Resolved the conflicts with mm_yyyy and date ranges and added the previously removed failing test cases.

Signed-off-by: Namrata Gachchi <ngachchi@nvidia.com>

Author: ngachchi

nemo_text_processing/text_normalization/hi/taggers/date.py

@@ -85,7 +85,7 @@ def __init__(self, cardinal: GraphFst):
         graph_mm_dd_yyyy += pynutil.insert(" preserve_order: true ")
 
         graph_mm_yyyy = (
-            months_graph + (delete_dash | pynini.accep("")) + years_graph + pynutil.insert(" preserve_order: true ")
+            months_graph + delete_dash + insert_space + years_graph
         )
 
         graph_year_suffix = era_graph
@@ -108,7 +108,7 @@ def __init__(self, cardinal: GraphFst):
             | graph_mm_dd
             | pynutil.add_weight(graph_dd_mm_yyyy, -0.001)
             | graph_mm_dd_yyyy
-            | graph_mm_yyyy
+            | pynutil.add_weight(graph_mm_yyyy, -0.2)
             | pynutil.add_weight(graph_year_suffix, -0.001)
             | pynutil.add_weight(graph_range, -0.005)
         )

nemo_text_processing/text_normalization/hi/verbalizers/date.py

@@ -39,9 +39,9 @@ def __init__(self):
 
         year = pynutil.delete("year: \"") + pynini.closure(NEMO_NOT_QUOTE, 1) + pynutil.delete("\"")
 
-        era = pynutil.delete("era: \"") + pynini.closure(NEMO_NOT_QUOTE, 1) + pynutil.delete("\"")
+        graph_era = pynutil.delete("era: \"") + pynini.closure(NEMO_NOT_QUOTE, 1) + pynutil.delete("\"")
 
-        range = pynutil.delete("text: \"") + pynini.closure(NEMO_NOT_QUOTE, 1) + pynutil.delete("\"")
+        graph_range = pynutil.delete("text: \"") + pynini.closure(NEMO_NOT_QUOTE, 1) + pynutil.delete("\"")
 
         graph_dd_mm = day + NEMO_SPACE + month
 
@@ -53,10 +53,6 @@ def __init__(self):
 
         graph_mm_yyyy = month + NEMO_SPACE + year
 
-        graph_era = era
-
-        graph_range = range
-
         optional_preserve_order = pynini.closure(
             pynutil.delete("preserve_order:") + delete_space + pynutil.delete("true") + delete_space
             | pynutil.delete("field_order:")

tests/nemo_text_processing/hi/data_text_normalization/test_cases_date.txt

@@ -14,6 +14,7 @@
 १०-२९-२०००~अक्टूबर उनतीस दो हज़ार
 ११-१४-११००~नवंबर चौदह ग्यारह सौ
 ०३-२०१०~मार्च दो हज़ार दस
+११-२०२४~नवंबर दो हज़ार चौबीस
 २०७०~दो हज़ार सत्तर
 २०२४~दो हज़ार चौबीस
 १२० ई. पू.~एक सौ बीस ईसा पूर्व