Support the CLDR 49 ordinal day of month pattern field

babel/core.py

@@ -1063,6 +1063,22 @@ def ordinal_form(self) -> PluralRule:
         """
         return self._data.get('ordinal_form', _default_plural_rule)
 
+    @property
+    def ordinal_day_formats(self) -> localedata.LocaleDataDict:
+        """Locale patterns for ordinal days of the month, e.g. "13th" in English.
+
+        The keys are `ordinal_form` plural categories, and ``{0}`` is replaced
+        with the day of the month.
+
+        Introduced in CLDR 49.
+
+        .. note:: The format of the value returned may change between
+                  Babel versions.
+
+        .. versionadded:: 2.19
+        """
+        return self._data.get('ordinal_day_formats', localedata.LocaleDataDict({}))
+
     @property
     def measurement_systems(self) -> localedata.LocaleDataDict:
         """Localized names for various measurement systems.

babel/dates.py

@@ -1444,7 +1444,7 @@ def __getitem__(self, name: str) -> str:
         elif char in ('w', 'W'):
             return self.format_week(char, num)
         elif char == 'd':
-            return self.format(self.value.day, num)
+            return self.format_day_of_month(num)
         elif char == 'D':
             return self.format_day_of_year(num)
         elif char == 'F':
@@ -1580,6 +1580,29 @@ def format_weekday(self, char: str = 'E', num: int = 4) -> str:
         context = "stand-alone" if char == "c" else "format"
         return get_day_names(width, context, self.locale)[weekday]
 
+    def format_day_of_month(self, num: int) -> str:
+        """
+        Return day of month from parsed datetime according to format pattern.
+
+        >>> from datetime import date
+        >>> format = DateTimeFormat(date(2016, 2, 28), Locale.parse('en_US'))
+        >>> format.format_day_of_month(1)
+        '28'
+
+        'd': Day of month - Use one or two for the minimum number of digits, or
+             three for the ordinal form (CLDR 49 and later), e.g. "28th" in
+             English. Locales without ordinal forms fall back to a single "d".
+        """
+        if num == 3:
+            ordinal_pattern = self.locale.ordinal_day_formats.get(
+                self.locale.ordinal_form(self.value.day),
+            )
+            if ordinal_pattern:
+                return ordinal_pattern.replace('{0}', str(self.value.day))
+            # No ordinal forms for this locale; format like a single "d" would.
+            num = 1
+        return self.format(self.value.day, num)
+
     def format_day_of_year(self, num: int) -> str:
         return self.format(self.get_day_of_year(), num)
 
@@ -1753,7 +1776,7 @@ def get_week_number(self, day_of_period: int, day_of_week: int | None = None) ->
     'Q': [1, 2, 3, 4, 5], 'q': [1, 2, 3, 4, 5],                         # quarter
     'M': [1, 2, 3, 4, 5], 'L': [1, 2, 3, 4, 5],                         # month
     'w': [1, 2], 'W': [1],                                              # week
-    'd': [1, 2], 'D': [1, 2, 3], 'F': [1], 'g': None,                   # day
+    'd': [1, 2, 3], 'D': [1, 2, 3], 'F': [1], 'g': None,                # day
     'E': [1, 2, 3, 4, 5, 6], 'e': [1, 2, 3, 4, 5, 6], 'c': [1, 3, 4, 5, 6],  # week day
     'a': [1, 2, 3, 4, 5], 'b': [1, 2, 3, 4, 5], 'B': [1, 2, 3, 4, 5],   # period
     'h': [1, 2], 'H': [1, 2], 'K': [1, 2], 'k': [1, 2],                 # hour

scripts/import_cldr.py

@@ -448,6 +448,7 @@ def _process_local_datas(sup, srcdir, destdir, force=True, dump_json=False):
             parse_calendar_quarters(data, calendar)
             parse_calendar_eras(data, calendar)
             parse_calendar_periods(data, calendar)
+            parse_calendar_ordinal_day_formats(data, calendar)
             parse_calendar_date_formats(data, calendar)
             parse_calendar_time_formats(data, calendar)
             parse_calendar_datetime_skeletons(data, calendar)
@@ -720,6 +721,25 @@ def parse_calendar_periods(data, calendar):
                     dest_dict[period_type] = str(day_period.text)
 
 
+def parse_calendar_ordinal_day_formats(data, calendar):
+    # Ordinal days of the month, e.g. "1st" in English (CLDR 49 and later).
+    # These are keyed by the ordinal plural categories of the locale.
+    ordinal_day_formats = data.setdefault('ordinal_day_formats', {})
+    for day_of_month_context in calendar.findall('dayOfMonths/dayOfMonthContext'):
+        if day_of_month_context.attrib.get('type') != 'format':
+            continue
+        for day_of_month_width in day_of_month_context.findall('dayOfMonthWidth'):
+            if day_of_month_width.attrib.get('type') != 'abbreviated':
+                continue
+            for elem in day_of_month_width.findall('dayOfMonth'):
+                ordinal = elem.attrib.get('ordinal')
+                if ordinal is None or elem.text is None:
+                    continue
+                if _should_skip_elem(elem, ordinal, ordinal_day_formats):
+                    continue
+                ordinal_day_formats[ordinal] = str(elem.text)
+
+
 def parse_calendar_date_formats(data, calendar):
     date_formats = data.setdefault('date_formats', {})
     for format in calendar.findall('dateFormats'):

tests/test_dates.py

@@ -553,6 +553,49 @@ def test_parse_pattern():
     assert dates.parse_pattern("hh' o''clock'").format == "%(hh)s o'clock"
 
 
+def test_parse_pattern_ordinal_day():
+    # CLDR 49 added the ordinal day of month field, e.g. "Feb 1st"
+    assert dates.parse_pattern("MMM ddd, y").format == '%(MMM)s %(ddd)s, %(y)s'
+
+
+def test_parse_pattern_rejects_unsupported_day_length():
+    with pytest.raises(ValueError):
+        dates.parse_pattern("dddd")
+
+
+def test_format_ordinal_day_of_month(monkeypatch):
+    # CLDR 49 and later provide one ordinal pattern per ordinal plural category
+    locale = Locale('en')
+    ordinal_day_formats = {
+        'one': '{0}st',
+        'two': '{0}nd',
+        'few': '{0}rd',
+        'other': '{0}th',
+    }
+    monkeypatch.setitem(locale._data, 'ordinal_day_formats', ordinal_day_formats)
+
+    def format_day(day: int) -> str:
+        return dates.format_datetime(datetime(2016, 2, day), "MMM ddd, y", locale=locale)
+
+    assert format_day(1) == 'Feb 1st, 2016'
+    assert format_day(2) == 'Feb 2nd, 2016'
+    assert format_day(3) == 'Feb 3rd, 2016'
+    assert format_day(4) == 'Feb 4th, 2016'
+    assert format_day(11) == 'Feb 11th, 2016'
+    assert format_day(22) == 'Feb 22nd, 2016'
+
+
+def test_format_ordinal_day_of_month_falls_back_without_forms():
+    # Locales without ordinal forms format the day like a single "d" would,
+    # without zero padding (CLDR 49 has no forms for German, for instance).
+    day = datetime(2016, 2, 1)
+    assert (
+        dates.format_datetime(day, "MMM ddd, y", locale='de')
+        == dates.format_datetime(day, "MMM d, y", locale='de')
+    )
+    assert dates.format_datetime(day, "dd.MM.y", locale='de') == '01.02.2016'
+
+
 def test_lithuanian_long_format():
     assert (
         dates.format_date(date(2015, 12, 10), locale='lt_LT', format='long') ==