Update tests and code for CLDR 48.2

- Unit IDs renamed in 48 still work (with deprecation warnings raised)

Author: akx

babel/core.py

@@ -1372,7 +1372,7 @@ def get_cldr_version() -> str:
     function is a string representing a version number, e.g. '47'.
 
     >>> get_cldr_version()
-    '47'
+    '48'
 
     .. versionadded:: 2.18
 

babel/numbers.py

@@ -741,7 +741,7 @@ def format_currency(
     >>> format_currency(1099.98, 'JPY', locale='en_US')
     '\\xa51,100'
     >>> format_currency(1099.98, 'COP', '#,##0.00', locale='es_ES')
-    '1.099,98'
+    '1.100'
 
     However, the number of decimal digits can be overridden from the currency
     information, by setting the last parameter to ``False``:

babel/units.py

@@ -1,11 +1,22 @@
 from __future__ import annotations
 
 import decimal
+import warnings
 from typing import Literal
 
 from babel.core import Locale
 from babel.numbers import LC_NUMERIC, format_decimal
 
+_DEPRECATED_UNIT_IDS: dict[str, str] = {
+    # Unit IDs deprecated in CLDR 48
+    "concentr-permillion": "concentr-part-per-1e6",
+    "concentr-portion": "concentr-part",
+    "concentr-portion-per-1e9": "concentr-part-per-1e9",
+    "permillion": "part-per-1e6",
+    "portion": "part",
+    "portion-per-1e9": "part-per-1e9",
+}
+
 
 class UnknownUnitError(ValueError):
     def __init__(self, unit: str, locale: Locale) -> None:
@@ -66,6 +77,19 @@ def _find_unit_pattern(unit_id: str, locale: Locale | str | None = None) -> str
     unit_patterns: dict[str, str] = locale._data["unit_patterns"]
     if unit_id in unit_patterns:
         return unit_id
+
+    # Only if a direct hit didn't work, try deprecated units...
+    new_id = _DEPRECATED_UNIT_IDS.get(unit_id)
+    if new_id is not None:
+        warnings.warn(
+            f"Unit identifier {unit_id!r} is deprecated; use {new_id!r} instead.",
+            DeprecationWarning,
+            stacklevel=4,
+        )
+        if new_id in unit_patterns:
+            return new_id
+
+    # ... and finally fuzzy matching.
     for unit_pattern in sorted(unit_patterns, key=len):
         if unit_pattern.endswith(unit_id):
             return unit_pattern

tests/test_core.py

@@ -413,4 +413,4 @@ def test_locale_parse_empty():
 
 
 def test_get_cldr_version():
-    assert core.get_cldr_version() == "47"
+    assert core.get_cldr_version() == "48"

tests/test_dates.py

@@ -613,10 +613,10 @@ def test_format_time(timezone_getter):
 def test_format_skeleton(timezone_getter):
     dt = datetime(2007, 4, 1, 15, 30)
     assert (dates.format_skeleton('yMEd', dt, locale='en_US') == 'Sun, 4/1/2007')
-    assert (dates.format_skeleton('yMEd', dt, locale='th') == 'อา. 1/4/2007')
+    assert (dates.format_skeleton('yMEd', dt, locale='th') == 'อาทิตย์ 1/4/2007')
 
     assert (dates.format_skeleton('EHm', dt, locale='en') == 'Sun 15:30')
-    assert (dates.format_skeleton('EHm', dt, tzinfo=timezone_getter('Asia/Bangkok'), locale='th') == 'อา. 22:30 น.')
+    assert (dates.format_skeleton('EHm', dt, tzinfo=timezone_getter('Asia/Bangkok'), locale='th') == 'อาทิตย์ 22:30 น.')
 
 
 @pytest.mark.parametrize(('skeleton', 'expected'), [
@@ -1194,8 +1194,8 @@ def test_issue_1192():
     # The actual returned value here is not actually strictly specified ("get_timezone_name"
     # is not an operation specified as such). Issue #1192 concerned this invocation returning
     # the invalid "no inheritance marker" value; _that_ should never be returned here.
-    # IOW, if the below "Hawaii-Aleutian Time" changes with e.g. CLDR updates, that's fine.
-    assert dates.get_timezone_name('Pacific/Honolulu', 'short', locale='en_GB') == "Hawaii-Aleutian Time"
+    # IOW, if the below "United States (Honolulu) Time" changes with e.g. CLDR updates, that's fine.
+    assert dates.get_timezone_name('Pacific/Honolulu', 'short', locale='en_GB') == "United States (Honolulu) Time"
 
 
 @pytest.mark.xfail

tests/test_lists.py

@@ -9,7 +9,7 @@
     (['string1', 'string2'], 'en', 'string1 and string2'),
     (['string1', 'string2', 'string3'], 'en', 'string1, string2, and string3'),
     (['string1', 'string2', 'string3'], 'zh', 'string1、string2和string3'),
-    (['string1', 'string2', 'string3', 'string4'], 'ne', 'string1,string2, string3 र string4'),
+    (['string1', 'string2', 'string3', 'string4'], 'ne', 'string1, string2, string3 र string4'),
 ])
 def test_format_list(list, locale, expected):
     assert lists.format_list(list, locale=locale) == expected

tests/test_numbers.py

@@ -40,7 +40,7 @@ def test_list_currencies():
 
     assert list_currencies(locale='pa_Arab') == {'PKR', 'INR', 'EUR'}
 
-    assert len(list_currencies()) == 307
+    assert len(list_currencies()) == 308
 
 
 def test_validate_currency():
@@ -297,7 +297,7 @@ def test_format_currency_format_type():
     assert (numbers.format_currency(1099.98, 'JPY', locale='en_US')
             == '\xa51,100')
     assert (numbers.format_currency(1099.98, 'COP', '#,##0.00', locale='es_ES')
-            == '1.099,98')
+            == '1.100')
     assert (numbers.format_currency(1099.98, 'JPY', locale='en_US',
                                     currency_digits=False)
             == '\xa51,099.98')
@@ -320,7 +320,7 @@ def test_format_compact_currency():
     assert numbers.format_compact_currency(123, 'EUR', locale='yav', format_type="short") == '€\xa0123'
     assert numbers.format_compact_currency(12345, 'EUR', locale='yav', format_type="short") == '€\xa012K'
     assert numbers.format_compact_currency(123456789, 'EUR', locale='de_DE', fraction_digits=1) == '123,5\xa0Mio.\xa0€'
-    assert numbers.format_compact_currency(123456789, 'USD', locale='ar_EG', fraction_digits=2, format_type="short", numbering_system="default") == '123٫46\xa0مليون\xa0US$'
+    assert numbers.format_compact_currency(123456789, 'USD', locale='ar_EG', fraction_digits=2, format_type="short", numbering_system="default") == '\u200f123٫46\xa0مليون\xa0US$'
 
 
 def test_format_compact_currency_invalid_format_type():

tests/test_support_format.py

@@ -53,7 +53,7 @@ def test_format_currency(ar_eg_format, en_us_format):
 def test_format_compact_currency(ar_eg_format, en_us_format):
     assert en_us_format.compact_currency(1099.98, 'USD') == '$1K'
     assert en_us_format.compact_currency(Decimal("1099.98"), 'USD') == '$1K'
-    assert ar_eg_format.compact_currency(1099.98, 'EGP') == '1\xa0ألف\xa0ج.م.\u200f'
+    assert ar_eg_format.compact_currency(1099.98, 'EGP') == '\u200f1\xa0ألف\xa0ج.م.\u200f'
 
 
 def test_format_percent(ar_eg_format, en_us_format):

tests/test_units.py

@@ -8,9 +8,9 @@
     ('speed-light-speed', 1, '1 světlo'),
     ('speed-light-speed', 2, '2 světla'),
     ('speed-light-speed', 5, '5 světel'),
-    ('concentr-portion-per-1e9', 1, '1 částice na miliardu'),
-    ('concentr-portion-per-1e9', 2, '2 částice na miliardu'),
-    ('concentr-portion-per-1e9', 5, '5 částic na miliardu'),
+    ('concentr-part-per-1e9', 1, '1 částice na miliardu'),
+    ('concentr-part-per-1e9', 2, '2 částice na miliardu'),
+    ('concentr-part-per-1e9', 5, '5 částic na miliardu'),
     ('duration-night', 1, '1 noc'),
     ('duration-night', 2, '2 noci'),
     ('duration-night', 5, '5 nocí'),
@@ -29,3 +29,9 @@ def test_new_cldr46_units(unit, count, expected):
 ])
 def test_issue_1217(count, unit, locale, length, expected):
     assert format_unit(count, unit, length, locale=locale) == expected
+
+
+def test_deprecated_unit_ids():
+    for id in ("concentr-permillion", "concentr-portion", "concentr-portion-per-1e9"):
+        with pytest.warns(DeprecationWarning, match=id):
+            format_unit(1, id, locale='en')