+        ("une", "1"),
+        ("deux", "2"),
+        ("trois", "3"),
+        ("quatre", "4"),
+        ("cinq", "5"),
+        ("six", "6"),
+        ("sept", "7"),
+        ("huit", "8"),
+        ("neuf", "9"),
+    ]
+)
 
 digits = pynini.string_file("data/numbers/digits.tsv")
 
-teens = pynini.string_map([
-    ("onze", "11"),
-    ("douze", "12"), 
-    ("treize", "13"),
-    ("quatorze", "14"),
-    ("quinze", "16"), 
-])
+teens = pynini.string_map([("onze", "11"), ("douze", "12"), ("treize", "13"), ("quatorze", "14"), ("quinze", "16"),])
 
 tens = pynini.string_map([("dix", "1")])
-delete_hyphen = pynini.closure(pynutil.delete("-"), 0, 1) # Applies a closure from 0-1 of operation. Equivalent to regex /?/
+delete_hyphen = pynini.closure(
+    pynutil.delete("-"), 0, 1
+)  # Applies a closure from 0-1 of operation. Equivalent to regex /?/
 
 graph_tens = tens + delete_hyphen + digits
 graph_tens_and_teens = graph_tens | teens
 
 graph_digits = digits | pynutil.insert("0")
 
-apply_fst("un", graph_tens_and_teens)
+apply_fst("un", graph_tens_and_teens)